You are on page 1of 531

Ancient Egyptian Chronology

HANDBOOK OF ORIENTAL STUDIES


SECTION ONE
THE NEAR AND MIDDLE EAST

Ancient Near East

Editor-in-Chief
W. H. van Soldt

Editors
G. Beckman • C. Leitz • B. A. Levine
P. Michalowski • P. Miglus

Middle East
R. S. O’Fahey • C. H. M. Versteegh

VOLUME EIGHTY-THREE
Ancient Egyptian Chronology

Edited by
Erik Hornung, Rolf Krauss, and David A. Warburton

BRILL
LEIDEN • BOSTON
2006
This book is printed on acid-free paper.

Library of Congress Cataloging-in-Publication Data

Ancient Egyptian chronology / edited by Erik Hornung, Rolf Krauss, and


David A. Warburton; with the assistance of Marianne Eaton-Krauss.
p. cm. — (Handbook of Oriental studies. Section 1, The Near and
Middle East ; v. 83)
Includes bibliographical references and index.
ISBN-13: 978-90-04-11385-5
ISBN-10: 90-04-11385-1
1. Egypt—History—To 332 B.C.—Chronology. 2. Chronology, Egyptian.
3. Egypt—Antiquities. I. Hornung, Erik. II. Krauss, Rolf. III. Warburton, David.
IV. Eaton-Krauss, Marianne.

DT83.A6564 2006
932.002'02—dc22
2006049915

ISSN 0169-9423
ISBN-10 90 04 11385 1
ISBN-13 978 90 04 11385 5
© Copyright 2006 by Koninklijke Brill NV, Leiden, The Netherlands.
Koninklijke Brill NV incorporates the imprints Brill, Hotei Publishing,
IDC Publishers, Martinus Nijhoff Publishers, and VSP.
All rights reserved. No part of this publication may be reproduced, translated,
stored in a retrieval system, or transmitted in any form or by any means,
electronic, mechanical, photocopying, recording or otherwise, without prior
written permission from the publisher.
Authorization to photocopy items for internal or personal use is granted
by Brill provided that the appropriate fees are paid directly to
The Copyright Clearance Center, 222 Rosewood Drive, Suite 910,
Danvers, MA 01923, USA.
Fees are subject to change.
printed in the netherlands
EDITORIAL

With this volume of the Handbook of Oriental Studies (Handbuch der


Orientalistik) a new editorial board for the ancient Near East takes
over from the board that has edited the HdO volumes for so many
years. The new editors have been chosen because of their expertise in
the fields that pursue the study of the civilizations of the ancient Near
East. These editors are: G. Beckman (Hittitology), C. Leitz (Egyptology),
B. Levine (Hebrew Bible/Ancient Israel), P. Michalowski (Sumerology),
P. Miglus (Near Eastern Archaeology), and W. H. van Soldt (Assyriology,
Editor-in-Chief ).
The board has planned a number of new volumes in the HdO
series and expresses the hope that with these volumes the series will
resume its original function of a handbook for Ancient Near Eastern
Studies. At the same time, the new board is committed to the publi-
cation of volumes that had already been planned and accepted and
which are scheduled to appear in the near future. However, it should
be pointed out that the present board members do not necessarily agree
with the entire content of a volume that they have ‘inherited’ from
their predecessors.
Finally, we would like to thank the members of the former editorial
board for all their work for the HdO series.

The Editors
CONTENTS

Acknowledgments .......................................................................... xi

Introduction .................................................................................... 1
Erik Hornung

PART I

EGYPTIAN CHRONOGRAPHICAL TRADITION AND


METHOD OF DATING

1. Royal Annals ............................................................................ 19


The Editors
2. The Royal Canon of Turin .................................................... 26
Kim Ryholt
3. King Lists and Manetho’s Aigyptiaka ...................................... 33
The Editors
4. Genealogy and Chronology .................................................... 37
Morris L. Bierbrier
5. Methods of Dating and the Egyptian Calendar .................... 45
The Editors

PART II

RELATIVE CHRONOLOGY

1. Predynastic—Early Dynastic Chronology .............................. 55


Stan Hendrickx
2. Dynasties 0–2 ............................................................................ 94
Jochem Kahl
3. Dynasty 3 .................................................................................. 116
Stephan J. Seidlmayer
viii contents

4. Dynasties 4 to 5 ...................................................................... 124


Miroslav Verner
5. Dynasties 6 and 8 .................................................................. 144
Michel Baud
6. First Intermediate Period ...................................................... 159
Stephan J. Seidlmayer
7. Middle Kingdom and the Second Intermediate Period ...... 168
Thomas Schneider
8. New Kingdom ........................................................................ 197
Erik Hornung
9. Dynasty 21 .............................................................................. 218
Karl Jansen-Winkeln
10. Third Intermediate Period .................................................... 234
Karl Jansen-Winkeln
11. Saite and Persian Egypt ........................................................ 265
Leo Depuydt
12. Nubian Kingdoms, Dyn. 25 through the Kingdom of
Meroe ...................................................................................... 284
Karola Zibelius-Chen
13. Chronological Links between the Cuneiform World of the
Ancient Near East and Ancient Egypt ................................ 304
Jörg Klinger

PART III

ABSOLUTE CHRONOLOGY

1. Radiocarbon Dating and Egyptian Chronology .................. 327


Sturt Manning
2. Luminescence Dating of Egyptian Artefacts ........................ 356
Christian Goedicke
3. Dendrochronology .................................................................. 361
Otto Cichocki
4. Dates relating to Seasonal Phenomena and Miscellaneous
Astronomical Dates ................................................................ 369
Rolf Krauss
5. Astronomy on the Horizon—A Tool for ancient Egyptian
Chronology? ............................................................................ 380
Juan Antonio Belmonte
contents ix

6. Lunar Days, Lunar Months, and the Question of the


Civil based Lunar Calendar .................................................. 386
Rolf Krauss
7. Long-term Variation in the Motions of the Earth and the
Moon ........................................................................................ 392
Kurt Locher
8. Lunar Dates ............................................................................ 395
Rolf Krauss
9. The Heliacal Rising of Sirius ................................................ 432
Teije de Jong
10. Egyptian Sirius/Sothic Dates and the Question of the
Sirius based Lunar Calender ................................................ 439
Rolf Krauss
11. Foundations of Day-exact Chronology: 690 BC–332 BC .... 458
Leo Depuydt

PART IV

CONCLUSIONS AND CHRONOLOGICAL TABLES

1. Conclusions and a Postscript to Part II, Chapter I ............ 473


Rolf Krauss & David A. Warburton
2. Chronological Table for the Dynastic Period ...................... 490
The Editors
3. Tables for Kushite Rulers ...................................................... 496
Karola Zibelius-Chen

Abbreviations & Bibliography ...................................................... 501

Index of Personal Names .............................................................. 509


Index of Place Names .................................................................. 514
General Index ................................................................................ 516
ACKNOWLEDGMENTS

It was Wolfgang Schenkel who suggested the project of a Handbook of


Egyptian Chronology to Brill, Academic Publishers, nearly a decade ago.
He will have asked himself in the intervening years whether the work
would ever appear.
Co-editor Erik Hornung developed the concept of the Handbook
and together we selected the authors for each section. The subsequent
addition of David Alan Warburton to the team provided an impetus
to overcome the inertia that had set in. He and Marianne Eaton-Krauss
took over the thankless task of translating and/or editing contributions
by authors whose mother tongue was not English, except for Chapters
II. 1 and 11, III. 7, 9, and 11.
I am indebted to the Stiftung Preussischer Kulturbesitz who trans-
ferred me from the Ägyptisches Museum to the Museum für Vor- und
Frühgeschichte in 2001 to do research on the Calendrics and Chronology
of ancient Egypt, the Near East and Europe. Although the move
deprived me of direct access to an Egyptological library, the Handbook
would not have seen the light of the day without this change of for-
tune. Thanks to Wilfried Menghin, Director of the MVF, I could con-
centrate on the Handbook during 2004 and 2005.
Last but not least, I thank Juán José Granados who encouraged me
to complete the unloved task.

R.K. February 2006


INTRODUCTION

Erik Hornung

It is widely accepted that chronology is the “spine” of history. The


ensuing image is slightly distorted since one can only extend or shorten
a spine using violence, stretching it or forcing it like the giant Procrustes
into a bed which is far too short. In recent years there have been
repeated attempts to cut an entire century off history, and not merely
in the “Dark Ages” of antiquity, but even now for the supposedly
“invented” Middle Ages. It would thus be more suitable to view time
as the clothing of history: a garment can be non-violently adjusted to
a changing body. The image of time as a piece of clothing, the habit
of mankind and the dress of the deity, can be found in the poem “Song
of Time” by the Pakistani poet Muhammad Iqbal (1877–1938),1 who
was doubtless inspired by Goethe’s expression, “the living dress of the
divinity” created by the spirit of the Earth “on the whirring loom of
time” (Faustus, verses 508f.).
But, first a word on the temporal horizon of the Egyptians.2 Even
'Ankhtifi of Mo'alla looked hopefully to a future of “millions of years”,
during which no one would approach his deeds.3 Indeed, before the
end of the FIP, these “millions of years” rapidly became a common-
place (being known from a wish for King Merykare' in Asyut),4 ulti-
mately becoming a synonym of the twin temporal concepts neheh and
djet. This is most clearly expressed in the representation of the Dyn.
21 illustrated papyrus of Khonsumes in Vienna, where all three stand
together, raised on yokes and thus divine,5 subordinated to the solar

1
In A. Schimmel, ed., Botschaft des Ostens (Tübingen & Basel, 1977), 149.
2
E. Hornung, “Zeitliches Jenseits im alten Ägypten”, Eranos-Jahrbuch 47 (1978),
269–307, esp. 297–298.
3
Schenkel, Memphis, 47.
4
F. Ll. Griffith, The Inscriptions of Siût and Dêr Rîfeh (London, 1889), pl. 13, IV 22.
In Dyn. 6, Pepy II was wished “millions of sed-festivals” (Urk. I, 115, 1).
5
H. von Demel, “Der Totenpapyrus des Chonsu-Mes”, Jahrbuch der Kunsthistorischen
Sammlungen in Wien, N.S. 13 (1944), 1–16; A. Piankoff & N. Rambova, Mythological Papyri
(New York: BS XL, 1957), no. 16; djet does not appear as a divinity, but rather form-
less, expressed in hieroglyphic writing.
2 erik hornung

orbit which constantly regenerates time while the sun itself is in the
tow of the hours: solar circuit and time are thus indivisibly woven
together. The memorial temples of the NK Pharaohs are thus “Houses
of Millions of Years” as they guarantee the mortuary cult for the rest
of the horizon of time6 just as the tomb is the place of “the fullness of
time (neheh).”7 The limited individual stream of life flows into this sea
of the “millions of years”, as Pharaoh ends his in “millions of sed-
festivals” in a specifically royal unit of time. In typical moderation,
Akhenaten wishes to be buried in the new residence Akhetaten “after
the millions of sed festivals which my father Aten has envisioned for
me”, while Nefertiti and the princess Meritaten are to be buried “after
millions of years”.8
In the Ramesside era, the horizon of the future grows even further,
to “millions of millions of years”, and thus exceeding our own astro-
nomical temporal horizons.9 The Inscription Dédicatoire in Abydos expresses
the Ramesside spirit: in wishing Ramesses II “millions more of years
than the lifetime of Re in heaven”,10 it thus exceeds the temporal dimen-
sions of the cosmos itself. The Egyptians probably had the feeling that
there was immeasurable time in the Beyond which was yet not beyond
the bounds of time. The dead do not belong to eternity, but to the
fullness of time benefiting from a new lifetime with every nightly pas-
sage of the sun-god. In the Beyond, however, all past time is accu-
mulated in the “Place of Annihilation” (hetemit): it is there that the hours
fall when they have passed, when they are “swallowed” by the time
which “gave birth” to them.
The greatest possible extreme appears in the post scriptum to Book
of the Dead, chapter 62, promising the deceased (in the role of the sun-
god Re!): “The fullness of time (neheh) without limits is given to me,
for I am the heir of Neheh to whom djet is given”. One immediately

6
Cf. M. Ullmann, König für die Ewigkeit—Die Häuser der Millionen von Jahren: eine
Untersuchung zu Königskult und Tempeltypologie in Ägypten (Wiesbaden: ÄAT 51, 2002).
7
References in P. Vernus, “La stèle C 3 du Louvre”, RdE 25 (1973), 217–234,
esp. 223–224 (i), also the occasional designation of the Necropolis. The tomb can also
be the “Horizon” of neheh, cf. F.-J. Schmitz, Amenophis I. (Hildesheim: HÄB 6, 1978),
213–214.
8
Murnane & Van Siclen, Stelae, 25, 41.
9
The statue of Bakenkhons in Munich, KRI III, 298; Hymn in P. Berlin 3049:
J. Assmann, Ägyptische Hymnen und Gebete (Zurich & Munich, 1975), 127 B. A first time
on the Cairo stele CG 34025 of Amenhotep III: Urk. IV, 1653, 16.
10
Inscription Dédicatoire, line 27 = 31; KRI II, 325, 7–8; cf. U. Luft, Beiträge zur
Historisierung der Götterwelt (Budapest: SA 4, 1978), 173.
introduction 3

recalls the dialogue of the Beyond in chapter 175 where Osiris asks
Atum about the lifetime in the Beyond and is giving the comforting
assurance that it is “millions of millions of years”.
This concerns the future. As far as the horizon of the past is con-
cerned, the Royal Canon of Turin11 gives totals for the dynasties of
the gods and demi-gods (“The Followers of Horus”) before Menes and
the beginning of the Dynastic Period which take us to nearly 37,000
years. The Early Dynastic kings are assigned long reigns, totalling almost
another 1000 years up to Izezi, so that the Egyptians of the NK could
look back towards a temporal horizon of nearly 40,000 years since the
start of the world. That is a great deal when compared with the figures
in the Christian and Jewish traditions. This matches statements made
by the ancient authors. Herodotus (2, 142) reports that the Egyptian
priests accounted for a human line of ancestors of 341 generations,
which he reckoned amounted to 11,340 years before which lay the age
of the gods. Greek historians, such as Hekataios of Abdera already
reached the gods in the 16th generation. Genealogies as long as those
given by Herodotus are generally encountered in the LP, the best known
being the family tree of Memphite priests of Dyn. 22 which includes
more than 60 generations.12 In another note, Herodotus (2, 145) dates
Hercules 900 years before his own day, and Dionysos “some 1000
years”.
Our principal witness for the ancient Egyptian chronology is Manetho,
and he must likewise have had similarly high numbers which were then
forced into a truly Procrustean bed by the Christian authors who thus
made it compatible with the Biblical chronology whose priority could
not be doubted. At the same time, however, some individual items were
systematically extended in order to achieve accord (particularly for the
period between Menes and Adam), or in order to make the Egyptian
past appear older.13 Even in pre-Christian times, there were efforts to
claim a chronological precedence for Greece at the expense of Egypt.
Eudoxus of Cnidus (d. 356 BC in Egypt) opposed Herodotus’ figures
using the trick of transforming years into months.14 From Manetho,

11
Gardiner, Canon; KRI II, 827–844.
12
Berlin 23673: Borchardt, Mittel, 96–100; a good illustration of the document will
be found in the catalogue A. Grimm, S. Schoske, D. Wildung, Pharao: Kunst und Herrschaft
im alten Ägypten (Munich, 1997), no. 89. l.
13
Cf. primarily Helck, Manetho, 76, 83; cf. also Krauss, Amarnazeit, 239 and Beckerath,
Chronologie, 38.
14
Cf. S. M. Burstein, “Images of Egypt in Greek historiography,” in: Antonio
Loprieno, ed., Ancient Egyptian Literature (Leiden, 1996), 591–604, esp. 596f.
4 erik hornung

Georgios the Synkellos (8th cent. AD) cites a figure of 11,985 years for
the reigns of the gods (and thus close to that of Herodotus), with
Hephaistos = Ptah assigned 9000 years. This Synkellos simply inter-
prets as months, which allows “only” 727¾ years for Ptah. From some-
time after the reign of Ramesses II, this Memphite god had usurped
the place of the Sun-god Re at the head of the divine dynasty although
Diodorus (I: 13, 2) still maintains the older tradition (with Helios as
the first king of Egypt).15 Eusebius claims that 13,900 years separated
Hephaestus and “Bites”, following which is another total of 11,025
years, and thus 24,925, which he then interprets as months like Synkellos,
reducing them to 2,206 years, which fitted perfectly with the Biblical
2242 years between Adam and the Flood, while radically cutting the
Egyptian temporal horizon.
In the traditional chronicle of the 30 dynasties, with 113 genera-
tions, the total number of years is named as 36,525, whereby the largest
share of 30,000 falls to Hephaestus; as the last native Pharaoh Nectanebos,
some 15 years before Alexander the Great, is assigned the Anno Mundi
5147. In his work on Life and Opinions of Philosophers, written around
200 AD, Diogenes Laertios goes further: he calculates that 48,863 years
separated the invention of philosophy by “Hephaestus, the son of the
Nile” and Alexander the Great—and he adds the number of solar and
lunar eclipses in this period. From Zoroaster to Xerxes, the Lydian
Xanthus reckoned 6000 years, and Plutarch gives a similar estimate in
De Iside 46 placing 5000 years between Zoroaster and the Trojan War.
This “older” Zoroaster, who belongs in the seventh mill. BC, plays an
important part in modern esoteric literature, frequently as the teacher
of Hermes Trismegistos.
The figure of the king “Menes” the Egyptians created a fictional
beginning for the historical era. With echoes of the gods Min and
Amun, it also served as a shortened version of the name of Thutmose
III, and thus served in an ideal fashion as a link between the world
of the gods and the Dynastic kings. It is not, however, a mere ques-
tion of the identity of the names as the Egyptians were frequently
satisfied with distant echoes. Every attempt to equate Menes with a
concrete name of the Archaic Period, particularly Narmer and 'Aha,
has been fruitless.16 With the fictive image of Menes, known since the

15
Immediately following this, however, he allows a variant that, “according to some
priests” it was Hephaestos.
16
On this issue, cf. Hornung & Staehelin, Skarabäen, 44–45, and H. Brunner, “Menes”,
introduction 5

time of Hatshepsut and placed at the head of the list of kings for the
first time by Sety I in the table of Kings at Abydos, the beginning of
history is given a form, as in Genesis the beginning of humanity is
attached to Adam as the first human. The Christian chronographers
perceived the analogy and attempted to bring the two into temporal
accord. As the founder of the Egyptian state, the deeds of Menes—
emptying the swamps and founding Memphis—make him a bearer of
culture.17
The universal historian Orosius reckoned some 5200 years separated
Adam from the birth of Christ, and the figures used by the other early
Christian authors generally fit into the same general framework. According
to Sextus Julius Africanus (ca. 217) the Creation dates to 5500 years
before the birth of Christ. Jerome (Hieronymus, 347–419) gives the
date of 5198 BC, Victorius of Aquitaine is exact with 25/III/5201 BC.
The medieval chroniclers, such as the Russian Nestor Chronicle from
the early 12th century or the continuation of the Chronicle of the
Frankish kings by Fredegar reckon that the world began around
5200/5500 BC, while the Jewish Calendar places it slightly later,
7/X/3761 BC, although the Kabbala of the 13th and 14th centuries
estimated the duration of the worldly ages at 7000 years each, the suc-
cession ending absolutely in the 50th millennium.18
This temporal horizon, with a maximum of six or seven thousand
years, was maintained through the Renaissance, and even to some
extent afterwards. The 18th century Freemasons began their calcula-
tions with the round figure of 4000 BC (which is also found in Zedler’s
Universal-Lexikon),19 athough in 1704 Père Pezron made Year 1 of the
world 5872 BC, and thus substantially earlier;20 whereby he also assigned
priority to the Assyrians and Chinese, before the Egyptians. It was at
about this time that Isaac Newton attempted to “improve” the Egyptian
chronology, in order to recover the priority for the Old Testament,
and because the Egyptians “in their vanity had made their monarchy

LÄ IV (1982), 46–48; there is also material from Nubian Sai, discussed by J. Vercoutter,
“A propos des Mni = Ménès”, in: S. Israelit-Groll, ed., Studies Lichtheim, II 1025–1032,
and the link with Memphis by J. P. Allen, “Menes the Memphite”, GM 126 (1992),
19–22.
17
H. Brunner, “Menes als Schöpfer”, ZDGM 103 (1953), 22–26.
18
G. Scholem, Die jüdische Mystik in ihren Hauptströmungen (Frankfurt, 1967), 195.
19
Johann Heinrich Zedler, Grosses vollständiges Universal-Lexikon (Halle & Leipzig, 1749)
LXI, col. 818.
20
Paul Pezron, Defense de l’antiquité des tems (Paris, 1704).
6 erik hornung

a few thousand years older than the world”.21 He could not imagine
that only a century later William Herschel would work with the hypoth-
esis that the world was two million years old, and that even Kant had
assumed a substantially greater temporal horizon.
It is a fascinating and still undescribed phenomenon how the great
voyages of discovery moved in parallel so that the Spanish, British,
Portuguese and other seafarers moved forth into hitherto unknown parts
of the world, bringing new continents into view, and thus the new dis-
coveries in space went hand in hand with the extension of time, ulti-
mately exploding the far too narrow temporal confines of the Christian
chronographers. And, again Egypt played a decisive role since the
ancient pre-Christian records were again put to use from the Renaissance
onwards. The decisive quantum jumps which continue up to the pre-
sent day began in the late 18th century as the study of geology began.
In 1778, Buffon estimated the age of the earth at 100,000 years, a
figure which was used by Goethe in Faustus when Mephistopheles
remarks on Faustus’s vision of ruling the sea:
This is naught new for me to explore
Knowing this a hundred thousand years of yore (verses 10210f.)
Similarly Cuvier, who in an elegant phrase of Heinrich Heine’s, “proved
in the most ungallant fashion that our mother earth is many thousands
of years older than she had hitherto admitted”,22 and Jean Paul is sur-
prised that “the earthy sphere . . . grows older by the day, backwards
(and not just forwards)”.23 It must have been a deeply wrenching expe-
rience for the times that the firmly established temporal horizon of the
Creation in the OT would be weakened with such terrifying speed.
With the temporal requirements of geology demanding ever greater
spans of time, the 19th century saw the final abandonment of the
chronology of the OT.24 In 1870, Lepsius still assumed that the most
ancient remains of mankind and their worked stones were 30 to 50,000
years old.25 However, millions of years were soon accepted; even Herschel,

21
Isaac Newton, The chronology of ancient kingdoms amended (London, 1728), 191.
22
Heinrich Heine, Französische Zustände, Artikel VIII, 27/V/1832.
23
Jean Paul, Der Komet oder Nikolaus Marggraf: eine komische Geschichte (1822), chapter
VI, note.
24
The trail-blazing three volume work of Charles Lyell, Principles of Geology (London,
1830–33), with numerous improved editions.
25
R. Lepsius, “Über die Annahme eines sogenannten prähistorischen Steinalters in
Ägypten”, ZÄS 8 (1870), 89–97, 113–121, esp. 90.
introduction 7

whom Jean Paul cites, reckoned with 2 million light-years for the most
distant stars, and thus a corresponding age for the universe; Kant refers
to “millions of years and centuries” in his General Natural History and
Theory of the Heavens. In the 20th century, astronomy would change this
to billions of years—whereby we come closer to the Ancient Egyptians.
In 1929, Hubble’s discovery of the expansion of the universe gave
another push. We have now all experienced an age during which the
cosmos gets to be billions of light years larger, and consequently older,
every couple of years. In fact, the process does not appear to be
finished—a fine example of how rapidly one can adjust to the extreme
acceleration of exploding time in history.
After this time travel to the further horizons, we can return to
Egyptian chronology, as we understand it today. From the beginning,
Egyptology had an intense preoccupation with chronology, and gener-
ally took the highest available figures of antiquity, which were once
again in vogue since the Renaissance—rather than the Biblical figures.
Champollion-Figeac placed the debut of the Dynastic Period at 5867
BC, and that of Dyn. 18 at 1822; John Gardner Wilkinson who always
preferred to orient himself on the OT used the values of 2320 and
1575; Lepsius lay between them at 3892 and 1591.26 One can see that
they rapidly came close to the presently accepted dates for the NK,
but that the earlier period was still the subject of great differences in
the 19th and early 20th centuries AD. Petrie consistently maintained
that the historical period began in the 6th millennium and even in
1935 Borchardt put Menes at 4056 BC whereas Eduard Meyer, fol-
lowed by Breasted, came close to our own estimates with 3315. Fun-
damental for the earlier chronology was the discovery of the Illahun
Papyri (ended in 1899) with their Sothic and lunar dates which offered
the first fixpoint before the NK. We can marginally note that even in
1870, Lepsius (immediately joined by Georg Ebers) energetically dis-
missed an Egyptian prehistoric stone age, and attributed all of the stone
tools to the historic period.27 It was only after 1890 that a window into
the deeper prehistoric past of Egypt was opened with a series of blows
beginning with the discovery of the Naqada culture and fundamental

26
A. Wiedemann provided a useful synopsis of the early propositions in Ägyptische
Geschichte (Gotha, 1884), 732f, and Borchardt, Annalen, 48–51, enlisting their “Main
Errors”.
27
Lepsius (n. 25); cf. Georg Ebers, “Über die Feuersteinmesser in Aegypten”, ZÄS
9 (1871), 17–22.
8 erik hornung

reflections of Georg Steindorff. After some initial uncertainty with a


broad spectrum of variations, the framework was gradually refined in
the course of the 20th century, and C-14 dates for the OK now lead
to considerable irritation when they lie a mere century above the other-
wise acceptable values.
Egyptian chronology is still the touchstone by which all of the other
chronologies in the ancient world are measured and the issue of its
reliability is thus central. A survey and examination of the chronolog-
ical prospects for the third and second millennia BC in the different
parts of the ancient world was the object of a series of meetings, begin-
ning with Gothenburg in August 1987 at the initiative of Paul Åström,28
and continued in August 1990 by Manfred Bietak at Schloss Haindorf,29
and in November 1996 at the same venue, and then in May 1998 in
Vienna. These last two already took place under the auspices of Bietak’s
major Sonderforschungsprojekt (Special Research Project) “Synchronization
of Civilizations in the Eastern Mediterranean in the 2nd millennium
BC”; another conference was held at Haindorf in May 2001.
Åström chose the title “High, Middle or Low” for his symposium
and thus placed his finger on the central question, which is still not
solved today: the choice between a longer, shorter, or medium position
for Hammurabi of Babylon. The conference in Gothenburg did close
with a formal vote on which of the three alternatives the participants
preferred: against 3 votes and 3 abstentions, the “Low” chronology was
adopted, and it is absolutely clear for Egypt that for the NK, this is
the only chronology with which we can live. There, I endeavoured to
avoid the astronomical problems when discussing the chronology of the
NK, and Kitchen also stresses in his most recent contribution that the
Egyptian chronology “is not based on these meagre astronomical data”.30
Helck was not a friend of astronomical data either.31
The apparent precision of astronomical and other sources from the
natural sciences (including the ice of Greenland) is always enticing, and
it is difficult to resist the charm. However, we should recall just how
long it was generally agreed that the original introduction of the Egyptian

28
Cf. Åström, ed., High, Middle or Low?
29
Published in Ä&L 3 (1993).
30
K. A. Kitchen, in: M. Bietak, ed., SCIEM 1996/98, 39.
31
He stressed in “Zur Chronologiediskussion über das Neue Reich”, Ä&L 3 (1992),
63–67, that the dated monuments should be awarded priority before the astronomical
data.
introduction 9

calendar produced Eduard Meyer’s “oldest date in world history” of


19/VII/4241 BC—until this was demolished by Otto Neugebauer and
Alexander Scharff in 1939. Even v. Beckerath’s “earliest absolute date
in Egyptian history” (Chronologie, 45) is an astronomical date (the heli-
acal rising of Sothis at Illahun) and cannot be maintained. The con-
troversy surrounding the Venus-dates in Mesopotamia, and the constantly
renewed and alternatively calculated eclipse-dates in Western Asia (which
do not touch Egypt) or the Sothis and lunar dates in Egypt have repeat-
edly demonstrated the problems of the astronomical dates and con-
tributed to the primacy of the purely historical dates. They are and
remain our most important source.
In an inscription in Karnak, the HP Osorkon proclaims that the cult
was regularly performed, “even as the moon in its course” due to his
efforts,32 and thus the moon and its regular phases would appear to be
the very pattern for precision. But lunar dates are repeated at quite
short intervals and are hardly useful if they cannot be embedded in a
fine mesh of other reference points, as is, for example, the case for the
Illahun dates. From 21 lunar dates in that archive, R. Krauss has now
calculated the alternative dates for the first year of Senwosret III at
1862/61 or 1837/36 BC, and here it is the large number of dates
which really brings weight since a single lunar date can bring no more
than one further confirmation for a date which has been calculated
using other means. Sothic dates with all of their countless parameters
and uncertainties are likewise only of value as confirmations, and can
never serve as the point of departure. And it almost never goes with-
out any editing.
In addition, the Egyptian scribes probably never aimed for such pre-
cision; rather in their administrative tasks they will have been satisfied
with rough approximations. It is thus that J. J. Janssen, who is pro-
foundly familiar with their work, notes the “accuracy of the Egyptian
scribes which is notoriously unreliable”.33 This will have applied to their
methods of measuring time, which was extremely successful despite all
its unreliability. The Egyptian calendar never depended upon interca-
lations and is of winning simplicity. This applies not only to the con-
stant length of the year, but also the schematic division of the year
(three seasons, best given their Egyptian names, Akhet, Peret and Shemu,

32
Reliefs III, pl. 16; Caminos, Chronicle, § 32 and 37f.
33
Janssen, Varia, 101.
10 erik hornung

each of four months of 30 days each). In Pharaonic times there was


never an era with a constant continuous numbering of years: with each
new Pharaoh, the count began anew.
Egyptian scribes thus had no inhibitions about copying older texts
in extenso without “modernizing” them, and the campaigns of a Pharaoh
could be used by another later pharaoh; royal constructions could be
newly dedicated with the addition of new names. But the framework
of Egyptian history can only be relative, using contemporary dates from
three millennia. Nevertheless here, encouraging progress has been made
in the primary task of establishing a correct sequence of kings.
As a festival celebrated (with the object of replenishing his powers)
in the 30th year of a king’s reign,34 the sed-festival is chronologically
relevant, and has frequently been included in chronological discussions.
The apparent exceptions to the 30-year rule led to the assumption that
this festival was regularly celebrated every 30 years, regardless of actual
reigns. However, most of the exceptions have been eliminated, and
there remain only a few uncertain cases (Hatshepsut, Akhenaten), which
can be explained in terms of particular circumstances. In the case of
Hatshepsut, the celebration can be linked to the accession of Thutmosis
I (or II); in Akhenaten’s case it can be anchored to the transition to
the worship of Aten as king. A rigid 30-year sequence—as has occa-
sionally been proposed—is improbable in the extreme.
The most certain and best documented cases of a sed festival which
was actually celebrated concern reigns where the chronology does not
offer any difficulties (Amenhotep III, Ramesses II, Ramesses III). In a
number of rather dubious cases, the alleged evidence of a sed-festival
has been used to argue a reign of more than 3 decades, even where
there is otherwise no evidence. This concerns Amenhotep II and
Thutmose IV in particular, but all of the “sources” for the sed-festivals
of these rulers are mere wishful thinking, although Wente and Van
Siclen refer to “jubilee evidence in favor of a long reign for Thutmose
IV”.35 They also take the alleged sed-festival of Hatshepsut as the basis
for proposing a reign of 13 years for Thutmosis II.36 A sed-festival is
also proposed with regard to the length of the reign of Senwosret III.37

34
On the issue of the date, etc., cf. Hornung & Staehelin, Sedfest.
35
Wente & van Siclen, “Chronology”, 230.
36
Wente & van Siclen, “Chronology”, 226.
37
J. W. Wegner, “The Nature and Chronology of the Senwosret III-Amenemhat
III Regnal Succession: Some Considerations Based on New Evidence from the Mortuary
introduction 11

Such conclusions are too risky, and this also applies to the refer-
ences to a “first time of the sed-festival” by kings of the MK and NK.
Were we to use this in, e.g., the case of Amenhotep I as a reference
demonstrating a reign of more than 3 decades, then we would come
into conflict with the documented reign of 21 years,38 and in the case
of Psammetichus II whose reign is certain to have lasted 6 years, the
“first time” cannot have any real meaning, and the same is true of
Shoshenq I, likewise with 21 regnal years.
With Hatshepsut, the “repetition” of sed-festivals, promised by Amun
and Thoth in their divine speeches in the chapelle rouge at Karnak cannot
be taken at face value. With Amenhotep II and Thutmose IV, “the first
time and repetition of the sed-festival” is merely the expression of a
wish, which cannot allow any historical conclusions. Only with a “sec-
ond time of the sed-festival”, known for Pepy II as well as the Archaic
kings Den and Qa-'a,39 do the sources depart from the wishful “repe-
tition”, and this also applies to the “3rd sed-festival” of Thutmose III
and Amenhotep III, as well as the multiple repetitions documented for
Ramesses II.
The abundance of sed-festival wishes, documented for virtually all the
dynasties, has no chronological importance. The sed-festival was merely
understood increasingly as an ideal measure of time and used in par-
allel with other concepts of time, precisely in order to wish Pharaoh
the longest possible regnal era, whereby the hoped for chain of jubilee-
festivals can be projected into the Beyond. Eduard Meyer already sug-
gested “that this festival cannot be used for chronological purposes”,40
and recent research gives us no reason to change this appraisal.
Aside from the astronomical dates, the physical sciences offer a num-
ber of other aids, among which C-14 radiocarbon dating has been

Temple of Senwosret III at Abydos”, JNES 55 (1996), 249–279, esp. 262–264, plead-
ing for a real celebration.
38
In the inscription of the “astronomer” Amenemhet, L. Borchardt, Die altägyptische
Zeitmessung (Berlin & Leipzig, 1920), Pl. 18. E. F. Wente, “Thutmose III’s Accession
and the Beginning of the New Kingdom”, JNES (1975), 265–272, esp. 271–272 sought
to use the references to a sed-festival and an alleged coregency with Ahmose to increase
the length of the reign.
39
For the references, cf. Hornung & Staehelin, Sedfest. For a new reference for Den
on a bowl from Abydos, see G. Dreyer, “Umm el-Qaab: Nachuntersuchungen im
frühzeitlichen Königsfriedhof 3./4. Vorbericht,” MDAIK 46 (1990), 53–90, esp. 80, with
Fig. 9. Against a 2nd sed-festival for 'Adj-ib (so A. Grimm “Ein zweites Sedfest des
Königs Adjib,” VA 1 [1985], 91–98), cf. Kamil O. Kuraszkiewicz, “Noch einmal zum
zweiten Sedfest des Adjib,” GM 167 (1998), 73–75.
40
E. Meyer, Geschichte des Altertums II, 1 (1953), 149, note 2.
12 erik hornung

widely used, and become indispensable for Prehistory and the Archaic
period. For the Dynastic period this procedure is, however, neither
sufficiently reliable nor sufficiently precise. Climatic or seasonal dates
such as harvests, the inundation, or preferred dates for expeditions in
the desert regions from either bank of the Nile can provide useful
checks, but we still lack a systematic and modern collection of the avail-
able dates. Even today, historical periods and events continue to be
explained in terms of climatic change,41 although a causal relationship
has never been demonstrated; the links with volcanic eruptions, such as
Thera, have triggered debates, but not led to any definitive conclusions.
Highly problematic are dates based on stylistic, linguistic and palaeo-
graphic criteria. Here there are amusing discoveries, such as the dat-
ing of a Hellenistic bronze of Socrates to “ca. 700 BC”, which I noted
in the Manchester Museum. The “Memphite Theology” has been
pushed back and forth across the entire history of ancient Egypt from
the Archaic period to the Ptolemies, like the unstoppable efforts to date
the NK Books of the Netherworld to the MK or even the OK. In
sculpture, there are still difficulties in the attribution of statues dating
to the MK and NK and the LP. The dating of the handwriting of the
Brooklyn oracle papyrus 47.218.3 can serve as an exemplary warning.
In this case, 50 higher officials and priests all personally signed a doc-
ument as witnesses, in Thebes on October 4, 651 BC ( Julian).42 Examined
individually, one would date the signatures quite differently, but the
date of the protocol clearly assigns them all to exactly the same single
day.
Synchronisms with the Near East are particularly useful when they
can be linked to the relatively reliable Assyrian chronology. By con-
trast, the numerous synchronizations with the Hittites are virtually use-
less as the Hittite sources cannot provide either dates or regnal lengths.
Our reliable point of departure remains the beginning of the reign of

41
So, e.g., Th. de Putter, “Les inscriptions de Semna et Koumma (Nubie): niveaux
de crues exceptionnelles ou d’un lac de retenue artificiel du Moyen Empire?”, SAK 20
(1993), 255–288 associates the frequently discussed extremely high Niles dating to the
end of Dyn. 12 with an “événement climatique”; for the movements of the Peoples
of the Sea, cf. J. Neumann & S. Parpola, “Climatic Change and the Eleventh–Tenth
Century Eclipse of Assyria and Babylonia”, JNES 46 (1987), 161–182; a drought leads
to famine and thus provokes the migrations. Cf. also S. J. Seidlmayer, Historische und
moderne Nilstände (Berlin, 2001).
42
R. Parker, A Saite Oracle Papyrus from Thebes in the Brooklyn Museum (Providence,
1962).
introduction 13

Taharqa in 690 BC43 As Depuydt shows in his contribution, the sequence


with a “day-exact chronology” begins on June 20, 688 BC ( Julian),
with the sale of a slave. The recently discovered inscription of Sargon
II in the Tang-i Var Pass in western Iran from the year 706 does not
offer absolute precision about his immediate predecessors, as had been
initially hoped, and thus alternatives remain (Shebitku as coregent or
sole ruler). And there remain many uncertainties in the TIP, as critics
such as David Rohl have rightly maintained; even our basic premise
of 925 for Shoshenq’s campaign to Jerusalem is not built on solid foun-
dations. Nevertheless, there is such a web of dates, genealogies and
relations between Egypt and the Near East that dramatic changes can
be excluded (whereas Rohl wanted to cut off 141 years), above all due
to the adjoining Ramesside era. In addition, there is archaeological
material (such as, e.g., coffins) which has generally not been exploited
for dating purposes, and there remains the prosopography of the officials
and priests. For the TIP, there remain also the aids offered by the
apis-bulls with their very precise data.
Already at Gothenburg, there was general agreement about the dates
for beginnings of the NK. Helck, Kitchen and Hornung/Krauss all
worked with the very narrow range of 1540 to 1530 BC for the start
of the reign of Ahmose, and after some debate, there is now general
acceptance for the reign of Ramesses II at 1279–1213 BC. Although
we must remain wary of confusing consensus with actual fact, for the
NK we now have such a fine mesh of relative dates which are them-
selves also woven into NE dates that major adjustments can probably
be excluded. While there is room for minor cosmetic corrections, we
are relatively confident about the framework. And now NE material
allows for the fall of Babylon to be set at ca. 1500 BC.44 We can now
trust that dendrochronology will provide greater precision—as the pre-
cise dating of the shipwreck of Uluburun with the Nefertiti scarab pro-
viding grounds for hope.45 The links in both directions—backwards to
Amarna and forwards to the Ramessides—mean that even for the
ancient crux of the length of the reign of Haremhab, those favoring a

43
Although L. Depuydt now also allows for the possibility of 691 BC, “Glosses to
Jerome’s Eusebios as a Source for Pharaonic History”, CdE 76 (2001), 30–31, note 1.
44
Gasche et al., Dating. They argue for 1499 BC, but allow alternatives for 1507
and 1491 BC.
45
P. I. Kuniholm, B. Kromer et al., Nature 381 (1996), 780–783. Construction of
the ship: 1316 BC.
14 erik hornung

shorter reign are so hard pressed for explanations that the controversy
would appear to have been silenced. In chronological issues we can
never take a single item and redate it while disregarding the larger
context in which it is fixed. Herein lies the great value of Manfred
Bietak’s endeavor to synchronize the civilizations of the 2nd mill. by
bundling and weaving them together.
After the departure of Parker and his “astronomically certain” dates
there was a certain aporia with regard to the chronology of the MK.
However, we will now have to take leave of Parker’s date for the start
of Dyn. 12 (1991 BC) and agree on a date around the middle of the
20th century (although ca. 36 years still separate v. Beckerath and
Krauss, with Kitchen in the middle). This offers encouraging possibil-
ities for the extension of the FIP, which has hitherto been cut short.
It is painful to recognize that the Near Eastern synchronism of Neferhotep
I will have to be dropped,46 but a certain degree of scepticism is always
required when dealing with synchronisms.
We are treading on very thin ice in the 3rd mill., even though fan-
tastically precise dates based upon astronomical orientations of the pyra-
mids are still being published.47 While the general orientation does offer
a certain framework, this is complemented by relative chronology. It
no longer seems necessary to raise the dates for the pyramid-builders
of Dyn. 4, to provide more building time and thus we can retain the
23–25 years of the king-lists. At the top we can begin with Dyn. 0 for
the start of the Dynastic period in the 4th mill., which is essential for
the synchronisms with the Near East.
The most recent large-scale summary of Egyptian chronology is Jürgen
von Beckerath’s Chronologie des pharaonischen Ägypten of 1997 which pro-
vides a balanced state-of-the-art picture of the foundations. In contrast
to Beckerath, we strive to separate clearly relative and absolute chronol-
ogy, and to provide more weight for the archaeological materials and
the factors derived from the physical sciences.
Chronology has always been an arena for radical hypotheses and dras-
tic moves. In antiquity, Christian chronographers manipulated Manetho’s
dates in order to achieve a convergence with the Biblical chronology.

46
Cf. C. Eder, Die ägyptischen Motive in der Glyptik des östlichen Mittelmeerraumes zu Anfang
des 2. Jts. v. Chr. (Louvain: OLA 71, 1996), 13; T. Schneider, Review of Eder, “Motive”,
ZDPV 114 (1998), 184.
47
Cf. the critical remarks by E. Aubourg, “Détermination de l’orientation de la
pyramide de Radjedef ”, Genava 49 (2001), 245–248.
introduction 15

At that time, the relative chronology was artificially lengthened whereas


the 20th century has been marked by efforts at radical shortening, by
eliminating dynasties or placing them in parallel with others; Velikovsky
aimed at an extremely bold “analysis of events” by means of which
similar events or historical constellations were simply slotted together—
in this fashion Hitler’s Russian campaign can be viewed as being the
same as Napoleon’s, and thus interpreted as the same event. . . . .
We will always be exposed to such attempts, but they could only be
taken seriously if not only the arbitrary dynasties and rulers, but also
their context, could be displaced. Were one to discover that Ramesses
II was really Necho II in disguise, and likewise Merneptah as Apries
and Ramesses III as Nectanebo I, one would still have to demonstrate
that in each and every case the two allegedly identical rulers were also
surrounded by the same officials, and that the religious and artistic
contexts were also entirely compatible. In the absence of such proofs,
we can hardly be expected to “refute” such claims, or even to respond
in any fashion. For Dyn. 20 we have such a fine mesh of dates, vir-
tually day-by-day, that the entire complex can hardly be assigned to
another dynasty as it would simply fail. Furthermore, the entire so-
called “dark ages” ca. 1200–700 BC are so well documented with
archaeological material from Egypt that there is not the slightest ground
for justifying a reduction. It is thus neither arrogance nor ill-will that
leads the academic community to neglect these efforts which frequently
lead to irritation and distrust outside of professional circles (and are
often undertaken with the encouragement of the media). These attempts
usually require a rather lofty disrespect of the most elementary sources
and facts and thus do not merit discussion. We will therefore avoid
discussion of such issues in our handbook, restricting ourselves to those
hypotheses and discussions which are based on the sources.
It is characteristic that the drastic reductions touching upon the Old
Testament play a very important role, as in antiquity. In order to make
Hatshepsut the Queen of Sheba and thus the contemporary of Solomon,
we are obliged to remove the 500 years which separate these two by
aligning a number of dynasties in parallel, rather than sequentially.
Here we face those ideological pressures which always have an unsuit-
ably powerful influence on the study of historical data.
We have referred to the typically Egyptian association of the highest
conceivable fullness of time with the solar orbit. “The lifetime of Re
in heaven” was the absolutely highest conceivable temporal horizon for
the ancient Egyptians, as the duration of being, comparable with the
16 erik hornung

lifetimes of galaxies in modern astronomy. According to the Litany of


Re (13th appellation) it was “greater than the West and its images”,
and thus actually greater than the duration of the Beyond. In the Book
of the Celestial Cow, this concept is transformed into an image where
neheh and djet “those two old and great gods” appear as the pillars of
heaven. As long as time endures, heaven will rest on its pillars and the
solar orbit will be maintained daily. Long before the Pythagoreans, the
Egyptians had postulated the concept of the eternal return, which was
finally given its impressive form by Nietzsche who was in turn draw-
ing on Goethe: “How often ‘tis repeated! will always be/repeated through
eternity” (Faustus verses 7012f.). But in Egypt, it was never the return
of the same, but rather return in a form transformed: the sun is every
day a new sun, reborn each day again by the goddess of heaven.
PART I

EGYPTIAN CHRONOGRAPHICAL TRADITION


AND METHOD OF DATING
I. 1 ROYAL ANNALS

The Editors

Royal Annals of Memphis

There are several fragments which almost certainly derive from more
than one stone slab, inscribed on both sides.1 The fragment with most
text preserved is the so-called Palermo stone; other pieces are in Cairo
and London. Nearly all of them were known and had been studied in
the first decades of the last century, and some specialists have proposed
reconstructions of the original text.2 The inscriptions which record events
in the reigns of Dyn. 1 through part of Dyn. 5 are arranged in a series
of rectangular compartments, set out in horizontal rows, reading from
right to left. Each compartment in row 1 contains only the name of a
mythical or early historical king. In the other rows, the hieroglyph for
“year” defines the right side of a compartment. Below the compart-
ments, the height of the annual Nile inundation for that year is recorded.3

Fig. I. 1.1. Rows 1 and 2 on the Palermo stone (Verso)

1
Wilkinson, Annals, 28.
2
Wilkinson, Annals, 29–36.
3
Wilkinson, Annals, 18–19.
20 the editors

The inscription of the first incomplete compartment of row 2 cites


“following of Horus” and “Birth of Anubis”. The second and third
compartments are separated by a dividing line that indicates a change
of reign. The second compartment lists “6 months 7 days”, presum-
ably the time elapsed in the last incomplete year of that reign. The
third compartment lists “4 months 13 days”, evidently with reference
to the new king, but the two intervals do not add up to a full year be
it lunar or solar. The compartment also records two rituals associated
with accession years, “the Unification of the Two Lands” and the
“Circuit of the Wall”.
For nearly all of the first three dynasties, the annals cite only occa-
sions that gave a year its name. Beginning with Dyn. 4, the annals
begin to resemble chronicles recounting a multiplicity of memorable
events in each year of a reign.4 The similarity of some so-called year
labels of Dyn. 1 to certain entries in the annals suggests that both
derive from the same archival institution.5 Specialists do not agree on
when the annals were compiled nor when the slab(s) were inscribed.
Palaeography and the rendering of the royal names of the Early Dynastic
Period favor of an OK date, but it cannot be excluded that the exist-
ing fragments come from later copies.6
As Wilkinson observed, most scholars who have studied the annals
have interpreted the information recorded at face value, instead of con-
sidering the cultural context.7 But even if cited events did not actually
occur, the year itself may be accepted as fact. Being a later compila-
tion, the annals are a less valid source for reconstructing regnal years
of Early Dynastic kings than contemporaneous data would have been.8
A definitive reconstruction of the Annals as a whole is not possible,
and their value for the Early Dynastic period is problematic. A revi-
sion of the most recent attempt at reconstruction is made further below.

South Saqqara Stone and Later Annals

Annals of Dyn. 6 were inscribed on both sides of a slab measuring ca.


226 × 92 cms that was reused as a sarcophagus lid. At that time the

4
Clagett, Science I, 48.
5
Wilkinson, Annals, 60.
6
Wilkinson, Annals, 23–24.
7
Wilkinson, Annals, 65.
8
Spalinger, “Texts”, 281.
royal annals 21

text was summarily erased. Baud and Dobrev analyzed the structure
of the text and suggested readings of details.9 Traces can be identified
of the titularies of the Dyn. 6 kings Teti, Userkare', Pepy I, and Merenre'
which show that the count was biennial during their reigns.
From later times there is a portion of the annals of Amenemhet II
containing chronicle-like entries.10 The so-called annals of Thutmose III11
report on a daily basis, whereas the others provide yearly overviews.12
The annals of Pami list offerings donated to the cult in Heliopolis dur-
ing his reign.13

Proposed Revision of Recent Reconstructions of the Memphite Annals Stone


(Fig. I. 1.2)

The latest reconstruction of the annals proposed by Beckerath,14 can


be improved by taking into account Baud’s corrections.15 Any attempt
at reconstruction must be based on a sound order of succession. This
is the case for the kings of Dyn. 1 and for Hetep-sekhemwy, Ra'-neb,
and Ny-netjer of earlier Dyn. 2, since Kahl established that Weneg is
the nsw bjt nb.tj name of Ra'-neb.16 The succession in the second half
of Dyn. 2 is unclear. Traces of a serekh with the Seth animal on Cairo
1 suggest that Per-ibsen followed Ny-netjer, but it is also possible that
Egypt was divided at that time. If so, the Annals might list the kings
in succession although they actually ruled simultaneously. It has recently
been confirmed that Kha-sekhemwy’s successor was Netjery-khet (Djoser),
not Nebka.17 The sequence Kha-sekhemwy: Djoser is recorded in row 5
of the Palermo stone. Djoser’s successor Sekhem-khet may be identifiable

9
M. Baud & V. Dobrev, “De nouvelles annales de l’Ancien Empire égyptien. Une
<Pierre de Palerme> pour la VIe dynastie”, BIFAO 95 (1995), 23–92; idem, “Le Verso
des annales de la VIe dynastie. Pierre de Saqqara-Sud”, BIFAO 97 (1997), 35–42; see
also below Baud, Chapter II. 5.
10
H. Altenmüller & A. M. Moussa, “Die Inschrift Amenemhets II. aus dem Ptah-
Tempel von Memphis”, SAK 18 (1991), 1–48.—J. Malek & S. Quirke, JEA 78 (1992),
13–18.
11
Urk. IV 645–673.
12
W. Helck, LÄ I, 279.
13
S. Bickel, M. Gabolde & P. Tallet, “Des annales héliopolitaines de la Troisième
Période Intermédiaire”, BIFAO 98 (1998), 31–56.
14
Beckerath, Chronologie, 174–179.
15
Baud, “Ménès”, 136–138.
16
See below Kahl, Chapter II. 2.
17
See below Seidlmayer, Chapter II. 3.
22 the editors

on the large Cairo fragment; the succession at the end of the dynasty
is unclear.
Maximum and minimum distances between the Palermo stone and
Cairo 1 can be determined on the basis of the preserved compartments
for [Djoser] in row 5. Beckerath’s change of reign after the first com-
pletely preserved compartment on Cairo 1 is spurious.18 Rather, there
are ½ + 9 compartments on Cairo 1 that belong to [Djoser]. If the
lost titulary was centred and had a width of 7 compartments, then at
least ½ + 9 + 7 + ½ + 9 = 26 compartments result for Djoser. If
the first of the 26 compartments lay to the left of the Palermo stone,
then the gap is 17 compartments or more. If the first of these com-
partments was identical with the first of [Djoser’s] preserved ½ + 4
compartments on the Palermo stone, then the gap measures at least
12 compartments in row 5, exceeding the 9 compartments deduced by
Wilkinson.19 A gap of 12 compartments in row 5 is not compatible
with a gap of 9 compartments in row 2. According to Kaiser the inter-
nal evidence of the Palermo stone and Cairo 1, including the corre-
spondence of rows 2 to 5, favours 9 missing compartments in row 2
as does the regularity of the “followings of Horus” and the six-yearly
bark festival in rows 2 to 4.20 We accept Kaiser’s arguments for the

Fig. I. 1.2. Revised reconstruction of the Annals stone (Verso)

18
Wilkinson, Annals, 53.
19
Wilkinson, Annals, 79.
20
W. Kaiser, ZÄS 86 (1961), 44.
royal annals 23

size of the gap in row 2 and the corresponding gap of 13 compart-


ments in row 5 in Beckerath’s reconstruction.
The left edge of row 2, and thus of the other rows on the recto (if
row 3 begins with “Serpent”) can be determined with reference to the
completely preserved titulary of Djer that spans 7 compartments. Taking
into account the gap of 9 compartments in row 2 between Cairo 1
and the Palermo stone, there were 20 compartments before and after
Djer’s centred titulary, for a total of 47 complete compartments.
The determination of the right edge of the inscribed field is more
difficult. Beckerath errs in identifying Kha-sekhemwy on the Palermo
stone as Nebka and in construing the right edge of row 5 by adding
the TC ’s 27 years of Bebti (<Kha-sekhemwy) to the right of **Nebka’s
compartments. By contrast, extrema for the right edge can be deter-
mined in row 6 by extrapolating the early years of Snofru. The min-
imum number of compartments is 7 (zp years 1 to 6 and a year zm#-t #wy),
and the maximum 12 (adding m-¢t zp years 1 to 5). Thus at least 16
and at most ca. 27 compartments belonged in row 2 to the right of
Djer’s reign. We opt for the maximum 75 compartments, correspond-
ing to 74 years, because the incomplete last year of ['Aha] and Djer’s
first one occupy two compartments. The recent discovery of a year
tablet of Nar-mer makes it possible that row 1 recorded not only 'Aha,
but also Nar-mer.21
The reign of Semer-khet is completely preserved in ½ + 7 + ½ com-
partments in row 3 on Cairo 1. Towards the left, ca. 24 compartments
can be reconstructed for Qa-'a as the successor of Semer-khet, if the
47th compartment for Djer marks the left edge of the inscribed field.
To the right of Semer-khet, half of a compartment and part of the last
complete compartment of 'Adj-ib are preserved without traces of his
titulary; Wilkinson postulates at least ½ +7 + ½ compartments for him.
Beckerath assumes that Den celebrated the Sed festival which is
attested on the Palermo stone in his 30th year. The assumption is unac-
ceptable, if only, because Qa-'a celebrated two Sed festivals according
to contemporaneous sources, whereas he ruled less than 30 years accord-
ing to Beckerath’s own reconstruction. The end of Den’s titulary is pre-
served on the Palermo stone and its beginning on Cairo 5.22 If Den’s
titulary had a width of 7 compartments, then at least ½ + 13 + 7 +

21
Wilkinson, Annals, 72.
22
Note, however, that Cairo 5 may be a forgery.
24 the editors

13 + ½ = 34 years are deducible, resulting in ½ + 11 + ½ years for


'Adj-ib. At most ½ + 17 + 7 + 17 + ½ compartments can be pro-
posed for Den, if 'Adj-ib reigned a minimum of 8 years.
If the right edge of the inscribed field is defined by 12 reconstructed
compartments for Snofru in row 6, then to the right of Den there
should be at least 8 and at most 12 compartments for “Serpent”. The
reconstruction results in at most ca. 90 compartments in row 3.
Row 4 preserves a series of counts, combined with “followings of
Horus” from the reign of Ny-netjer whose titulary is partially preserved.
The example of Djoser in row 5 shows that the first count could occur
as late as the second year. If so, then ca. ½ + 16 + 7 + 16 + ½ =
40 compartments are possible. If the width of the centred titulary
amounted to 8 compartments, then 41 compartments result. The recon-
struction yields a gap of 2 compartments at most between the 41st
compartment of Ny-netjer and the following reign on Cairo 1. The
dilemma cannot be solved by making the distance between the Palermo
stone and Cairo 1 smaller, because we already presume the minimum
distance.
The position of the titulary of Per-ibsen (?) as Ny-netjer’s successor
implies that the former’s reign spanned about 10 compartments. The
remainder of row 4 towards the left edge amounts to ca. 22 compart-
ments (with the width of the Cairo 1 compartments), to be distributed
among the kings of Dyn. 2.
Between Ny-netjer’s reconstructed first year and the right edge of
row 4 as defined by Snofru’s reconstructed first compartment, there
remain ca. 29 compartments (width of Palermo stone compartments)
for Ra'-neb and Hetep-sekhemwy. A reconstruction along these lines
results in a total of at least ca. 104 compartments in row 4.
In row 5 the reconstruction of the preserved counts that are com-
bined with “followings of Horus”, allows the reconstruction of at least
10 lost compartments of [Kha-sekhemwy] and of another 1 + ½, if
the first count and “following of Horus” occurred in the second reg-
nal year. The result is 18 compartments for [Kha-sekhemwy], leaving
ca. 17 compartments towards the right edge of row 5 for kings of late
Dyn. 2. For [Djoser] ½ + 26 + ½ compartments can be reconstructed
and 7 compartments for [Sekhem-khet] as his successor. The remain-
ing ca. 16 compartments at the left end of row 5 belong to the last
three kings of Dyn. 3. The reconstruction yields altogether 86 or 87
compartments at most.
royal annals 25

About 19 compartments should follow after Snofru’s rnpt zp 8 on the


Palermo stone, if Djer’s 47th compartment defines the left edge of the
inscribed field. Reconstructed row 6 contains a maximum of ca. 34
compartments for Snofru.
Rows 7 and 8 of the recto of the Annals are mostly destroyed. Cairo
fragment 3 shows Djedefre' occupying the last third of row 8, whereas
the rest of it and all of row 7 must have belonged to [Cheops]. The
short reign of Bicheris might have been accommodated at the end of
row 8. If rows 7 and 8 amounted to more than 30 years for Cheops
and Djedefre', then row *9 would not be long enough for Khephren,
and Beckerath is right in postulating a row *10. (Row 1 would yield
at least 100, and at most 123 compartments, possibly corresponding to
the 115 mythical kings listed in the TC before Menes.)
Altogether we estimate ca. 75 compartments (= 74 years) in row 2,
ca. 90 compartments in row 3, ca. 104 compartments in row 4, and
ca. 87 compartments in row 5, corresponding to ca. 164 years for Dyn.
1, ca. 141 years for Dyn. 2 (if the reigns recorded were successive),
and ca. 50 years for Dyn. 3, or a total of ca. 355 years for Dyns. 1
to 3. The result is virtually the same as Kaiser’s figure of 359 years,
but markedly less than Beckerath’s 393 years. The reliability of any
sum is, however, open to question; whereas the order of succession as
preserved in the Annals conforms to contemporaneous data, its rele-
vance for counting regnal years remains conjectural.
I. 2 THE TURIN KING-LIST OR SO-CALLED
TURIN CANON (TC ) AS A SOURCE FOR CHRONOLOGY1

Kim Ryholt

Introduction

The King-list or “Royal Canon of Turin” is the only true king-list pre-
served from ancient Egypt before the Ptolemaic period. It is a “true
king-list” in the sense that the compiler of the document aimed at
recording all of the kings of Egypt along with their reign-lengths. This
stands in striking contrast to the other lists, such as the cultic assem-
blages of deceased kings engraved on walls in the temples of Sety I
and Ramesses II at Abydos and in the tomb of the priest Tjuloy at
Saqqara,2 even if these lists preserve the royal names in forms that are
superior to those of the more or less contemporaneous TC.
Only two studies on the nature of the TC as such have been pub-
lished; the first by Redford in 1986,3 and the other by myself in 1997.4
Studies of the chronological implications and attempts at reconstruc-
tions of the list have been more numerous. The most significant advances
in the reconstruction touch the Late OK, FIP and the SIP.5
Around 1820 Bernardino Drovetti, French Consul in Egypt, acquired
the TC, which eventually passed to the Turin Museum. The script is
hieratic, written on the reverse of a discarded tax register dating to the
reign of Ramesses II. Thus the king-list could have been written at the
earliest in the time of Ramesses II himself. The TC can be accurately

1
For detailed discussion of the Turin king-list, see K. Ryholt, Ä&L 14 (2004),
135–155.
2
See Redford, Annals, 18–24.
3
Redford, Annals, 1–18.
4
Ryholt, Situation, 9–33.
5
Late OK: K. Ryholt, ZÄS 127 (2000), 87–100. FIP: J.v. Beckerath, ZÄS 93 (1966),
18–19. SIP: Ryholt, Situation, 69–75, 94–97, 118–119, 151–159, 163–165; for Frg. 134,
cf. J. P. Allen, BASOR 315 (1999), 49–50.—W. Helck, SAK 19 (1992), 150–216, pro-
posed a series of new arrangements, none of them based on autopsy. I could verify
none of them whereas several can definitely be rejected, cf. Ryholt, Situation, 21. Note
that Beckerath, Chronologie, 20, follows Helck’s reconstruction.
the royal canon of turin 27

described as a copy, drawn up in a somewhat careless manner on waste


paper, from a damaged and imperfect original. The scribe apparently
did not attempt to supplement missing information from other sources
which may have been available at the time.6 Apart from a section cut
off in antiquity, the papyrus was presumably intact upon discovery. It
has since been reduced to more than 300 fragments, apparently due
to rough handling. Since Champollion first saw and described it in
1824, the papyrus has deteriorated considerably. However, thanks to
the efforts of several scholars over the past 150 years, most of the larger
fragments have now been joined. Yet, many of the very small scraps
remain unpublished, and the position of many important fragments
remains conjectural.
The papyrus has a large format, measuring 42 cm in height and
about 1¾ m in length. There are now 11 columns; one or two were
lost in antiquity when the piece of papyrus was cut off. It could have
included Dyns. 17, 18 and part of Dyn. 19. What follows refers to a
new reconstruction, in progress (Table I. 2.1).7

Table I. 2.1. Concordance between Gardiner’s edition


and the new reconstruction

Column 1 Gardiner col. I


Column 2 Includes Frg. 41–42 (Gardiner col. IX) and Frg. 150–152
and Frg. 22+unnumbered fragment (Gardiner col. X)
Column 3 Gardiner col. II
Column 4 Gardiner col. III
Column 5 Gardiner col. IV
Column 6 Gardiner col. V
Column 7 Gardiner col. VI
Column 8 Gardiner col. VII
Column 9 Gardiner col. VIII
Column 10 Includes Frg. 105+108 (Gardiner, col. IX), Frg. 138
(Gardiner, col. X) and unnumbered fragments (Gardiner,
col. X.13–21)
Column 11 Gardiner col. XI

6
Thus, for instance, the names of ten late OK kings are lost from TC and the
damaged writings of the names of two following kings are intact in the Abydos List.
7
For the new arrangement of the SIP section (columns 7–11) see Ryholt, Situation,
69–75, 94–97, 118–119, 151–159, 163–165.
28 kim ryholt

Groupings of Kings; Headings and Totals

The TC lists three main categories of “rulers”:


(1) god and demigod c. 1½ columns Cols. 1-2 (bottom)
kings (nΔr.w):
(2) spirit kings (#¢.w): c. ½ column Cols. 2 (bottom)-3
(top half )
(3) historical or human 8½ columns Cols. 3 (bottom half )-11
kings (~rmΔ.w):8
Each ruler’s name and the length of reign, either precisely in years,
months and days, or in years alone, was recorded.9 The variations in
detail provided for individual kings indicates that the Vorlage made use
of a number of sources with different formats (Table I. 2.2). The most
information is provided for the archaic kings; their reigns were recorded
in years, months and days and their ages at death were also noted. But
the royal names in this section are the most severely corrupted. The
details are therefore not necessarily indicative of superior source material.

Table I. 2.2. Division of the Turin King-list into sections


Section Period Details provided about kings

A Archaic Period Dyns. 1–2 Years, months and days + Age at death
B OK Dyns. 3–6 Years alone10
C Late OK & Dyns. 7–8 Years, months and days
FIP (Herakleopolis) Dyns. 9–10
D FIP (Thebes) Dyn. 11 Years alone
E MK & Dyn. 12 Years, months and days
SIP Dyns. 13–16

Explicit information about the nature of a given group of kings is pro-


vided by headings, most damaged. The coherent bits of what remains
may be translated as follows:11

8
This division of the kings is based on Manetho’s terminology, cf. Waddel, Manetho,
2–19.
9
By contrast to Manetho gender goes unremarked. The TC included at least one
female ruler, viz. Nofrusobk (7.2). Nitocris of Dyn. 8, who has hitherto been consid-
ered a woman on the basis of later tradition, seems to have been male; see K. Ryholt,
ZÄS 127 (2000), 92–93, 99–100.
10
There are two anomalies within this section where both months and days are
recorded as well, i.e. TC 4.7 (a difficult royal name, or perhaps rather the record of
a lacuna, recorded between Sekhem-khet and Huni) and TC 5.1 (Teti).
11
Redford, Annals, 11–13, and Helck, SAK 19 (1992), 151–216, offer different recon-
the royal canon of turin 29

Table I. 2.3

Heading for Dyns. 1–5 (3.10)


“[Kings of the house of (?)] King Menes.”
Totals for Dyns. 1–5 (4.26)
“Total of kings from Menes until [Wenis: x amounting to 767 (or 768) years.]”12
Totals for Dyns. 6–8 (5.14–15)
“[Total of ] kings [until Neferirkare': x] amounting to 181 years, 6 months,
3 days, and a lacuna of 6 (years). Total: 1[87 years, 6 months, and 3 days].”
Totals for Dyns. 1–8 (5.15–17)
“[Total of ] kings13 [from] Menes; their kingship, their years, and a lacuna
[thereto]: 9[4]9 years and 15 days, and a lacuna of 6 years. Total: [x kings
amounting to] 955 years and 1[5] days.”
Totals for Dyns. 9–10 (6.10)
“Total: 18 kings . . .”—rest lost
Heading for Dyn. 11 (6.11)
“Kings of . . .”—rest lost
Totals for Dyn. 11 (6.18)
“[Total:] 6 kings who ruled 1[36 years] and a lacuna of 7 (years). Total 143
years.”
Heading for Dyn. 12 (6.19)
“[Kings of ] the residence ’IΔ-t #wy.”
Totals for Dyn. 12 (7.3)
“Total of kings of the residence [’IΔ-t #wy]: 8 who ruled 213 years, 1 month
and 17 days.”
Heading for Dyn. 13 (7.4)
“Kings [who were] after the children (?) [of Dual] King [Sehet]epibre'.”14
Totals for Dyn. 15 (c. 10.29)
“[Total:] 6 [Hyk]sos. They ruled 1[0]8 [years].”

structions and translations. On the reconstruction of summations for Dyns. 6–8 and
1–8, see W. Barta, MDAIK 35 (1979), 13–14, and K. Ryholt, ZÄS 127 (2000), 91,
94–96.
12
There is not space enough for this figure to have included years, months and
days; possibly even the years were excluded and simply the number of kings recorded.
The number of years is based on the summations for Dyns. 6–8 and 1–8.
13
Lit. “kingship”; clearly an error.
14
Reading [˙r]-s# §rd[.w], see Ryholt, Ä&L 14 (2004), 142, n. 38. An alternative
reading [˙r]-s# ms[.w], which produces the same sense, was proposed by J. P. Allen at
the colloquium The Second Intermediate Period: Current research, future prospects at the British
Museum, 2004.
30 kim ryholt

There is no heading for the kings of Dyn. 6, nor for the Herakleopolitan
kings or those of Dyn. 14. Chronological details important from a
modern perspective go unmentioned, above all, information on over-
lapping dynasties and coregencies. When two groups of kings overlap,
they are simply listed consecutively. For example, the first Theban kings
of Dyn. 11 follow the last Herakleopolitan rulers with whom they were
contemporaneous.
The reign-lengths of Dyns. 3–6 and 9–10 recorded in full years alone
pose a special problem. It remains unclear how these figures were
rounded off. The reliability of the totals provided for groups of kings
is intimately related to the accuracy with which the individual reigns
were recorded. The totals provided for Dyns. 6–8 and 1–8 are subject
to the same factors. The former total again includes a lacuna and both
include a series of kings whose reigns are expressed only in years while
others have reigns recorded in years, months and days. The total for
Dyns. 1–8 includes no less than 26 kings recorded by years alone,
resulting in a maximum margin of error of a quarter century. The part
which appears to have suffered the most is the Late OK section (Dyns.
7–8). As noted above, a group of ten consecutive kings is entirely lost
and the names of the next two are only partly preserved.15 The lost
kings are accounted for by the word wsf (“lacuna”) in both the total
for Dyns. 6–8 and that covering all of Dyns. 1–8.
The total of Dyn. 11 is recorded in full years only, obviously because
the kings reigns were so recorded. Since the fractions of individual
reigns are omitted, the total is inevitably approximate, with the mar-
gin of error amounting to few years. The error might be more significant
if “Mentuhotep I”, who never actually reigned as king, was assigned a
fictitious reign-length. Finally, there is the lacuna referring to Mentuhotep
IV and the question of the accuracy of the 7 years ascribed to him in
the total for Dyn. 11.
There is a certain amount of evidence that several Dyn. 12 kings
reigned jointly and that some coregencies could have lasted up to a
decade. Yet the scribe who calculated the sums for the individual groups
of kings was not aware of (or did not take account of ) the overlap-
ping reigns. He simply added the full lengths of individual reigns to
calculate the total for the dynasty. The overlapping reigns probably

15
J.v. Beckerath, JNES 21 (1962), 140–147; K. Ryholt, ZÄS 127 (2000), 96–99.
the royal canon of turin 31

accounted for a bit more than 2 decades. Furthermore the TC ’s infor-


mation about Dyn. 12 is disquietingly incompatible with the ample
contemporaneous sources.16
In the sections for Dyns. 13 and 14, the notations of at least two
lacunae are preserved. One is recorded after the entry for Sonbef (7.6)
and the other after Nebsenre' (9.14). It is not clear how many kings
were lost, but at least one king can be identified by name in relation
to the first lacuna, viz. Nerikare'.17 Further kings may have been lost
in the same sections.18 There are two instances where reign-lengths are
partially damaged. One is #w-ôb-r' of Dyn. 14 whose reign is recorded
as “[x years], lacuna, 18 days” (9.12) and the other is swsr-n-r' of Dyn.
16 with “12 years, lacuna, [x] days” (11.8).

Relation to Manetho’s King-list

It is significant that the five sections A–E in Table I. 2.2 all correspond
to the dynastic arrangements Manetho adopted. Section A equates to
Manetho’s Dyns. 1–2, Section B to his Dyns. 3–6, Section C to his
Dyns. 7–10, Section D to his Dyn. 11, and Section E to his Dyns.
12–16. The correspondence becomes even more striking when the tex-
tual division of the kings into groups is taken into account. The TC
groups the following kings: Dyns. 1–5, Dyn. 6, Dyns. 7–8, and Dyns.
9–10, followed by Dyns. 11, 12, 13, 14, 15 and 16. The main difference
from Manetho’s list is that the TC does not split up Dyns. 1–2 (Archaic
Period), Dyns. 3–5 (Memphis), Dyns. 7–8 (Memphis), and Dyns. 9–10
(Herakleopolis).19 The few reigns of the OK that were not corrupted
beyond recognition in Manetho seem to be either rounded up or down
in comparison to the TC. Hence Manetho’s figures seem to be based
on a tradition with reigns recorded in years and months if not also in
days.

16
See below Schneider, Chapter II. 7.
17
Ryholt, Situation, 318.
18
Ryholt, Situation, 70 (Dyn. 13), 94–95 (Dyn. 14). Alternative interpretations have
been presented by C. Bennett, GM 159 (1997), 11–17 and J. P. Allen, BASOR 315
(1999), 50–51.
19
Cf. below Seidlmeyer, Chapter II. 3.
32 kim ryholt

Conclusion

Despite its immense historical value, the TC is in various ways far


removed from an ideal source. The incomplete state of preservation
and inadequate publication inhibit full access both to the information
it once contained and to what remains. There are also several intrin-
sic features that detract from its value as chronological source. If the
scribe’s priorities were historical, he was clearly not primarily concerned
with either absolute or relative chronology. This is obviously not an
ideal point of departure, and it warns us that there may be errors that
we are presently unable to verify lurking in the document. Accordingly,
the text should be treated with circumspection.
I. 3 KING-LISTS AND MANETHO’S AIGYPTIAKA

The Editors

Comprehensive king-lists are engraved on walls in the temples of Sety


I and Ramesses II at Abydos, at Karnak and in the tomb of the priest
Tjuloy at Saqqara.1 They list kings in an apparently correct sequence
and render their names in forms superior to those of the more or less
contemporaneous TC. In the Abydos list, the names of ten kings lost
from the Late OK section of the TC and two damaged names that
follow are recorded intact.2 By contrast, the king-list at Karnak from
the time of Thutmose III, though mentioning a number of rulers omit-
ted in other lists, does not give the names in historical sequence. Since
all these lists served the cult of the deceased kings, the names sufficed,
and further information, such as regnal years, was not necessary. None
mentions kings considered to be illegitimate, for example Hatshepsut
or the Amarna rulers. The later king-lists of Greek historians seem to
derive from the annalistic tradition, rather than from such cultic king-
lists.
In Book II of his History, Herodotus reported the names and correct
reign lengths for the kings of Dyn. 26, and he stated that there were
330 earlier kings, including 18 Ethiopians and Queen Nitocris, but his
information on the period prior to Dyn. 26 is virtually useless for recon-
structing Egyptian history and chronology.
A king-list with 38 + 53 names and regnal figures is attributed to
Eratosthenes.3 Apollodoros preserved 38 names that were copied by
George the Monk, known as Syncellus. No. 29 is easily recognizable:
XVMAEFYA < Sty mr.n Pt˙, i.e. Sety I who is given 11 regnal years,
as is no. 36 SIFOAS < *SIFYAS, i.e. Siptah with 5 regnal years.4 The
names are accompanied by their secondary Greek translations, added
only after noticeable corruption of the text occurred.

1
For the Abydos and Saqqara lists, see Redford, Annals, 18–24.
2
Cf. above, Chapter I. 2.
3
Waddell, Manetho, 213–225.
4
Krauss, Amarnazeit, 274–276, with additional literature.
34 the editors

Diodorus utilized Herodotus and other sources for the chapters on


Egypt in his Bibliotheca Historica.5 As a rule he did not list regnal years,
except for Kheops and Khephren where he follows Herodotus. Diodorus
mentions five ruling Egyptian queens instead of Herodotus’s single
example.
Manetho’s Aigyptiaka is purportedly the work of a priest who was a
contemporary of Ptolemy II.6 The Aigyptiaka displays remarkable simi-
larity to the Turin King-list: rulers, including illegitimate kings recorded
by name with reign length, arranged in groups, and listed in a sequence.7
But there are two features which are specific to the Aigyptiaka and which
have no counterpart in pharaonic tradition: glosses and antisemitism.
The former reflect primarily Herodotus and Diodorus,8 which suggests
that the list may have been compiled after Diodorus. The antisemitism
is manifest in the story of the enemies of king Amenophis, i.e. the lep-
ers of Egyptian origin and the descendants of the Hyksos as common
forefathers of the Jews.9 Antisemitism is not attested before the Macca-
beans;10 therefore the story cannot be ascribed to Manetho in the 3rd
century BC, but should rather be dated to the 1st century AD. The
traditional explanation that Manetho made use not only of authentic
records, but also of popular romances devoid of historical value,11 does
not take into account the motive of antisemitism and its history.
It is a fact that the Aigyptiaka was cited by none of the great com-
pilers of antiquity like Pliny, Diodor, and Strabo, nor was it used by
any known Alexandrian scholar.12 When Apion cited the destruction of
Avaris by 'Ahmose according to the Egyptian historian Ptolemy of
Mendes,13 he did so without mentioning the differing version in the

5
E. Murphy, The antiquities of Egypt: a translation with notes of book I of the Library of
history of Diodorus Siculus, rev. and expanded (New Brunswick, 1990).
6
G. Verbrugghe & J. M. Wickersham, Berossos and Manetho (Ann Arbor, 1999),
95–120.
7
For Manetho’s king-list, see Waddel, Manetho, and F. Jacoby, Die Fragmente der
Griechischen Historiker, III C (Berlin & Leiden, 1958), no. 609.
8
L. Stern, “Die Randbemerkungen zu dem manethonischen Königscanon”, ZÄS
23 (1885), 87–96.
9
M. Stern, Greek and Latin authors on Jews and Judaism I–III ( Jerusalem, 1974–1984),
66–83.
10
I. Heinemann, “Antisemitismus”, in: Pauly-Wissowa, RE, Supplement vol. 5 (1931),
3–43.
11
Meyer, Chronologie, 78–79; Gardiner, Egypt, 47; Beckerath, Chronologie, 35.
12
G. F. Unger, Chronologie des Manetho (Berlin, 1867), 3; 116.
13
A. Dihle, “Ptolemäus von Mendes”, in: Pauly-Wissowa, eds., RE, Reihe 1, Vol.
23.2 (1959), 1861.
king-lists and manetho’s AIGYPTIAKA 35

Aigyptiaka. The first to cite the Aigyptiaka was Flavius Josephus in Contra
Apionem; subsequently, the christian chronographers Africanus and
Eusebius cited it; still later Syncellos contributed greatly to the trans-
mission.14 The silence of authors earlier than Josephus and the anachro-
nisms in the text arouse suspicion that it is a pseudepigraphic work,
based on the authentic tradition of the annalistic king-list, but com-
piled after Diodorus.15
Gardiner’s comments on the Aigyptiaca can be paraphrased as follows:16
In Manetho’s work the entire history of Egypt, after the reigns of the
gods and demi-gods, was divided into 31 dynasties of royal families,
beginning with Menes and ending with Alexander the Great’s conquest
in 332 BC. In spite of all the defects this division into dynasties exhibits,
it has taken so firm a root in the literature of Egyptology that there
is but little chance of its ever being abandoned. In the forms in which
the book has reached us there are inaccuracies of the most glaring
kind, these finding their climax in Dyn. 18, where the names and true
sequence are now known from contemporary sources. Africanus and
Eusebius often do not agree; for example Africanus assigns nine kings
to Dyn. 22, while Eusebius has only three. Sometimes all that is vouch-
safed to us is the number of kings in a dynasty (so in Dyns. 7–10, 20)
and their city of origin. The royal names are apt to be so incredibly
distorted, that of Senwosret I of Dyn. 12, for instance, being assimilated
in the form of Sesonchosis to that of Shoshenk of a thousand years
later. The lengths of reigns frequently differ in the versions of Africanus
and Eusebius, as well as often showing wide departures from definitely
ascertained figures. When textual and other critics have done their best
or worst, the reconstructed Manetho remains full of imperfections.
None the less, Manetho did preserve details which have been con-
ceded as historical by Egyptologists if only very late. An example is
the name of a king Nephercheres, whom Manetho had placed in Dyn.
21; in the 1930s a small object bearing the name was found among

14
See in detail Beckerath, Zwischenzeit, 11–20; idem, Chronologie, 35–38, assuming
that the Aigyptiaka is authentic.
15
Waddell, Manetho, VIII, XIV; E. Hengstenberg, Manetho und die Hyksos. Beilage zu:
Die Bücher Moses und Ägypten (Berlin, 1841); R. Krauss, “Manethos Ägyptische Geschichte—
eine ptolemäische oder römische Kompilation?”, in: Timelines. Studies in honour of Manfred
Bietak, E. Czerny et al., eds., (Leuven: OLA 149.3, 2006), 227–234.
16
Gardiner, Egypt, 46–47.
36 the editors

the grave goods of Psusennes I in Tanis.17 Another example is the rul-


ing queen Akencheres (< 'Ankhetkheprure') of Dyn. 18, the daughter
of a king, whose historical existence was archaeologically first confirmed
in the 1970s.18 Manetho’s reign lengths are at times correct to the
month, for example in the case of Ramesses Miamun, i.e. Ramesses
II, who reigned for 66 years and 2 months according to Josephus’s
copy of Manetho. On the other hand, Manetho’s pre-Dyn. 18 regnal
figures often seem to be systematically distorted by multiples of ten,19
probably in an attempt to bring Egyptian chronology into line with a
version of biblical chronology.

17
Cf. below, Chapter II. 9.
18
Cf. below, Chapter II. 8.
19
Helck, Manetho, 81–83.
I. 4 GENEALOGY AND CHRONOLOGY

Morris L. Bierbrier

Genealogy, the study of family pedigrees, can provide a useful fix on


proposed chronological schemes for ancient Egypt. Obviously the pas-
sage of time can be limited or, in certain instances, can be expanded,
by the knowledge that an individual, his family, or a stated number of
generations of a family, lived within a set period. It has been calcu-
lated that a generation might comprise 25–35 years on average, but,
in the surviving ancient Egyptian documentation, we are not neces-
sarily dealing in averages.1 Thus the possibility remains that in certain
cases an elderly father might produce a son who in turn might live to
a great age and so throw any calculations off. A complex and inter-
linked genealogy might overcome this possibility but such genealogies
are rare.
The use of genealogies in chronological research is limited by sev-
eral factors. The most important of these is the lack of documentation.
This is especially true of those periods when the chronology is most
uncertain. In other periods, abundant documentation can lead to con-
fusion as the same names are used repeatedly and identification of
different generations may be uncertain. The documentation normally
only refers to the elite families although occasionally at such places as
Kahun or Deir el-Medina, information on lower-ranking families is
forthcoming. It is now clear that the Egyptians practised some form of
ancestor worship and records were certainly kept by some families of
their ancestral lines although such information was not necessarily
recorded on stone and thus has not survived to posterity.2 The census
records which survive give detailed information of families with both
parents’ names. It is of course true that Egyptians, like all other peo-
ple down to this day, could be tempted to falsify their genealogies or

1
D. Henige, “Generation-counting and late New Kingdom chronology”, JEA 67
(1981), 182–184.
2
M. L. Bierbrier, The Tomb-builders of the Pharaohs (London, 1982), 95–6; J. Keith-
Bennett, “Anthropoid Busts II”, BES 3 (1981), 43–72.
38 morris l. bierbrier

at any rate to attach themselves erroneously to a good family line so


particularly illustrious ancestries should be treated with caution. The
genealogical information becomes fuller as time progresses and is espe-
cially voluminous from the late NK to the LP. Some pedigrees are
known extending up to 13 generations, but it is usually difficult to fix
these to a certain chronological range.3 At least two pedigrees claim a
genealogical link back to the MK. However, the inscription of Khnumibre'
jumps abruptly from Dyn. 19 to Imhotep of Dyn. 3 and so cannot be
taken as a serious link.4 Similarly the inscription of 'Ankhefensakhmet
which names 60 generations, mostly holders of the high priesthood of
Ptah, going back to the MK with the contemporary ruler for each, has
too many gaps and inconsistencies to be accepted at face value.5 It
seems to have been put together from various lists of high priests (prob-
ably unrelated) and other ancillary material. It can only be used when
independent information is available from other sources.
Another problem with genealogy is the lack of proper genealogical
terms in the Egyptian language. “Father” jt, “Mother” mwt, “Son” z #,
“Daughter” z#t, “Brother” sn, and “Sister” snt are used, but there are
no words for other relationships so “uncle” is “brother of my father”
or “nephew” is “son of my brother”.6 To avoid such long circumlo-
cutions, jt / mwt could be used for “father- / mother-in-law”, while sn
might also mean “brother-in-law”, “nephew” or even” uncle”, or might
conceivably be no relation at all but a good friend and contemporary.
The term jt can also be used to specify an ancestor and not just a
father.7 One could disentangle the exact relationships if proper docu-
mentation were available for all problems, but it is usually not.
With regard to maternal relationships, it is not always clear that the
wife of an official is the mother of his children unless specifically stated.

3
J.-C. Goyon, “Les cultes d’Abydos à la basse époque d’après une stèle du Musée
de Lyon”, Kemi 18 (1968), 29–44; R. El-Sayed, “Deux statues inédites du Musée du
Caire”, BIFAO 83 (1983), 135–143; L. M. Leahy & A. Leahy, “The Genealogy of a
Priestly Family from Heliopolis”, JEA 72 (1986), 133–147.
4
Couyat & Montet, Ouâdi Hammâmât, Nos. 91–93; Wildung, Rolle, Doc. XVI.130,
83–84.
5
Borchardt, Mittel, 96–100; C. Maystre, Les grands prêtres de Ptah de Memphis (Freiburg
& Göttingen, 1992), 93–97.
6
G. Robins, “The Relationships specified by Egyptian Kinship Terms of the Middle
and New Kingdoms”, CdE 54 (1979), 197–217; M. L. Bierbrier, “Terms of Relationship
at Deir el-Medina”, JEA 66 (1980), 100–107.
7
G. A. Gaballa, The Memphite Tomb-chapel of Mose (Warminster, 1977), 22; Allam,
Ostraka, 44.
genealogy and chronology 39

Such a relationship is usually assumed but may not be correct. The


guardian Penbuy of Dyn. 19 is known to have had two wives—
Amentetwosret and Iretnefer. A stela shows Penbuy, Iretnefer and his
adult son Amenmose, but another shows him with Amentetwosret and
an infant Amenmose. In the first stela, Amenmose must be shown with
his step-mother, but the stela gives no indication of this.8 The dissolu-
tion of marriages by death or divorce would not have been uncom-
mon, yet plurality of marriages is rarely documented.
For genealogies to have a major impact on chronological problems,
it is essential that there must be a clear and uncontroversial link of
genealogy to some fixed chronological point, usually the reign of a
King. If links can be found with successive generations, the effectiveness
of the genealogy with regard to the order of succession and more impor-
tantly the maximum passage of time becomes even more crucial. If the
genealogy in question is that of the royal family itself, then the fixed
links are self-evident. One important genealogy of this nature is that
of Pasenhor of Dyn. 22 who traces his ancestry through 16 generations
including four kings.9
Very little documentation, let alone genealogical information, sur-
vives from the development from writing ca. 3400 BC to the end of
Dyn. 3. One key document, recently discovered at Abydos, is the dynas-
tic seal of King Den of Dyn. 1 which lists his immediate predecessors:
Nar-mer, 'Aha, Djer, and Wadj. The genealogical content is provided
at the end with the name of the King’s mother as Merytneith.10 In
view of her appearance, it is highly probable that not only do we have
a dynastic listing but also a genealogical listing—father-to-son. Unfor-
tunately, the length of time per generation can only be estimated. A
second seal of the last king of Dyn. 1 lists all his predecessors, but
leaves out Merytneith. It would be optimistic to assume a straight father-
to-son succession throughout the whole dynasty. No such documenta-
tion survives for Dyns. 2–3 and here the order and genealogy of the
kings remains uncertain.

8
M. L. Bierbrier, Hieroglyphic Texts from Egyptian stelae etc. 10 (London, 1982), pl. 72;
M. L. Bierbrier & H. de Meulenaere, “Hymne à Taouêret sur une stèle de Deir el-
Médineh,” in R. Holthoer & T. Linders, eds., Sundries in honour of Torgny Säve-Söderbergh
(Uppsala, 1984), 23–29.
9
Kitchen, TIP1, 488, Table 19.
10
G. Dreyer, “Ein Siegel der frühzeitlichen Königsnekropole von Abydos”, MDAIK
43 (1987), 33–43; G. Dreyer, “Umm el-Qaab”, MDAIK 52 (1996), 72–73.
40 morris l. bierbrier

While the amount of information about private families increases


from Dyn. 4, the tomb inscriptions do not normally name the parents
of the deceased but more usually only his wife and children. Thus it
is extremely difficult to put together a lengthy and coherent genealogy
apart from those of the royal families of Dyns. 4–6. There too many
gaps remain and many of the reconstructions are speculative. What is
certain is that the genealogical succession to the throne when known
did not run smoothly and did not follow in generation sequence in all
cases so the king-lists alone do not reflect generations; for example,
Khephren succeeded his brother Ra'djedef in Dyn. 4 and Ra'neferef
was probably followed by his brother Neuserre' in Dyn. 5, while
Nemtyemzaf I was followed by his brother Pepy II in Dyn. 6. Chronology
here is indicated rather by the careers of these few officials who list
the sovereigns that they served and so limit the time which has elapsed.
Thus Sekhemkare' son of Khephren lived through the reigns from
Khephren of Dyn. 4 to Sahure' of Dyn. 5 while Ptahshepses was
brought up under Menkaure' and lived into the reign of Neuserre'. It
has been suggested that Dyns. 3–4 and 4–5 were linked by marriages
of royal heiresses, but there is no proof of this.11
Unfortunately, genealogy is not much of an aid in determining the
chronology of the FIP. The genealogies of the royal families are uncer-
tain and those of the nomarchs do not extend for enough time or pre-
sent enough fixed links to prove crucial. The list of previous nomarchs
in the tomb of Ukhhotep of Meir is unfortunately not complete or tied
to any fixed points. Nor is it clear that the list is in any way genealogical.12
When adequate documentation occurs in the MK, there is still not
enough material to present long coherent genealogies apart from the
royal family of Dyn. 12 itself.13 The papyri from Kahun allow a few
families to be reconstructed and the succession of office holders, in one
case father-to-son, confirms the dating order of some papyri.14 However
these fragments are not crucial to the relative chronology of Dyn. 12
which can be fixed by other methods.

11
CAH I/2, 145–189; M. Verner, Forgotten Pharaohs, lost pyramids: Abusir (Prague, 1994),
134–155 on Ra'neferef.
12
A. Blackman, The Rock Tombs of Meir (London, 1915), III, 16–21.
13
Franke, Personendaten.
14
U. Luft, “Illahunstudien I: Zu der Chronologie und den Beamten in den Briefen
aus Illahun”, Oikumene 3 (1982), 101–156.
genealogy and chronology 41

Similarly no consecutive genealogies can be found during the SIP.


Some detailed genealogies of short-lived royal families can be put
together, but the links between them remain tenuous and open to
different interpretations. It is possible that some kings indicated their
filiation in their royal names which would provide a genealogical sequence
but this view has perhaps been pushed too far.15 The later genealogies
which claim to reach back to the MK cannot be taken seriously and
seem merely to look to well-known historical figures without any genealog-
ical consistency.
The advent of the NK led to an increasing amount of genealogical
material in the form of statues, stelae, tomb inscriptions and papyri,
both official and unofficial. The pedigrees of the royal families of Dyns.
18–19 can be constructed in some detail apart from the confusion
which occurs at the end of the dynasties. The regnal years known from
other sources fix a chronology which does not conflict with the royal
genealogies. The private genealogies and careers of officials again confirm
the standard chronology that there are not too many years missing
from the known regnal years. For example, 'Ahmose Penekheb served
from 'Ahmose I to Hatshepsut so guaranteeing that the reigns of
Amenhotep I and Thutmose I–II cannot extend over too long a period.16
Until recently no family could be traced which extended from Dyn.
18 to Dyn. 19 with the slight exception of the parents of Ramesses I
who must have lived then but are only fixed in relation to their son
and the dubious genealogy of the high priests of Ptah. New research
has revealed that the high priest of Amun Wennefer lived at the end
of Dyn. 18, certainly during the reign of Haremheb and possibly that
of Tut'ankhamun. His younger son Amenmose is well attested in the
reign of Ramesses II, being in charge of the Ramesseum probably from
early in the reign when doubtless adult. The career of father and son
prove that the contentious reigns of Haremheb and Sety I should not
be unduly long but are still too imprecise to determine for certain a
long or short reign for Haremheb.17
From Dyn. 19 onwards much more genealogical information becomes
available concerning the royal court, the officials, and even humbler

15
Ryholt, Situation, esp. 207–289.
16
CAH II/1, 295.
17
F. Kampp-Seyfried, “Die Verfemung des Namens P #-rn-nfr”; D. Raue, “Ein Wesir
Ramses’ II.”, in: Stationen, 303–319 and 341–345.
42 morris l. bierbrier

folk such as the workmen of Deir el-Medina. The royal inscriptions


now name in detail the king’s sons, daughters and wives, while the
tombs of the officials and workmen give the names of the parents,
grandparents, in-laws, and other relations. This new information can
be crucial in determining the length of uncertain periods or the max-
imum extent possible of reign lengths. For example, some uncertainty
remains at the end of Dyn. 19 and the beginning of Dyn. 20 when it
has been speculated in the past that an interregnum took place. The
maximum year dates are known: Ramesses II, 66 years, 2 months (his
last); Merneptah, 10 years, Sety II 6 years (his last), but it is not clear
if the reigns of Amenmesses, 3 years; Siptah 7 years; Twosre 8 years;
and Sethnakhte of Dyn. 20, 2 years are consecutive or contemporary
with others.18 However, certain individuals are attested who survive
through this period into the reign of Ramesses III such as the vizier
Hori, a great-grandson of Ramesses II19 and the workman Nekhemmut
son of Khons who is attested under Ramesses II and was a foreman
in year 13 of Ramesses III.20 The scribe Kenherkhepeshef is appar-
ently attested at Deir el-Medina from year 30 of Ramesses II until the
reign of Siptah, and his widow Naunakhte (who must have been a
great deal younger than him) remarried, had eight children, and sur-
vived until the reign of Ramesses V.21 In view of this evidence, it seems
logical that Merneptah could not have reigned much longer than ten
years. It is highly probable that Amenmesse’s years are to be totally
contained within those of Sety II and that Twosre backdated her reign
to the death of Sety II, thus encompassing that of Siptah and so reign-
ing two years not eight. In view of Sethnakhte’s recently discovered
stela, he may well have dated his reign from the death of Siptah or
shortly thereafter and so was contemporary and not consecutive with
Twosre. More importantly, there can have been no interregnum and
so the time elapsed from the death of Sety II to the accession of
Ramesses III was probably not much more than ten years if that.

18
If the highest known regnal date is the last, then the reign could have been as
much as 11 months shorter. For the latest views on Siptah’s reign see R. Drenkhahn,
Die Elephantine-Stele des Sethnacht und ihr historischer Hintergrund (Wiesbaden, 1980); Jansen,
Village Varia, 116.
19
H. de Meulenaere, “Le vizier ramesside Hori”, Annuaire de l’Institut de philologie et
d’histoire orientales et slaves 20 (1968–72), 191–98.
20
Bierbrier, Kingdom, esp. 30–33, corrected in M. L. Bierbrier, “The Family of Sen-
nedjem”, CdE 59 (1984), 199–213.
21
Bierbrier, Kingdom, 26–29.
genealogy and chronology 43

A great deal of documentary evidence is available for the regnal


dates of Dyn. 20, and genealogical information confirms that few extra
years should be added to those known, certainly not more than a
decade. The case of Naunakhte has already been mentioned, but other
families from Deir el-Medina can be traced from Dyn. 19 to the begin-
ning of Dyn. 21. The family of Sennedjem runs from Dyn. 19 to the
beginning of Dyn. 21, notably Nekhemmut junior who is attested in
office from Ramesses IV to Ramesses IX. The family of Kaha can be
traced from early Dyn. 19 to Ramesses XI. More importantly, the
scribe Amennakhte was appointed to office in year 16 of Ramesses III
and survived until Ramesses VI. His son is attested from Ramesses III
to Ramesses IX, while his grandson appears from Ramesses VI to
Ramesses IX. Three further generations are known through to Smendes
of Dyn. 21.22 The genealogies thus confirm that the extent of Dyn. 20
is most likely correct, and there are no substantial gaps in chronology.
From Dyn. 22 onwards a large body of texts on statuary and coffins
record the genealogies of the royal families and the priestly class. Some
of these extend back to Dyn. 21 and one even to Dyn. 19 although
that genealogy is doubtful.23 Although these genealogies are not pre-
cise enough to solve the various chronological cruxes of the period,
they do link Dyn. 21 to Dyn. 26, and even with the provision of a
genealogical jump or two, they limit the time period so that it is unlikely
that the entire time span can extend further then has been postulated
by Kitchen. It may be slightly abbreviated, but again the genealogies
make clear that the time span cannot be radically less than the gen-
erally accepted chronology of the period. For example, the fourth
prophet of Amun Djedkhonsuiuefankh, a contemporary of Osorkon I,
has a great-great-grandson the fourth prophet of Amun Nakhtefmut,
who was a contemporary of Osorkon III and whose granddaughter
married into the Besenmut family which is attested from Dyn. 22 to
Dyn. 26.24

22
Bierbrier, Kingdom, 19–44.
23
Bierbrier, Kingdom, 51–53, where the line between Ipuy under Merneptah and
Ankhefenkhons under Osorkon I is doubtful.
24
For royal and priestly families of this period see Kitchen, TIP; Bierbrier, Kingdom,
45–108, with revisions in Bibliotheca Orientalis 36 (1978), 306–309; G. Vittmann, Priester
und Beamte im Theben der Spätzeit (Vienna, 1978); J. Taylor, “A priestly family of the
25th Dynasty,” CdE 59 (1984), 27–57; D. Aston & J. Taylor, “The Family of Takeloth
III and the ‘Theban’ Twenty-third Dynasty,” in Leahy, Libya, 131–154.
44 morris l. bierbrier

From Dyn. 26, the chronology of Egypt becomes firmly fixed to that
of Persia, Greece, and ultimately Rome. Genealogy thus becomes for
the most part unnecessary to determine chronological conundrums.
However, family information can still prove useful in resolving some
minor points at issue. Unfortunately, much of the genealogical infor-
mation is patchy and uncertain with the most information from Thebes
which was no longer the centre of major activity. Genealogies and indi-
vidual careers can be constructed from surviving papyri and have been
useful in confirming the dates of some of the obscure rebel pharaohs
such as Harwennefer and 'Ankhwennefer.25 The survival of coherent
genealogical material from the Roman Period is severely limited and
is no longer of any practical aid in settling minor matters of chronology
which remain.

25
C. A. R. Andrews, Catalogue of Demotic Papyri in the British Museum IV. Ptolemaic Legal
Texts from the Theban Area (London, 1990), passim for genealogies; P. W. Pestman, “A
family archive which changes history”, as well as, P. W. Pestman, “Haronnophris and
Chaonnophris”, in: Hundred-Gated Thebes: Acts of a Colloquium on Thebes and the Theban
Area in the Graeco-Roman Period, S. P. Vleeming, ed., (Leiden: PLB 27, 1995), 91–137.
I. 5 METHODS OF DATING AND THE
EGYPTIAN CALENDAR

The Editors

During Dyn. 1 years were not yet counted, but simply named, as
exemplified by the so-called year labels of Dyn. 1 that were attached
to oil vessels. Since the reign of king “Serpent”, labels were addition-
ally inscribed with the hieroglyph for “year” (rnpt).1 Presumably epony-
mous occasions were selected at the outset of or early in the year, being
scheduled or symbolic and only coincidentally historical.2 The labels
are the earliest evidence for reckoning time by years,3 but the type of
year, that was in any case civil, not regnal, is not known. The annals
attest that the “months” and days of ['Aha’s] last year and [Djer’s]
first year, add up to 10 “months” and 20 days, less than either a full
lunar or solar year. The figures cannot be explained until further infor-
mation comes to light; perhaps they are just a scribal error.
Rnpt zm#-t #wy designated an incomplete accession year at least as early
as [Djer] down through Dyn. 8. How the following years were expressed
changed over time. At least from the penultimate year of ['Aha] and
through the end of Semer-khet’s reign, i.e. for most of Dyn. 1, the
annals record the biennial occurrence of “following of Horus”, but there
was no successive numbering of them.4 Between Ny-netjer and Khase-
khemwy at the end of Dyn. 2 the annals document a dating system
that was based upon a regular biennial census (Δnwt, jpt),5 numbered
successively within a reign and coinciding with the biennial “following
of Horus”. Whereas the “following of Horus” is still attested for Djoser,
the biennial census is not. When it reappears under Snofru, it is not
a regular biennial event. According to the annals, the 8th occurrence

1
The earliest known “year label” dates from the reign of Narmer, see G. Dreyer,
MDAIK 54 (1998), 138–139.
2
Wilkinson, Annals, 63–64.
3
Helck, Thinitenzeit, 144–175.
4
Wilkinson, Annals, 90–91.
5
Edel, Grammatik, 179–180.
46 the editors

of the census followed the 7th occurrence without an intervening cen-


sus-free year. Contemporaneous dates from Snofru’s pyramid at Maidum
attest a series of annual counts, recorded as rnpt zp x (year of the xth
count),6 interrupted now and then by a rnpt m-¢t zp (year after the xth
count).7 Scholars are at odds about whether there was a regular bien-
nial count thereafter during the OK; the annals of earlier Dyn. 6 indi-
cate a biennial count.8
Beginning in Dyn. 11 successive calendar years during each reign
were counted. Gardiner presumed that during the MK the kings dated
their second regnal year from the New Year’s Day (I Akhet 1) follow-
ing the actual day of their accession, so that their first year consisted
merely of some months and days after the demise of their predecessor.9
This may have been the case, since pBerlin 10055 from Illahun attests
that year 1 of [Amenemhet III] followed directly on year 19 of [Senwosret
III],10 which may reflect the accession of Amenemhat III after Senwosret
III had died in the course of year 19 or 20. But it may also signify
that a coregency began after year 19.
In Dyn. 18 a new system was adopted which continued until at least
ca. 800 BC.11 Year 1 began on the actual day of accession and the
following years were counted, accordingly, from the anniversary in the
civil year, so that the civil year now always spanned parts of two reg-
nal years. From Dyn. 26 through the Roman period the antedating
system was introduced so that regnal year 2 began on I Akhet 1 after
the accession.12
During the entire pharaonic period the concept of an “era” was
employed only once. The w˙m mswt era began in 19 Ramesses XI and
lasted at least 10 and perhaps as much as 12 years.13 The so-called
400 year era of Seth Nubti14 should also be mentioned, as well as the
attribution of the Amarna pharaohs’ reigns to Haremhab.15 During the

6
For the reading rnpt zp, see E. Edel, JNES 8 (1949), 35–39; Gardiner read ˙#t zp,
see his Grammar, Excursus C.
7
See below, Chapter II. 4.
8
See below, Chapter II. 5.
9
Gardiner, “Years”, 21–23, followed by Beckerath, Chronologie, 10.
10
When Gardiner wrote, the date in pBerlin 10055 was mistakenly associated with
Senwosret II and III.
11
Cf. below Jansen-Winkeln, Chapter II. 10.
12
A. Leahy, JEA 74 (1988), 187.
13
See below, Chapter II. 8.
14
Cf. Beckerath, Zwischenzeit, 161.
15
See below, Chapter II. 8.
methods of dating and the egyptian calendar 47

MK, the nomarchs counted their years of office independently from


royal years,16 and it is quite possible that the Theban HP’s of early
Dyn. 21 counted their pontifical years.17

The Solar Calendar of 365 Days

There were 12 months of 30 days in the Egyptian calendar, supple-


mented by five additional days, the so-called epagomenal or “added”
days, for a total of 365 days. A tendency to regard the year as amount-
ing to only 360 days is evident, for example when the daily income of
a temple is stated to be one 360th of the yearly revenue.18 The well-
known disregard of the epagomenai in calendar schemes seems to be
another consequence of this tendency.19
Dates within a year were expressed in terms of the three seasons of
four months each: Akhet (#¢t), “inundation”; Peret ( prt) “winter”, pre-
sumably the season of the “emergence” ( pr) of the fields from the flood;
Shemu (“mw) “summer”.20 Originally months were numbered, not named,
but a few month names are known from the MK.21 Varying series of
names are documented for the months in the NK from which the
Greek, Aramaic, and Coptic names of the civil months derive; Ptolemy
used the Greek names.22
The heliacal rising of the fixed star Sirius (Egyptian: Sothis) was the
only astronomical event which occurred on a certain day in the 365-
day civil calendar (see below, chapters III. 9.10). Because the calendar
did not provide for an extra day every 4 years, the date of the rising
of Sirius in the civil calendar shifted accordingly. The concept of a
“Sothic cycle” for a complete shift of 1460 years (i.e., 365 days × 4)
is first documented in the Hellenistic Period. The earliest examples for
dating in the civil calendar of 365 days come from Djoser’s reign,23 or
ca. 150 years after its introduction.

16
Cf. for example W. K. Simpson, JARCE 38 (2001), 7–8, regnal year 43 of Senwosret
I corresponding to year 25 of the nomarch Amenemhet.
17
See below, Chapter II. 9.
18
Cf. G. A. Reisner, “The Tomb of Hepzefa, Nomarch of Siut”, JEA 5 (1918), 84.
19
Cf. for example, Leitz, Studien, 5–6; A. J. Spalinger, OLZ 87 (2002), 25.
20
Gardiner, Grammar, Excursus C.
21
J. P. Allen, The Heqanakht Papyri (New York: PMMAEE 27, 2002), 135–137.
22
See below, Chapter III. 11.
23
J. Kahl, N. Kloth & U. Zimmermann, Die Inschriften der 3. Dynastie (Wiesbaden:
ÄA 56, 1995), 70–71.
48 the editors

Calendar Adjustment

R. Weill has played with the idea that the SIP was shorter than the
standard chronology allows, implying that at a calendar adjustment
days had been cut out of the 365-day year.24 Parker argued against
the feasibility of such an adjustment;25 his arguments were reinforced
by Depuydt.26
It is indeed correct that Pharaonic sources do not attest an interca-
lated day.27 But what if the Egyptians added or lost a calendar day in
the course of their history, be it by intercalation or by mistaken day
counting? The close agreement between late Egyptian and general
chronology precludes such an error or deliberate intercalation after the
beginning of the Saite Period. The loss or gain of a calendar day would
shift the absolute date of a recorded lunar date by either exactly –11
(–14) years or by exactly +14 (+11) years.28 Shifts of this size are not
compatible with the standard chronology of the NK. Furthermore the
supposition of an intercalated day would break the link between
the lunar date and a Sothic date implicit in the astronomical dates of
the Illahun archive.
It is well known that in 238 BC Ptolemy III attempted to introduce
an improvement in the Egyptian calendar by adding a 366th day every
four years, anticipating the later Julian calendar reform. But his edict
went unheeded. A second attempt under Augustus succeeded in impos-
ing a form of the Julian calendar on Egypt. This so-called Alexandrian
calendar remained in use in Egypt until the Arab conquest, and it still
survives today as the liturgical calendar of the Coptic and Ethiopian
churches.
Alongside the 365 day calendar, a lunar calendar was also used.
Specialists are not in accord on how it functioned in detail (see below,
chapters III. 6.10).

24
R. Weill, CdE 24/47 (1949), 13–18.
25
R. A. Parker, RdE 9 (1952), 101–108.
26
L. Depuydt, “On the Consistency of the Wandering Year as Backbone of Egyptian
Chronology” JARCE 32 (1995), 43–58.
27
Parker, Calendars, 39.
28
It is left to the reader to figure that out.
methods of dating and the egyptian calendar 49

Conversion of Dates

Astronomers reckon ancient astronomical events terms of the Julian,


not the Gregorian calendar. The correlation between the Egyptian and
the Julian calendars is implicit in Ptolemy’s dating of astronomical obser-
vations according to the Egyptian calendar and to the 365 day Egyptian
year of the so-called Ptolemaic Royal Canon which began with year
1 of the Babylonian king Nabonassar.29 During the first years of the
Era Nabonassar the correlation between the Egyptian and the corre-
sponding Julian calendar dates was:
Royal Canon, year 1 : Thoth 1 = February 26, 747 BC
The Julian calendar year 745 BC was a leap year and so it follows
that:
Royal Canon, year 4 : Thoth 1 = February 25, 744 BC
Royal Canon, year 8 : Thoth 1 = February 24, 740 BC
On this basis it is possible to convert any Egyptian date in a given or
chosen year into the Julian calendar equivalent, with due consideration
that the Julian calendar has a leap year and the Egyptian calendar
does not. Tables for conversion have been constructed by Neugebauer.30

Egyptian Calendar Day

The word for calendar day used in dates was sw, not hrw (day as
opposed to night).31 Parker presented a circumstantial argument in favor
of the beginning of the calendar day at dawn: “It is obvious that, when
the [lunar] month begins, the first day of the month also begins.”32
Most Egyptologists accepted the validity of Parker’s argument.33 However,
in the 1980s Leitz and Luft revived the arguments presented by Sethe
in the 1920s that the calendar day began with sunrise.34 They were

29
See below, Chapter III. 11.
30
P. V. Neugebauer, Astronomische Nachrichten, No. 6261 (Kiel, 1937).
31
Edel, Grammatik, 182.
32
Parker, Calendars, 10.
33
The validity of the argument was denied by Grzybek (n. 35), 142–143, who other-
wise accepted the beginning of the calendar day at dawn.
34
K. Sethe, “Die Zeitrechnung der alten Ägypter III”, NAWG (1920), 130–138.
U. Luft, Altorientalische Forschungen 14 (1987), 3–11; Leitz, Studien, 1–5.
50 the editors

countered by scholars like Grzybek, Spalinger and Wells who defended


Parker’s standpoint.35
Actually, August Böckh had solved the question in 1863 when he
used Ptolemy’s Almagest to demonstrate that the Egyptian calendar day
began at dawn before sunrise.36 In the Almagest Ptolemy twice recorded
observations of Mercury, first with double dates:
Hadrian, year 18, Epiphi 18 to 19, at dawn;37
Antoninus, year 4, Phamenoth 18 to 19, at dawn.38
Subsequently, he gave single dates for the same observations:
Mercury as morning star: Hadrian, year 18, Epiphi 1939
Mercury as morning star: Antoninus, year 4, Phamenoth 1940
Mercury is only observable at dawn as morning star or at dusk as
evening star. Ptolemy could have substituted single dates for double
dates only if he reckoned dawn as the beginning of the Egyptian cal-
endar days Epiphi 19 and Phamenoth 19. Accordingly, the Egyptian
calendar day began during the observability of Mercury as morning
star, i.e. at dawn before sunrise. Hieroglyphic sources also exist that
imply the beginning of the calendar day at dawn—notably an entry in
the Calendar of Lucky and Unlucky Days, describing the appearance
of Seth in the bow of the solar bark which led Malinine to conclude
that the calendar day began before sunrise, during ˙≈-t #.41 The time
when the god appeared at the bow of the solar bark, according to an
entry in the Calendar of Lucky and Unlucky Days.42 His argument is
supported by the identification of the god Seth with planet Mercury
here, rising shortly before the sun and thus visible at dawn.43

35
E. Grzybek, Du calendrier macédonien au calendrier Ptolémaique (Basel: SBA 20, 1990),
147–151; A. J. Spalinger, OLZ 87 (1982), 25; R. A. Wells, BiOr 49 (1992), 723.
36
A. Böckh, Über die vierjährigen Sonnenkreise der Alten, vorzüglich über den Eudoxischen,
(Berlin, 1863), 298–308.—Cf. R. Krauss, SAK 32 (2004), 275–286.
37
J. G. Toomer, Ptolemy’s Almagest, translated and annotated (Princeton, 1984), 449.
38
Toomer (n. 37), 450.
39
Toomer (n. 37), 456.
40
Toomer (n. 37), 455–456.
41
For ˙≈-t # as dawn, see K. Jansen-Winkeln, SAK 23 (1996), 201–202, and D. Kurth,
GM 108 (1989), 34–35.
42
M. Malinine, “Nouveaux fragments du calendrier égyptien des jours fastes et
néfastes”, in: Mélanges Maspero I, 2–3, Orient ancien (Cairo: MIFAO 66, 1935–1938),
887–888, 898.
43
R. Krauss (n. 36), 284–285.—For Seth as planet Mercury see Neugebauer &
Parker, EAT III, 180.
methods of dating and the egyptian calendar 51

D-1 dawn D dawn

d d+1

midnight midnight
Fig. I. 5.1

Because an Egyptian calendar day (D) begins at dawn before sunrise,


it overlaps two Julian calendar days (d) and (d + 1) that begin at mid-
night as Fig. I. 5.1 shows.
Neugebauer’s tables yield d as equivalent of D. The assertion that
an Egyptian calendar day D equals the Julian calendar day d, means
that the bright part of D corresponds to the bright part of d. It is
implied that the early hours of d are not contained in D, and that the
last hours of D correspond to the first hours of the Julian calendar day
(d + 1).
PART II

RELATIVE CHRONOLOGY
II. 1 PREDYNASTIC—EARLY DYNASTIC CHRONOLOGY1

Stan Hendrickx

Introduction

There exists quite some confusion in the terminology used for the cul-
tural phases and the relative chronology of Egypt during the 4th mil-
lennium, generally known as the Predynastic period. Already the term
“Predynastic” in itself may cause a problem. Although it evidently refers
to all periods previous to the dynastic history of Egypt, its use is in
reality reserved for the Naqada culture of (Upper) Egypt which dates
to the 4th millennium BC and economically represents a late Neolithic
culture with increasing social complexity which is at the origin of the
pharaonic civilisation. Most authors also call the preceding Badari cul-
ture “Predynastic”, although it has also been considered “Neolithic”,
together with the 5th millennium finds from the Fayum, Merimde Beni
Salama and el-Omari in LE.2
The terminology for the relative chronology of the Predynastic period
is also far from consistent (Table II. 1.1). Originally Petrie distinguished
three chronologically distinct cultures which he called respectively
Amratian, Gerzean and Semainean, after type sites that he excavated.3
By doing so he stressed the material differences, especially those between
the Amratian and Gerzean which according to Petrie were caused by
the arrival of a “new race”. This view has already been abandoned
for a long time, but the terms Amratian and Gerzean in particular
continue to be used by some scholars up to the present day. It is how-
ever already a long time since the word “Naqada culture” has been
used, divided in two or three phases (cf. below). This has the advan-
tage of indicating the continuous development that took place during
the 4th millennium BC.

1
I wish to thank Werner Kaiser, Edwin van den Brink, Nathalie Buchez and
Christiana Köhler for comments and information on various aspects of the problems
concerned with this contribution. Joanne Rowland most kindly perused the English.
2
E.g. B. Midant-Reynes, The Prehistory of Egypt. From the First Egyptians to the First
Pharaohs (London 2000).
3
Petrie, Diospolis Parva; idem, Prehistoric Egypt (London 1920).
56 stan hendrickx

Table II. 1.1


Petrie 1920: 46–50 Scharff 1931: 16–30 Kaiser Hendrickx Hassan 1988
Brunton 1937, 1948 Kantor 1944, 1965, 1957, 1990 1996, 1999
Vandier 1953 1993
Baumgartel 1970

Semainean Protodynastic Naqada Stufe Naqada Terminal


IIIa–c IIIA–D Predynastic
Gerzean Naqada II Naqada Stufe Naqada Late Predynastic
IIa–d IIC–D
Amratian Naqada I Naqada Stufe Naqada Middle Predynastic
Ia–c I–IIB
Badarian Badarian Badarian Badarian Early Predynastic

In LE the local 4th millennium culture identified at Maadi and Buto,


for example has recently been the focus of research.4 The original name
“Maadi culture/Maadian” is still used at present,5 but it has meanwhile
also been termed “Buto-Maadi culture”6 and “Maadi-Buto culture”.7
As up to now the site of Maadi has provided the largest amount of
material available, it seems obvious that the terminology should stress
the importance of that site. On the other hand, the position of “Buto”
in the Maadi-Buto terminology indicates that the terminal stage(s) of
this culture at Buto (and many other contemporary Delta sites) are no
longer present at the key site of Maadi itself. On a more general level,
it has been called “predynastic” because it is contemporaneous with
part of the Naqada culture in UE,8 but was also referred to as “chal-

4
For recent overviews, see D. Faltings, “Ergebnisse der neuen Ausgrabungen in
Buto. Chronologie und Fernbeziehungen der Buto-Maadi-Kultur neu überdacht”, in:
Stationen, 35–45; idem, “Recent Excavations in Tell El-Fara’in/Buto: New Finds and
their Chronological Implications”, in: Eyre, ed., Proceedings, 365–375; idem, “The
Chronological Frame and Social Structure of Buto in the Fourth Millennium BCE”,
in: Interrelations, 165–170.
5
E.g. Midant-Reynes (n. 2).
6
E.g. K. Schmidt, “Comments to the Lithic Industry of the Buto-Maadi Culture
in Lower Egypt”, in: L. Krzyzaniak, M. Kobusiewicz & J. A. Alexander, eds., Environmental
Change and Human Culture in the Nile Basin and Northern Africa (Poznan 1993), 267–277;
J. Wunderlich, T. von der Way & K. Schmidt, “Neue Fundstellen der Buto-Maadi-
Kultur bei Esbet el-Qerdahi”, MDAIK 45 (1989), 309–318; Faltings, in: Stationen, 35–45.
7
S. Hendrickx, “La chronologie de la préhistoire tardive et des débuts de l’histoire
de l’Egypte”, Archéo-Nil 9 (1999), 13–81; T. E. Levy & E. C. M. van den Brink,
“Interaction Models, Egypt and the Levantine Periphery”, in: Interrelations, 3–38.
8
Midant-Reynes (n. 2).
predynastic—early dynastic chronology 57

colithic”9 or “late prehistoric”.10 More exceptionally it has been termed


“Early Bronze”,11 mainly for comparison with the Levant.
There is no obvious change in the material culture marking the tran-
sition between Predynastic and Early Dynastic in UE. Here one should
distinguish between cultural chronology based on material evidence and
the historical chronology based on written documents. The latter how-
ever are very exceptional outside of elite sites such as Abydos and
Saqqara. Because of this, the end of the Naqada III culture is, from
the archaeological point of view, to be placed within or at the end of
Dyn. 2. The term Naqada culture will therefore also be used for the
Early Dynastic period.

Tasian

A culture preceding the Badarian (cf. below) was originally identified


by Brunton at Deir Tasa, and accordingly labelled by him as Tasian.12
One of the main archaeological characteristics are narrow beakers
with flaring rim and incised decoration. Subsequently, the Tasian has
been discredited by Baumgartel as a separate cultural entity and was
considered part of the Badarian.13 This view has been accepted for a
long time, but more recently Kaiser has considered the Tasian as an
entity distinct from the Badarian.14 For him, the Tasian represents the

9
F.-J. De Cree, “‘Mutatis Mutandis’ Egyptian Relations with Palestine in the
Chalcolithic and Early Bronze Age I–IV” GM 124 (1991), 21–42; K. Schmidt, “Tell
el-Fara’in/Buto and el-Tell el-Iswid (south): The Lithic Industries from the Chalcolithic
to the Early Old Kingdom”, in: E. C. M. van den Brink, ed., The Nile Delta in Transition:
4th.–3rd Millennium B.C. ( Jerusalem, 1992), 31–42; idem, “Lower and Upper Egypt in
the Chalcolithic Period. Evidence of the Lithic Industries: A View from Buto”, in: L.
Krzyzaniak, K. Kroeper & M. Kobusiewicz, eds., Interregional Contacts in the Later Prehistory
of Northeastern Africa (Poznan, 1996), 279–289; S. P. Tutundzic, “Chalcolithic Canaan
and Egypt: Reinvestigations and Considerations”, Journal of the Serbian Archaeological Society
12 (1996), 25–33.
10
T. von der Way, Untersuchungen zur Spätvor- und Frühgeschichte Unterägyptens (Heidelberg:
SAGA 8, 1993).
11
A. R. Schulman, “At the Fringe: The Historiography and Historicity of the
Relations of Egypt and Canaan in the Early Bronze Age I”, in: P. R. de Miroschedji,
ed., L’urbanisation de la Palestine à l’âge du bronze ancien (Oxford: BAR 527, 1989), 433–453.
12
G. Brunton, Mostagedda and the Tasian Culture (London, 1937).
13
E. J. Baumgartel, The Cultures of Prehistoric Egypt I (London, 1955), 20–21.
14
W. Kaiser, “Zur Südausdehnung der vorgeschichtlichen Deltakulturen und zur
frühen Entwicklung Oberägyptens”, MDAIK 41 (1985), 61–88.
58 stan hendrickx

transmitter of LE elements to UE. However, this seems most unlikely


because of the presence of beakers strongly resembling those of the
Tasian in Sudanese neolithic contexts, for example at Kadero and
Kadada, dated to late 5th–early 4th millennium cal. BC.15 Most recently,
Tasian beakers have also been published for the Eastern Desert.16 In
the Western Desert, Tasian burials have been identified in the Wadi
el-Hol, on the desert road between Luxor and Farshut17 and at Gebel
Ramlah, near Nabta Playa.18 The new evidence, together with a rein-
terpretation of older information, allows rather for the consideration of
the Tasian as a desert phenomenon, which however also had exten-
sive contacts with the Nile valley.19 The finds from Wadi Attula in the
Eastern Desert have been radiocarbon dated between 4940 and 4455
cal BC, a surprisingly early date.20 These are the only dates available
at present and although confirmation from other sites remains desir-
able, this indicates not only a cultural difference with the Badarian but
eventually also a chronological. It is however to be noted that the finds
from the Nile valley occurring in a Badarian context, indicate that the
Tasian may have started before the Badarian, but was at least also
partially contemporaneous with it.

Badarian

The Badari culture has originally been identified in the region of Badari
(Qaw el-Kebir, Hammamiya,21 Mostagedda22 and Matmar),23 where a
number of small cemeteries, containing in total about 600 tombs, and

15
E.g. M. Ch∑odnicki, “New Types of the Neolithic Pottery in Kadero (Sudan)”,
CRIPEL 17,2 (1997), 29–35; J. Reinold, Kadruka and the Neolithic in the Northern
Dongola Reach”, Sudan & Nubia 5 (2001), 2–10.
16
Friedman & Hobbs, “A ‘Tasian’ Tomb in Egypt’s Eastern Desert”, in: Gifts,
178–91.
17
J. C. Darnell & D. Darnell, “Opening the Narrow Doors of the Desert: Discoveries
of the Theban Desert Road Survey”, in: Gifts, 132–155.
18
M. Kobusiewicz, J. Kabacinski, R. Schild, J. D. Irish & F. Wendorf, “Discovery
of the first Neolithic cemetery in Egypt’s Western Desert”, Antiquity 78 (2004), 566–578;
R. Schild et al., “Gebel Ramlah Playa”, in: Tides of the Desert—Gezeiten der Wüste,
Jennerstrasse 8, eds., (Cologne, 2002), 117–23.
19
Friedman & Hobbs (n. 16), 189.
20
Friedman & Hobbs (n. 16), 178.
21
G. Brunton & G. Caton Thompson, The Badarian Civilisation and Prehistoric Remains
near Badari (London, 1928).
22
Brunton (n. 12).
23
G. Brunton, Matmar (London, 1948).
predynastic—early dynastic chronology 59

40 poorly documented settlement sites. However, characteristic Badarian


finds have also been made much further to the south (Mahgar Dendera,24
Armant,25 Elkab,26 and Hierakonpolis)27 and also in the Wadi Ham-
mamat.28 Besides Mahgar Dendera most of these finds are unfortu-
nately limited in extent. 29
Until recently, most authors tended to consider the Badari culture
as a chronologically separated unit, out of which the Naqada culture
developed. However, the situation is certainly far more complex. Since
more Badarian finds, or Badari related finds, have been made south
of Badari, the Badari culture might well have been present between at
least the Badari region and Hierakonpolis. Regional differences may
have existed, the unit in the Badari region itself being the only well
documented one up to now.
The chronological position of the Badari culture is still subject for
discussion. Its relative position as being older than the Naqada culture
has already for a long time been shown by the stratified site at Hamma-
miya.30 From TL-dating, the culture might already have existed by
5000 cal BC. However, based on the available radiocarbon dates, only
the period around 4400–4000 cal BC can be confirmed for certain.31
Additional information on the chronological position of the Badarian
may be gained by searching for its origin. Recent investigations have
shown strong links between the ceramic industry of the Badarian and
the Bashendi B unit from the Dakhleh oasis, 32 and with the Late/Final
Neolithic from the Nabta-Kiseiba area.33 The Bashendi B culture can

24
S. Hendrickx, B. Midant-Reynes & W. Van Neer, Mahgar Dendera 2 (Haute Egypte),
un site d’occupation Badarien (Leuven, 2001).
25
O. H. Myers & H. W. Fairman, “Excavations at Armant 1929–1931”, JEA 17
(1931), 228–229.
26
P. M. Vermeersch, Elkab II. L’Elkabien, Epipaléolithique de la Vallée du Nil Egyptien
(Leuven, 1978), 139–143.
27
M. A. Hoffman, “A Preliminary Report on 1984 Excavations at Hierakonpolis”,
NARCE 132 (1986), 3–14.
28
F. Debono, “Expéditon archéologique royale au désert oriental (Keft—Kosseir):
Rapport préliminaire sur la campagne 1949”, ASAE 51 (1951), 74.
29
For a more detailed overview of Badarian sites, see S. Hendrickx & E. C. M.
van den Brink, “Inventory of Predynastic and Early Dynastic Cemetery and Settlement
Sites in the Egyptian Nile Valley”, in: Interrelations, 346–399.
30
Brunton & Caton Thompson (n. 21), 73–8; D. L. Holmes & R. F. Friedman,
“Survey and Test Excavations in the Badari Region, Egypt”, PPS 60 (1994), 105–142.
31
Hendrickx (n. 7), 19.
32
C. A. Hope, “Early and Mid-Holocene Ceramics from the Dakhleh Oasis: Traditions
and Influences”, in: Gifts, 39–61.
33
See especially site E-75–8, K. Nelson, “Ceramic Assemblages of the Nabta-Kiseiba
60 stan hendrickx

be dated between 5650/5400 and 3950 cal BC,34 and the Late/Final
Neolithic at site E-75–8, roughly between 5500 and 4700 cal BC. From
these dates a very early start for the Badarian seems distinctly possi-
ble, contrarily to the opinion of the present author expressed only a
few years ago.35 Nevertheless it remains a fact that a duration of a
thousand years or even more for the Badarian does not seem to be
supported by the limited number of cemeteries and tombs known in
the Badari region, where research is supposed to have been as exhaus-
tive as possible. If originally based in its present desert regions, the
Badarians might only at a later stage have used the Nile valley for per-
manent living.

Naqada Culture—History of Research—Petrie’s Sequence Dating

In 1895 a huge cemetery of previously unknown type was discovered


by W. M. F. Petrie at Naqada.36 At first it was thought to date from the
FIP, but Jacques de Morgan soon realised the prehistoric nature of the
cemetery,37 which was later confirmed as “predynastic” by Petrie. A
number of important similar cemeteries were excavated in UE during
the beginning of the last century. The original study on the relative
chronology of the Naqada culture goes back to the early years of the
20th century, when Petrie worked out his Sequence Dating,38 the first
attempt at what is now known as seriation. The history of this remark-
able relative chronology has in recent years already been presented and
discussed many times39 and will only be presented briefly here.

Area”, in: K. Nelson & Associates, Holocene Settlement of the Egyptian Sahara. Volume 2.
The Pottery of Nabta Playa (New York—Boston—Dordrecht, 2002), 34–35.
34
M. M. A. McDonald, “The Late Prehistoric radiocarbon Chronology for Dakhleh
Oasis within the wider environmental and cultural Setting of the Egyptian Western
Desert”, in: C. A. Marlow & A. J. Mills, eds., The Oasis Papers 1: The Proceedings of the
First Conference of the Dakhleh Oasis Project (Oxford, 2001), 26–42; idem, “Dakhleh Oasis
in Predynastic and Early Dynastic Times: Bashendi B and the Sheikh Muftah Cultural
Units”, Archéo-Nil 12 (2002), 109–120.
35
Hendrickx (n. 7), 19.
36
W. M. F. Petrie & J. E. Quibell, Naqada and Ballas (London, 1896); E. J. Baumgartel,
Petrie’s Naqada Excavation: A Supplement (London, 1970).
37
J. de Morgan, Recherches sur les origines de l’Egypte. I. L’age de la pierre et des métaux
(Paris, 1896).
38
W. M. F. Petrie, “Sequences in Prehistoric Remains”, JRAI 29 (1899), 295–301;
idem (n. 3), Diospolis Parva, 4–12; idem, Prehistoric Egypt (n. 3), 3–4.
39
J. C. Payne, “The Chronology of Predynastic Egyptian Decorated Ware”, Eretz-
predynastic—early dynastic chronology 61

The Sequence Dating is based on the grave goods from the ceme-
teries excavated by Petrie and his assistants at Naqada, Ballas40 and
Diospolis Parva.41 As a first step, the pottery was arranged in a cor-
pus of “predynastic” pottery, consisting of nine classes of pottery and
over 700 types (Table II. 1.2).42 Next, all of the objects from each grave
were noted on a slip of card. Finally, the cards were arranged in a
relative chronological order based on the resemblance of types. In this
stage of his work, Petrie used only 900 relatively intact graves con-
taining five or more different pottery types, out of over 4000 excavated
graves. The chronological order was defined by two main principles.
Firstly, an earlier and a later phase were distinguished through the
observation that the classes of White Cross-lined pottery on one hand,
and Decorated and Wavy Handled pottery on the other hand, hardly
ever occurred together. Secondly, it was accepted that there had been
an evolution of the shape of the Wavy Handled types, going from glob-
ular to cylindrical shapes, while at the same time the handles had
changed from two functional handles to a continuous decorative line.43

Table II. 1.2. Pottery Classes Distinguished by W. M. F. Petrie, Corpus of


Prehistoric Pottery and Palettes (London, 1921) and Their Characteristics
Class Defined by Class relations Fabric

Black-Topped B Firing technology Nile silt


Red-Polished P Surface treatment Nile silt
Fancy F Shape mainly Red-Polished mainly Nile silt
White Cross-Lined C Decoration Red-Polished Nile silt
Incised Black N Decoration Nile silt
Decorated D Decoration mainly Late Marl clay
Wavy-Handled W Morphological detail Marl clay
Rough R Fabric/surface treatment Nile silt, organic
temper
Late L Fabric/chronology some Rough mainly Marl clay

Israel 21 (1990), 77–82; D. C. Patch, The Origin and Early Development of Urbanism in
Ancient Egypt: A Regional Study, U.M.I. (Ann Arbor / Pennsylvania, 1991), 153–170;
B. Midant-Reynes, Préhistoire de l’Egypte. Des premiers hommes aux premiers pharaons (Paris,
1992), 240–243; idem (n. 2), 257–259; B. Adams, Ancient Nekhen. Garstang in the City of
Hierakonpolis (New Malden, 1995), 21–26; T. A. H. Wilkinson, State Formation in Egypt.
Chronology and Society (BAR 651, Oxford, 1996), 9–11; S. Hendrickx, “The Relative
Chronology of the Naqada Culture: Problems and Possibilities”, in: A. J. Spencer, ed.,
Aspects of Early Egypt (London, 1996), 36–43; idem (n. 7), 20–25.
40
Petrie & Quibell (n. 36).
41
Petrie (n. 3), Diospolis Parva.
42
W. M. F. Petrie, Corpus of Prehistoric Pottery and Palettes (London, 1921).
43
Cf. Petrie (n. 3), Diospolis Parva, pl. II; B. Adams, Predynastic Egypt (Aylesbury,
1988), 27; Hendrickx (n. 7), 31, fig. 9.
62 stan hendrickx

This evolution was corroborated by the fact that the “early” and “late”
Wavy Handled types did not occur in the same tomb. In addition, the
“late” Wavy Handled types and an important number of types from
the Late class occurred in tombs which could be dated by inscriptions
to Dyn. 0 or 1.
When all of the grave cards had been arranged in order, Petrie
divided the cards into fifty equal groups, each of them consisting of 18
graves, numbering them as Sequence Dates from thirty to eighty. By
choosing to start at SD 30, he left space for earlier cultures, which he
thought were still to be discovered.44 Finally the fifty sequence dates
were divided into three groups which he considered to be archaeolog-
ically, culturally and chronologically different. The “cultures” were
named Amratian (SD 30–37), Gerzean (SD 38–60) and Semainean (SD
60–75), after important predynastic cemetery sites.
The Sequence Dates were continued with a second typological cor-
pus, for the “protodynastic” pottery.45 This is almost exclusively based
on material from the extensive cemeteries at Tarkhan.46 This time the
number of types reached 885 and no classes of pottery were distin-
guished, which makes the corpus in some cases difficult to use. The
“protodynastic” corpus partially overlaps with the most recent types of
the “predynastic” corpus, as a result of which the Sequence Dates for
the “protodynastic” corpus start already from SD 76 and continue to
SD 86, which should mark the beginning of Dyn. 3. However, the SDs
83–86 remained almost completely theoretical because of the lack of
Dyn. 2 material at Tarkhan. The distinction between the individual
Sequence Dates is not carried out in the way in which it was done for
the “predynastic” corpus. This time, however, the transition to a new
SD is based on typological breaks which Petrie defined mainly through
the development of the Wavy Handled types. Finally, Petrie connected
the Sequence Dating with the historically dated pottery types and other
objects from the royal tombs of the earliest dynasties at Abydos.47

44
When he discovered the Badarian, Brunton tried to apply Sequence Dating
(Brunton & Caton Thompson (n. 21), 26; Brunton (n. 12), 50–51), but this was from
the beginning considered problematic and never found its way into the scientific
literature.
45
W. M. F. Petrie, Corpus of Proto-Dynastic Pottery (London, 1953).
46
W. M. F. Petrie, Tarkhan I and Memphis V (London, 1913); idem, Tarkhan II
(London, 1914).
47
Petrie (n. 46), Tarkhan I and Memphis V, 3.
predynastic—early dynastic chronology 63

The development of the Sequence Dates certainly represents one of


the major intellectual performances in the study of predynastic Egypt,
and most of the basic observations by Petrie, such as the evolution of
the Wavy Handled types, were never contradicted. Nevertheless, a num-
ber of methodological shortcuts and possible errors concerning the prac-
tical elaboration of the Sequence Dates were subsequently pointed out
by several authors.48 Petrie makes no clear distinction between typol-
ogy and chronology. This, for example, is obvious from the heteroge-
neous manner by which his pottery classes have been defined (cf. Table
II. 1.2). Furthermore, the definition of the individual types within these
classes is not bound by strict rules.49 Even more important is the fact
that Petrie only included tombs with five or more objects, resulting in
the under-representation of the earlier period.50 The lack of attention
to eventual regional differences is also to be noted.
The most striking omission in Petrie’s manner of working remains
that he never took the horizontal distribution of the graves into con-
sideration. This, despite the fact that he noted for instance that none
of the cemeteries from Diospolis Parva covered the whole of the SDs,
but that “early” and “late” cemeteries could be distinguished.51 Strangely
enough, Petrie does not mention spatial distribution within the ceme-
teries of Naqada, Ballas or Diospolis Parva, although it is hardly imag-
inable that he did not notice any clustering of tombs with similar
funerary equipment. On the occasion of later excavations, by former
assistants of Petrie, the existence of groups of chronologically related
graves, and therefore the differences in the spatial distribution of objects,
were noticed several times and at different sites,52 but no attempts were
made to use these observations for chronological purposes.

48
G. F. Legge, “New Light on Sequence-dating”, PSBA 35 (1913), 101–113; A. Scharff,
Das vorgeschichtliche Gräberfeld von Abusir el-Meleq (Leipzig, 1926), 71–74; H. J. Kantor,
“The Final Phase of Predynastic Culture, Gerzean of Semainean?”, JNES 3 (1944),
110–136; Baumgartel (n. 13), 2; idem, Predynastic Egypt (CAH I, IXa, London, 1970),
3–5; W. Kaiser, “Stand und Probleme der ägyptische Vorgeschichtsforschung”, ZÄS 81
(1956), 87–109; Hendrickx (n. 39), 37–38; idem (n. 7), 21–22.
49
Petrie (n. 42), 5.
50
Hendrickx (n. 39), 37.
51
Petrie (n. 3), Diospolis Parva, 31–32.
52
D. Randall-McIver & A. C. Mace, El Amrah and Abydos. 1899–1901 (London,
1902), 3; E. R. Ayrton & W. L. S. Loat, Pre-dynastic Cemetery at El-Mahasna (London,
1911), 2; T. E. Peet, The Cemeteries of Abydos. Part II. 1911–1912 (London, 1914), 18;
Brunton & Caton Thompson (n. 21), 50–1.
64 stan hendrickx

Finally, Petrie’s wish for a very detailed relative chronology in 50


Sequence Dates causes a fundamental problem. As a basic principle,
the definition of the original Sequence Dates was made in a manner
so as to minimise the chronological dispersion of each type of pottery.
This results in a compromise between the competing claims of all pot-
tery types for closer proximity. However, the “perfect balance” obtained
by Petrie is purely artificial, since, whenever new graves will be added
to the system, the range of Sequence Dates for a number of types will
have to be expanded, and the accuracy suggested by the Sequence
Dating system becomes purely hypothetical. The integration of new
cemeteries over time made the whole system more and more problematic.

Kaiser’s Stufen Chronology

W. Kaiser was the first to reinvestigate the relative chronology of the


predynastic period in a fundamental manner.53 He used the horizontal
distribution of pottery classes and types of objects within cemetery
1400–1500 at Armant as point of departure.54 Three spatial zones were
distinguished by the relative percentages of Black-Topped, Rough and
Late Wares, each of them dominating one zone. These zones are con-
sidered to represent chronological stages. Within these three periods,
subdivisions, called Stufen, were recognised according to the clustering
of types of objects, these being almost exclusively pottery. The results
of the analysis of the Armant cemetery are completed with a limited
investigation of cemeteries for which the publication was less detailed,
but where pottery types occur which are not represented at Armant.
In this manner Kaiser distinguished three main stages of the develop-
ment of the Naqada culture, each with their subdivisions. All in all,
11 Stufen are identified, the two earliest and the two most recent of
which are not represented at Armant. The archaeological description
of the Stufen is based on types of objects, according to the Petrie typol-
ogy, which Kaiser accepts as characteristic for a particular Stufe. Material
from cemeteries other than Armant is also included. The characteristics
of the Stufen are also used by Kaiser to study the geographical distri-
bution of the Naqada culture. The chronological expansion which can

53
W. Kaiser, “Zur inneren Chronologie der Naqadakultur”, Archaeologia Geographica
6 (1957), 69–77.
54
R. L. Mond & O. H. Myers, Cemeteries of Armant I (London, 1937).
predynastic—early dynastic chronology 65

Table II. 1.3. Correlation between Petrie’s Sequence Dating and Kaiser’s Stufen
Kaiser SD Petrie SD

Zeitstufe III SD 63–80 Semainean SD 63–76


Zeitstufe IIcd SD 40/45–63 Gerzean SD 38–62
Zeitstufe IIab SD 38–40/45
Zeitstufe I SD 30–38 Amratian SD 31–37

be observed from UE both towards the north and the south, is con-
sidered further proof of the validity of the Stufen chronology.
The Stufen have also been compared with the Sequence Dating.55
Although Kaiser defined three main periods, the SD’s attributed to
them show that they are largely, but not completely, identical to the
Amratian, Gerzean and Semainean distinguished by Petrie (Table II.
1.3). When compared to the Sequence Dating, Kaiser’s system has the
advantage of including not only information from the typological appa-
ratus, but also from the spatial distribution of the objects. Furthermore
it does not give the idea of extreme accuracy, but by defining periods,
it escapes largely, although not completely, the problem of becoming
increasingly meaningless when new data are added.
However, this does not mean that the Stufen system is without its
problems. Although Kaiser included data from a number of cemete-
ries besides the one at Armant, it remains a basic fact that data from
only a single cemetery are used for the description of the Naqada cul-
ture throughout UE. Nevertheless, Kaiser is well aware of the possi-
bilities for regional differentiation, and has noticed regional phenomenon
at Mahasna for instance.56 The problem caused by using the cemetery
at Armant becomes even more complicated because the earliest phase
of the Naqada culture is not present at Armant, and even the most
recent phases are very sparsely documented or absent. Therefore, the
definition of the Stufen Ia and Ib is merely based on hypothesis, although
examples from other cemeteries besides Armant are given. The descrip-
tion of Stufe IIIb, though less hypothetical than Stufen Ia and Ib, is also
based on information from other cemeteries. In most cases it was not
possible to study the spatial development of these cemeteries, and there-
fore Kaiser’s description of the Stufen Ia–b and IIIb depends largely on

55
Kaiser (n. 48), 109.
56
Kaiser (n. 53), 74.
66 stan hendrickx

the theoretical evolution of pottery types, especially from the Wavy-


Handled class as already accepted by Petrie.
A very practical problem is that Kaiser’s study was only published
in abridged version as an article. Because of spatial considerations within
the publication, Kaiser was unable to provide details on his analytical
method. The plates illustrating the article visualise the description of
the Stufen by presenting for each of them the most important and
characteristic types of objects.57 Unfortunately these plates have been
used by some as absolute guidelines, despite the fact that the plates are
only intended to be considered as an idealised outline of the develop-
ment of the Stufen. This created a false idea of secure dating, especially
for the Stufen IIIa–b. More recently Kaiser mentioned in an article the
extension of his Stufen chronology into Dyn.1.58 However, the manner
in which this was done remains until now unpublished. The extension
was nevertheless used by several authors.
Kaiser’s distinction of three phases within the Naqada culture reflects
Petrie’s original division. At first view there seems to be no problem
because several cemeteries belonging to the Naqada culture bear evi-
dence for the presence of three groups of graves, dominated respec-
tively by the presence of Black-Topped, Rough and Late/Wavy Handled
pottery. The moment at which the transitions are placed is however
not beyond dispute. This and other more specific problems concern-
ing the Stufen chronology will be discussed when the individual subdi-
visions of the Naqada culture are presented.

Computer Seriation

Computer seriation has also been applied on predynastic cemeteries to


study their relative chronology. A pioneer attempt made by E. M.
Wilkinson is at present only of historical interest.59 Far more important
is Kemp’s seriation by multi-dimensional scaling of the graves within
cemetery B at el-Amrah and the cemetery of el-Mahasna.60 However,

57
Kaiser (n. 53), Pl. 21–24.
58
W. Kaiser, “Zur Entstehung des gesamtägyptischen Staates”, MDAIK 46 (1990),
Abb. 1.
59
E. M. Wilkinson, “Techniques of Data Analysis. Seriation Theory”, Technische und
Naturwissenschaftliche Beiträge zur Feldarchäologie. Archaeo-Physika 5 (1974), 1–134.
60
B. J. Kemp, “Automatic Analysis of Predynastic Cemeteries: A New Method for
an Old Problem”, JEA 68 (1982), 5–15.
predynastic—early dynastic chronology 67

this seriation is not used for the evaluation of Kaiser’s Stufen chronol-
ogy, but for Petrie’s Sequence Dating.
A far more elaborate study of the relative chronology using seriation
has been made by T. A. H. Wilkinson.61 Eight Predynastic—Early
Dynastic cemeteries were seriated.62 For this purpose, 1420 out of 1542
types from Petrie’s corpus which occurred in the eight cemeteries were
condensed into 141 groups.63 This approach, of course, carries the risk
of producing strongly heterogeneous types, and as Wilkinson himself
notes, some of his groups “bring together types with broad similarities
but some significant differences”.64 There are indeed problems with the
majority of the newly defined groups, especially for the plates, cups
and bowls (P 001, P 004, P 033, P 034 etc.). Among the most notable
difficulties is the grouping of Nile silt and marl clay bowls and even
the very characteristic Nile silt bread moulds in the groups P 094,
P 095 and P 103. Several groups of jars are also very heterogeneous
(e.g. P 013, P 019, P 029, P 128, P131, P138). One can also question
the validity of distinguishing two groups of wine jars (P 107 and P 108)
by the broadness of their shoulders, while very distinctive elements such
as the applied ridges and the wavy decorations are completely ignored.
Furthermore, it is difficult to understand why in group P 052 cylin-
drical jars both with and without incised decoration have been grouped.
This is all the more disturbing because this element has been used as
a chronological indicator by Kaiser for the development of his Stufen
chronology. It is obvious that the manner in which Petrie’s types have
been grouped, makes it impossible to arrive at the same results as those
which Kaiser obtained for his Stufen chronology.
The inevitable conclusion is that when Petrie’s types are grouped
into only 141 new groups, these become heterogeneous to a degree no
longer consistent with the concept “type”. While this does not render
Wilkinson’s seriations totally meaningless, the many anomalies involved
introduce a disturbing element of uncertainty in the results.65 Whenever

61
T. A. H. Wilkinson, “A New Comparative Chronology for the Predynastic—Early
Dynastic Transition”, JACF 7 (1994–1995), 5–26; idem (n. 39).
62
Tarkhan Hill and Valley Cemeteries, Turah, Matmar Cemeteries 200/3000–
3100/5100, Mostagedda Cemetery 1600–1800, Mahasna, el-Amrah Cemetery b, Armant
Cemetery 1400–1500, Hierakonpolis Fort Cemetery.
63
Wilkinson (n. 39), 25.
64
Wilkinson (n. 39), 23.
65
For a more extensive discussion of Wilkinson, State Formation in Egypt (n. 39), see
the review by S. Hendrickx in JEA 85 (1999), 241–245.
68 stan hendrickx

possible, the result of the seriations has also been plotted on a plan of
the cemetery, to facilitate discussion of the chronological development
of the cemeteries involved. The horizontal distribution of graves and
pottery types is, however, only used as a method of control and not
as a primary source of information. The use of a general typology for
all eight cemeteries allowed Wilkinson direct comparison of the seri-
ation results. Each sequence distinguished was compared against Kaiser’s
Stufen chronology. Significant differences emerged, most notably in
Kaiser’s demarcation of the three major Naqada culture phases.

Distribution Studies

A number of studies on the relative chronology of the Naqada period


have started from the spatial distribution of objects within the ceme-
teries rather than from seriation. Once more, the problem of group-
ing types, which was the main problem for seriation, will occur, albeit
to a lesser extent. The first study of this kind concerns an unpublished
M.A. thesis by R. Friedman on the spatial distribution and relative
chronology at Naqa ed-Deir cemetery 7000.66 Comparison is made with
Kaiser’s Stufen chronology. Spatially distinguished groups of graves with
objects characteristic for the Stufen Ic–IId are also represented at Naqa
ed-Deir, but some differences in the characteristic pottery types for the
individual Stufen have also been observed.
J. C. Payne applied Kaiser’s chronology to the available information
for the Main Cemetery at Naqada.67 She concludes that the same Stufen
can be distinguished both at Armant and at Naqada and also that the
differences in the archaeological description of the Stufen remain very
limited, the most important being situated in Stufe IIb.68

66
R. F. Friedman, Spatial Distribution in a Predynastic Cemetery: Naqa ed Dêr 7000 (unpubl.
MA. diss., Berkeley, 1981). The chronological framework established by Friedman was
afterwards used by S. H. Savage, Descent, Power and Competition in Predynastic Egypt: Mortuary
Evidence from Cemetery N 7000 at Naga-ed-Dêr (U.M.I., Ann Arbor/Arizona State University,
1995); idem, “Descent Group Competition and Economic Strategies in Predynastic
Egypt”, JAnAr 16 (1997), 226–68. For the latter study see however P. Delrue, “The
Predynastic Cemetery N7000 at Naga ed-Dêr. A Re-evaluation”, in: H. Willems, ed.,
Social Aspects of Funerary Culture in the Egyptian Old and Middle Kingdoms (Leuven, 2001),
21–66.
67
J. C. Payne (n. 39), 77–82; idem, “Predynastic Chronology at Naqada”, in: R. F.
Friedman & B. Adams, eds., The Followers of Horus. Studies Dedicated to Michael Allen
Hoffman (Oxford, 1992), 185–192.
68
Payne (n. 39), 81.
predynastic—early dynastic chronology 69

The most extensive study based on spatial distribution is a still incom-


pletely published Ph.D. dissertation by the present author.69 The limited
number of “Predynastic” cemeteries for which both a map and a grave
register, be it eventually incomplete, are available served as a starting
point.70 For the Early Dynastic period information came from a number
of “Protodynastic” and “Archaic” cemeteries in LE.71 Methodologically,
there is not much difference to the method already developed by Kaiser.
This implies that the distinction of related groups of graves is not only
based on their contents but at the same time on their spatial distribu-
tion within the cemetery. As a result, a conflict of interests will arise
between the search for closer chronological proximity of all examples
of one pottery type on one hand, and the definition of spatially well
defined groups of graves on the other hand. Neither one of these two

69
S. Hendrickx, De grafvelden der Naqada-cultuur in Zuid-Egypte, met bijzondere aandacht
voor het Naqada III grafveld te Elkab. Interne chronologie en sociale differentiatie (unpubl. PhD.
diss., Leuven, 1989); idem (n. 39); idem (n. 7).
70
The following cemeteries, from north to south, are involved: Matmar, cemetery
2600–2700 [Brunton (n. 23), pl. VIII–IX, XIX], el Badari, cemeteries 3700 and 3800
[Brunton & Caton-Thompson (n. 21), pl. III, XXXII–XXXIII], Hammamiya, ceme-
tery 1500–1800 [G. Brunton, Qau and Badari I (London, 1927), pl. VI, X–XI; Brunton
& Caton-Thompson (n. 21), pl. XXX–XXXI]; Qaw el-Kebir, cemetery 100 [idem, pl.
III, XXX], el-Salmany [A. el Sayed, “A Prehistoric Cemetery in the Abydos Area”,
MDAIK 35 (1979), 249–301], Naqada Main Cemetery [Petrie & Quibell (n. 36), pl.
LXXXII–LXXXIII; Baumgartel Petrie’s Naqada Excavation: A Supplement; S. Hendrickx,
“Predynastische objecten uit Naqada en Diospolis Parva (Boven-Egypte)”, Bulletin van
de Koninklijke Musea voor Kunst en Geschiedenis 57 (1986), 31–44; J. C. Payne, “Appendix to
Naqada Excavations Supplement”, JEA 73 (1987), 181–190], Armant cemeteries 1300
and 1400–1500 [Mond & Myers, n. 54], Hierakonpolis, locality 27, ‘Fort Cemetery’
[B. Adams, The Fort Cemetery at Hierakonpolis (London—New York, 1987)], Elkab
[S. Hendrickx, Elkab V. The Naqada III Cemetery (Brussels, 1994)], Kubbaniya, South
cemetery [H. Junker, Bericht über die Grabungen von der Akadamie der Wissenschaften in Wien,
auf den Friedhöfen von El Kubanieh—Sud. 1910–1911 (Wien, 1919)].
71
The following cemeteries, from north to south, are involved: Abu Roash, ceme-
teries 0, 300, 400–500, 800–900 & M [A. Klasens, “The Excavations of the Leiden
Museum of Antiquities at Abu-Roash. Report of the First Season: 1957. Part I”, OMRO
38 (1957), 58–68; idem, “The Excavations of the Leiden Museum of Antiquities at
Abu-Roash: Report of the First Season 1957. Part II”, OMRO 39 (1958), 20–31; idem,
“The Excavations of the Leiden Museum of Antiquities at Abu-Roash: Report of the
Second Season 1958. Part I”, OMRO 39 (1958), 32–55; idem, “The Excavations of
the Leiden Museum of Antiquities at Abu-Roash: Report of the Second Season 1958.
Part II. Cemetery 400”, OMRO 40 (1959), 41–61, idem, “The Excavations of the
Leiden Museum of Antiquites at Abu-Roash: Report of the Third Season 1959. Part
I”, OMRO 41 (1960), 69–94; idem, “The Excavations of the Leiden Museum of
Antiquities at Abu-Roach: Report of the Third Season 1959. Part II. Cemetery M”,
OMRO 42 (1961), 108–28], Saqqara, cemetery west of Serapeum [R. Macramallah,
Un cimetière archaïque de la classe moyenne du peuple à Saqqarah (Le Caire, 1940)], Turah
[ Junker, Turah], Tarkhan, Valley cemetery, cemeteries A, F, G, H, J & Q [Petrie
(n. 46), Tarkhan I; idem (n. 46), Tarkhan II].
70 stan hendrickx

elements can be accepted as prevailing over the other. Thus, most


unfortunately, it seems impossible to establish clearly defined, “objec-
tive” rules for the definition of archaeological complexes representing
relative chronological periods within the Naqada culture. This implies
that although a relative chronology defined in this manner is of course
based on the seriation principle, it nevertheless depends considerably
on the personal interpretation of the researcher.
Most cemeteries of the Naqada culture were used over several cen-
turies, during which they expanded horizontally, but not in a pre-
meditated manner. For the Naqada I–II period in particular this results
in cemeteries consisting of “patches” of simultaneous tombs, as can be
shown for the cemetery at Adaïma for example.72 During the Naqada
III period, the cemeteries tend to develop in a more linear manner,
as can be seen at Elkab73 and Tura.74 By comparing the cemeteries
that were analysed, it becomes clear that similar groups exist for different
cemeteries. In this manner, 11 groups of graves, an equal number to
Kaiser’s Stufen, are distinguished and their relative chronological order
defined through their mutual position in the cemeteries, and through
the evolution of the pottery classes and types of objects. However, com-
paring groups of related objects from geographically different cemeter-
ies does not have to imply that they are also contemporaneous in
absolute chronological terms. Unfortunately, this question cannot be
answered because of the limited number of C14 dates available for the
Naqada cemeteries from UE. For this reason, and since related groups
of archaeological objects can be distinguished at several sites, until
proven otherwise, we may as well accept the simultaneity of archaeo-
logical groups with a strong resemblance, suggesting that the same
chronological periods may well have existed for the different cemeteries.
In a further stage, the data from cemeteries without published maps
were integrated and the possibilities of regional variability investigated.
This eventually allows for an archaeological description of each of the
relative chronological periods. The general observations made by Kaiser
for cemetery 1400–1500 at Armant are not fundamentally contradicted,

72
E. Crubézy, T. Janin & B. Midant-Reynes, Adaïma II. La nécropole prédynastique (Le
Caire, 2002), 415–417.
73
Hendrickx (n. 70, Elkab V), 205–216.
74
Hendrickx (n. 39), 57–59.
predynastic—early dynastic chronology 71

and therefore the number of relative chronological periods is equal to


the number of Stufen distinguished by Kaiser, although in some cases
important differences occur in their archaeological description (cf. below).
It was therefore decided to follow Kaiser’s work as closely as possible,
but to replace the word “Stufe” by “Naqada” and at the same time
change the letter indication into capital letters, which results in “Naqada
IA” etc.75

Description of Naqada Periods—Naqada I

The descriptions presented here are mainly based on the distribution


studies made for a number of cemeteries for which both a map and
tomb inventory are available.76 Only the main developments and most
characteristic types of objects are mentioned (cf. Table II. 1. 4a–b).
Although the archaeological characteristics of the Badarian are strongly
related to those of the early Naqada I period, it is nevertheless to be
noted that at present no cemeteries are known which show a contin-
uous use from the Badarian into the Naqada period. The burials are
however of the same type and organised in the same manner. The
differences are mainly to be found in the material culture, for which
the pottery presents the most obvious evidence. Not only will the char-
acteristic rippled surface of the Badarian only occur most exception-
ally during Naqada IA, but more importantly, the three principal pottery
fabrics of the Badarian differ from those of the Naqada culture.77 Naqada
I sites occur only in UE, from the Badarian region in the north to a
few sites south of Aswan, in Lower Nubia. All of the Naqada I period
is dominated by Black-Topped pottery, which makes up over 50% of
the assemblage. Red-Polished and White Cross-Lined are the second
most important categories.

75
Cf. Levy & van den Brink, in: Interrelations, 9.
76
Hendrickx (n. 69); see notes 70–71. For a more detailed overview of the relative
chronological periods represented at individual Naqada sites, see Hendrickx & van den
Brink, in: Interrelations, 346–399.
77
R. F. Friedman, Predynastic Settlement Ceramics of Upper Egypt: A Comparative Study of
the Ceramics of Hemamieh, Nagada and Hierakonpolis (U.M.I., Ann Arbor / Berkeley, 1994).
72

Table II. 1.4a. Chronological evolution of Predynastic—Early Dynastic types of objects (Hendrickx, n. 69). Included
are: Predynastic typology: Abydos cem. E., U, X, F, Frankfort; el-Ahaiwa; el-Amrah cem. A, B; el-Badari cem.
3600, 3700, 3800, 3900, 4000, 4600; Elkab; Hammamiya cem. 1500–1800, 1900–2000; el-Mahasna; Matmar cem.
200, 2600–2700, 3000–3100, 5100; Mesaid; Mustagedda cem. 200, 300, 400, 1200, 1600–1800, 4000, 5200, 10000,
11700; Qaw cem. 0, 100, 200; Salmany. Protodynastic typology: Badari cem. 3100, 3200, 5500, 6000; Hammamiya
cem. 1600–1800, 2000; Matmar cem. 200, 600, 900–1000, 2000; Mustagedda cem. 1600–1800; Qaw cem. 400–500,
600, 700, 800; Tarkhan Hill and Valley Cemeteries; Turah. The cemeteries from Naqada have been excluded
because of the incomplete data set that would distort the general overview.
IA IB IC IIA IIB IIC IID1 IID2 IIIA1 IIIA2 IIIB IIIC1 IIIC2 IIID ? total

B 16 68 224 217 256 159 38 30 – – 1 – – – 38 1047


C 4 7 30 6 – – – – – – – – – – 2 49
N – – – 1 1 – 1 – – – – – – – – 3
F 1 2 22 20 24 30 5 1 3 1 – – – – 10 119
P 2 18 73 53 69 231 119 107 15 3 2 – – – 35 727
D – – 1 1 15 104 58 21 16 9 5 – – – 11 241
R – 1 16 56 252 715 385 432 218 352 232 50 77 30 121 2937
stan hendrickx

L – – 1 1 12 55 50 99 141 873 798 221 905 177 66 3399


W – – – – – 49 29 112 66 1096 872 146 71 – 6 2447
stone 1 2 10 12 3 20 16 12 3 124 190 37 442 45 27 944
palet – 1 37 21 27 41 25 19 13 158 105 22 4 – 52 525
flint – 1 9 2 10 18 1 8 1 5 2 – 5 1 20 83

total 24 100 423 390 669 1422 727 841 476 2621 2207 476 1504 253 388 12521

# tombs 11 45 121 115 170 264 118 145 81 608 448 109 267 73 225 2761

average 2,18 2,22 3,49 3,39 3,93 5,38 6,16 5,80 5,88 4,31 4,93 4,36 5,63 3,46 1,72 4,53
Table II. 1.4b
IA IB IC IIA IIB IIC IID1 IID2 IIIA1 IIIA2 IIIB IIIC1 IIIC2 IIID ? total

B 66,7 68,0 53,0 55,6 38,3 11,2 5,2 3,6 – – 0,0 – – – 9,8 8,4
C 16,7 7,0 7,1 1,5 – – – – – – – – – – 0,5 0,4
N – – – 0,3 0,1 – 0,1 – – – – – – – – 0,0
F 4,2 2,0 5,2 5,1 3,6 2,1 0,7 0,1 0,6 0,0 – – – – 2,6 1,0
P 8,3 18,0 17,3 13,6 10,3 16,2 16,4 12,7 3,2 0,1 0,1 – – – 9,0 5,8
D – – 0,2 0,3 2,2 7,3 8,0 2,5 3,4 0,3 0,2 – – – 2,8 1,9
R – 1,0 3,8 14,4 37,7 50,3 53,0 51,4 45,8 13,4 10,5 10,5 5,1 11,9 31,2 23,5
L – – 0,2 0,3 1,8 3,9 6,9 11,8 29,6 33,3 36,2 46,4 60,2 70,0 17,0 27,1
W – – – – – 3,4 4,0 13,3 13,9 41,8 39,5 30,7 4,7 – 1,5 19,5
stone 4,2 2,0 2,4 3,1 0,4 1,4 2,2 1,4 0,6 4,7 8,6 7,8 29,4 17,8 7,0 7,5
palet – 1,0 8,7 5,4 4,0 2,9 3,4 2,3 2,7 6,0 4,8 4,6 0,3 – 13,4 4,2
flint – 1,0 2,1 0,5 1,5 1,3 0,1 1,0 0,2 0,2 0,1 – 0,3 0,4 5,2 0,7

total 100,0 100,0 100,0 100,0 100,0 100,0 100,0 100,0 100,0 100,0 100,0 100,0 100,0 100,0 100,0 100,0
predynastic—early dynastic chronology
73
74 stan hendrickx

Naqada IA

The distinction made by Kaiser between Stufe Ia and Ib cannot be


confirmed beyond doubt for the cemeteries for which a map is available.78
This was also noted by Friedman for cemetery 7000 at Naqa ed-Deir,
where tombs predating Stufe Ic are present, but could not be distin-
guished in groups matching Kaiser’s Stufe Ia and Ib.79 Only in the east-
ern part of the Naqada Main Cemetery was it possible to distinguish
spatially a number of Naqada IA–B tombs, but a clear distinction
between IA and IB could not be made, although this may be due to
the incomplete data available. A number of individual tombs from the
cemeteries E and U at Abydos, as well as tombs from el-Amrah, can
be attributed to Naqada IA because of the presence of simple Black-
Topped cups and beakers belonging to Petrie’s types B 18 d, B 21 b
and B 22 b/d/f. Also frequently occurring in tombs attributed to Naqada
IA is White Cross-Lined pottery, the types of which are mainly sim-
ple convex round based bowls and plates (C 10 e/l/n), although restricted
cups (C 64 b/n) and slender restricted jars (C 76 h) are also repre-
sented. Restricted shapes and lip rims are however very exceptional.

Naqada IB

Besides the tombs already mentioned from Naqada Main Cemetery, a


small group of Naqada IB tombs occurs south of the centre of ceme-
tery 2600–2700 at Matmar. In the ceramic assemblage, there is a noted
increase of Red-Polished pottery. The type variation increases both for
the Black-Topped and White Cross-Lined class. Besides the simple
Black-Topped cups and beakers with straight wall already mentioned
for Naqada IA, types with inflected wall, some of them large in size
(B 18 b/c, B 21 c/d2, B 22 j, B 25 b/c, B 26 b), become characteristic.
Additional types for the White Cross-Lined are oval plates (C 4 h,
C 5 d/m, F 11 a) and exceptionally flat based bowls with concave wall
(C 35). The frequency of slender restricted jars increases. The Red-
Polished types consist mainly of small bowls and plates (P 1 a,
P 11 a, P 17), as well as slender jars (C 75 b, C 76 d, C 76 w), the

78
Hendrickx (n. 39), 41.
79
Friedman (n. 66).
predynastic—early dynastic chronology 75

shapes of which also occur among the Black-Topped and White Cross-
Lined pottery.

Naqada IC

Groups of Naqada IC tombs can be spatially identified at several ceme-


teries (Matmar, Salmany, Naqada and Armant). Both cups and beakers
with straight and flaring walls occur, but the latter group shows increas-
ing variability. Among the tall beakers, the importance of high, slen-
der types increases (B 27 a, B 27 f, B 35 a) and types with explicit
concave upper part (B 26 a–c). Hardly ever attested before are restricted
regularly curved jars with large aperture (B 55 b, B 57 a/b, B 58 b/c,
B 62 b/d, B 77 a), or similar shouldered jars (B 71 b, B 74 a/b, B
79 a). Also new are flat based bottles (B 92 a/b), which also occur
among the Red-Polished pottery (P 56 a/b). As before, the majority
of the Red-Polished pottery however reflects shapes known for the
Black-Topped.
The majority of the White-Cross lined bowls and plates have, by
this stage, concave walls and a flat base (C 24, C 26 l, C 27 n, C 30
h) or inflected walls with concave upper part (C 53–54). Although the
decoration still primarily consists of geometric motifs (probably imita-
tion of basketry), figurative representations become more frequent. The
limited amount of Rough pottery which occurs from now on consists
mainly of small bowls and plates. Footed stone vessels, almost exclusively
from basalt,80 and rhomboidal or fishtail flint knives remain exceptional
and are of the same types as during Naqada IA-B. There is, on the
other hand a marked increase in palettes, mainly rhomboidal in shape.

Naqada II

The definition of the transition between Naqada I and II is most prob-


lematic. According to Kaiser’s general principles, Stufe I should be dom-
inated by Black-Topped pottery, which is indeed a fact, and Stufe II

80
Cf. L. M. Mallory-Greenough, J. D. Greenough & J. V. Owen, “The Stone Source
of Predynastic Basalt Vessels: Mineralogical Evidence for Quarries in Northern Egypt”,
JAS 26 (1999), 1261–1272.
76 stan hendrickx

by Rough pottery. This however, is not so for Stufe IIa at the Armant
cemetery, where Black-Topped pottery remains dominant and even for
Stufe IIb the prevalence of Rough over Black-Topped pottery remains
limited.81 The differences between Stufe IIa and IIb, when the domi-
nant class of pottery changes from Black-Topped to Rough, and espe-
cially between Stufe IIb and IIc, with the introduction of Wavy Handled
pottery and a number of new Decorated types, are much more impor-
tant than the difference between Stufe Ic and IIa. It is to be noted,
however, that the Rough pottery does not appear out of the blue at
a certain moment in the evolution of the Naqada culture. It is more
than obvious from settlement excavations that the Rough pottery makes
up the vast majority of pottery since the beginning of the Naqada cul-
ture and even already during the Badarian,82 but the Rough ware finds
its way only slowly to the cemeteries.
There are certainly enough reasons to make a distinction between a
first and a second period within the development of the Naqada cul-
ture, but it seems more logical to draw the line between Stufe IIa and
IIb or perhaps even between Stufe IIb and IIc. However, for the revi-
sion of Kaiser’s Stufen chronology presented here, the division between
Stufe I and II has been maintained between Naqada IC and IIA, the
archaeological characteristics of which are largely identical with Kaiser’s
Stufen Ic and IIa. This is in order not to introduce yet another com-
pletely different system which would render the use of older literature
difficult.

Naqada IIA

As for the Naqada IC period, spatially distinguished groups of tombs


can be identified at the cemeteries of Matmar, Salmany, Naqada and
Armant. The expansion into Nubia has continued and A-group ceme-
teries contemporaneous with Naqada IIA–B can be found south of
those already identified for Naqada I. Black-Topped pottery remains
dominant but White Cross-Lined ware disappears largely while Rough
pottery for the first time makes up a relevant part of the ceramic assem-
blage. The distinction between Naqada IC and IIA does not only

81
Hendrickx (n. 39), 39–40.
82
E.g. Brunton (n. 12); Hendrickx et al. (n. 24).
predynastic—early dynastic chronology 77

depend on the representation of the wares. Of great importance is the


appearance, during Naqada IIA, of a number of pottery types, espe-
cially small bag-shaped Rough types with pointed base (R 65 b/c, R
66 a/p, R 69 c/d), which were not yet present during Naqada IC.
With regards to the Black-Topped pottery, the very early cups and
beakers with straight wall disappear almost completely. For the remain-
ing Black-Topped types there is little difference from Naqada IC,
although the occurrence of modelled rims is new (e.g. B 35 b, B 37
b, B 38 c, B 50) as well as large flat based, regularly curved jars with
strongly marked rim (B 53 a/b). Once again, the shapes of the Red-
Polished pottery are related to those of the Black-Topped class.
Rhomboidal palettes still occur regularly but the number of fish shaped
palettes increases.

Naqada IIB

Spatially distinguished groups of tombs can once again be recognised


at the cemeteries of Matmar, Salmany, Naqada and Armant, already in
use from at least the Naqada IC period. The quantitative importance
of Black-Topped pottery starts to diminish, although it is still the best
represented class. Beakers now occur only with inflected walls and
mainly exclusively with slender shapes (groups B 25, B 27, B 35) and
often with modelled rim. The majority of the Black-Topped types con-
sists of regularly curved jars, a limited number of which has pointed
bases (groups B 41, B 44) but the majority is flat based (groups B
56–58, B 62–68). The diversity of Red-Polished types increases and
is not as closely linked to the Black-Topped types as was previously
the case.
The presence of Rough pottery has more than doubled. This is
reflected in a more important diversity of cups and bowls (R 1–36),
the previously occurring small bag shaped jars (R 62–69) with pointed
base and also large storage jars (R 80–86) which will become very
characteristic for the Naqada IIC–IID2 period. White Cross-Lined pot-
tery has disappeared completely and is replaced (?) by a limited num-
ber of Decorated types, the decoration of which is limited to spirals,
undulating lines and “scales”. That the limited number of Late types
is indeed made from marl clay cannot be confirmed beyond doubt
given the confusion within this class of pottery, where no clear dis-
tinction has been made between Nile silt and marl clay pottery.
78 stan hendrickx

There are some indications for regional diversity during Naqada IIB.
Friedman observed a strong presence of Black-Topped ware for Stufe
IIb at Naqa ed-Deir,83 and a similar phenomenon was also noted by
Kaiser for el-Mahasna.84 For Naqada, Payne noted differences with Stufe
IIb as described by Kaiser.85 All of this seems to indicate that the tran-
sition between Naqada IIA/B and Naqada IIC did not happen in the
same manner and/or at the same moment for the whole of UE.

Naqada IIC

Cemeteries and settlements of this period are to be found over a larger


area than those of the previous period. Besides at the cemeteries already
mentioned for the previous periods, Naqada IIC was also spatially
identified at Badari, Hammamiya, Naqa ed-Deir and the Hierakonpolis
Fort Cemetery. Furthermore, Naqada IIC occurs at cemeteries in the
neighbourhood of the Fayum (Gerza,86 Haraga,87 and Abusir el-Meleq),88
and in the eastern Nile delta the important cemetery at Minshat Abu
Omar may have started towards the end of Naqada IIC but probably
rather during Naqada IID.89 In Nubia, A-group cemeteries with strong
Egyptian influences can be found.
With the Naqada IIC period, major changes occur in the pottery
assemblage. The importance of the Black-Topped class drops dramat-
ically and at the same time there is a marked increase of Rough pot-
tery. The appearance of Wavy-Handled pottery is also very important.
Although a limited number of the already known types still occurs,
Black-Topped pottery is now dominated by shouldered jars with a small
base and modelled rim (B 38 a/c, B 53 a–c) and similar shapes with
pointed base (B 39 a–b). The Red-Polished pottery already in use dur-
ing the previous period continues to be used. The most important
difference is a strong increase of regularly curved jars with a small base
(P 40 group). The diversity within the Rough class increases, mainly

83
Friedman (n. 66), 74–5.
84
Kaiser (n. 53), 74.
85
Payne (n. 39), 81.
86
W. M. F. Petrie, The Labyrinth, Gerzeh and Mazguneh (London, 1912).
87
R. Engelbach & B. Gunn, Harageh (London, 1923).
88
Scharff (n. 48).
89
W. Kaiser, “Zum Friedhof der Naqada-Kultur von Minshat Abu Omar”, ASAE
71 (1987), 119–126.
predynastic—early dynastic chronology 79

because of the differentiation within previously occurring groups. The


large jars (R 81–86), which are a very important part of the Rough
class, are totally dominated by the type R 81.
The new Wavy-Handled class is dominated by W 19, but the ear-
liest examples of this class (W 1–3) also belong to this period although
their provenance and position is sometimes unclear.90 The Decorated
pottery is dominated by relatively small flat based regularly curved jars
(e.g. D 43) and broad round based vessels (e.g. D 61), both with lug
handles. Besides the decorative patterns already present, there a now
very characteristic figurative representations (D 40–48). Marl clay pot-
tery is certainly represented from Naqada IIC onwards. Both the Wavy-
Handled and the Decorated class are made from this fabric, as is
the large majority of the Late pottery. The latter is primarily used for
cups and bowls, as well as a number of medium sized shouldered jars
(L 53 a–b). The types of stone vessels that had occurred up until now
have disappeared and are replaced by shapes inspired by the Decorated
pottery. Rhomboidal palettes become rare, and in their place we find
fish shapes and palettes with antithetic bird’s head.

Naqada IID1

Cemeteries and settlements are now probably also found in the delta
(Minshat Abu Omar, Kafr Hassan Daoud) and the gradual replace-
ment of the Maadi-Buto culture can be observed at Buto (cf. below).
Black-Topped pottery becomes rare and only a number of regularly
curved and shouldered jars occur (B 38 c, B 39 a–b, B 53 a–b), all of
them with modelled rim. For the Red-Polished pottery there is hardly
any difference from Naqada IIC, both in frequency and typology. A
similar observation can be made for the Late class, despite its slowly
increasing importance. Rough pottery continues to represent over half
of the assemblage. The most important changes are to be found in the
group of large storage jars where type R 81 is gradually replaced by
the types R 84 and especially R 85 h and R 86 d, meaning that vessels
with a small or pointed base are preferred to round based examples.
The Wavy-Handled types become smaller and more slender (W 25,
W 44) compared to the types characteristic for Naqada IIC. The two

90
Kaiser (n. 48), 87–109.
80 stan hendrickx

Decorated types characteristic for Naqada IIC (D 43, D 61) continue


to be used but figurative representations occur less frequently. The stone
vessel types remain identical to those of Naqada IIC, while rhombic
palettes have almost completely disappeared.

Naqada IID2

The definition of Kaiser’s Stufen IId2 and IIIa1 causes a problem because
they share the same Wavy-Handled types and differ mainly through
the presence or absence of Black-Topped types, the number of which
is at any rate minimal, and through their Decorated types, which also
only account for a very small part of the assemblage. Another char-
acteristic should be the transition from R 84–86 to L 30 b,c but the
latter types are in reality very similar to R 84 and R 84 c and the
difference consists mainly of a less well cared for product. It is there-
fore to be feared that the attribution of a vessel to one of these types
by excavators other than Petrie may have been rather arbitrary. Also
the spatial distribution at Armant Cemetery 1400–1500 easily allows
for a different clustering of graves, by which the group defined by
Kaiser as Stufe IId2, no longer exists.91 Furthermore, the Wavy Handled
types always seem to display the fastest evolution of shape, and it would
be very strange if this would not have been so during Stufe IId2–IIIa1.
For all these reasons the archaeological description of the Stufen IId2
and IIIa1 cannot be maintained in the way in which it was defined
by Kaiser and the description given here for Naqada IID2 therefore
differs to some extent from Kaiser’s Stufe IId2. The distribution of ceme-
teries and settlements remains the same as for Naqada IID1.
Wavy-Handled and Late pottery increases in importance while all
other classes decrease (lightly). The types represented for Black-Topped
and Red-Polished pottery nevertheless remain almost identical. This is
not the case for Late pottery, where not only the previously known
cups, bowls and jars (L 53 group) show a greater diversity but fur-
thermore a new type of large, round based shouldered jars (especially
L 36 n) occurs. The latter will become increasingly popular from Naqada
IIIA1 onwards. The large majority of the Rough types present during
Naqada IID1 remains so, but with less diversity. A few new types occur,

91
Hendrickx (n. 69), 274–276, pl. 87–89.
predynastic—early dynastic chronology 81

which are mainly imitations of Late types (e.g. R 57 b-c). Among the
large storage jars, the decline of type R 81 already noted for Naqada
IID1 continues, for the benefit of R 84 (= L 30 b) and R 86 p (= L
30 –). There is a marked change for the Decorated ware. Of the two
shapes typical from Naqada IIC onwards, the small flat based regu-
larly curved jars (e.g. D 43) hardly ever occur any more, while the
broad round based vessels (e.g. D 61) still do. Figurative representa-
tions are almost completely missing. For the Wavy-Handle class, the
tendency towards smaller vessels and narrow shapes continues (groups
W 43 and W 47), while the handles loose their functionality.

Naqada III

The transition from Kaiser’s Stufe II to Stufe III is not without prob-
lems. The difference between them is made up by the Late class, which
takes over from the Rough class as numerically the most important
group. However, Kaiser’s view of the spatial distribution of the Rough
and Late pottery at Armant92 does not take into account the fact that
an important number of the Late types are in reality made in the
Rough fabric (especially the types belonging to the L 30 series), although
he is well aware of the problem.93 Counting these with the Rough class,
shows that at Armant no part of the cemetery is dominated by marl
clay pottery. However, this does not mean that groups of graves dom-
inated by marl clay pottery do not occur during the Naqada culture.
On the contrary, large groups of graves at Elkab and Hierakonpolis
for instance, and even entire cemeteries such as those of Tarkhan, Tura
and Abu Roash are completely dominated by marl clay pottery. Only
the transition in dominance from Rough to Late pottery should be
placed more recently than suggested by Kaiser. This raises the ques-
tion of whether a transition from Naqada II to III should be situated
at this moment. As we are dealing with a gradual evolution and not
with sudden changes, this question is only of limited importance. A
change of main period is not supported by the characteristics of the
funerary equipment of the tombs. However, as it is certain that at least
from the beginning of the Naqada III period that the Naqada culture

92
Kaiser (n. 53), Pl. 15 B–C.
93
Kaiser (n. 53), 76, note 9.
82 stan hendrickx

has spread over the whole of Egypt and as the earliest writing and his-
torical documents occur during the Naqada IIIA1 period (cf. below),
it seems possible nevertheless to maintain the division between Naqada
II and III, albeit not for the original reasons. Naqada III sites occur
throughout Egypt and Nubia and their distribution can no longer be
used as a possible element for chronological information.

Naqada IIIA1

The definition of Kaiser’s Stufen IIIa1 causes a particular problem. The


number of tombs at Armant for Stufe IIIa1 is very limited,94 and the
Wavy-Handled types found in them are very similar to those of Stufe
IId2. As mentioned already there is no domination by marl fabric pot-
tery that would make a difference from Stufe IId2. It therefore seems
impossible to retain Stufe IIIa1 in the way defined by Kaiser and the
description given here for Naqada IIIA1 will differ strongly. The major-
ity of Stufe IIIa1 tombs is included in Naqada IID2, while part of Stufe
IIIa2 is considered a separate entity and redefined as Naqada IIIA1
(cf. below).
Black-Topped pottery no longer occurs and the importance of Red-
Polished pottery has strongly diminished. For most types of the latter
class, parallels can be found among the Late pottery and it is very pos-
sible that confusion has occurred. Unfortunately it has not yet been
possible to check whether the vessels identified as Red-Polished are
indeed made from Nile silt and not from the marl fabric characteris-
tic of the Late class. The large Rough storage jars are now dominated
by slender types (L 30 g/k, L 31 a) which are, however, taller than
the Naqada IID types. Among the round based shouldered marl clay
storage jars the types L 36 n/s are particularly well represented. There
are no important changes in the remaining part of the Late pottery,
although the diversity and quantity increases. The Decorated pottery
characteristic of Naqada IIC–IID2 no longer occurs. The types of ves-
sels, which are now decorated also, occur without decoration among
the Late class. The decoration is largely reduced to series of undulat-
ing lines (groups D 20–21, D 24–25). The Wavy-Handled types again
become more slender (W 49–50) and the handles that had already lost

94
Only 4 tombs. It is to be noted that on the distribution plan (Kaiser (n. 53), Pl.
20 C), the symbols for Stufe IIIa1 and IIIa2 have been confused.
predynastic—early dynastic chronology 83

their functionality are, in a number of cases, changed into a continu-


ous ornamental band (W 51 a, W 56 a, W 56 g). Finally it is to be
noted that palettes show less diversity in shape and are dominated by
simple rectangular types.

Naqada IIIA2

Kaiser’s definition of Stufe IIIa2 at Armant is based on a very limited


number of graves and the majority of the types of objects presented
as characteristic for Stufe IIIa2,95 are not represented at the Armant
cemetery. The description of Stufe IIIa2 is therefore largely theoretical.
When studying the spatial distribution of the Naqada III cemetery of
Elkab two groups could be distinguished within material characteristic
for Kaiser’s Stufe IIIa2.96 A similar observation could be made for the
Hierakonpolis Fort Cemetery,97 which was also not yet published at the
time when Kaiser made his study of the relative chronology. Because
of this and the above mentioned problems with the Stufen IId2–IIIa1,
the earliest group distinguished within Kaiser’s Stufe IIIa2 was read-
justed to Naqada IIIA1. The validity of this was later confirmed by
the horizontal development of the elite tombs at cemetery U at Abydos
and especially the position of tomb U-j and the types of objects found
in it (cf. below).
The Naqada IIIA2 period is characterised by a spectacular decrease
of the Rough class in favour of Wavy-Handled pottery. It is to be
noted however, that this picture is strongly influenced by the Tarkhan
Valley cemetery from which the majority of the Naqada IIIA2 tombs
known comes. The possibility for regional variation, especially between
UE and LE, can therefore not be excluded. The Nile silt Red-Polished
pottery has now been completely replaced by marl clay Late ware, the
variety and quantity of which once again increases. The large round
based shouldered storage jars are now dominated by somewhat broader
types (L 36 a/g2/k).98 The large Rough jars consist mainly of very
slender pointed jars (L 31 a), while the series R 84–86 no longer occurs.

95
Kaiser (n. 53), Pl. 24 B.
96
Hendrickx (n. 70, Elkab V), 205–16.
97
See the review by S. Hendrickx, BiOr 47 (1990), col. 643–646.
98
Naqada IIIA2 tombs have been published both according to Petrie’s predynastic
and protodynastic typology (Petrie (n. 42), 1921, (n. 45), 1953). The most important
84 stan hendrickx

Decorated pottery diminishes in quantity, but for the remaining exam-


ples both the vessel types and the decoration are identical to Naqada
IIIA1. It is however to be noted that the amount of Decorated vessels
would be much higher if the Wavy-Handled jars of type W 62 with
net pattern decoration would be included. These are most character-
istic for Naqada IIIA2 and are part of the Wavy-Handled jars with
their maximum diameter located at the rim, relatively broad base and
continuous decorative band (W 55, W 58, W 61, W 62).99 The stone
vessels consist mainly of calcite bowls and plates. Animal shaped palettes
have almost completely disappeared. The vast majority are rectangu-
lar with incised lines around the edges.

Naqada IIIB

Kaiser’s Stufe IIIb does not occur at the Armant cemetery, on which
his chronological framework is based, and is therefore mainly an inter-
pretation of the theoretical evolution of the Wavy-Handled types. More
recently, Kaiser divided Stufe IIIb in two subdivisions, IIIb1 and IIIb2
and added three Stufen, IIIc1, IIIc2 and IIIc3.100 With the late types of
the Wavy Handled class as main characteristics, the chronological stages
distinguished by Kaiser are summarised in Table II. 1.5.101 The dis-
tinction made between Stufe IIIb1 and IIIb2 does not seem justified,
since at Tarkhan, for instance, pottery types characteristic of the Stufen

concordances for the marl fabric storage jars are: 60 g = L 36 n, 60 j = L 36 k,


L 60 m = L 36 a.
99
For the Naqada IIIA2 Wavy-Handled jars, the most important concordances
between Petrie’s predynastic and protodynastic typologies are: 46 b = W 62 –, 46 d =
W 58, 46 f = W 58 –.
100
Kaiser (n. 58). In an earlier stage of research, Kaiser distinguished three peri-
ods, called Horizonte [W. Kaiser, “Einige Bemerkungen zur ägyptischen Frühzeit. III.
Die Reichseinigung”, ZÄS 91 (1964), 92–96; W. Kaiser & G. Dreyer, “Umm el-Qaab.
Nachuntersuchungen im frühzeitlichen Königsfriedhof. 2. Vorbericht”, MDAIK 38 (1982),
260–269]. For discussion of a number of problems related to the Horizonte, see E. C. M.
van den Brink, “The Incised Serekh-signs of Dynasties 0–1, Part I: Complete Vessels”,
in: Spencer (n. 39), 148–150.
101
Kaiser (n. 58) gives no archaeological description for these new Stufen, neither
does he discuss the way in which they have been distinguished. Therefore, Table II.
1.5 is based on personal information kindly supplied by W. Kaiser (Poznan symposium
1992, letter 30 Oct. 1993). The following correlation with the early kings of Egypt can
be made: Stufe IIIb2 = Iry-Hor and earlier; Stufe IIIc1 = Ka—Narmer; Stufe IIIc2 =
Hor Aha—Djer; Stufe IIIc3 = Djet/Den until the end of Dyn. 1.
predynastic—early dynastic chronology 85

Table II. 1.5. Relative chronological periods as distinguished by Kaiser (n. 53);
idem (n. 58), Abb. 1 and Hendrickx (n. 69), idem (n. 39), 36–43, illustrated
by the types of Wavy Handled/Cylindrical jars
Kaiser 1957, 1990 Stufe Hendrickx 1989, 1996 Naqada

– – no cylindrical jars IIID


50 t IIIc3 50 b-c / h-t IIIC2
50 d IIIc2 50 d-g IIIC1
48 s-t, 49 d, 50 d IIIc1 – –
48 s-t, 49 d/l IIIb2 – –
47 IIIb1 47 r-t, 48 s, 49 d/g IIIB
W 50–51 a, 55, 56 g, 61–62 IIIa2 W 55, 58, 60–62 IIIA2
– – W 49–51, 56 a/g IIIA1
W 41, 43 b, 47 g IIIa1 – –
W 41, 43 b, 47 g IId2 W 41–42, 43 b, 47 a/g/m IID2
W 24, 25 IId1 W 24–25, 27 IID1
W 3, 19 IIc W 3, 19 IIC

IIIb1 and IIIb2 are very often present together in the same tomb.102
Furthermore, the spatial distribution of the two groups of types shows
no obvious patterning at Tarkhan. Also, the very obvious spatial dis-
tribution of the Turah cemetery does not support the idea of a chrono-
logical difference between the above mentioned types.103 Therefore, the
description of Naqada IIIB as presented here is more or less equiva-
lent to both Stufe IIIb1 and IIIb2.
In terms of percentages, the distribution of the pottery classes remains
almost identical to Naqada IIIA2. There are however a number of
important changes in the types of vessels. The slender Rough jars with
pointed base have almost completely disappeared. The other Rough
types hardly change. For the marl clay storage jars, two size classes
can be distinguished. The smaller types (L 36 b, L 38 a) occur more
frequently than the large ones (L 36 a, L 36 k). A rare but notable
new type is large jars with applied rope decoration, generally consid-
ered to be wine jars (protodynastic 76).104 On the one hand, the few
Decorated vessels remaining are identical to those of the Naqada IIIA1

102
Hendrickx (n. 39), 58–59.
103
Hendrickx (n. 39), 59.
104
For this type of jars, see van den Brink “The Incised Serekh-signs of Dynasties
0–1, Part I: Complete Vessels”; idem, “The Pottery-Incised Serekh-Signs of Dynasties
0–1. Part II: Fragments and Additional Complete Vessels”, Archéo-Nil 11 (2001), 23–100.
His group I–II occurs mainly during Naqada IIIB, group III–IV during Naqada IIIC1.
86 stan hendrickx

period. Net paintings on Wavy Handled jars, on the other, have almost
completely disappeared. The Wavy Handled class is now represented
by cylindrical jars with a small band of incisions instead of an applied
decorative band (W 71 a, W 80 = 47 r-t, 48 s, 49 d/g). A remark-
able change is the strong increase in stone vessels. Among them are
imitations of the cylindrical jars and a large amount of bowls and plates.

Naqada IIIC1

Kaiser’s Stufe IIIc1 consists of types which are partially characteristic


for Stufe IIIb2 and partially for Stufe IIIc2. The existence of this kind
of “transitional period” can of course not be denied, but it is not appro-
priate to distinguish a period, which has no types of objects charac-
teristic for that period only. This is especially so since the archaeological
description of the Stufen is often used for dating individual graves or
even objects. It therefore seems better to distinguish less periods, and
admittedly have eventually a slightly less detailed idea of the chrono-
logical evolution of a cemetery. A limited amount of the types men-
tioned for Stufe IIIc1 is therefore in the present description integrated
in Naqada IIIB, but the large amount is joined with Kaiser’s Stufe IIIc2
types into Naqada IIIC1.
The Late class now becomes more important than the Wavy-Handled.
The marl fabric storage jars still consist of large (protodynastic 60 b/d)
and smaller types (59 group), but there is also a new group of still
smaller jars (65 group).105 Very large jars with decorative band (76
group) already occurred occasionally during Naqada IIIB, but now
become more frequent. Large Rough jars with pointed base on the
other hand no longer occur and the same applies to Decorated vessels.
The Wavy-Handled jars have now evolved into cylindrical jars with
slightly curved wall and without decorative band (W 90 = 50 d–g).
The previously occurring stone vessel types continue to be used but an
increase in squat jars and restricted cups can be observed.

105
See S. Hendrickx et al., “Milk, Beer and Bread Technology during the Early
Dynastic Period”, MDAIK 58 (2002), 284–286.
predynastic—early dynastic chronology 87

Naqada IIIC2

Naqada IIIC2 largely coincides with Kaiser’s Stufe IIIc3. The amount
of Wavy-Handled cylindrical jars diminishes drastically, but the amount
of stone vessels shows an almost identical increase. As the latter group
consists for a considerable part of cylindrical jars, it is obvious that this
is merely a change of material and not of vessel type. The pottery
cylindrical jars without decorative band have straight walls and are nar-
row (protodynastic 50 b–c/h–t) compared to those of Naqada IIIC1.
The large marl fabric storage jars become far less frequent than pre-
viously, but the occurrence of the small (protodynastic group 59) and
even smaller types (group 65) strongly increases. The large wine jars
with applied band decoration continue to be present. Palettes have
almost completely disappeared from the tomb equipment.

Naqada IIID

Naqada IIID is less well defined than the other phases of the Naqada
culture which have been described. This is due to the limited number
of tombs identified as Naqada IIID, the majority of which come from
Qaw and Abu Roash.106 This could however be improved as the recent
excavations at Adaïma and Helwan have yielded tombs apparently dat-
ing to Naqada IIID. In future, it should be possible to distinguish sub-
divisions within Naqada IIID and probably also regional differences.
The link with the early Dyn. 3 pottery is probably closer than has been
suspected before, but this also needs further investigation. It is also
problematic that an important characteristic of Naqada IIID is defined
in a negative manner, that is to say the absence of cylindrical jars.
Both of the smaller types of marl fabric storage jars continue to be
important. The large wine jars with applied band decoration have devel-
oped into “torpedo” shapes with almost vertical walls and high rims
(protodynastic 76 s/u/y). In addition, the vessels are made with less
care. This tendency apparently becomes more explicit as Naqada IIID
advances. Most important is the appearance of new types of vessels,
which will develop further during the OK. Among them bowls with

106
The Abu Roash tombs are not integrated in Table II. 1.4a–b because they are
not published according to Petrie’s typologies.
88 stan hendrickx

internal rim,107 early Maidum bowls and beer jars.108 Among the stone
cylindrical jars, nearly all of them in calcite, types without decorative
band are now more frequent than those with. New is the frequent
occurrence of dummy calcite cylindrical jars.

Correlation between Naqada III and Dyns. 0–2

Before discussing the actual correlation between Naqada III and Dyns.
0–2, the use of the concepts “Dyn. 0” and “Dyn. 00” should be dis-
cussed. The first notion was already used on occasion by Petrie and
Quibell but has only far more recently gained wide acceptance after
its reintroduction by Kaiser.109 Dyn. 0 has however been used with
different meanings and the only consistency is the inclusion of Iry-Hor
and Ka. Perhaps its use can be defended for referring to the line of
kings from Abydos identified through inscriptions and tombs at the
cemeteries U and B. Dyn. 00 on the other hand has only occasionally
been used, partially overlapping the meaning of Dyn. 0. Unless it’s
significance would be clearly defined and generally accepted, it seems
better to avoid the use of Dyn. 00. A number of elite tombs can be
connected by inscriptions or seal impressions with the Late Predynastic—
Early Dynastic kings.110 The relative chronological position of these
tombs was in some cases defined when studying the spatial distribution
within the Naqada III cemeteries of LE (Tarkhan, Turah). However,
the royal tombs from Abydos and the elite mastabas from Saqqara are
mainly attributed to a particular Naqada period by the characteristic
objects they contained (Table II. 1.6). It is furthermore possible to inte-
grate the spatial distribution of the elite tombs from cemetery U at
Abydos, which can be linked to the local late predynastic kings.111

107
D. Raue, “Ägyptische und nubische Keramik der 1.–4. Dynastie”, in: W. Kaiser
et al., “Stadt und Tempel von Elephantine 25./26./27. Grabungsbericht”, MDAIK 55
(1999), 173–189.
108
Hendrickx et al. (n. 105).
109
Kaiser (n. 14), 71.
110
Hendrickx (n. 39), 59–61.
111
G. Dreyer, Umm el-Qaab: Nachuntersuchungen im frühzeitlichen Königsfriedhof.
3./4. Vorbericht. MDAIK 46 (1990), 61–62; idem, Umm el-Qaab. Nachuntersuchungen
im frühzeitlichen Königsfriedhof. 5./6. Vorbericht. MDAIK 49 (1993), 36–37; idem,
Umm el-Qaab I. Das prädynastische Königsgrab U-j und seine frühen Schriftzeugnisse (Mainz,
1998), 18–19; G. Dreyer et al., Umm el-Qaab. Nachuntersuchungen im frühzeitlichen
Königsfriedhof 7./8. Vorbericht. MDAIK 52 (1996), 29–30; R. Gundlach, Der Pharao
predynastic—early dynastic chronology 89

Table II. 1.6. Correlation between Naqada IIIA1–IIID and Dyns. 0–1
King Naqada Abydos Saqqara Tarkhan Turah

Qa-a IIID Q S 3120, S 3121,


S 3500, S 3505 – –
Semerkhet IIIC2/D U – – –
Adjib IIIC2 X S 3038, S 3111,
S 3338 – –
Den IIIC2 T S 3035, S 3506,
S 3507, S X – –
IIIC1/2 – S 3036 – –
Djet IIIC2 Z S 3504 1060 –
Djer IIIC1 O S 3471, S 3503 – 235
Hor-Aha / Djer IIIC1/2 – – 300 –
Hor-Aha IIIC1 B 10/15/19 S 3357 – –
Narmer IIIC2 – – 1982 –
IIIC1 B 17/18 (?) – 414, 415, 1100 –

Ka IIIC1 – – 261 –
IIIB/C1 – – 1627, 1651
IIIB B 7/9 – – –
Iry-Hor IIIB B 1/2 – – –
– IIIB U-t, U-x, U-y, – 1702 54, 64, 89
U-z
– IIIA2 U-g, U-h, U-s, – – –
U-u, U-v
Scorpion I IIIA1 U-j – – –
– IIIA1 U-a, U-k, U-o, – – –
U-r, U-qq

In the southern part of cemetery U at Abydos, a group of mud brick


lined tombs clearly dates to Naqada IIIA1. It is even possible within
this group to recognise a west-east evolution from older tombs (U-a,
U-o, U-qq) to more recent tombs (U-r, U-k, U-j, U-i).112 The latter
are continued further to the east by Naqada IIIA2 tombs (U-f (?),
U-g, U-h). Starting from tomb U-s, which still dates to Naqada IIIA2,
a series of tombs sets of, linking cemetery U to cemetery B, where the
tombs of Iry-Hor up to Hor-'Aha are located. Unfortunately the evi-
dence published for these tombs in the preliminary reports is limited.
Eventually tombs U-u and U-v still date to Naqada IIIA2. U-t prob-
ably dates to Naqada IIIB, and because of their position in the ceme-
tery, this would also have been the case for U-y, U-z and U-x. The
earliest tombs of cemetery B, those of Irj-Hor (B1–2) and Ka (B7/9)
can be dated to Naqada IIIB on more reliable basis.

und sein Staat. Die Grundlegung der ägyptischen Königsideologie im 4. und 3. Jahrtausend (Darmstadt,
1998), 54, Abb. 54.
112
Dreyer et al. (n. 111, 1996), 29–30.
90 stan hendrickx

The pottery found in the tombs of Narmer (B 17–18), Hor-'Aha


and Djer is characteristic for Naqada IIIC1, making this apparently a
relatively short period. This seems to be confirmed by the limited num-
ber of Naqada IIIC1 tombs at Tarkhan compared to those of Naqada
IIIB and also IIIC2.113 The tombs dating to the reigns of “Serpent”,
Den and 'Adj-ib can be attributed to Naqada IIIC2. The position of
the tomb of Semer-khet, who apparently only reigned for a short period,
remains uncertain because hardly any objects from his tomb are known,114
and no mastabas dating to his reign have been identified at Saqqara.
The tombs dating to the time of Qa-'a fit well within Naqada IIID as
described above, but although there are certainly similarities with the
tombs of late Dyn. 2, there are also differences with the limited amount
of pottery known for the tombs of Per-ibsen115 and Kha-sekhemwy116
at Abydos. As mentioned before, it is at present impossible to make
well defined subdivisions within Naqada IIID but this will certainly be
possible in the future.

Radiocarbon Chronology

Radiocarbon dates for Naqada cemeteries are limited in number, and


the majority of them have been made a long time ago, resulting in
questionable dates and important deviations. Furthermore, the calibra-
tion curves for the (second half of the) 4th millennium BC show impor-
tant fluctuations with long possible data ranges as a consequence. It is
generally considered a “bad period” for radiocarbon dating.117 It is
therefore impossible to link the phases of the Naqada culture distin-
guished to an absolute chronology. The limited number of dates avail-
able, at any rate, does not allow for the distinguishing of chronological

113
Naqada IIIA2: 488 tombs; IIIB: 306 tombs; IIIC1: 73 tombs; IIIC2: 206 tombs.
114
G. Dreyer et al., “Umm el-Qaab. Nachuntersuchungen im frühzeitlichen Königs-
friedhof 11./12. Vorbericht”, MDAIK 56 (2000), 120–122.
115
Petrie, Abydos I, pl. VII.
116
Petrie, Abydos I, pl. VII; E.-M. Engel, “Abydos. Umm el-qa’ab, Grab des
Chasechemui. Deutsches Archäologisches Institut Kairo”, BCE 20 (1997), 25–28; idem,
Abydos, Umm el-Qa’ab, Grab des Chasechemui”, BCE 21 (2000), 50–58.
117
Cf. B. Midant-Reynes & P. Sabatier, “Préhistoire égyptienne et radiocarbone”,
Archéo-Nil 9 (1999), 83–107.
predynastic—early dynastic chronology 91

phases.118 By also integrating the dates from settlement sites it is nev-


ertheless possible to confirm the outline defined by Fekri Hassan.119
Radiocarbon dates from historically dated royal and elite tombs at
Abydos and Saqqara have been used in an attempt to date the reigns
of individual kings accurately.120 This however is not without problems
as is shown by the more recently published dates from Abydos,121 indi-
cating that Scorpion I (tomb U-j) and Hor-'Aha would be more or less
contemporaneous. Because of the characteristics of the funerary equip-
ment and the location within the cemetery, this is however to be
excluded. Eventually the use of old wood could account for this type
of aberration, something, which might quite easily have occurred in
the well-organised elite cemeteries. The correlation presented here (Table
II. 1.7) between the Naqada periods, the Dyns. 0–2 kings and the avail-
able radiocarbon dates is therefore to be considered preliminary and
approximate only. A major problem concerning the interpretation of
radiocarbon dates is the discrepancy between the historical chronology
and the radiocarbon chronology. The radiocarbon dates are consis-
tently older with by least 100 years,122 resulting in a very long time
span for Dyn. 2. As this dynasty is unfortunately poorly known, this
question has to remain open for the time being. Also, the correlation

118
Contra S. H. Savage, “AMS Radiocarbon Dates from the Predynastic Egyptian
Cemetery, N.7000, at Naga-ed-Dêr”, JAS 25 (1998), 235–249, cf. A. R. Miljard &
T. A. H. Wilkinson, “Comment on ‘AMS Radiocarbon Dates from the Predynastic
Egyptian Cemetery, N7000, at Naga-ed-Dêr by S. H. Savage’”, JAS 26 (1999), 339–341.
119
F. A. Hassan, “Radiocarbon Chronology of Archaic Egypt”, JNES 39 (1980),
203–207; idem, “Radiocarbon Chronology of Predynastic Naqada Settlements, Upper
Egypt”, Current Anthropology 25 (1984), 681–683; idem, “Radiocarbon Chronology of
Neolithic and Predynastic sites in Upper Egypt en the Delta”, AAR 3 (1985), 95–116;
F. A. Hassan & S. W. Robinson, “High-precision Radiocarbon Chronometry of Ancient
Egypt, and Comparisons with Nubia, Palestine and Mesopotamia”, Antiquity 61 (1987),
119–135. See also Hendrickx (n. 7), 13–81; Midant-Reynes & Sabatier, Archéo-Nil 9
(1999), 83–107.
120
Hassan, JNES 39 (1980), 203–207.
121
R. M. Boehmer, G. Dreyer & B. Kromer, “Einige Frühzeitliche 14C-Datierungen
aus Abydos und Uruk”, MDAIK 49 (1993), 63–68; J. Görsdorf, G. Dreyer & U. Hartung,
“14C Dating Results of the Archaic Royal Necropolis Umm el-Qaab at Abydos”,
MDAIK 54 (1998), 169–175; idem, “New 14C Dating of the Archaic Royal Necropolis
Umm el-Qaab at Abydos (Egypt)”, Radiocarbon 40 (1998), 641–647; Dreyer (n. 111,
Umm el-Qaab I ), 17–19.
122
Görsdorf, Dreyer & Hartung, MDAIK 54 (1998), 175. See also H. Haas et al.,
“Radiocarbon Chronology and the Historical Calendar in Egypt”, in: O. Aurenche,
J. Evin & F. Hours, eds., Chronologies in the Near East. Relative Chronologies and Absolute
Chronologie 16,000–4,000 B.P. (Oxford, 1987), 585–606.
92 stan hendrickx

Table II. 1.7. Absolute chronology


cal. BC

Naqada IIID from ca. 2920 onwards [Semerkhet]/Qa-a – Dyn. 2


Naqada IIIC2 ca. 3000–2920 Djet – Adjib
Naqada IIIC1 ca. 3150–3100 Narmer – Djer
Naqada IIIB U-t, Iry-Hor – Ka
Naqada IIIA2 ca. 3350–3150 U-g,h,s,u,v
Naqada IIIA1 U-a,k,o,r,qq – Scorpion I
Naqada IIC–IID2 ca. 3600–3350
Naqada IA–IIB ca. 4000/3900–3600

Table II. 1.8. Concordance between the Buto stratigraphy and the relative
chronologies of the Naqada culture and the Southern Levant (after E. C. Köhler,
Tell el-Fara’in—Buto III. Die Keramik von der späten Naqada-Kultur bis zum frühen Alten
Reich (Schichten III bis VI) (Mainz, 1998); Faltings, in: Interrelations, 165–170; Levy
& van den Brink, in: Interrelations, 3–38)
Buto Maadi Naqada period Southern Levant

Buto V Naqada IIIC2–IIID EB II


Buto IIIf–IV Naqada IIIB–IIIC1 EB IB (late)
Buto IIId–e Naqada IIIA2 EB IB (middle)
Buto IIIb–c Naqada IIIA1
Buto IIIa Naqada (IID1)–IID2 EB IB (early)
Buto IIb Naqada IIC–IID1
Buto IIa Wadi Digla II Naqada IIA–B EB IA
Buto Ib Wadi Digla I Naqada IC
Buto Ia (Badari) – Naqada IA–B Late Chalcolithic

between radiocarbon dates from the southern Levant and Egypt is


equally problematic.123

Maadi-Buto Culture124

The importance of this culture has only been realised during the last
decade. The Maadi-Buto culture, dating mainly to the 4th millennium
BC, is no longer regarded as a regional culture in the neighbourhood

123
E. Braun, “Proto, Early Dynastic Egypt and Early Bronze I–II of the Southern
Levant: Uneasy 14C Correlations”, Radiocarbon 43 (2001), 1279–1295.
124
Only a short overview is presented here. More detailed archaeological descrip-
tions can be found in the literature mentioned.
predynastic—early dynastic chronology 93

of Maadi, today a modern suburb of Cairo, because it was apparently


present over a large part or perhaps even the whole of LE, especially
during Naqada IIC/D.125 This, however, does not necessarily implicate
a political unity of LE.126 The connection between the different chrono-
logical phases, early, middle and late Maadi-Buto, recognised at Maadi
and Buto between themselves and in relation with the relative chronol-
ogy of the Naqada culture,127 poses considerable problems and the
earlier phases in particular have been revised several times over recent
years. At present they can nevertheless be summarised with some
precision.128
Of great importance is the presence in the lowest layer of occupa-
tion at Buto, Stratum Ia/b, of a large quantity of locally made south
Levantine Chalcolithic-style pottery, known from the Beersheba culture
in the southern Levant.129 Stratum Ia, containing about one-third of
locally made Chalcolithic pottery, is to be dated before 3700 cal BC,130
contemporaneous with Naqada IA/B. The remaining part consists of
LE Neolithic pottery. The amount of south Levantine Chalcolithic-style
pottery diminishes in Stratum Ib, which could be contemporaneous
with Naqada IC, although this is mainly based on the intermediate
position of this stratum between Buto Ia and IIa for which more reli-
able links with the Naqada culture are available. Buto II represents the
“classical” stage of the Maadi-Buto culture and the two layers distin-
guished, Buto IIa and b, can be dated to respectively Naqada IIA–B
and Naqada IIC–IID1. Buto Stratum IIIa, corresponding mainly to
Naqada IID2, represents a transitional phase between the LE Maadi-
Buto culture and the Naqada culture which by that time has started
spreading northward. The influence of the Naqada culture in the Delta
becomes far more important during Buto Stratum IIIb–c, correspond-
ing roughly to Naqada IIIA1. During Buto Stratum IIId–f hardly any
elements of the Maadi-Buto culture remain. (See also page 487, first
footnote).

125
Levy & van den Brink, in: Interrelations, 11–13.
126
See however Levy & van den Brink, in: Interrelations, 8.
127
Levy & van den Brink, in: Interrelations, 13, Table I.4.
128
Faltings, in: Interrelations, 165–170; Levy & van den Brink, in: Interrelations, 19,
Table I.8.
129
Faltings, in: Stationen, 35–45; idem (n. 6, Proceedings); idem, in: Interrelations
165–170. See also C. Commenge & D. Alon, “Competitive Involution and Expanded
Horizons: Exploring the Nature of Interaction between Northern Negev and Lower
Egypt (c. 4500–3600 BCE)”, in: Interrelations, 139–153.
130
Faltings, in: Interrelations, 168.
II. 2 INSCRIPTIONAL EVIDENCE FOR THE RELATIVE
CHRONOLOGY OF DYNS. 0–21

Jochem Kahl

Sources

Contemporaneous king-lists intended as historical records are not at


our disposal for reconstructing the chronology of the earliest dynasties
nor for establishing the order of the kings. That such lists existed is
shown by the Royal Annals.2 The only contemporaneous sources are
of a different nature, viz. a) kings’ names inscribed on ritual or festival
vessels or on administrative seals (and in one case, on a non-royal
statue), useful in reconstructing the succession of kings; b) administra-
tive labels citing eponymous events of specific years and stone vessels
inscribed with festival notes which aid in determining the length of
reigns. Later sources used for this study are restricted to the OK. It
must be borne in mind that all sources, whether contemporaneous or
later, may contain scribal errors,3 and that some, such as the Royal
Annals, were subject to ideological influence.

Kings before Nar-mer

The earliest writing from Egypt preserves the names of several rulers
who preceded Nar-mer, here considered the first king of Dyn. 1.4 Two

1
I would like to thank Eva-Maria Engel and Barbara Kneißler for information and
for help in preparing the documents.
2
Wilkinson, Annals; M. Baud, “Les frontières des quatre premières dynasties. Annales
royales et historiographie égyptienne,” BSFE 149 (2000), 32–46; idem, “Ménès”, 109–
147.—Schäfer, Annalen, remains the fundamental publication for the Palermo Stone,
the largest fragment of annals that has survived.
3
E.g. the labels Petrie, RT I, pl. 15: 16–17 where the meaning of the eponymous
event was misunderstood—see G. Dreyer et al., “Umm el-Qaab, Nachuntersuchungen
im frühzeitlichen Königsfriedhof, 11./12. Vorbericht,” MDAIK 56 (2000), 116 n. b; or,
for the reign of Den, an entry in the Annals (Cairo Fragment 5, recto, lower part, 5)
mentioning the planning (?) of a building which must have been erected under Djer,
cf. Wilkinson, Annals, 246–247.
4
Recent scholarship favours Narmer over 'Aha for the role of Menes, first king of
dynasties 0–2 95

of them are documented in the necropolis of This at Umm el-Qaab,


Abydos: Iry-Hor and Sekhen/Ka.5 Archaeological evidence makes
Sekhen/Ka the predecessor of Nar-mer and thus a successor of Iry-
Hor.6 The presence of the element Hor in the name Iry-Hor can be
cited in support of this interpretation, since the names of some other
“kings” attested before Nar-mer also include the god’s name: Ny-Hor,
Hat-Hor, and Pe-Hor. These “kings,” as well as some others, are pre-
sumed to have been local rulers or rulers who opposed the Thinite
elite. Information for determining the chronological relationship of these
rulers is provided solely by archaeological evidence.7 The following
names are attested:
Ny-Hor, at Tura8
Hat-Hor, at Tarkhan9
“Trio” (three circles surmounting vertical strokes), from the eastern
Delta,10 and perhaps also at Tura11
Pe-Hor (alternatively read Iry-Hor and thus assignable to him), at
Qustul12
Ny-Neit(?), at Helwan13
“Crocodile”, at Tarkhan14
“Bird and vertical sign”, at Tarkhan15

the First Dynasty; cf. Kitchen, RITA II, 533–534. Support for this interpretation is
provided by Docs. 1 and 2 (see infra); cf. L. Morenz, “Gegner des Narmer aus dem
Papyrus-Land: NW and W '-”, GM 189 (2002), 88. Baud, “Ménès”, 109–110, provides
a summary of the arguments pro and contra both Nar-mer and 'Aha.
5
Both readings are possible; see Kahl, System, 38–40.
6
W. Kaiser & G. D. Dreyer, “Umm el-Qaab, Nachuntersuchungen im frühzeitlichen
Königsfriedhof, 2. Vorbericht,” MDAIK 38 (1982), 238.
7
See J. Kahl, “Hieroglyphic Writing during the Fourth Millennium BC: An Analysis
of Systems,” Archéo-Nil 11 (2001), 106, fig. 3, and cf. W. Kaiser & G. Dreyer, MDAIK
38 (1982), 260–69; T. von der Way, Untersuchungen zur Spätvor- und Frühgeschichte Unterägyptens
(Heidelberg: SAGA 8, 1993), 101.
8
Junker, Turah, 147, fig. 57; Kaiser and Dreyer (n. 6), 260–69.
9
Kaiser & Dreyer (n. 6), 260–69.
10
H. G. Fischer, “Varia Aegyptiaca,” JARCE 2 (1963), 44–47.
11
Junker, Turah, 46–47, fig. 57; cf. Fischer (n. 10); von der Way (n. 7), 101.
12
B. B. Williams, The A-Group Royal Cemetery at Qustul: Cemetery L (Chicago: OINE
III, 1986), 163, pl. 76.
13
Identified in an inscription on a jar uncovered by Z. Y. Saad at Helwan in 1949
or 1950: E. C. Köhler & E. C. M. van den Brink, “Four Jars with Incised Serekh-Signs
from Helwan Recently Retrieved from the Cairo Museum,” GM 187 (2002), 65–66,
76 fig. 1: 2, 77 fig. 2: 2.
14
G. D. Dreyer, “Horus Krokodil, ein Gegenkönig der Dynastie O,” in: The Followers
of Horus. Studies Dedicated to Michael Allen Hoffmann 1944–1990, R. Friedman & B. Adams,
eds. (Oxford: ESAP 2, 1992), 259–63.
15
W. M. F. Petrie, Tarkhan I and Memphis V (1913), pl. 31:71; Dreyer (n. 14), 260.
96 jochem kahl

“Scorpion”, at Hierakonpolis16
a ruler with an obscure name, at Buto17
Not included in this list is a group of signs consisting of two falcons
on a serekh (“Double Falcon”), known from Abydos, Tura, Beda, and
the Sinai.18 Whether it represents a king’s name or is symbolic of royal
authority per se, remains open. Several groups of signs on labels and
in inscriptions on vessels from Tomb U-j at Umm el-Qaab, as well as
signs on the Min colossi from Coptos, on the Libya Palette and on
some other small finds, have been understood as kings’ names.19 But
this interpretation is problematic.20 The groups may be place names
and/or the names of gods instead.21

Nar-mer to Qa-'a: The Succession

Inscriptions preserved in seal impressions and on stone vessels have


established a highly reliable model for the succession during Dyn. 1.
Impressions of two different seals associated with the administration of
the necropolis were found in the royal cemetery of Umm el-Qaab,
Abydos. Both seals listed kings who were buried there. One dates from
the time of Den or 'Adj-ib (Doc. 1); the other is temp. Qa-'a or Hetep-
sekhemwy (Doc. 2). According to Werner Kaiser, whose interpretation

16
Quibell, Hierakonpolis I, pl. 25 (bottom), 26c; cf. G. Dreyer, “Ein Siegel der
frühzeitlichen Königsnekropole von Abydos,” MDAIK 43 (1987), 41–42; W. Kaiser,
“Zum Siegel mit frühen Königsnamen von Umm el-Qaab,” MDAIK 43 (1987), 116–17;
Helck, Thinitenzeit, 92; W. Kaiser, “Zur Entstehung des gesamtägyptischen Staates,”
MDAIK 46 (1990), 291 n. 23.
17
Von der Way (n. 7), 99, fig. 22: 6.
18
For documentation and discussion see E.-M. Engel, “Ein weiterer Beleg für den
Doppelfalken auf einem Serech,” in press; Junker, Turah, 47 fig. 57; J. Cledat, “Les
vases de El-Béda,” ASAE 13 (1914), 119 figs. 3–4, 120 fig. 6; E. D. Oren, “Sinai,” The
New Encyclopedia of Archaeological Excavations in the Holy Land, E. Stern, A. Levinson Gilboa
& J. Aviram, eds., vol. 4 ( Jerusalem, 1993) 1388; Kaiser & Dreyer (n. 6), 260–269.
19
G. Dreyer, Umm el-Qaab I. Das prädynastische Königsgrab U-j und seine frühen Schriftzeugnisse
(Mainz: AV 86, 1998), 178–80.
20
See B. J. Kemp, “The Colossi from the Early Shrine at Coptos in Egypt,” Cambridge
Archaeological Journal 10 (2000), 211–242; J. Kahl, “Das Schlagen des Feindes von Hu:
Gebel Tjauti Felsinschrift 1,” GM 192 (2003), 47–54. J. Baines, “The earliest Egyptian
writing: development, context, purpose”, in: S. D. Houston, ed., The First Writing. Script
Invention as History and Process (Cambridge, 2004), 150–189.
21
So J. Kahl, “Die frühen Schriftzeugnisse aus dem Grab U-j in Umm el-Qaab,”
CdE 78 (2003), 112–135.
dynasties 0–2 97

is followed here, the metal cylinder seal used to make the impressions
subsumed under Doc. 1 was cut during the reign of Den and altered
after his death.22 This accounts for the unsymmetrical sequence of
kings’ Horus names (written without serekhs)23 and the name of the god
Khent-imenty. The royal names are arranged from left to right:
Nar-mer—'Aha—Djer—“Serpent”—Den—Meret-neit. (The sequence
Djer—“Serpent”—is confirmed by Doc. 3, see infra.). Social, rather
than chronological reasons dictated that Meret-neit follows Den; as
king’s mother, her status was lower than her son’s.24 This interpreta-
tion is supported by the material (limestone)25 of the stelae for her tomb
at Umm el-Qaab and by seal impressions found there which are closer
to those from the time of “Serpent” than to those temp. 'Adj-ib.26 That
Meret-neit is presumed to have served as regent for her son accounts
for her burial among the kings at Umm el-Qaab.27 Accordingly, the
chronological order should be Nar-mer—'Aha—Djer—“Serpent”—
Meret-neit—Den.
Several impressions from sealings of leather bags facilitated the recon-
struction of a second cylinder seal, Doc. 2. Its design is similar to that
of Doc. 1. Again, there are no serekhs. The sequence of kings’ names
from Nar-mer to Qa-'a is reversed, with mention of Khent-imenty, per-
haps as tutelary deity of the necropolis.28 Meret-neit is omitted, prob-
ably because of her lower status.29 The royal names, arranged from left
to right, are: Qa-'a—Semer-khet—'Adj-ib—Den—“Serpent”—Djer—
'Aha—Nar-mer.
Inscriptions on stone vessels corroborate the succession Den—'Adj-
ib—Semer-khet—Qa-'a (Docs. 4–8) or segments of it (Docs. 9–13,
15–17). Paleographical analysis shows that kings’ names were added
from reign to reign. Sometimes 'Adj-ib’s name was erased (Docs. 10–11;

22
Kaiser (n. 16, 1987), 119.
23
Dreyer (n. 16), 35, argues that this was intended to designate the rulers in ques-
tion as deceased.
24
So both Dreyer (n. 16), 37, and Kaiser (n. 16, 1987), 118 n. 13.
25
Not one of the hard stones (grano-diorite, granite, or basalt) used for kings’ ste-
lae since the reign of Den; cf. Fischer (n. 10), 41–43.
26
Kaplony, Inschriften I, 495–496.
27
A seal impression from Saqqara, tomb S 3503 may name Djer and Meret-neit;
cf. W. B. Emery, Tombs II, 169 (2), fig. 226; Kaplony, Inschriften II, 1183 (730); III,
fig. 730; Helck, Thinitenzeit, 101.
28
Dreyer et al., “Umm el-Qaab. Nachuntersuchungen im frühzeitlichen Königsfriedhof,
7./8. Vorbericht,” MDAIK 52 (1996), 73.
29
So Dreyer et al. (n. 28), 72.
98 jochem kahl

cf. Doc. 15). Only a few of the inscriptions on these vessels use Horus
names; in most cases they mention instead the nsw bjt nb.tj title +
name.30 Because other contemporaneous inscriptions give both names
of a king, it is possible to equate Horus Den with nsw-bjt Khasty,31
Horus 'Adj-ib with nsw bjt nb.tj Mer-pi-bia,32 Horus Semerkhet with nsw
bjt nb.tj Iry-netjer33 and also with a second unreadable nsw bjt nb.tj
name,34 and, finally, Horus Qa-'a with nsw bjt nb.tj Sen,35 nsw bjt nb.tj
Sehetep36 and nsw bjt nb.tj Qa-'a.37
The three nsw bjt nb.tj names of Qa-'a can be interpreted as indica-
tive of chronologically different periods of his reign. According to this
proposal, Sen is the oldest of the three. On year labels of Qa-'a,38 Sen
is associated with the official Henu-ka39 who is documented under
Qa-'a’s predecessor Semer-khet.40 The nsw bjt nb.tj name Sen and the
official Henu-ka are both mentioned in connection with a “Sixth
Occasion of Inspection”; another document citing the same event men-
tions the nsw bjt nb.tj Sehetep.41 Therefore, Sehetep will have replaced

30
For the controversy about whether the element nb.tj belongs to the name or to
the title, cf. Wilkinson, Egypt, 206–207.
31
Label, Umm el-Qaab, Tomb T: Petrie, RT I, 22, 40–42, pls. 11: 14, 15: 16;
Helck, Thinitenzeit, 101, 123, 159, 169–170, 172–174, 188, 234; seal impression, Umm
el-Qaab, Tomb T: Petrie, RT II, pl. 19: 151; Kaplony, Inschriften I, 127; II, 807 (730);
seal impression, Umm el-Qaab, Cemetery T (?): Kaplony, Inschriften II, 1104 (83); III,
fig. 83; seal impression, Umm el-Qaab, Tomb T: Petrie, RT II, 25, 49, pl. 7: 5–6;
Kaplony, Inschriften II, 1142 (364); III, fig. 364; label, Umm el-Qaab: G. Godron, Études
sur le Horus Den et quelques problèmes de l’Égypte archaique (Geneva, 1990), pls. 1: 1—3: 6;
seal impression, Saqqara, Tomb S 3506: Emery, Tombs III, 68–69 (18), pl. 79: 18;
Kaplony, Inschriften II, 1118 (196); III, fig. 196; Helck, Thinitenzeit, 191; seal impression,
Abu Roash, Tomb M XII: P. Montet, “Tombeaux de la Ire et de la IVe dynasties à
Abou-Roach (deuxième partie)—Inventaire des objets,” Kêmi 8 (1946), 205–12, pl. 14;
Kaplony, Inschriften I, 135 (W); III, fig. 195; Helck, Thinitenzeit, 191.
32
Seal impression, Umm el-Qaab, Tomb X: Kaplony, Inschriften III, fig. 245.
33
Seal impression, Umm el-Qaab, Tomb U: Kaplony, Inschriften III, fig. 229.
34
Label, Umm el-Qaab, Tomb Q: E.-M. Engel, Das Grab des Qa'a in Umm el-Qa’ab:
Architektur und Inventar (Diss., microfiche, Göttingen 1997), 437, fig. 217: 5; label, Umm
el-Qaab, Tomb Q: Dreyer (n. 28), 73–74, pl. 14d.
35
Three labels, Umm el-Qaab, Tomb Q: Petrie, RT I, pl. 17: 29; Petrie, RT II,
pl. 8: 3, 12: 6; Petrie, Abydos I, pl. 11: 11.
36
Label, Umm el-Qaab, Tomb Q: Dreyer (n. 28), 74–75, pl. 14e.
37
Stone vessel, prov. not known: Kaplony, Steingefässe, 26–32, pls. 4, 20 (32).
38
Labels from Umm el-Qaab, Tomb Q: Engel (n. 34), 455, figs. 221: 1 and 4.
39
Cf. Kaplony, Inschriften I, 579.
40
Cf. labels from his reign: Engel, (n. 34), 437, fig. 217: 1; Petrie, RT II, pl. 8: 5.
41
Two labels, Umm el-Qaab, Tomb Q: Engel (n. 34), 455, figs. 221: 1 and 4 (Sen);
label, Umm el-Qaab, Tomb Q: Dreyer (n. 28), 74–75, pl. 14e (Sehetep).
dynasties 0–2 99

Sen in that year. The “Sixth Occasion of Inspection” could not have
taken place before the king’s sixth year.42 The new nsw bjt nb.tj name
Qa-'a for the king is the youngest because it is associated with his sed-
festivals.43
The style and content of inscriptions attesting two enigmatic kings
(Horus Senefer-ka44 and Horus “Bird”)45 date them to the time of Qa-
'a or slightly later.46 Three explanations are possible: (a) Senefer-ka and
“Bird” were rivals of Qa-'a. At the beginning of his reign, Qa-'a had
the “peaceful” name Sen, “the one who fraternizes.” The change to
Sehetep, “the one who pacifies” and to Qa-'a “the one with raised
arm” reflect political developments, viz. Qa-'a opposition to and even-
tual victory over two opponents. This alternative is favoured here. (b)
The names Senefer-ka and “Bird” are also names of Qa-'a; i.e., he
also changed his Horus name in the course of his reign.47 (c) The names
belong to rulers who reigned after Qa-'a died. The brevity of their
reigns did not permit either to arrange for Qa-'a’s burial nor were any
seals cut. Seal impressions found in Tomb Q at Umm el-Qaab leave
no doubt that Hetep-sekhemwy buried Qa-'a there.48

Evidence for the Lengths of Reigns during Dyn. 1

During the Early Dynastic Period a regnal year was not numbered but
identified by one or more specific significant events occurring in its
course.49 This is inferred from the existence of different names for the
same year in contemporaneous sources (labels and stone vessels) and
confirmed by the information provided by the Royal Annals. Perhaps
the use of more than one event resulted from the necessity to “name”
a year when it began, at a time when only scheduled festivals and

42
Presuming that there was only one “occasion” in any given year.
43
Stone vessel, Umm el-Qaab, Tomb Q: Petrie, RT I, pl. 9: 8; stone vessel, pri-
vate collection: Kaplony, Steingefässe, 26–32, pls. 4, 20 (12).
44
Lauer, Pyramide IV.1, pl. 17: 86; Emery, Tombs III, pl. 38: 1; Kaplony, Steingefässe,
33 (13).
45
Petrie, RT II, pl. 8A: 6; Lauer, Pyramide IV.1, pl. IV: 17.
46
For the sake of completeness, mention should be made of a seal impression from
Tomb S 3505 at Saqqara which preserves traces of an otherwise unknown Horus
name; see Kaplony, Inschriften I, 147, 149; III, fig. 742.
47
Cf. Lauer, Pyramide IV.1, 15 (86), with reference to the name Senefer-ka.
48
Dreyer (n. 28), 71.
49
Cf. Baud, “Ménès”, 109–147; idem, BSFE 149 (2000), 32–46.
100 jochem kahl

ritual or cultic events could be selected to identify the year. At year’s


end, events unforeseen at its beginning, such as expeditions or cam-
paigns, could have been chosen and cited retrospectively.50 In general,
isolated eponymous events are unsuitable for determining the sequence
of regnal years, by contrast to sed festivals and other recurring events.
Even if sed festivals were celebrated for the first time before regnal
year 30,51 mention of one suggests a point later, rather than earlier
in a given reign. A sed festival is documented for Den52 and 'Adj-ib;53
Qa-'a celebrated a second.54 Other eponymous events which took place
repeatedly are also suggestive of a long reign. Examples are the “Sixth
Inspection” mentioned above and a “Second Running of Apis” during
the reign of Qa-'a.55 However, it is not known whether these events
occurred at regular intervals. Thus they provide only a vague idea of
reign length at best.
The Royal Annals, which survive on fragments in Palermo, Cairo
and London,56 list entries for every year of Semer-khet’s reign (cf. infra).
For other kings, only some years are preserved.

50
Similarly, Dreyer, MDAIK 56 (2000), 116 n. a.
51
Hornung & Staehelin, Sedfest, 62–63.
52
Label, Umm el-Qaab, Tomb T: Petrie RT, I, 21–22, 40–41, pls. 11: 5, 14: 12;
Helck, Thinitenzeit, 71, 123, 160, 169–70, 215; label, Umm el-Qaab, Tomb T: G.
Dreyer, MDAIK 46 (1990), 80, pl. 26a. The mention of a second sed festival without
a king’s name on the fragment of a stone vessel from Umm el-Qaab, Tomb T (Dreyer,
MDAIK 46 (1990), 80, fig. 9 and pl. 26d) may refer to Den; alternatively, the frag-
ment may be an intrusive find from the tomb of 'Adj-ib, of Semer-khet or—most
probably—of Qa-'a.
53
Stone vessel, Step Pyramid, Gallery H: Lauer, Pyramide IV.1, pl. III: 7; IV.2, 19–20
(35); Helck, Thinitenzeit, 123–124; stone vessel, Umm el-Qaab, Tomb X, Q, U: Lauer,
Pyramide IV.1, pl. III: 6; IV.2, 20; Petrie, Abydos I, pl. 5 (upper left); idem, RT I, 20–21,
39–40, pls. 6: 2, 7: 5 and 10, 8: 11; Helck, Thinitenzeit, 123, 192, 228; stone vessel,
Saqqara, Tomb S 2446: Quibell, Archaic Mastabas (Cairo, 1923), 13, 41, pl. 33: 5;
Lauer, Pyramide IV.1, pl. III: 4; IV.2, 20. Cf. K. O. Kuraszkiewicz, “Noch einmal zum
zweiten Sedfest des Adjib,” GM 167 (1998), 73–75.
54
Sed festival: stone vessel from Umm el-Qaab, Tomb Q: Petrie, RT I, 20–21, 40,
pl. 8: 7–7a; second stone vessel from Tomb Q: Petrie, RT I, pl. 9: 8; stone vessel from
the Step Pyramid, outside Galleries H and B: Lauer, Pyramide IV.1, pl. IV: 4; IV.2,
24–25 (42); stone vessel in a Swiss private collection: Kaplony, Steingefässe, 34–38 (16),
pl. 5: 22.
Second sed-festival: stone vessel from Saqqara, Step Pyramid, outside Galleries H
and B: Lauer, Pyramide IV.1, pl. IV: 5; IV.2, 25 (43); another stone vessel from the
Step Pyramid, Gallery B: Lauer, Pyramide IV.1, pl. 8: 41; IV.2, 24 (41); stone vessel in
a Swiss private collection: Kaplony, Steingefässe, 26–32, pls. 4, 20 (12).
55
Label, Umm el-Qaab, Tomb Q: Engel (n. 34), 464, fig. 224.
56
See note 2, above.
dynasties 0–2 101

'Aha: highly probable x+2 years (end of his reign)57 and probably
x+5+y years (later years of his reign)58
Djer: highly probable 10+y years (beginning of his reign)59 and x+9+y
years (middle of his reign)60
Den: x+6+y years (middle of his reign)61 and highly probably x+14+y
years (later years of his reign)62
'Adj-ib: x+2 years (end of his reign)63
Semer-khet: 9 years (his complete reign)64
Qa-'a: 2+y years (beginning of his reign)65
Reconstructions of the Annals differ widely and must be considered
highly speculative. Here statements about hypothetical reign lengths are
disregarded.66 The possibility cannot be ignored that Nar-mer’s reign
was also recorded in the annals, especially now following on the dis-
covery of a label at Umm el-Qaab citing one of his years.67

Table II. 2.1. The kings from Nar-mer to Qa-'a


Horus Names nsw bjt nb.tj Names Sed Festivals Other Rulers

Nar-mer
'Aha
Djer
« Serpent »
Meret-neit
Den Khasti one
'Adj-ib Mer-pi-bia one
Semer-khet Iry-netjer and an
unreadable name
Qa-'a Sen
Sehetep Senefer-ka, “Bird”
Qa-'a two

57
Palermo Stone, recto, II: 1–2; cf. Wilkinson, Annals, 90–91, fig. 1.
58
Cairo fragment 5, recto, upper register; cf. Wilkinson, Annals, 238–40, fig. 10.
59
Palermo Stone, recto, II: 3–12; cf. Wilkinson, Annals, 90, 92–103, fig. 1.
60
Cairo fragment 1, recto, II; cf. Wilkinson, Annals, 186–93, fig. 4.
61
Cairo fragment 5, recto, lower register; cf. Wilkinson, Annals, 240–47, fig. 10.
62
Cf. Wilkinson, Annals, 103–19, fig. 1; Baud, BSFE 149 (2000), 37.
63
Cairo fragment 1, recto, III: 1–2; cf. Wilkinson, Annals, 193–94, fig. 4.
64
Cairo fragment 1, recto, III: 3–11; cf. Wilkinson, Annals, 194–201, fig. 4.
65
Cairo fragment 1, recto, III: 12–13; cf. Wilkinson, Annals, 201–202, fig. 4.
66
For an instructive overview, see Wilkinson, Annals, 256–57.
67
Dreyer, MDAIK 54 (1998), 139, fig. 29 and pl. 5c.
102 jochem kahl

Hetep-sekhemwy to Netjery-khet:68 The Succession

The sequence of three Dyn. 2 kings is secure: Hetep-sekhemwy, who


buried Qa-'a at Umm el-Qaab69—Ra'-neb—Ny-netjer. The inscription
on the shoulder of CG 1 (Doc. 18), a statue depicting a kneeling man
(presumably a priest), lists these Horus names in that order. Additional
support for the sequence is provided by inscriptions on stone vessels in
different hands mentioning Hetep-sekhemwy and Ra'-neb (Docs. 19,
20).70 In an inscription mentioning the ka-house of Hetep-sekhemwy on
another stone vessel from the Step Pyramid (Doc. 21), the name Ny-
netjer is written over an erased name. This document, along with the
inscribed fragment of a stone vessel from the tomb of Per-ibsen (Doc.
22), substantiates the sequence on the statue Doc. 18. (During Ny-net-
jer’s reign, Ra'-neb’s name was erased several times; see Doc. 20, 21
(?), 22.)71
nsw bjt nb.tj Weneg72 is attested only by inscriptions on stone vessels
found in the Step Pyramid and in Tomb S 3014.73 Weneg’s exact posi-
tion, as well as the identification of his Horus name among those known,
has remained open until now.74 A long-known inscription from Tomb
P at Umm el-Qaab (Doc. 22) provides the key to solving some of the
problems associated with Weneg.75 In the inscription the nsw bjt nb.tj
name Ny-netjer faces the opposite direction from the name of Ra'-neb
and that of his palace (Fig. II. 2.1). Ra'-neb’s name is partially erased.

68
Here and below the Horus name Netjery-khet is used to identify the first king of
Dynasty 3, in preference to Djoser, since the latter is not documented in contempo-
raneous texts.
69
See n. 48, above.
70
For the sequence Hetep-sekhemwy—Ra'-neb, cf. the arguments of H. G. Fischer,
“An Egyptian Royal Stela of the Second Dynasty,” Artibus Asiae 24 (1961), 47–48 with
n. 11.
71
A stone vessel inscribed with the names of Qa-'a and Ra'-neb (Kaplony, Steingefässe,
34–38 (16), pl. 5: 22) does not help to reconstruct the order of these kings.
72
For the reading, cf. B. Grdseloff, “Notes d’épigraphie archaique,” ASAE 44 (1944),
288–291.
73
Lauer, Pyramide IV.1, pls. V: 4, 19: 105, 20: 101–103 and 106–107; IV.2, 50–53.
74
Helck, Thinitenzeit, 103, proposed to identify him with the enigmatic Horus Sa,
known from the mention of his Ka-house in inscriptions on stone vessels from the Step
Pyramid; cf. Lauer, Pyramide V, 7–8, pls. 6–7; Helck, “Die Datierung der Gefässaufschriften
aus der Djoserpyramide,” ZÄS 106 (1979), 124.
75
See the excellent photo published in the exhibition catalogue Kemet alle sorgenti del
tempo, A. M. Donadoni Roveri & F. Tiradritti, eds. (Milan, 1998), 251.
dynasties 0–2 103

Fig. II. 2.1. Inscription on stone vessel fragment BM EA 35556


(drawing after photo: E.-M. Engel)

Scrutiny of the inscription reveals that the name Ny-netjer is written


over Weneg. Traces of the plant sign used to write Weneg are dis-
cernible, as are the enigmatic strokes to the upper left and right of it
(Fig. II. 2.2) which are also attested in another of his inscriptions.76
Thus Ny-netjer must have been Weneg’s successor, and the original
inscription referred to the palace of Horus Ra'-neb and to nsw bjt nb.tj
Weneg. Therefore the Horus name of nsw bjt nb.tj Weneg should be
Ra-neb.
nsw bjt nb.tj Nub-nefer is attested only on two stone vessels from the
Step Pyramid (Docs. 23, 24).77 Both inscriptions mention him in con-
nection with a building called Ówt-mn.t-'n¢. Another stone-vessel inscrip-
tion associates this structure with the Gold name Ren.78 The Palermo

76
Lauer, Pyramide IV.1, pl. 20: 104; IV.2, 50.
77
Lauer, Pyramide IV.1, pl. VI: 3–4.
78
Lauer, Pyramide IV.1, pl. 19: 98.
104 jochem kahl

Fig II.2.2. Reconstruction of the king’s name as originally written on BM EA 35556


(drawing: E.-M. Engel)

Stone gives Ren as Ny-netjer’s Gold name.79 Nub-nefer’s reign should


therefore be in proximity to Ny-netjer’s. Currently there are two options
for his dynastic position. Either Nub-nefer was Ra'-neb’s nsw bjt nb.tj
name80 or he was an ephemeral ruler who occupied the throne briefly
after Ny-netjer’s death.81 The evidence just presented for identifying
Weneg as Ra'-neb’s nsw bjt nb.tj name shows that the second alterna-
tive is correct. Kaiser’s suggestion that nsw bjt Sened was the last king
to reign over UE and LE before Kha-sekhemwy is very plausible.82
Circumstantial evidence for this sequence is the survival of the funer-
ary cult of Sened into Dyn. 4.83 nsw bjt Sened or nsw bjt nb.tj Nub-nefer

79
Recto, IV.
80
So Gunn, “Inscriptions from the Step Pyramid site III. Fragments of inscribed
Vessels,” ASAE 28 (1928), 156 n. 2; idem, ASAE 44 (1944), 292; cf. Beckerath, Handbuch, 48.
81
Helck (n. 74, Datierung), 131–32; Wilkinson, Egypt, 89.
82
“Zur Nennung von Sened und Peribsen in Sakkara B 3,” GM 122 (1991), 49–55.
83
In the tomb of Shery, Saqqara B 3: A. Mariette, Les mastabas de l’ancien empire
(Paris, 1889), 92–93; A. Moret, “Fragments du mastaba de Shery, prêtre des rois
Peribsen et Send,” Monuments Piot 25 (1921/22), 273–98.
dynasties 0–2 105

might be identical with Horus Sa, who is known from the mention of
his ka-house in inscriptions on stone vessels from the Step Pyramid.84
It is not clear whether the next two names—Horus Sekhem-ib and
Seth Per-ibsen—belonged to a single ruler or to two different kings.
Per-ibsen certainly claimed to rule over all of Egypt,85 but the sources
do not confirm this. Contemporaneous evidence for Seth Per-ibsen is
restricted to UE between Elephantine and Beit Khallaf, just north of
Abydos,86 except for his funerary cult in association with nsw bjt Sened
at Saqqara.87 Sekhem-ib is attested at Abydos and Saqqara. Seal impres-
sions mentioning Sekhem-ib come from the tomb of Per-ibsen, while
at Saqqara, stone vessels with Sekhem-ib’s name were found in the
Step Pyramid. But this does not prove that Sekhem-ib exercised influence
in the Memphite region, since these vessels could have been brought
to Saqqara from Abydos after Sekhem-ib’s death. Theories about the
relationship between the names Sekhem-ib and Per-ibsen88 suggest that
(a) Sekhem-ib and Per-ibsen were names borne simultaneously by a
single king;89 (b) Horus Sekhem-ib was the older name of Seth Per-
ibsen;90 (c) Horus Sekhem-ib buried Seth Per-ibsen and was thus his
successor.91 Down to the present there is no compelling argument favour-
ing one alternative over the others.
The last king of Dyn. 2 was Horus-Seth Kha-sekhemwy. His power
base seems to have been Hierakonpolis where he is attested as victor
over northern enemies under the name Horus Kha-sekhem.92 Presumably

84
Cf. note 74 above and M. J. Raven, “Les fouilles de Leyde dans la tombe de
Mérytneith à Saqqara. Campagnes 2001–2002,” BSFE 155 (2002), 31.
85
Kaplony, Inschriften III, fig. 368; cf. Kahl, System, 83–84.
86
Cf. Kahl, System, 356–58 (Quellen 2887–2914).
87
In the tomb of Shery, see note 83, above. Kaiser considers dynastic and political
reasons that might have led to the institution of Per-ibsen’s cult under Kha-sekhemwy.
88
Cf. Schneider, Lexikon, 405; Wilkinson, Egypt, 90–91.
89
E.g., Grdseloff, ASAE 44 (1944), 295.
90
E.g., E. Drioton & J. Vandier, L’Égypte4 (Paris, 1962), 164.
91
E.g., Helck, Thinitenzeit, 103–104.
92
Cf. the statues Cairo JE 32161 (Quibell, Hierakonpolis I, 11, pls. 40 (upper), 41
(left); Quibell & Green, Hierakonpolis II (London, 1902), 27–28, 44) and Oxford Ashmolean
Museum E 517 (Quibell, Hierakonpolis I, 11, pls. 39, 40 (bottom); Quibell & Green,
Hierakonpolis II, 31, 43–44. See now H. Sourouzian, “L’iconographie du roi dans la
statuaire des trois premières dynasties,” in: Kunst des Alten Reiches. Symposium im Deutschen
Archäologischen Institut Kairo am 29. und 30. Oktober 1991 (Mainz: DAIKS 28, 1995),
141–43, pl. 51. Cf. also the following stone vessels: Cairo CG 14724: Quibell, Hierakonpolis
I, pl. 38 (upper); P. Kaplony, “Bemerkungen zu einigen Steingefässen mit archaischen
Königsnamen,” MDAIK 20 (1965), 26 (53), pl. 6; Helck, Thinitenzeit, 72, 106–107;
J. Baines, Fecundity Figures. Egyptian Personification and the Iconology of a Genre (Warminster,
106 jochem kahl

later in his reign, Kha-sekhem took the Horus-Seth name Kha-sekhemwy


to demonstrate that peace and harmony had been restored through his
actions.93 Considerable circumstantial evidence exists for Horus Netjery-
khet as the successor of Kha-sekhemwy: (a) a seal with the names of
officials used once together with a seal of Kha-sekhemwy and then with
a seal of Netjery-khet (Doc. 25); (b) Ny-maat-Hep,94 mother of the royal
children and king’s mother (of Netjery-khet) is attested on seal impres-
sions in Kha-sekhemwy’s tomb at Umm el-Qaab as well as in Tomb
K1 at Beit Khallaf (Doc. 26a–b); (c) many seal impressions naming
Netjery-khet which come from Kha-sekhemwy’s tomb.95
The following Horus names of Dyn. 2 can be associated with nsw
bjt nb.tj names or nbw-names:
Horus Hetep-sekhemwy: nsw bjt nb.tj Hetep96
Horus Ra'-neb: nsw bjt nb.tj Weneg (see P. 102–103, above)
Horus Ny-netjer: nsw bjt nb.tj Ny-netjer,97 nbw Ren98
Horus Sekhem-ib(-Per-en-maat): nsw bjt nb.tj Sekhem-ib-Per-en-maat99
Seth Per-ibsen: nsw bjt/nsw bjt nb.tj Per-ibsen100
Horus-Seth Kha-sekhemwy(-Hetep-netjerwy-imef ): nsw bjt nb.tj Kha-
sekhemwy-Hetep-netjerwy-imef 101/Kha-sekhemwy-Nub-khetsen102

1985), 245; B.G. Aston, Ancient Egyptian Stone Vessels: Materials and Forms (Heidelberg:
SAGA 5, 1994), pl. 2a; Philadelphia, Penn. Univ. Mus. E 3958: Quibell, Hierakonpolis
I, pl. 38 (bottom); Helck, Thinitenzeit 72, 106–107; Baines, Fecundity Figures, 245; Oxford,
Ashm. Mus. No. not known: Quibell, Hierakonpolis I, pl. 37 (middle, right); Helck,
Thinitenzeit, 72, 106–107. Presumably also from Hierakonpolis are: Saqqara, no. 662:
Lauer, Pyramide IV.1, pl. 3: 18; IV.2, 8 (18); Helck (n. 74, Datierung), 132 n. 7; pri-
vate collection: Kaplony, MDAIK 20 (1965), 24, 26, fig. 54.
93
Cf., e.g., Wilkinson, Egypt, 91–92.
94
See S. Roth, Die Königsmütter des Alten Ägypten von der Frühzeit bis zum Ende der 12.
Dynastie (Wiesbaden: ÄAT 46, 2001), 59–67.
95
See, for the moment, Dreyer, “Der erste König der 3. Dynastie,” Stationen, 31–34.
96
Seal impressions: Kaplony, Inschriften III, figs. 281–82.
97
Presuming identical Horus and nsw bjt nb.tj names imply identity; for the latter,
see Lauer, Pyramide IV.1, pl. 15: 73.
98
Palermo Stone, recto, IV.
99
Again, presuming identical Horus and nsw bjt nb.tj names imply identity; for the
latter, see Lauer, Pyramide IV.1, pl. 18: 87–94.
100
As the preceeding. For nsw bjt Per-ibsen, cf. Kaplony, Inschriften III, fig. 368; for
nsw bjt nb.tj Per-ibsen, cf. Petrie, RT I, pl. 4: 7.
101
As the preceeding. For nsw bjt nb.tj Hetep-netjerwy-imef, cf. Kaplony, Inschriften
III, fig. 214
102
As the preceeding. For this nsw bjt nb.tj name, cf. W. M. F. Petrie, Tombs of the
Courtiers and Oxyrhynkhos (London, 1925), pl. 8: 18.
dynasties 0–2 107

The Lengths of the Reigns of Dyn. 2-kings

Contemporaneous sources yield comparatively little information about


the duration of reigns during Dyn. 2. Some stone vessels from the Step
Pyramid bear inscriptions citing specific events.103 One of them men-
tions the “Fourth Occasion of the Sokar Festival” (probably year 24);104
another, the “Seventeenth Occasion (of the cattle count)” (probably
year 34).105 These dates have been ascribed to Ny-netjer, since he seems
to have been the only ruler of the dynasty to have reigned more than
30 years.106 For the same reason, inscriptions mentioning a sed festival
are thought to refer to him.107 The Annals preserve information about
three kings of Dyn. 2: Ny-netjer, years 6–21108 and perhaps x + 9 years
at the end of his reign;109 Per-ibsen, 6 + y years from the beginning
of his reign;110 Kha-sekhemwy, perhaps years 3–6111 and most proba-
bly years 12–18 towards the end of his reign.112

Chronologically Significant Inscriptions, Dyns 1–3

Doc. 1 5 clay impressions of a cylinder seal


Dyn. 1, reign of Den—'Adj-ib
a–c: Abydos, find nos. Ab.K 300, 301a–b
d: Abydos
e: London, UC 188 (provisional no., assigned by Kaplony)

103
Lauer, Pyramide V, 88–90, (nos. 273–75), figs. 172–74.
104
Lauer, Pyramide V, 88 (no. 273), fig. 172; Helck (n. 74, Datierung), 128; Wilkinson,
Egypt, 85–86.
105
Lauer, Pyramide V, 88–89 (no. 274), fig. 173; Helck (n. 74, Datierung), 128;
Wilkinson, Egypt, 86.
106
Cf. Helck (n. 74, Datierung), 128. Wilkinson, Annals, 120, estimates a 39-year
reign for Ny-netjer.
107
Cf. Helck (n. 74, Datierung), 130 and Lauer, Pyramide V, 3–7, 59, figs. 6–13: 91.
But cf. I. Regulski, who assumes a sed festival of Kha-sekhemwy (I. Regulski, “Second
Dynasty Ink Inscriptions from Saqqara Paralleled in the Abydos Material from The
Royal Museums of Art and History (RMAH) in Brussels”, in: S. Hendrickx, R. F.
Friedman, K. M. Cialowicz & M. Chlodnicki, eds. Egypt at its Origins (Leuven: OLA
138, 2004), 960–967.
108
Palermo Stone, recto, IV. 1–16; cf. Wilkinson, Annals, 119–29, fig. 1.
109
Cairo fragment 1, recto, IV. 1–9; cf. Wilkinson, Annals, 202–206, fig. 4.
110
Cairo fragment 1, recto, IV. 10–15; cf. Wilkinson, Annals, 202–203, 207–208,
fig. 4.
111
London fragment, recto, upper part; cf. Wilkinson, Annals, 248–51, fig. 11.
112
Palermo Stone, recto, V. 1–7; cf. Baud, BSFE 149 (2000), 36–38; Wilkinson,
Annals, 130–36, fig. 1.
108 jochem kahl

Table. II. 2.2. The kings from Hetep-sekhemwy to Netjery-khet


Horus and/or Seth nsw bjt nb.tj Name Gold Name Sed Festivals
Name

Hetep-sekhemwy Hetep
Ra'-neb Weneg
Ny-netjer Ny-netjer Ren one(?)
Sa(?) Nub-nefer
Sened
Sekhem-ib Sekhem-ib
(Per-en-maat) / (Per-en-maat) /
Per-ibsen (one ruler?) Per-ibsen (one ruler?)
Kha-sekhem / Kha-sekhemwy
Kha-sekhemwy (Hetep-netjerwy-imef /
(Hetep-netjerwy-imef ) Nub-khetsen)
Netjery-khet Netjery-khet Nub(?)

Prov.: a-d, Umm el-Qaab, Tomb T; e, not known but prob-


ably also Tomb T.
Comment: See Pp. 96–97, above
Bibliography: Kaplony, Inschriften III, fig. 809; Dreyer, MDAIK 43 (1987),
33–43, pls. 3–5; Kaiser, MDAIK 43 (1987), 115–19; Dreyer, MDAIK
49 (1993), 61

Doc. 2 several clay impressions of a cylinder seal


Dyn. 2, reign of Hetep-sekhemwy
Abydos, find. nos. Ab.K 1486–1500
Prov.: Umm el-Qaab, Tomb Q
Comment: See Pp. 96–97, above
Bibliography: Dreyer, MDAIK 52 (1996), 72–73, fig. 26, pl. 14b–c

Doc. 3 2 fragments of a serpentine bowl


Dyn. 1, reign of Djer—“Serpent”
Berlin, Egyptian Mus. 15453 and Abydos, find no. Ab.K
5089
Prov.: Umm el-Qaab Tomb Z and Tomb T, respectively
Comment: These fragments bear incised Horus names attesting the seg-
ment Djer—“Serpent” of the sequence established by Doc. 1.
Bibliography: Petrie, RT II, 25, pl. 7: 1; V. Müller et al., “Umm el-
dynasties 0–2 109

Qaab. Nachuntersuchungen im frühzeitlichen Königsfriedhof. 11./12.


Vorbericht”, MDAIK 56 (2000), 109–10, fig. 22a, pl. 10a.

Doc. 4 schist bowl


Dyn.1, reign of Den, with inscriptions added under 'Adj-
ib—Qa-'a
Cairo, Egyptian Museum JE 88345
Prov.: Saqqara, Step Pyramid, Gallery B
Comment: The bowl bears four incised nsw bjt/nsw bjt nb.tj names:
Khasty, <Mer>-pi-bia, Iry-netjer, and Qa-'a.
Bibliography: Lauer, Pyramide III, pl. 19: 2; IV.1, 10, pls. I: 7, 4: 21;
IV.2, 10–12.

Doc. 5 6 fragments of a schist plate


Dyn. 1, reign of Den, with inscriptions added under 'Adj-
ib—Qa-'a
Prov.: Saqqara, Step Pyramid, Gallery H
Comment: The inscriptions record the nsw bjt/nsw bjt nb.tj names Khasty,
[Mer-pi-]bia, Iry-netjer and <Qa->'a.
Bibliography: Firth and Quibell, The Step Pyramid II (Cairo, 1935), pl.
105: 3; Lauer, Pyramide IV.1, 9–10, pl. 4: 20; IV.2, 9–12; Helck,
Thinitenzeit, 113, 187.

Doc. 6 2 fragments of a schist bowl


Dyn. 1, reign of Den, with inscriptions added under 'Adj-
ib—Qa-'a
Cairo, Egyptian Museum JE 55254–55255
Prov.: Saqqara, Step Pyramid, outside Galleries H and B
Comment: The adjoining fragments bear four incised nsw bjt/nsw bjt
nb.tj names, Khasty, Mer-pi-bia, Iry-netjer and Qa-'a.
Bibliography: Gunn, ASAE 28 (1928), 156, 158, pl. I: 2; Firth and
Quibell, Step Pyramid II, pl. 88: 1; Lauer, Pyramide IV.1, 10, pl. 4: 19;
IV.2, 9–12; Helck, Thinitenzeit, 113, 187.

Doc. 7 calcite bowl


Dyn. 1, reign of Den, with inscriptions added under 'Adj-
ib—Qa-'a
Michailidis collection
Prov.: not known
110 jochem kahl

Comment: The bowl bears the nsw bjt/nsw bjt nb.tj names Khasty, Mer-
pi-bia, Iry-netjer and Qa-'a.
Bibliography: Kaplony, Beschriftete Kleinfunde in der Sammlung Georges Michai-
lidis. Ergebnisse einer Bestandsaufnahme im Sommer 1968 (Istanbul, 1973), 6
(25), pl. 7: 25.

Doc. 8 “pyroxen-syenit” bowl


Dyn. 1, reign of Den, with inscriptions added under 'Adj-
ib—Qa-'a
Swiss private collection
Prov.: not known
Comment: The bowl shows four incised nsw bjt/nsw bjt nb.tj names:
Khasty, Mer-pi-bia, Iry-netjer and Qa-'a.
Bibliography: Kaplony, Steingefässe, 20–24 (9), pls. 2, 17, 18

Doc. 9 “black stone” (diorite?) bowl


Dyn. 1, reign of Den, with added inscriptions of 'Adj-ib and
Semer-khet
New York, MMA 58.125.2
Prov.: not known (ex. Michailidis collection)
Comment: The bowl shows three inscised nsw bjt/nsw bjt nb.tj names:
Khasty, Mer-pi-bia and Iry-netjer.
Bibliography: Kaplony, MDAIK 20 (1965), 13 (22), pls. 3: 22, 4: 22.

Doc. 10 fragment of a rock crystal bowl


Dyn. 1, reign of Den, with added inscriptions of 'Adj-ib and
Semer-khet
London, BM EA 49278
Prov.: Umm el-Qaab, Tomb T
Comment: Mer-pi-bia, originally incised near the name Khasty (untouched),
was erased and replaced by Iry-netjer, documenting the sequence Khasty,
Mer-pi-bia, Iry-netjer.
Bibliography: E. Naville, The Cemeteries at Abydos I (London, 1914), 35,
pls. 8 (lower right), 14: 1; Spencer, Objects, 42 (271), pls. 23: 271, 26: 271.

Doc. 11 fragment of a stone vessel


Dyn. 1, reign of Den, with added inscriptions of 'Adj-ib and
Semer-khet
dynasties 0–2 111

Paris, Louvre E. 11035


Prov.: Umm el-Qaab, Tomb T
Comment: As Doc. 10.
Bibliography: E. Amélineau, Les nouvelles fouilles d’Abydos 1895–1896,
(Paris 1899), pl. 42; K. Sethe, “Die ältesten geschichtlichen Denkmäler
der Ägypter”, ZÄS 35 (1897), 3; Lauer, Pyramide IV.2, 10.

Doc. 12 fragment of a quartz-crystal bowl


Dyn. 1, reign of Den, with added inscription of 'Adj-ib
Philadelphia, Univ. Mus. E 06847
Prov.: Umm el-Qaab, Tomb U
Comment: The names Khasty and Mer[-pi]-bia are incised on the
fragment.
Bibliography: Petrie, RT I, 19, 38–39, pl. 5: 9; II, pl. 47: 31.

Doc. 13 fragment of a red limestone bowl


Dyn. 1, reign of Den, with added inscription of 'Adj-ib
Cairo, Egyptian Museum JE 34378
Prov.: Umm el-Qaab, Tomb T
Comment: The vessel may have originally shown more names than
Khasty and Mer-pi-bia.
Bibliography: Petrie, RT I, 19, 38–39, pl. 5: 12.

Doc. 14 three adjoining fragments of a calcite vessel


Dyn. 1, reign of Den with added inscription of 'Adj-ib
London, BM EA 32659
Prov: Umm el-Qaab, Tomb X or Z113
Comment: 'Adj-ib’s Horus name is written in front of the nsw bjt name
Khasty.
Bibliography: Petrie, RT I, 19, 38, pl. 5: 11; Spencer, Objects, 41 (268),
pl. 25: 268.

Doc. 15 three adjoining fragments of a schist bowl


Dyn. 1, prior to the reign of Semer-khet, with inscriptions
added naming Semer-khet and Qa-'a

113
Cf. Spencer, Objects, 41 (268).
112 jochem kahl

Cairo, Egyptian Museum JE 55268 (one fragment only)


Prov.: Saqqara, Step Pyramid, outside Galleries H and B
Comment: The field in the serekh has been cut down, i.e. the name
Semer-khet has been substituted for another.
Bibliography: Gunn, ASAE 28 (1928), 158, pl. 1: 5; Lauer, Pyramide
IV.1, pls. IV: 3, 8: 38; IV.2, 22, (38); Helck, Thinitenzeit, 101–102.

Doc. 16 part of a schist bowl


Dyn. 1, reign of 'Adj-ib, with added inscription of Qa-'a
Prov.: Saqqara, Step Pyramid, Great South Court
Comment: The two nsw bjt nb.tj names incised on the vessel are Mer-
pi-bia and Qa-'a; Iry-netjer was not mentioned.
Bibliography: Lauer, Pyramide IV.1, pl. 8: 36; IV.2, 21 (36).

Doc. 17 schist bowl


Dyn. 1, reign of Semer-khet with added inscription of Qa-'a
Cairo, Egyptian Museum JE 88344
Prov.: Saqqara, Step Pyramid, Gallery B
Comment: The treatment of the serekhs shows very well that different
hands inscribed the two Horus names Semer-khet and Qa-'a.
Bibliography: Lauer, Pyramide IV.1, pl. 8: 39; IV.2, 22; Kaplony, Inschriften
I, 593.

Doc. 18 granite statue of a kneeling man


Dyn. 2 or 3
Cairo, Egyptian Museum CG 1
Prov.: Mit Rahineh
Comment: The sign preceding the names Hetep-sekhemwy, Ra'-neb
and Ny-netjer incised in that order on the right shoulder blade of the
figure has been interpreted to read divine ancestor (Helck), falcon (Fischer),
or phoenix (Moret).
Bibliography: PM III (2nd ed.), 864; A. Moret, “L’influence du décor
solaire sur la pyramide,” in: Mélanges Maspero I (Cairo, 1961), 624, fig. 1;
Fischer, Artibus Asiae 24 (1961), 45–46; E. L. B. Terrace & H. G. Fischer,
Treasures of Egyptian Art from the Cairo Museum (London, 1970), 25–28;
Helck, Thinitenzeit, 240.
dynasties 0–2 113

Doc. 19 diorite bowl


Dyn. 2, reign of Hetep-sekhemwy, with added inscription of
Ra'-neb
Cairo, Egyptian Museum JE 65413
Prov.: Saqqara, Step Pyramid, Gallery H
Comment: The Horus names of Hetep-sekhemwy and Ra'-neb, incised
by different hands, face the goddess Bastet.
Bibliography: Lauer, Pyramide IV.1, pls. II: 8, 11: 58; IV.2, 31–32;
Fischer, Artibus Asiae 24 (1961), 46–47.

Doc. 20 flint bowl


Dyn. 2, reign of Hetep-sekhemwy (or earlier) with inscrip-
tions naming Hetep-sekhemwy and Ra'-neb
Cairo, Egyptian Museum JE 41981
Prov.: Giza, Valley Temple of Mycerinus
Comment: Both Horus names, incised by different hands, face the god-
dess Bastet. Hetep-sekhemwy is incised over an erasure, while Ra'-neb,
partially erased, is written behind Hetep-sekhemwy. Reisner argued that
Ra'-neb was succeeded by Hetep-sekhemwy on the basis of the partial
erasure of Ra'-neb’s name, but Docs. 21 and 22 show that Ra'-neb
was Hetep-sekhemwy’s successor.
Bibliography: G. A. Reisner, Mycerinus. The Temples of the Third Pyramid
at Giza (Cambridge/Mass., 1931), 102 (1), 179 (1), pl. 70c.; Fischer,
Artibus Asiae 24 (1961), 46–47; Helck, Thinitenzeit, 72, 103.

Doc. 21 footed schist bowl


Dyn. 2, reign of Hetep-sekhemwy or Ra'-neb, with added
inscription of Ny-netjer.
Prov.: Saqqara, Step Pyramid, Gallery B
Comment: The bowl provides evidence for the reign of Ny-netjer being
later than that of Hetep-sekhemwy. A reference to the “ka-house of
Horus Hetep-sekhemwy” follows the name Ny-netjer. As Lacau and
Lauer noted, Ny-netjer could have been substituted for the name of
another king (Hetep-sekhemwy or Ra'-neb.)
Bibliography: Lauer, Pyramide IV.1, pl. 15: 74; IV.2, 36 (74); Helck,
Thinitenzeit 195–196.
114 jochem kahl

Doc. 22 fragment from the rim of a bowl of volcanic ash


Dyn. 2, reign of Ra'-neb, with added inscription of Ny-
netjer
London, BM EA 35556
Prov.: Umm el-Qaab, Tomb P
Comment: See Pp. 102–103, above
Bibliography: Petrie, RT II, 26, 51, pl. 8: 12; R. Weill, La IIe et la IIIe
dynasties (Paris, 1908), 150; Spencer, Objects, 42 (275), pl. 26: 275; Helck,
Thinitenzeit, 103, 195–96 (n. 130); Spencer, in: Kemet (see n. 75, above),
251.

Doc. 23 fragment of a schist bowl


Dyn. 2, reign of Ny-netjer or slightly later
Cairo, Egyptian Museum JE 55268
Prov. Saqqara, Step Pyramid, outside Galleries H and B
Comment: nsw bjt Nub-nefer is mentioned in connection with the build-
ing Ówt-mn.t-'n¢, also attested from the reign of Ny-netjer.
Bibliography: Lauer, Pyramide IV.1, 6, pl. VI: 3; IV.2, 48–49.

Doc. 24 fragment from the rim of a schist bowl


Dyn. 2, reign of Ny-netjer or slightly later
Cairo, Egyptian Museum JE 55294
Prov.: Saqqara, Step Pyramid, outside Galleries H and B
Comment: nsw bjt Nub-nefer is mentioned in connection with the build-
ing ˙wt-mn.t-'n¢; cf. Doc. 23.
Bibliography: Lauer, Pyramide IV.1, 6, pl. VI: 3; IV.2, 48–49.

Doc. 25 clay seal impressions


Dyn. 2, reign of Kha-sekhemwy
Oxford, Ashm. Mus. A 1909.1118A-D, F-O; London, UC
112–114 (nos. assigned by Kaplony)
Prov.: Abydos, Shunet ez-Zebib
Comment: The seal (Kaplony, Inschriften III, fig. 346) with names of
officials, which made these impressions was used in tandem with a seal
of Kha-sekhemwy as well as with another naming Netjery-khet; cf.
P. 106, above.
Bibliography: Kaplony, Inschriften I, 164–66; II, 869 n. 1011; III, figs.
269, 346 and 801.
dynasties 0–2 115

Doc. 26a clay seal impressions


Dyn. 2 or 3, reign of Kha-sekhemwy or Netjery-khet
Cairo, Egyptian Museum CG 11106–112, 11143, 11145;
Châteaudun 38(2) (no. assigned by Kaplony); Paris, Louvre
no no.; Toronto, ROM B 2324; London, UC 92(2), 95, 96
(nos. assigned by Kaplony); Abydos (cf. Dreyer, infra)
Prov.: Umm el-Qaab; Tomb V
Comment: These impressions from the tomb of Kha-sekhemwy men-
tion Queen Ny-maat-Hep as mother of the royal children.
Bibliography: Kaplony, Inschriften I, 161; II, 855 (983), 1137 (325); III,
fig. 325; Dreyer, Stationen (see n. 95 above), 33.

Doc. 26b clay seal impressions


Dyn. 3, reign of Netjery-khet
London, UC 149–52 (nos. assigned by Kaplony)
Prov.: Beit Khallaf, Tomb K1
Comment: These impressions from a tomb dating to the reign of Netjery-
khet name Queen Ny-maat-Hep as king’s mother. Therefore one can
conclude for a sequence Kha-sekhemwy—Netjery-khet.
Bibliography: Kaplony, Inschriften I, 167; II, 873 (1070), 1137 (326); III,
fig. 326.
II. 3 THE RELATIVE CHRONOLOGY OF DYNASTY 3

Stephan J. Seidlmayer

To reconstruct the chronology of Dyn. 3,1 the number, names and


sequence of its kings must first be established. Very little information
can be gleaned from the Manethonian tradition since the surviving
epitomes are unfortunately marred by erratic repetitions and inflated
reign lengths.2 But the kinglists of the NK3 do provide crucial evidence,
while pWestcar preserves the names and sequence of two Dyn. 3 rulers.4
The data from these sources, summarized in the table below, must be
correlated with each other, and with OK evidence:

Westcar TC Saqqara list Abydos list Manetho Manetho


(Africanus) (Eusebius)

Nb-k # (19 yrs.) Nb-k # Necherôphês Necherôchis


(28 yrs.)
˛sr ˛srjt (19 yrs.) ˛sr ˛sr-z # Tosorthros Sesorthos
(29 yrs.)
˛srtj (6 yrs.) ˛srttj Ttj Tyreis (var. six more kings
Tyris; 7 yrs.)
Ów-≈f # (6 yrs.) S≈s Mesôchris
(17 yrs.)
Nb-k # Nb-k#-R' Nfr-k#-R' Sôyphis
(16 yrs.)
Tosertasis
(19 yrs.)

(continued on next page)

1
Recent discussions include: Baud, “Ménès”; idem, Djéser et la IIIe dynastie (Paris,
2002); Beckerath, Chronologie, 160–163; G. Dreyer, “Der erste König der 3. Dynastie”,
in: Stationen, 31–34; S. Seidlmayer, “Die staatliche Anlage der 3. Dynastie”, in M. Bietak,
ed., House and palace in Ancient Egypt (Vienna, 1996), 198–200 note 14.
2
Waddell, Manetho, 40–45. For the relationship between Manetho and the kinglists
of the NK, see Helck, Manetho, 19–24 and Beckerath, Chronologie, 160.
3
TC, col III, lines 4–8; Saqqara list, nos. 12–15; Abydos list, nos. 15–19.
4
A. M. Blackman, The Story of King Kheops and the Magicians, Repr., W. V. Davies,
ed., (Reading, 1988), col. I.14, col. I.19 and passim.
dynasty 3 117

Table (cont.)
Westcar TC Saqqara List Abydos List Manetho Manetho
(Africanus) (Eusebius)

Ównj (24 yrs.) Ównj Achês


(42 yrs.)
Sêphuris
(30 yrs.)
Kerpherês
(26 yrs.)
total: 214 yrs. total: 198 yrs.
(aram: 197 yrs.)

Associating names from the lists with contemporaneous monuments—


above all, with royal mortuary complexes—provides additional infor-
mation about the number and sequence of kings. In OK sources, the
names of only two Dyn. 3 kings were written in cartouches: Nebka
and Huni. Normally, the kings of the dynasty were identified in con-
temporaneous sources by their Horus names, not by their personal
names. Therefore, the problem of correlating Dyn. 3 attestations with
the names of the later lists arises.

Five Horus names are known with certainty from Dyn. 3 contexts:
Netjery-khet, Sekhem-khet, Kha'ba, Zanakht and Qahedjet. Two other
names were considered by Swelim.5 Ink inscriptions on some stone ves-
sels from the galleries below the Step Pyramid,6 i.e., in a Dyn. 2 con-
text,7 attest the ka-house of za. Equating za with Horus Zanakht and
identifying him as Djoser’s predecessor and the owner of the original
mastaba below the Step Pyramid8 seems unfounded.9 Za and Zanakht

5
Swelim, Problems, 181–183.
6
Lacau & Lauer, Pyramide V, 7; B. Gunn, “Inscriptions from the Step Pyramid Site”,
ASAE 28 (1928), 168.
7
W. Helck, “Die Datierung der Gefässaufschriften aus der Djoserpyramide”, ZÄS
106 (1979), 120–132.
8
So J.-P. Lauer, Observations sur les pyramides (Cairo: BdE 30, 1960), 82; Kaplony,
Inschriften I, 409; J.-P. Lauer & H. Altenmüller, in: C. Vandersleyen, Das Alte Ägypten
(Berlin: Propyläen Kunstgeschichte 15, 1975), 113.
9
Doubts were expressed by: Beckerath, Handbuch, 177; Kaplony, Steingefässe, 7 n. 8,
and Wildung, Rolle, 55; Helck rejected these ideas, initially in “Datierung” (n. 7), 130,
and subsequently in Thinitenzeit, 108.
118 stephan j. seidlmayer

are clearly different; the common phonetic element za is written with


a different sign in each name. Since za never occurs alone or in a
serekh, it is doubtful that it represents a king’s name.10 Reports on an
alleged Horus B#, mentioned by Swelim, cannot be verified; even if he
could be shown to exist, there are no grounds for assigning him to
Dyn. 3.
Reviewing the available documentation, a number of problems are
evident which are taken up in turn below. Both TC and the Abydos
List name Nebka as the first ruler of the dynasty. But in pWestcar, a
king Nebka is a successor of Djoser, while a basically similar name
occurs in the Saqqara List as the penultimate entry and as the last in
the Abydos List. It was always suspected that there was only a single
king Nebka whose name was duplicated (in the Abydos List) or shifted
from its correct position. Since all attestations for Nebka from the OK
can be shown to refer to a king who reigned near the end of Dyn. 3,
the existence of a like-named king at the beginning of the dynasty was
always doubtful. Recent excavations at Abydos revealed unequivocal
evidence that Horus Netjery-khet buried Kha'-sekhemwy, the last king
of Dyn. 2,11 there, making it certain that no reign (and especially, a
chronologically significant one as shown in TC) could have intervened
between them.
Djoser, the second name in the list and, in fact, the first king of
Dyn. 3, is well attested. Later inscriptions confirm his identity with
Horus Netjery-khet, the owner of the Step Pyramid complex at Saqqara.12
His successor Djoserti/Djoser-teti is certainly Horus Sekhem-khet, the
owner of the second, smaller step pyramid complex at Saqqara. Evidence
for the equation includes the morphological similarity of the two archi-
tectural complexes, their geographic proximity and the fact that the
Nebti-Name of Sekhem-khet is attested as Djosert(i)-'ankh on an ivory
plaque from his pyramid.13
Data for the three remaining kings are less easily brought into line.
The so-called Layer Pyramid of Zawyet el-Aryan is the only other pyra-

10
See J.-P. Lauer, “A propos de l’invention de la pierre de taille”, in: Fs G. Mokhtar
(Cairo: BdE 97.2, 1985), 62–63.
11
Dreyer, in: Stationen, 31–33.
12
The relevant sources are the Famine stela at Sehel, temp. Ptolemy V (P. Barguet,
La stèle de la famine à Séhel, Cairo: BdE 24, 1953), a statue dedicated by Ahmose (Berlin
14765; Wildung, Rolle, 79–83) and numerous graffiti left by NK visitors to the Step
Pyramid complex (Wildung, Rolle, 65–72).
13
M. Baud, Djéser (n. 1), 67–68. Doubts expressed by Helck, “Das Kleidertäfelchen
aus der Pyramide des S¢m-§t”, WZKM 54 (1957), 72–76, do not seem justified.
dynasty 3 119

mid complex datable to Dyn. 3.14 The “great excavation” at the same
site is not contemporaneous with the Layer Pyramid, but belongs instead
to later Dyn. 4.15 The brickwork-complex of el-Dêr, discussed by Swelim,16
is probably not a pyramid at all (although it is difficult to judge this
monument on the basis of available documentation).17 If it were a pyra-
mid, its comparatively small size would associate it with the series of
small step pyramids in UE. Finally, evidence currently available sup-
ports the assignment of the initial building phase of the pyramid at
Maidum to Snofru (and Dyn. 4), not to Huni.18
Architecturally, the Layer Pyramid of Zawyet el-Aryan is very close
to the step pyramids at Saqqara—for example, the extensive subter-
ranean magazines are a common feature—and quite different from the
earlier phase of the Maidum pyramid. Accordingly, the structure is
most likely assignable to the successor of Horus Sekhem-khet/Djoserti;
in any case, it should not belong to Huni. The archaeological record
preserves no evidence of mortuary complexes for two later Dyn. 3
kings, a fact which may not be due simply to accident of preservation.
Their absence may reflect instead a temporary restructuring of arrange-
ments for the royal mortuary cult related to the appearance at the end
of the dynasty of the series of small step pyramids in UE associated
with the names of kings Nebka, Huni and Snofru.19 This line of rea-
soning supports an early date for the Layer Pyramid.
The name of the Layer Pyramid’s owner, Djoser’s second successor,
is listed as missing (S≈s, Ów-≈f #) in TC. Several stone vessels from

14
PM III, 313.
15
Convincingly established by J.-P. Lauer, “Sur l’âge et l’attribution de l’excavation
monumentale à Zaouiêt el-Aryân”, RdÉ 14 (1962), 21–36. Thus the much-discussed
royal name which appears in the masons’ graffiti from this monument (e.g. J. Cerny,
“Name of the King of the Unfinished Pyramid at Zawjet el-Aryan”, MDAIK 16 (1958),
25–29, and A. M. Dodson, “King [Bjk-k#]”, ZÄS 108 (1981), 171) is irrelevant for the
history and chronology of Dyn. 3.
16
Swelim, Problems, 337.
17
Cf. the archaeological situation as presented in the original publication, M.R.
Macramallah, “Une forteresse du moyen empire (?) à Abou Rawâch”, ASAE 32 (1932),
161–173. Macramallah’s opinion on the dating of the site is very tentative.
18
See R. Stadelmann, “Snofru und die Pyramiden von Maidum und Dahschur”,
MDAIK 36 (1980), 443–446.
19
W. Kaiser & G. Dreyer, “Zu den kleinen Stufenpyramiden Ober- und Mittel-
ägyptens”, MDAIK 36 (1980), 43–59; Seidlmayer, in “Anlage” (n. 1), 205–209; idem,
“Town and State in the Early OK, a view from Elephantine”, in Aspects of Early Egypt,
A. J. Spencer, ed., (London, 1996), 108–127. Horus Zanakht/Nebka is represented by
a seal impression in the context of the royal complex attached to the pyramid at
Elephantine and the name Huni occurs in an inscription on a block from the pyra-
mid; for Snofru as the owner of the pyramid of Seila see J. Leclant, Orientalia 57 (1988),
336 Fig. 40, and Orientalia 58 (1989), 368.
120 stephan j. seidlmayer

mastaba Z500 in the cemetery of Zawyet el-Aryan20 bear the Horus


name Kha'ba who might well be the owner of the pyramid. But, the-
oretically, Kha'ba could also be a successor of the Layer Pyramid’s
owner. As the next ruler, the NK lists name Nebka-Re' or Nefer-ka-
Re', both corrupted from Nebka, the name attested in pWestcar and
twice in the OK. The earlier OK attestation occurs in inscriptions from
the mastaba of Akhet'a who held a priestly title in the king’s cult.21
Helck identified Akhet'a with a person whose titles (but neither the title
relating to the cult of Nebka nor the name) appear among the inscrip-
tions on stone vessels from the Step Pyramid, to conclude that both
the person and King Nebka date to early Dyn. 3.22 But this equation
is neither obvious nor convincing. Even if Helck were correct, Akhet'a
could have nevertheless outlived Djoser and his immediate successors
to serve in the cult of a later king Nebka. The style of the reliefs from
Akhet'a’s tomb and the overall similarity of its plan to the tomb chapel
of Metjen clearly point to a late Dyn. 3/early Dyn. 4 date for the
owner. The fact that the king’s name in Akhet'a’s title is written in a
cartouche clearly precludes the possibility that the tomb could have
been decorated under Djoser.
The second OK attestation of King Nebka occurs in the name of a
funerary domain associated with the cult of Neusserre'.23 The context
links Nebka to late Dyn. 3/early Dyn. 4 when the system of funerary
domains is first attested, under Huni and Snofru.24 Thus the pre-NK
sources substantiate the presence of a king Nebka towards the end of
Dyn. 3, a conclusion quite in keeping with the fact that the existence
of a like-named ruler at the very beginning of Dyn. 3 can be excluded.
Significantly, Nebka can be identified with Horus Zanakht on the
basis of a seal impression from mastaba K2 at Beit Khallaf where the
names occur in parallel.25 The king is attested as Horus Zanakht at a

20
For the inscriptions see D. Dunham, Zawjet el-Aryan. The Cemeteries of the Layer
Pyramid, (Boston, 1978), 29–34, pl. XIIA. G. A. Reisner, Mycerinus, The Temples of the
Third Pyramid at Giza, (Cambridge Mass., 1931), 103, classifies Z500 as a stairway
mastaba, making Swelim’s idea (Problems, 78–89) that it could be the mortuary tem-
ple of the Layer Pyramid unlikely.
21
Ziegler, Catalogue, 96–103.
22
W. Helck, “Datierung” (n. 7), 129; idem, Thinitenzeit, 107.
23
Borchardt, Ne-user-Re, 79 Fig. 54.
24
H. K. Jacquet-Gordon, Les noms des domaines funéraires sous l’Ancien Empire Égyptien
(Cairo: BdE 34, 1962), 7–10.
25
Seidlmayer, in “Town” (n. 19), 121, and Pl. 23.
dynasty 3 121

number of other sites, in the rock inscriptions of Wadi Maghara, and


by seal impressions from the mortuary temple of the Step Pyramid and
from Elephantine.26 One sealing from Elephantine derives from layers
of refuse in the royal complex attached to the small step pyramid at
the site establishing a connection between Nebka and the series of small
step pyramids in UE. Unfortunately, the stratigraphic position of the
sealing is too equivocal to allow the relationship to be defined more
precisely.27
As the last ruler of Dyn. 3, the TC and the Saqqara List cite Huni,
a corrupted form of nswt Ówj.28 Huni was apparently the first king reg-
ularly identified in contemporaneous sources by his personal name, pref-
aced nswt (by contrast to the later usage of nswt-bjt), written in a
cartouche. To date, no document gives Huni’s Horus name in associ-
ation with his personal name. There is, however, one monument which
needs to be discussed in this context: a relief slab of Horus Qahedjet
depicting him embraced by Horus of Heliopolis.29 Allegedly from
Dahshur, it should derive from a Dyn. 3 royal mortuary complex at
the site. In fact, two alabaster sarcophagi from a shaft tomb in the
vicinity of the mortuary precinct of Senwosret III provide evidence for
the presence of a Dyn. 3 royal complex at Dahshur.30 Accordingly,
Qahedjet is an ideal candidate for Huni’s Horus name. It should not
be overlooked, however, that the correlation is not absolutely certain.
If Kha'ba is not the name of the owner of the Layer Pyramid at Zawyet
el-Aryan, but rather of one of his successors, then Kha'ba could qualify
as Huni’s Horus name. (Because of its early architectural type, it is

26
Discussed by Seidlmayer, in W. Kaiser et al., “Stadt und Tempel von Elephantine,
9./10. Grabungsbericht”, MDAIK 38 (1982), 304 with n. 82.
27
Ibidem, 304–305, idem, in “Anlage” (n. 1), 198–200.
28
This reading was established by L. Borchardt, “König Huni?”, ZÄS 46 (1909),
12–13, and corroborated by H. Schäfer, “König Huni”, ZÄS 52 (1914), 98–100. The
discussion triggered by H. Goedicke’s article “The Pharaoh Ny-swt˙”, ZÄS 81 (1956),
18–24, and continued by E. Meltzer, “A reconsideration of [njswt Ówj]”, JEA 57 (1971),
202–203, W. Barta, “Zum altägyptischen Namen des Königs Aches”, MDAIK 29 (1973),
1–4, and W. Helck, “Der Name des letzten Königs der 3. Dyn. und die Stadt Ehnas”,
SAK 4 (1976), 125–134, remains unconvincing. Why should not the first king who reg-
ularly had his personal name inscribed in a cartouche include the title nswt inside the
ring, in contrast to what became convention only subsequently? Actually this would
parallel the usage attested much more frequently with the epithet z#-R' which was
sometimes included within the cartouche.
29
J. Vandier, “Une stèle égyptienne portant un nouveau nom royal de la troisième
dynastie”, CRAIBL (1968), 16–22; Ziegler, Catalogue, 54–57.
30
PM III, 885. I am grateful to Dieter Arnold for pointing this out to me.
122 stephan j. seidlmayer

unlikely that the pyramid is Huni’s tomb). If so, the name Qahedjet
could be the Horus name of Djoser’s second successor whose personal
name is lost from the NK lists. Since Qahedjet probably owned a mor-
tuary complex at Dahshur, the Layer Pyramid would necessarily belong
to Horus Zanakht/Nebka. This reconstruction is not in itself impossi-
ble, but it seems to suit the available evidence less well. The results of
these deliberations may be summarized as follows:

Name in cartouche Horus name Mortuary complex


˛sr NΔrj-§.t great step pyramid, Saqqara
˛srtj S¢m-§.t lesser step pyramid, Saqqara
(unknown) ›'-b # layer pyramid, Zawyet el-Aryan
Nb-k # Z #-n¢t (unknown)
nswt Ówj Q #-˙≈.t mortuary cult place of unknown type
at Dahshur

Data currently available are sufficient neither for determining the length
of reigns for each king nor the length of the dynasty as a whole. Nor
does contemporaneous evidence exist for estimating a minimum length
of reign for any king. Only TC provides more or less useful data.
However, King Nebka was displaced in this document to the begin-
ning of the dynasty; perhaps the 19-year reign accorded him does not
represent original information but simply duplicates the figure given for
Djoser. The lengths of reigns in the Manethonian tradition are arbi-
trarily inflated; it would be guesswork to attempt to reduce them just
as arbitrarily.
The Old Kingdom Annals unfortunately can provide little assistance.31
The Palermo fragment preserves the end of Kha'-sekhemwy’s reign and
the beginning of Djoser’s in line 5. In addition, Kaiser noticed that,
taking all the available criteria into account, the distance between Cairo
1 and the Palermo fragment can be reconstructed so that the distance
between the two changes of reign preserved on both fragments matches
the combined reign lengths of Djoser and Djoserti as recorded in the
TC.32 While the line separating the two reigns is not actually visible on

31
See Beckerath, Chronologie, 174–179; Baud, “Ménès”, 135–138; idem, Djéser (n. 1),
50–52; Wilkinson, Annals.
32
W. Kaiser, ZÄS 86 (1991), 44 with note 8.
dynasty 3 123

the stone,33 such a reconstruction would still seem possible, in view of


the worn state of the surface at this place. Therefore, the evidence of the
Annals at least does not necessarily contradict the testimony of the TC.
Any reconstruction of the lengths of the reigns for later Dyn. 3 rulers
depends on estimating the distance between Cairo fragment 1 and the
left edge of the original stone slab. Beckerath’s recent calculation allows
only 24 years for all three remaining kings, in clear contradiction of
TC which gives 24 years to Huni alone plus six more for “Hu-djefa”.
In addition, an unknown number of years has to be taken into account
for Nebka. In view of the extremely fragmentary state of the evidence,
it would be hasardous to put excessive trust into any attempt to recon-
struct the original slab. Given the current state of knowledge, it is wiser
to regard the total of 50 years for the entire dynasty, which emerges
from Beckerath’s reconstruction, only as a minum figure. However,
even if the 19 years for Nebka of the TC be considered reliable, the
total length of the dynasty would not exceed 75 years.

33
Wilkinson, Annals, 53.
II. 4 CONTEMPORANEOUS EVIDENCE FOR THE
RELATIVE CHRONOLOGY OF DYNS. 4 AND 5

Miroslav Verner

The crucial problem for reconstructing the relative chronology of Dyns.


4 and 5 is our ignorance of how the system of dating according to the
“year/occasion of the (cattle) count” (rnpt zp) was employed. Ever since
Gardiner’s paper,1 specialists have acknowledged that this census formed
the basis for counting regnal years during the OK. The existence and
use of the term rnpt m-¢t zp (“year after the count”) was taken as clear-
cut evidence for a biennial census. However, the preponderance of rnpt
zp over rnpt m-¢t zp years among the burgeoning number of dates recov-
ered in recent years from excavations, taken in conjunction with an
entry on the Palermo Stone attesting the cattle count in two succes-
sive years of Snofru, indicates that the situation is much more com-
plex. Nowadays, some Egyptologists maintain that the census was biennial
during Snofru’s reign with the sole exception of the seventh and eighth
counts which were conducted in successive years.2 Others are of the
opinion that a biennial system was not employed under Snofru,3 while
yet others equivocate.4 For subsequent reigns, opinions fluctuate from
the presumption of a biennial system5 to the assumption that on
certain occasions an “odd” count could have been ignored.6 Finally,
there is also the theory that annual cattle counts became more and
more frequent during the OK until they became the rule by the end
of Dyn. 6.7
Obviously the existence of the census per se is not at issue, but rather
its regularity during the OK. However, a statistical review of documented

1
Gardiner, “Years”, 11–28.
2
E.g. R. Stadelmann, “Beiträge zur Geschichte des Alten Reiches”, MDAIK 43
(1987), 229–239; Baud, “Ménès”, 121.
3
E.g. R. Krauss, “Pyramid”, 47–50.
4
E.g. Spalinger, “Texts”, 281.
5
E.g. Baud, “Ménès”, 122–123, 128–129.
6
Spalinger, “Texts”, 318.
7
Beckerath, Chronologie, 147.
dynasties 4 to 5 125

dates, even if the list is incomplete and the attribution of some con-
tested, allows some conclusions. For example, from the beginning of
Dyn. 4 to the end of Dyn. 5, the years of the count and those fol-
lowing the count occurred in succession, and rnpt zp years were con-
sistently more frequently documented than rnpt m-¢t zp years. Moreover,
no clear-cut tendency can be observed towards a marked decrease in
the number of years following the count throughout the period.8 On
the contrary, the evidence for rnpt zp and rnpt m-¢t zp from Snofru’s
reign at the beginning of Dyn. 4 to Djedkare'’s at the end of Dyn. 5
supports the opposite conclusion.
Could a cattle count take place in the year of a king’s accession to
the throne? Until quite recently, the opinion prevailed that the first cattle
count of any given king’s reign occurred during the first full year after
his accession. But an entry among the annals inscribed on the stone
recently discovered at South Saqqara casts doubt on this assumption,
since the text explicitely mentions a (cattle) count of the year of the
“Unification of the Two Lands” at the beginning of the reign of Merenre'.9
The potential usefulness of the series, though incomplete, of rnpt zp
and rnpt m-¢t zp dates for Dyns. 4 and 5 is demonstrated by analysis
of the data for both Snofru and Djedkare'. Down to the present, the
highest recorded number of rnpt zp years for Snofru is 24; only half of
them (viz., rnpt zp 2, 7, 8, 12, 13, 14, 15, 16, 18, 23 and 24) are doc-
umented among the preserved dates. Can we assume that about the
same proportion—i.e., ca. half of the evidence for the intervening rnpt
m-¢t zp years—remains to date unattested? If so, the number (three) of
such currently documented dates—rnpt m-¢t zp 10, 13 and 18—should
be doubled. An estimate for the length of Snofru’s reign based on these
data would be 24 + 6 = 30 years. Using other arguments, Krauss10
and, independently, Barta11 arrived at nearly the same length for the
king’s reign.
Djedkare's highest documented census is the 21st (or possibly the
22nd). Thirteen other “occasions” are known: 1, 3, 4(?), 5, 6, 8, 9, 10,
11, 14, 15, 16, 17, and 18. In this series, seven (and possibly eight)

8
Contra Beckerath, Chronologie.
9
Baud & Dobrev, “Annales”, 47.
10
R. Krauss, “The Length of Sneferu’s reign and how long it took to build the
Red Pyramid”, JEA 82 (1996), 48.
11
W. Barta, “Die Chronologie der 1. bis 5. Dynastie nach den Angaben des rekon-
struierten Annalensteins”, ZÄS 108 (1981), 21.
126 miroslav verner

rnpt zp or about one-third of the total is still unattested. There are


seven rnpt m-¢t zp dates preserved: 1, 3, 4, 7(?), 10, 14, and 17. Adding a
third of this amount, or about two years, to the number of rnpt zp (21
or 22?) and rnpt m-¢t zp (7) years actually attested yields 30 (or possibly
31) years for the reign. It must be emphasized, however, that for the
present these calculations, since speculative, must be treated with reserve.
No matter how cautiously conclusions be drawn, the available data
indicate that the dating system was irregular during Dyns. 4 and 5,
and not principally biennial with few exceptions. In practice, annual
cattle counts apparently prevailed. The theory that intervening years
were omitted from the record under certain circumstances12 is contra-
dicted by the so-called masons’ inscriptions which consistently refer only
to rnpt zp years. These short texts associated with the construction pro-
jects of the state are the most frequently preserved dated documents
from Dyn. 4 and 5. Why should these inscriptions regularly omit every
second year from the administrative record?
If an irregular dating system pertained during Dyns. 4 and 5, what
economic and/or administrative necessity determined its irregularity? It
is possible, for instance, that during the formative period of the bureau-
cracy the frequency of the census was linked to the financing of large
projects—buildings, reclamation of land from the marshes, etc. Were
consecutive census years occasioned by funding shortfalls? Were factors
influencing the decision to organize the census annually or biennally
always the same or did they differ?
A special problem is the discrepancy between the available contem-
poraneous evidence and the reign lengths recorded of Dyns. 4 and 5
rulers in the TC. By contrast to the Manethonian tradition, the TC
was long considered by specialists to be the standard against which
contemporaneous data should be measured. Reign lengths given in the
papyrus were sometimes used as evidence for annual cattle counts and
at other times, for a biennial census. The compiler was presumed to
have omitted one or more signs from some figures and to have mis-
takenly duplicated entries. Obviously, comparison of data from the very
damaged papyrus with contemporaneous evidence can hardly be expected
to provide a definitive version of OK chronology. The names of only
three Dyn. 4 and 5 kings survive in the papyrus out of a presumed
17; three more partially-preserved names can be reconstructed. The

12
Posener-Kriéger, “Graffiti”; see also Spalinger, “Texts”, 318.
dynasties 4 to 5 127

remaining eleven are lost. Furthermore, the sequences at the end of


Dyn. 4 and in mid-Dyn. 5 are by no means certain. Nor does evi-
dence from contemporaneous documents always inspire confidence. For
example, there is only a single case where the precise date of a king’s
death and the accession of his successor are known, viz. for Sahure'
followed by Neferirkare'. Regardless, the exact length of Sahure'’s tenure
still cannot be established, since we do not know how regularly the
census was taken during his reign.
Disappointing as this may be, analysis of contemporaneous dates, both
rnpt zp and rnpt m-¢t zp, offers some stimulating insights. Assuming that
the census was irregular throughout Dyns. 4 and 5, the minimum length
of a given king’s reign should equal the total of the highest attested cen-
sus year (rnpt zp) and the number of attested intervening years (rnpt
m-¢t zp). A comparison of the results of such calculations with the data
recorded in the TC is represented in the table that follows:

Contemporaneous Evidence TC

Dyn. 4
Snofru 27 + x years 24 years
Cheops (Khufu 13 + x 23
Ra'djedef 11 (10?) + x 8
Khephren (Ra'kha'ef ) 15 + x ?
Bicheris not attested ?
Mycerinus (Menkaure') 14 (?) + x 18
Shepseskaf 2 + x 4
Thamphthis not attested 2
Dyn. 5
Userkaf 4 + x 7
Sahure' 8 + x 12
Neferirkare' 5 + x ?
Shepseskare' not attested 7
Ra'neferef 1 + x 1
Neuserre' 8(?) + x 11 + x
Menkauhor not attested 8
Djedkare' 28 (29 ?) + x 28
Wenis 9 + x 30

As can be seen, the estimate for the length of Snofru’s reign exceeds
the figure provided by the papyrus; the same is probably true for the
reigns of Ra'djedef and Djedkare'. On the other hand, the entry in
128 miroslav verner

the TC for Shepseskare', if correctly associated with that king, does not
reflect the contemporaneous record, since virtually no clear-cut evi-
dence for this mysterious ruler has yet surfaced.13 Such discrepancies
call into question the credibility of the papyrus for Dyns. 4 and 5.14

List of Contemporaneously Documented Dates, Dyns. 4–5

SNOFRU
TC III.9: 24 years
Manetho: Sóris—29 years
Beckerath, Chronologie, 159: 35 years (?)
rnpt zp
• 2 (n) Δnwt; Cairo frg no. 415
• 7 (n) Δnwt; Palermo Stone, recto 616
• 7, #bd III; Maidum pyramid17
• 8 (18 ?), #bd III “mw, sw 2; Maidum pyramid18
• 8 (n) Δnwt; Palermo Stone, recto 719
• 12, #bd IV “mw, sw 1 (?); Maidum pyramid20
• 13, . . . prt (?), sw 10; Maidum pyramid21
• 13 (16 ?), #bd I “mw, sw . . .; Maidum pyramid22
• 13, . . . “mw, sw . . .; Maidum pyramid23
• 13 (16 ?), . . . “mw, sw . . .; Maidum pyramid24
• 13; Maidum pyramid25

13
M. Verner, “Who was Shepseskare, and when did he reign?”, in: M. Bárta,
J. Krej‘í, eds., Abusir and Saqqara in the year 2000 (Prague, 2000), 581–602.
14
Verner, “Remarks”.
15
H. Gauthier, “Quatre nouveaux fragments de la pierre de Palerme”, in: G. Maspero,
Le Musée égyptien III (Cairo, 1915), 50–52.
16
Schäfer, “Annalen”, 30.
17
Posener-Kriéger, “Graffiti”, 20, pl. 8, A.20.
18
Posener-Kriéger, “Graffiti”, pl. 8, A.27.
19
Schäfer, “Annalen”, 31.
20
Posener-Kriéger, “Graffiti”, pl. 8, A.33.
21
Posener-Kriéger, “Graffiti”, pl. 7, A.1.
22
Posener-Kriéger, “Graffiti”, pl. 7, A.11.
23
Posener-Kriéger, “Graffiti”, pl. 7, A.5.
24
Posener-Kriéger, “Graffiti”, pl. 7, A.9.
25
Posener-Kriéger, “Graffiti”, pl. 7, A.2.
dynasties 4 to 5 129

• 13; Maidum pyramid26


• 14 (17 ?), #bd II “mw, sw . . .; Maidum pyramid27
• 14 (17 ?), . . . prt; Maidum pyramid28
• 14 (17 ?), #bd I + x; Maidum pyramid29
• 15, #bd II prt, sw 14; Dahshur, Red Pyramid30
• 15 (?), #bd III prt; Maidum pyramid31
• 15, #bd III “mw, sw 10 + x; Maidum pyramid32
• 15 (?), #bd IV “mw (?), sw 10; Maidum pyramid33
• 15 (?), . . . “mw (?), sw . . .; Maidum pyramid34
• 15; Dahshur, Red Pyramid35
• 16, #bd I #¢t, sw 13; Dahshur, quarry mark36
• 16, #bd III #¢t; Dahshur, Red Pyramid37
• 16, #bd IV #¢t, sw 14; Maidum pyramid38
• 16, #bd II (?) “mw, sw 12; Maidum pyramid39
• 16 (?), #bd . . . prt, sw 2; Dahshur, Red Pyramid40
• 16 (?), #bd I prt; Maidum pyramid41
• 16 (?), #bd III prt; Maidum pyramid42
• 16; Maidum pyramid43
• 17, #bd II prt, sw 10 + x; Maidum pyramid44
• 17, #bd I prt, sw 20; Maidum pyramid45

26
Posener-Kriéger, “Graffiti”, pl. 8, A.31.
27
Posener-Kriéger, “Graffiti”, pl. 8, A.23.
28
Posener-Kriéger, “Graffiti”, pl. 8, A.24.
29
Posener-Kriéger, “Graffiti”, pl. 8, A.26.
30
R. Stadelmann (n. 2), 234–235, fig. 2.
31
Posener-Kriéger, “Graffiti”, pl. 7, A.4.
32
W. M. F. Petrie, E. J. H. Mackay & G. A. Wainwright, Meydum and Memphis III
(London, 1910), 9, pl. V, 6.
33
Posener-Kriéger, “Graffiti”, pl. 7, A.6.
34
Posener-Kriéger, “Graffiti”, pl. 7, A.9.
35
Stadelmann (n. 2), 233–5, fig. 1.
36
LD II, I g.
37
Stadelman (n. 2), 234–235, fig. 2.
38
A. Rowe, The Museum Journal 22 (1931), 26, pl. 38, fig. 2.
39
Posener-Kriéger, “Graffiti”, pl. 7, A.3.
40
H. Sourouzian, MDAIK 38 (1982), 389–390, fig. 5.
41
Rowe (n. 38), 26.
42
Posener-Kriéger, “Graffiti”, pl. 8, A.22.
43
Posener-Kriéger, “Graffiti”, pl. 7, A.7.
44
Petrie et al. (n. 32), 9, pl. 5, 2 left.
45
Posener-Kriéger, “Graffiti”, pl. 7, A.13.
130 miroslav verner

• 17, #bd I prt, sw 22; Maidum pyramid46


• 17, #bd III prt, sw 'rq; Maidum pyramid47
• 17, #bd III prt, sw . . .; Maidum pyramid48
• 17, #bd III #¢t, . . .; Maidum pyramid49
• 17, . . . prt; Maidum pyramid50
• 17, (#bd) I + x prt; Maidum pyramid51
• 17, (#bd) I + x; Maidum pyramid52
• 17, #bd . . .; Maidum pyramid53
• 17; Maidum pyramid54
• 18, #bd I prt, sw 21; Maidum pyramid55
• 23, #bd II “mw; Maidum pyramid56
• 24, #bd II #¢t, . . . (?); Dahshur, Red Pyramid57
• 24, #bd . . . prt, . . .; Dahshur, Red Pyramid58
Damaged evidence of rnpt zp
• 10 + x, #bd IV “mw; Maidum pyramid59
• 10 + x; Maidum pyramid60
• (1)6 (?), #bd I + x, sw 12; Maidum pyramid61
• (1)6 (?), #bd I + x, “mw (?), sw 2; Maidum pyramid62
• (1)7 (?), IV “mw, sw 21; Maidum pyramid63
• . . ., #bd IV . . ., sw . . .; Maidum pyramid64

46
Posener-Kriéger, “Graffiti”, pl. 7, A.14.
47
Petrie et al. (n. 32), 9, pl. 5,4.
48
Petrie et al. (n. 32), 9, pl. 5, 3.
49
Posener-Kriéger, “Graffiti”, pl. 7, A.12.
50
Posener-Kriéger, “Graffiti”, pl. 7, A.16.
51
Posener-Kriéger, “Graffiti”, pl. 7, A.17.
52
Posener-Kriéger, “Graffiti”, pl. 7, A.18.
53
Posener-Kriéger, “Graffiti”, pl. 8, A.21.
54
Posener-Kriéger, “Graffiti”, pl. 7, A.19.
55
Posener-Kriéger, “Graffiti”, pl. 8, A.29.
56
Posener-Kriéger, “Graffiti”, pl. 9, A.42.
57
Reconstruction of a mason’s mark in LD Text I, 206 by Stadelmann (n. 2), 234–236,
fig. 3.
58
Stadelmann (n. 2), 239–240, fig. 4.
59
Posener-Kriéger, “Graffiti”, pl. 8, A.34.
60
Posener-Kriéger, “Graffiti”, pl. 8, A.35.
61
Posener-Kriéger, “Graffiti”, pl. 7, A.8.
62
Posener-Kriéger, “Graffiti”, pl. 7, A.10.
63
Posener-Kriéger, “Graffiti”, pl. 7, A.15.
64
Posener-Kriéger, “Graffiti”, pl. 9, A.36.
dynasties 4 to 5 131

• . . ., #bd III “mw, sw . . .; Maidum pyramid65


• . . ., #bd III “mw, . . . 10 + x; Maidum pyramid66
rnpt (m)- xt zp67
• 10, #bd 1 + x; Maidum pyramid68
• 13, . . .; Maidum pyramid69
• 18, #bd IV “mw, sw (?) 5; Maidum pyramid70
damaged evidence of rnpt (m-)¢t zp
• . . .; Maidum pyramid71
attested rnpt zp: 2; 7; 8; 12; 13; 14; 15; 16; 17; 18; 23; 24
attested rnpt (m-)¢t zp: 10; 13; 18
rnpt zp: rnpt (m-)¢t zp—12: 3
CHEOPS (KHUFU)
TC III.10 (?): 23 years
Manetho: Súphis (I)—63 years
Beckerath, Chronologie, 159: 23 years
rnpt zp
• 4 (?), #bd . . .; G 2130, Khentika72
• 5, . . . “mw (?), sw 5; G 120373
• 8, #bd I prt, . . . (?); a loose (?) block found at the upper end of the
causeway, near the entrance to the king’s mortuary temple74
• 8, #bd III “mw, sw 20; G 4000, Hemiunu75
• 10, #bd IV prt, sw 23 (or 24); G 4000, Hemiunu76

65
Posener-Kriéger, “Graffiti”, pl. 9, A.37.
66
Posener-Kriéger, “Graffiti”, pl. 9, A.41.
67
For the transcription of m-¢t see Edel, Grammatik, 180.
68
Posener-Kriéger, “Graffiti”, pl. 8, A.30.
69
Posener-Kriéger, “Graffiti”, pl. 8, A.32.
70
Posener-Kriéger, “Graffiti”, pl. 8, A.28.
71
Posener-Kriéger, “Graffiti”, pl. 9, A.38.
72
Attributed to Khufu by Smith, “Evidence”, 118 fig. 6; 127 no. 4; so also Y. Harpur,
Decoration in Egyptian Tombs of the Old Kingdom (London, 1987), 269.
73
Attributed to Khufu by Smith, “Evidence”, 118, fig. 6; 127 no. 2, and by G. A.
Reisner, A History of the Giza Necropolis I (Cambridge, Mass., 1942), 76 n. 2 and 391/19/.
74
Attributed to Khufu by Smith, “Evidence”, 119 fig. 7; 126f. no. 1; originally,
A. Rowe read this date “year 13”, see Reisner (n. 73), 71.
75
Attributed to Khufu by Junker, Giza I, 159, fig. 24/10/, 161.
76
Junker, Giza I, 161, no. 12.
132 miroslav verner

• 10, #bd I “mw, sw 10 + x; G 4000, Hemiunu77


• 10, #bd II “mw, sw 10 + x; G 4000, Hemiunu78
• 12, #bd II “mw; G 2120, Seshatsekhentiu79
• 12, #bd II . . .; G 7130–40, Khufukhaf I80
rnpt (m-)¢t zp
• 1381
attested rnpt zp: 4, 5, 8, 10, 12
rnpt (m-)¢t zp: 13
rnpt zp: rnpt (m-)¢t zp—5: 1
RA'DJEDEF
TC III.11 (?): 8 years
Beckerath, Chronologie, 159: 9 years
rnpt zp
• 1, #bd III prt, . . .; pyramid of Ra'djedef 82
• 11 (or 10 ?), #bd I pr(t), sw 24 (?); boat pit, south side of Khepos’s
pyramid83
attested rnpt zp: 1, 11 (10 ?)
rnpt (m-)¢t zp: not attested
rnpt zp: rnpt (m-)¢t zp: 2: 0 (?)

77
Junker, Giza I, 158, 160; 159 fig. 24/1/.
78
Junker, Giza I, 159 fig. 24/2/; 160.
79
Attributed to Khufu by Smith, “Evidence”, 118, fig. 6; 127 no. 3, and by Spalinger,
“Texts”, 285; according to N. Strudwick, The Administration of Egypt in the Old Kingdom
(London, 1985), 117 no. 6, the reign of Khephren is also possible.
80
Attributed to Khufu by Smith, “Evidence”, 119, fig. 7; 127 no. 8, and by W. K.
Simpson, The Mastabas of Kawab, Khafkhufu I and II (Boston, 1978), 9. This dating also
accords with Stadelmann’s theory that Khufukhaf I assumed the name Khephren after
succeeding Ra'djedef, see SAK 11 (1985), 165–172.
81
K. P. Kuhlmann, in: Tides of the Desert—Gezeiten der Wüste. Contributions to the
Archaeology and Environmental History of Africa in Honmour of Rudolph Kuper (Köln: Africa
Praehistorica 14, 2002), 125–138.
82
M. Vallogia, in: Études sur l’Ancien Empire et la nécropole de Saqqara dédiés à Jean-
Philippe Lauer (Montpellier, 1997), 419.
83
According to I. E. S. Edwards, in: The Unbroken Reed: Studies in the Culture and
Heritage of Ancient Egypt in honour of A. F. Shore ( London, 1994), 101, 105 n. 20, Posener-
Kriéger read “year 10” (i.e. “year of the 10th cattle count”); see also R. Krauss,
Orientalia 66 (1997), 3 n. 16. Both readings are possible; however, the reading of the
left column with the date is not doubt-free, and the right column of the graffito is
almost illegible, see A. M. Abubakr & A. Y. Mustafa, The Funerary Boat of Khufu, in:
BÄBA 12 (1971), 11, fig. 6 bottom left.
dynasties 4 to 5 133

KHEPHREN (RA'KHAEF)
TC III.12: 20 + x years
Manetho: Suphis (II) 66 years
Beckerath, Chronologie, 159: 26 years
rnpt zp
• 1, #bd IV #¢t, sw 5; ostracon from Helwan tomb 299 H 284
• 5, #bd III prt, sw 22; ostracon from Helwan tomb 335 H 285
• 7, #bd IV prt, sw 10; G 7530–40, Meresankh III86
• 7, #bd IV prt, sw 20; G 7530–40, Meresankh III87
• 10, #bd III “mw, sw 24; ostracon Leiden J 42988
• 10 (?), #bd III “mw, . . .; G 7350, Hetepheres II (?)89
• 12, #bd II “mw, sw 10; G 7650, Akhtihotep and his wife Meretites90

84
Z. Saad, Royal Excavations at Saqqara and Helwan (1941–1945), Suppl. ASAE no. 3
(Cairo, 1947), 106, pl. 42 a left; as Spalinger notes, “Texts”, 287, Khephren’s car-
touche in the inscription clinches its assignment to his reign.
85
Saad (n. 84), 107, pl. 43 at right; probably temp. of Khephren.
86
Attributed to Khephren by D. Dunham & W. K. Simpson, The Mastaba of Mersyankh
III (Boston, 1974), 3 fig. 1; see also Smith, “Evidence”, 127 no. 9, 119 fig. 7.
87
Attributed to Khephren by Dunham & Simpson (n. 86), 3 fig. 1; see also Smith,
“Evidence”, 127 no. 9, 119 fig. 7. Anthropological examination of Meresankh III’s
bones put her age at death at about fifty, see V. G. Callender, Egypt in the Old Kingdom
(Melbourne, 1998), 172–173.
88
According to H. Goedicke, JEA 54 (1968), 24, 28–29, pl. 5 no. 4, the ostracon
is of the same date as those of Helwan; thus rnpt zp 10 probably refers to Khephren.
Cf. also, idem, Old Hieratic Palaeography (Baltimore, 1988), pl. 16.
89
According to Reisner (n. 73), 73 n. 2; see also Smith, “Evidence”, 119 fig. 7, 127
no. 9. The date was inscribed on the back of a block from G 7350, purportedly built
for Hetepheres II. However, Smith identified the figures ín the relief on the front as
Hetepheres II (?) and Meresankh III (?), and he dated it to the time of Shepseskaf
(HESPOK, 1946, 164–165, 302, pl. 45 a). The attribution of G 7350 to Hetepheres II
is not based on textual evidence. P. Jánosi, ZÄS 123 (1996), 56–57, has questioned
Smith’s conclusions. He suggests that the relief might have originally shown an anony-
mous prince followed by his mother and his wife. No doubt the attribution of the date
is fraught with difficulties, regardless of the fact that a high date and Shepseskaf ’s
reign are mutually exclusive, which leaves either Khephren or Menkaure'. Since
Meresankh III was probably buried in the tomb intended originally for her mother at
the beginning of Menkaure'’s reign (see the dates rnpt zp 1 and rnpt (m-)¢t zp 1 sub
Menkaure'), it would be surprising to find her represented with her mother in a tomb
built as late as Menkaure'’s reign. When considering the data from G 7350 and G
7530–40 and with due circumspection in view of the complex stratigraphy and unclear
chronology of cemetery G 7000, one is inclined to assign the date to Khephren.
90
Attributed to Khephren by Smith, “Evidence”, 119 fig. 7, 127–128 no. 11 b; see
also Reisner (n. 73), 73 n. 1. For doubts about the attribution to Khephren, and a
still higher date of rnpt zp 13 (see below) from the tomb of Akhtihotep, see P. Jánosi,
Giza in der 4. Dynastie. (Wien 2005), 71–73, 443.
134 miroslav verner

• 12 Tnwt . . .; LG 87, Nikaure91


• 13, #bd IV . . .; G 7650, Akhtihotep and his wife Meretites92
rnpt (m-)¢t zp
• 4 Tnwt, #bd II “mw, sw 3; ostracon from Helwan tomb 305 H 293
• 4 Tnwt, #bd II “mw, sw 4; ostracon from Helwan tomb 305 H 294
• 5, #bd II “mw, sw 8; ostracon from Helwan tomb 322 H 295
• 5, #bd III prt, sw 22; ostracon from Helwan tomb 335 H 296
attested rnpt zp: 1, 5, 7, 10, 12, 13
attested rnpt (m-)¢t zp: 4, 5
rnpt zp: rnpt (m-)¢t zp—6: 2
BICHERIS
TC III.13 (?): . . .?. . . years
Manetho: 22 years (Bicheris, preceded by Ratoises, was inserted by
Manetho between Mycerinus and Shepseskaf.)
Beckerath, Chronologie, 159: 7 years

MYCERINUS (MENKAURE')
TC III.14 (?): 18 (28 ?) years
Manetho: Mencherés—63 years
Beckerath, Chronologie, 159: 28 years.

91
Urk. I, 16.14; see also H. Goedicke, Die privaten Rechtsinschriften aus dem Alten Reich
(Wien, 1970), 21–22. There is no unanimity on the dating of the will of Khephren’s
son Nikaura. For instance, Strudwick (n. 79; 107) concluded that Nikaura should have
been born in Khephren’s reign, “and thus would be no older than twenty-two at the
end of his father’s reign”. Consequently, rnpt zp 12 should then apply to Khephren’s
successor Menkaure'. With reference to art historical criteria and the replacement of
Tnwt by ipt in the date (the former being supposed by Goedicke, Rechtsinschriften, 22,
to have disappeared by the beginning of Dyn. 5), Spalinger (“Texts”, 294) opts for
Menkaure', too. But Baud, Ménès, 128, argues that Khephren’s name occurs in Nikaura’s
tomb with such an insistence that the date should refer to this king. Jánosi (n. 90),
too, does not exclude the attribution of the date to Khephren, provided that Nikaura
was born before his father ascended the throne. Taking all the arguments in account,
including the possibility that Khephren might have become king later in his life, one
is inclined to assign this date to him.
92
Attributed to Khephren by Smith, “Evidence”, 119 fig. 7, 128 no. 11.
93
Saad (n. 84), 106–107, pl. 42 b right. For the translation and interpretation, see
H. G. Fischer, Orientalia 29 (1960), 187–190; Spalinger, “Texts”, 287.
94
Saad (n. 84), 106–107, pl. 42 b left. For translation and the reference of the ostra-
con, see Fischer (n. 93), 187–90; see also Spalinger, “Texts”, 287.
95
Saad (n. 84), 107, pl. 43 a right; see Spalinger, “Texts”, 288.
96
Saad (n. 84), 106–107, pl. 43 b right; see Spalinger, “Texts”.
dynasties 4 to 5 135

rnpt zp
• 2, #bd II prt, sw 22; G VI S97
• 2, #bd IV “mw, sw 22; G 7530–40, Meresankh III98
• 11, #bd . . ., sw 10 + x; G VI S99
The following dates from the Gebelein papyri can probably be attrib-
uted to Mycerinus:100
rnpt (m-)¢t zp
• 2, #bd . . . #¢t, sw 20; Gebelein, frag. A101
• 3, #bd III prt, sw 26; Gebelein, frag. B102
• 11, . . .; Gebelein, rouleau IV103
attested rnpt zp: 2, 11
attested rnpt (m-)¢t zp: 2(?), 3(?), 11(?)
rnpt zp: rnpt (m-)¢t zp—2: 3(?)
SHEPSESKAF
TC III.15 (?): 4 years
Manetho: Sebercherés—7 years
Beckerath, Chronologie, 159: 5 years

97
Junker, Giza X, 75, fig. 35.9, 78, no. 10. With regard to the occurrence of this
date on some blocks at the site, Junker attributed the date to Menkaure' because of
the presence of his crew names.
98
Dunham & Simpson, (n. 86), 3, fig. 1 e; see also Smith, “Evidence”, 116 fig. 4,
126 no. 2. The date is inscribed to the north of the subsidiary niche, eastern façade,
of the mastaba (from the date on the south side, only . . . prt, sw 17 survived). Reisner
attributed the date to Khephren, see Smith, “Evidence”; Spalinger, “Texts”, 286, accepts
this dating, though with some hesitation. However, Reisner’s dating can be seriously
questioned. If the tomb was built around the 7th census of Khephren, as indicated by
two masons’ inscriptions (Simpson & Dunham (n. 86), 3, Fig. 1 b, c), the lower date
found on the mastaba’s façade can hardly be earlier. But to which event did the date
refer? The attribution of the date to Menkaure' seems to be, therefore, more proba-
ble. For the complex, and the problem of its history, see Jánosi (n. 90), 500 and idem,
ZÄS 123 (1996), 46–62.
99
Junker, Giza X, 75 fig. 35.10, 77 no. 9; because Menkaure'’s crew names were
found on some blocks at the site, Junker attributed the date to his reign.
100
P. Posener-Kriéger, RdÉ 27 (1975), 216 and idem in: Fs Elmar Edel: 12. März 1979
(Bamberg, 1979), 318–331.
101
Posener-Kriéger (n. 100; 1979), 318–331.
102
Posener-Kriéger, RdE 27 (1975), 215–216.
103
Posener-Kriéger, RdE 27 (1975), 215–216.
136 miroslav verner

rnpt zm# t #wy


• Palermo Stone104
• #bd II “mw, sw 10;105
• #bd II “mw, sw 10; G 5552106
• #bd III “mw, sw . . .; G 7450107
• #bd IV “mw, sw 4; G 7450108
rnpt zp
• 1, #bd 1 “mw, sw 21 (G 753040, Meresankh III)109
rnpt (m-)¢t zp
• tpy, #bd II prt, sw 28110
• 1 (n) ipt (!) ô˙ 'wt nbt; edict of Shepseskaf for the pyramid of Mycerinus111
attested rnpt zp: 1
attested rnpt (m-)¢t zp: 1
rnpt zp: rnpt (m-)¢t zp—1: 1
THAMPHTHIS
TC III, 16 (?): 2 years
Manetho: Thamphthis—9 years
Beckerath, Chronologie, 159: 2 years
USERKAF
TC III.17: 7 years
Manetho: Usercherés—7 years
Beckerath, Chronologie, 159: 8 years

104
Schäfer, “Annalen”, 32–33.
105
Attributed to Shepseskaf by Helck, in: Fs Goedicke, 107.
106
Attributed to Shepseskaf by Helck (n. 105), 107–108.
107
Attributed to Shepseskaf by Helck (n. 105), 107–108.
108
The date to the right of the entrance to Meresankh III’s chapel refers to the
preparation of the queen’s burial. Reisner attributed it to Shepseskaf, see Smith, “Evi-
dence”, 126, fig. 4, 118; Jánosi (n. 90), 501 concurs. But Dunham & Simpson (n. 86),
8, pl. 2 a, fig. 2, and also Spalinger, “Texts”, 288–289, assign it to Menkaure'.
109
The date, inscribed on the left side of the entrance to Meresankh III’s chapel
and referring to the completion of the queen’s burial, was attributed by Reisner to
Shepseskaf, see Smith, “Evidence”, 126, fig. 4 on p. 118. Jánosi (n. 90), 501, attrib-
utes the date to Shepseskaf, too. On the other hand, Dunham & Simpson (n. 86), 8,
pl. 2 a, and also Spalinger, “Texts”, 289, attribute the date to Menkaure'.
110
G. A. Reisner, Mycerinus: The temples of the third Dynasty at Giza (Cambridge, Mass.,
1931), 278 no. 1; see also H. Goedicke, Königliche Dokumente aus dem Alten Reich (Wiesbaden:
ÄA 4, 1967), 16–17, fig. 1 and P. Jánosi, GM 141 (1994), 49–54.
111
Schäfer, “Annalen”, 34.
dynasties 4 to 5 137

rnpt zp
• 3; Palermo Stone, verso 2112
• 3, #bd III prt, sw . . .; sun temple of Userkaf 113
rnpt (m)-xt zp
• 1 (n) Δnwt; Cairo frag. no. 1 recto 2114
attested rnpt zp: 3
attested rnpt (m)-xt zp: 1
rnpt zp: rnpt (m)-xt zp, 1: 1
SAHURE'
TC III 18 (?): 12 years
Manetho: Sephrés—13 years
Beckerath, Chronologie, 155: 13 years
rnpt zp
• 1; Cairo Frg. no. 1 verso 2115
• 2, #bd I “mw, sw 20; mason’s inscription, mortuary temple of Sahure'116
• 4, #bd IV #¢t, sw 12; masons’ inscription, mortuary temple of Sahure'117
• 5, #bd I #¢t; sun temple of Userkaf, tablet A118
• 5, #bd III prt; sun temple of Userkaf, tablet B119
• 5, #bd III “mw; sun temple of Userkaf, tablet C120
rnpt (m)-xt zp
• 2 (Palermo Stone, verso 3)121
• 5, #bd II prt; sun temple of Userkaf, tablet D122
• 6; Palermo Stone, verso 4123

112
Schäfer, “Annalen”, 34.
113
Haeny, in: BÄBA 8 (1969), 41–42 no. 6.
114
Gauthier (n. 15), 45–46, pl. 26.
115
Gauthier (n. 15), 47.
116
L. Borchardt, Das Grabdenkmal des Königs Sahure I, (Leipzig, 1910), 88, M 26.
117
Borchardt (n. 116), 89, M 29.
118
Probably to be attributed to Sahure', see Verner, “Remarks”, 386–390.
119
Probably to be attributed to Sahure', see Verner, “Remarks”, 386–390.
120
Probably to be attributed to Sahure', “Remarks”, 386–390.
121
Schäfer, “Annalen”, 36–37.
122
Probably to be attributed to Sahure', see Verner, “Remarks”, 386–390.
123
Schäfer, “Annalen”, 38–39. To date, this is the highest contemporaneous date
attested for Sahure'’s reign. A. Roccati, La littérature historique sous l’Ancien Empire (Paris,
1982), 48, for instance, read the worn signs as 7 and, provided the census was bien-
nial, evidence for year 15 of Sahure'’s reign. However, as pointed out by Wilkinson
(Annals, 168), 6 is a more likely reading than 7; in case of a biennial census, it would
refer to year 13.
138 miroslav verner

attested rnpt zp: 1, 2, 4, 5


attested rnpt (m)-xt zp: 2, 5, 6
rnpt zp: rnpt (m)-xt zp, 4: 3
NEFERIRKARE'
TC III.19 (?): length of reign lost
Manetho: Nefercherés—20 years
Beckerath, Chronologie, 155: 20 years
rnpt zm# t#wy; Palermo Stone verso 4124
rnpt zp
• 5; Palermo Stone verso 5125
• 5, #bd IV #¢t, sw 4; mason’s inscription, pyramid of Khentkaus II126
• 5, #bd IV; mason’s inscription, pyramid of Neferirkare'127
attested rnpt zp: 5
rnpt m-¢t zp: not attested
rnpt zp: rnpt m-¢t zp—1: 0
SHEPSESKARE'
TC III.20 (?): 7 years
Manetho: Sisirés—7 years
Beckerath, Chronologie, 155: 7 years
RA'NEFEREF
TC III.21 (?): 1 year
Manetho: Cherés—20 years
Beckerath, Chronologie, 155: 11 years
rnpt zp
• tpy, #bd IV #¢t sw 4 + x; pyramid of Ra'neferef 128
attested rnpt zp: 1
rnpt (m-)¢t zp: not attested
rnpt zp: rnpt (m-)¢t zp—1: 0

124
Schäfer, “Annalen”, 39.
125
Schäfer, “Annalen”, 40.
126
Probably to be attributed to Neferirkare', see Verner, ZÄS 107 (1980), 159,
fig. 3; idem The Pyramid Complex of Khentkaus (Prague, 1995), 43–45.
127
L. Borchardt, Das Grabdenkmal des Königs Nefer-ir-ke #-re' (Leipzig, 1909), 46 and n. 6.
128
Corrected copy: Verner, ZÄS 126 (1999), 76, fig. 6.
dynasties 4 to 5 139

NEUSERRE'
TC III.22 (?): 11 (+ x years ?)
Manetho: Rathurés—44 years
Beckerath, Chronologie, 155: 31 years; cf. ibidem 208, where the figure
30—or 20 ? + 1 or 5 ?—is given
rnpt zm3 (t3wy)
• unpublished potsherd no. 763/I/84–x, mortuary temple of Ra'neferef
rnpt zp
• 1, #bd I #¢t, sw 10 + x; unpublished potsherd no. 531/I/82, mortu-
ary temple of Ra'neferef
• 1, #bd tpy #¢t, . . .; unpublished potsherd no. 763/I/84–e, mortuary
temple of Ra'neferef
• 2, #bd III “mw, sw 10; a potsherd found by Borchardt129 in (or to the
west of ?) the mastaba of Djadjamankh in Abusir
• 5 (?), #bd III prt, (sw) wpw; masons’ inscription, mastaba of Ptahshepses130
• 7, #bd III #¢t, sw 1 (or 7 ?); jar for beef fat no. 531/I/82, mortuary
temple of Ra'neferef
rnpt (m-)¢t zp
• 2, #bd 3 #¢t, sw 24; masonry block, found in the southern “Eckbau”
of the mortuary temple of Neuserre'131
attested rnpt zp: 1, 2, 5(?), 7
attested rnpt (m-)¢t zp: 2
rnpt zp: rnpt (m-)¢t zp—4(?): 1
MENKAUHOR
TC III.23: 8 years
Manetho: Mencherés—9 years
Beckerath, Chronologie, 155: 9 years
DJEDKARE'
TC III.24: 28 years
Manetho: Tancherés—44 years
Beckerath, Chronologie, 155: 38 years

129
L. Borchardt, Das Grabdenkmal des Königs Ne-user-r' (Leipzig, 1907), 139.
130
To be attributed to Neuserre': M. Verner, Baugraffiti der Ptahschepses Mastaba (Prague,
1992), 110, graffito no. 194.
131
Borchardt (n. 129), 145.
140 miroslav verner

rnpt zm# t#wy


• #bd III prt, sw 29; tomb of Wepemneferet132
rnpt zp
• 1, #bd II #¢t, . . .; unp. pap., (mortuary temple) archive of Ra'neferef,
pl. 51 A
• rnpt tpy (sic), . . .; unp. pap., archive of Ra'neferef, pl. 76 A
• 3, #bd IV #¢t, sw 25; papyrus archive, temple of Neferirkare'133
• 5, #bd IV #¢t, . . . unp. pap., archive of Ra'neferef, pl. 76 C
• 6, #bd IV prt, sw 22; wooden box for linen found in the tomb of Nefer
and Kahay134
• 8, #bd IV “mw; papyrus archive, mortuary temple of Neferirkare'135
• 9; rock stela, Sinai136
• 10, #bd IV . . ., sw 24; papyrus archive, mortuary temple of Neferirkare'137
• 11, #bd II #¢t, sw 11; papyrus archive, mortuary temple of Neferirkare'138
• 14, #bd tpy “mw (#bd II “mw); papyrus archive, mortuary temple of
Neferirkare'139

132
The date, mentioned in Wepemneferet’s will on a wall of his tomb (S. Hassan,
Giza II, Cairo, 1936, fig. 219), was attributed by Spalinger (“Texts”, 302, with a ref-
erence to K. Baer, Rank and Title in the Old Kingdom, Chicago, 1960, 66) to Wenis.
However, the persons mentioned in the tomb include a craftsman named Ra'neferef-
ankh. If born in the reign of Ra'neferef, which seems probable, he could have been
about 30 to 40 years old at the beginning of Djedkare'’s reign. If this date be ascribed
to Wenis, he would have been very old at the time of that king’s acession. The date
should, therefore, refer to Djedkare' rather than Wenis.
133
Posener-Kriéger & de Cenival, Abusir Papyri, pls. 13, 13 A; Posener-Kriéger, Archives
II, 480.
134
H. Altenmüller & A. Moussa, The Tomb of Nefer and Kahay (Mainz: AV 5, 1971),
18, 43–44, fig. 11. The excavators assigned the date to Djedkare'. However, Spalinger,
“Texts”, 302 suggested either Ra'neferef or Neuserre'. Surely Ra'neferef can be excluded
(see above sub Ra'neferef ). The tomb seems to have been built in the time of Neuserre',
as the excavators surmised; see also, e.g., N. Cherpion, Mastabas et Hypogées de l’ancien
Égypte (Bruxelles, 1989), 135. However, as pointed out by Altenmüller & Moussa, the
burial in shaft no. 8, where the box with the date was found, was the last one made
in the tomb and should be contemporary with Nefer’s children. The dating of the
inscription to the time of Djedkare is, therefore, very plausible.
135
Posener-Kriéger, de Cenival, Abusir Papyri, pls. 69, 69 A; Posener-Kriéger, Archives
II, 490.
136
A. H. Gardiner, T. E. Peet & J. Cernÿ, The Inscriptions of Sinai I (London, 1952),
pl. VIII, no. 14; II, 61.
137
Posener-Kriéger & de Cenival, Abusir Papyri, pls. 72, 72 A; Posener-Kriéger, Archives
II, 490.
138
Posener-Kriéger & de Cenival, Abusir Papyri, pls. 53, 53 A; Posener-Kriéger, Archives
II, 490.
139
Posener-Kriéger & de Cenival, Abusir Papyri, pls. 2, 2 A; Posener-Kriéger, Archives
II, 490.
dynasties 4 to 5 141

• 15, #bd IV prt, (sw) wpw; papyrus archive, mortuary temple of Nefe-
rirkare'140
• 15, #bd IV #¢t, sw 27; unp. pap., archive of Ra'neferef, pl. 20 B
• 15, #bd IV #¢t, sw 28; unp. pap., archive of Ra'neferef 141
• 15, #bd I, . . .; unp. pap., archive of Ra'neferef, pl. 21 L
• 15 (n) T(nwt); unp. pap., archive of Ra'neferef, pl. 3 A
• 16; papyrus archive, mortuary temple of Neferirkare'142
• [1]6, #bd IV “mw, sw 28143
• 17, #bd III; unp. pap., archive of Ra'neferef, pl. 8 D
• 18, #bd III “mw, sw 'rq; unp. pap., archive of Ra'neferef, pl. 45
• 18, #bd IV “mw, sw 'rq; unp. pap., archive of Ra'neferef, pl. 63 A
• 21 (22 ?), #bd IV #¢t, sw 12; papyrus archive, mortuary temple of
Neferirkare'144
damaged evidence of rnpt zp
• 2 + x, #bd I . . ., sw . . .; unp. pap., archive of Ra'neferef, pl. 76 B
• 10 + x; unp. pap., archive of Ra'neferef, pl. 85 C
• 11 + x Tnw(t) ô˙ ( 'wt nb); unp. pap., archive of Ra'neferef, pl. 1 A.
rnpt (m-)¢t zp
• 1, #bd IV #¢t, sw 'rq; unp. pap., archive of Ra'neferef, pl. 77 A
• 1, #bd IV “mw, sw 1; unp. pap., archive of Ra'neferef, pl. 77 B
• tpy; unp. pap., archive of Ra'neferef, pl. 82 0
• 3; rock stela, Sinai.145
• 4, #bd III “mw, sw 15; unp. pap., archive of Ra'neferef, pl. 76 C.
• 4, #bd III “mw; unp. pap., archive of Ra'neferef, pl. 69 A
• 7 (?), #bd I #¢t; unp. pap., archive of Ra'neferef, pl. 76 D
• 10, #bd IV “mw, sw 21; papyrus archive, mortuary temple of Neferir-
kare'146

140
Posener-Kriéger & de Cenival, Abusir Papyri, pls. 47, 47 A; Posener-Kriéger, Archives
II, 490.
141
Posener-Kriéger, in: Mélanges Mokhtar II (Cairo, 1985), 195–210.
142
Posener-Kriéger & de Cenival, Abusir Papyri, pls. 1, 1 A; Posener-Kriéger, Archives
II, 490.
143
Urk I, 63.11; W. S. Smith, “Evidence”, 113 n. 2; see also E. Eichler, SAK 18
(1991), 146–147 (letter of Izezi to Senedjemib).
144
Posener-Kriéger & de Cenival, Abusir Papyri, pls. 41, 41 A; Posener-Kriéger, Archives
II, 490. Posener-Kriéger transcribed the numeral following rnpt zp as 21. There is,
however, the trace of another vertical stroke which allows the reconstruction ‘22’.
145
Gardiner et al. (n. 136), I, pl. VII, no. 13; II, 60.
146
Posener-Kriéger & de Cenival, Abusir Papyri, pls. 14, 14 A; Posener-Kriéger, Archives
II, 490.
142 miroslav verner

• rnpt zp (m)-xt zp (sic) 14, #bd I #¢t, sw 28; unp. pap., archive of Ra'ne-
feref, pl. 66 B
• 14, #bd . . . #¢t; unp. pap., archive of Ra'neferef, pl. 76 J
• 17, #bd I “mw (?), sw 23; sarcophagus of Idu, Abusir147
damaged evidence of rnpt (m)-xt zp
• 4 + x, #bd I #¢t; unp. pap., mortuary temple archive of Ra'neferef,
pl. 77 I
• 10 + x, . . .; unp. pap., mortuary temple archive of Ra'neferef, pl.
76 E
documents dating from the time of either Djedkare' or Wenis:
• rnpt zp 4, #bd IV prt, sw 2; papyrus archive, mortuary temple of
Neferirkare'148
• rnpt zp 4, #bd I “mw, sw 'rq; papyrus archive, mortuary temple of
Neferirkare'149
• rnpt zp 11, #bd III prt, sw 3 (?); mason’s inscription, tomb of Rawer
II150
attested rnpt zp: 1, 3, 4(?), 5, 6, 8, 9, 10, 11, 14, 15, 16, 17, 18, 21
(22?)
attested rnpt (m)-xt zp: 1, 3, 4, 7 (?), 10, 14, 17
rnpt zp: rnpt (m)-xt zp, 15(?): 7(?)
WENIS
TC III.25: 30 years
Manetho: Onnos—33 years
Beckerath, Chronologie, 155: 20 years

147
M. Verner, SAK 8 (1980), 258–260, pl. 16. A new examination of the badly dam-
aged inscription showed that the date should be read rnpt (m-)¢t zp 17 rather than rnpt
zp 14, as suggested shortly after the discovery of the tomb.
148
Posener-Kriéger & de Cenival, Abusir Papyri, pls. 11, 11 A; Posener-Kriéger, Archives
II, 491, was hesitant about the attribution of this date to either Djedkare' or Wenis.
149
Posener-Kriéger & de Cenival, Abusir Papyri, pls. 11, 11 A; Posener-Kriéger, Archives
II, 491 was hesitant about the attribution of this date to either Djedkare' or Wenis.
150
Junker, Giza III, 223–235; idem Giza VIII, 39f. Though a sealing bearing the
name of Djedkare' was found in the tomb, the attribution of the date to him is some-
what uncertain. Nevertheless Baer (n. 132), 98, assigned the tomb to the end of Dyn.
5, while Harpur (n. 72), 213, dates it mid-Djedkare' to Wenis.
dynasties 4 to 5 143

rnpt zp
• 3, #bd IV #¢t, sw 11; mason’s inscription, anonymous mastaba “L”,
Abusir151
• 6, #bd II “mw, sw 28
• 8, #bd IV “mw; papyrus archive, mortuary temple of Neferirkare'152
rnpt (m)-xt zp
• 4, #bd I prt153
Concerning several dates whose attribution is uncertain (Djedkare' or
Wenis), see above under Djedkare'.
attested rnpt zp: 3, 6, 8
attested rnpt (m)-xt zp: 4
rnpt zp: rnpt (m)-xt zp, 3: 1

Postscript.—Unfortunately, the edition of this volume has been delayed


by four years. In the meantime, there has appeared some new infor-
mation relating to the subject of my article to which I could not respond.
For instance, a work by J. S. Nolan (The Original Lunar Calendar
and Cattle Counts in Old Kingdom Egypt in: AH 17, 2003, 75–97)
offering the explanation of the imbalance between the “years of an
occasion” and “years after an occasion” in the contemporaneous Old
Kingdom documents. Moreover, some new conclusions concerning the
dated documents of the late Fifth Dynasty eventuated as a result of an
examination of the papyri from Raneferef ’s mortuary temple archive
(see P. Posener-Kriéger, M. Verner, H. Vymazalová, The Pyramid
Complex of Raneferef. The Papyrus Archive, in press). It is thus a
matter of some regret that this article could not be as comprehensive
as I would have liked.

151
M. Verner & V. Callender, Djedkave’s Family Cemetery (Prague, 2002), 103.
152
Posener-Kriéger & de Cenival, Abusir Papyri, pls. 54 c, 54 A c; Posener-Kriéger,
Archives II, 491.
153
Posener-Kriéger & de Cenival, Abusir Papyri, pls. 50, 50 A; Posener-Kriéger, Archives
II, 491.
II. 5 THE RELATIVE CHRONOLOGY OF
DYNASTIES 6 AND 8

Michel Baud

Although data are quite abundant for Dyn. 6 and derive from a vari-
ety of sources (royal annals and decrees on stone, administrative doc-
uments on papyrus, expedition graffiti), the interpretation of the dating
system used by the monarchy remains controversial. For this period,
the dogma of the biennial census has been challenged in the most
recent studies, especially in the compilations of Spalinger1 and Helck2
prior to the publication of the South Saqqara Stone, with the royal
annals of Dyn. 6.3 In theory, the discovery of such a monument might
be expected to clarify chronological questions, since the text is orga-
nized in year-compartments ending with the citation of the year, either
of the census type (rnpt zp) or post-census type (rnpt m-¢t zp). But unfor-
tunately, the inscriptions were quite systematically erased prior to the
reuse of the slab as a sarcophagus lid, so that neither the demarcation
of the compartments (which may in any case have been painted, rather
than carved) nor most of the dates are preserved. Documents of
significantly later date, such as the TC, do not provide any help for
evaluating the dynasty’s duration, or the lengths of individual reigns.
Most, if not all, of the figures preserved are at odds with contempo-
raneous OK data, despite repeated efforts to reconcile the two.4 Therefore,
the value of the papyrus lies more in the realm of historiography than
in chronology.5

1
Spalinger, “Texts”, 275–319.
2
W. Helck, “Gedanken zum Mord an König Teti”, in: Fs Goedicke, 108–109.
3
Baud & Dobrev, “Annales”, 30–31; Baud & Dobrev, “Le verso des annales de la
VIe dynastie, Pierre de Saqqara-Sud”, BIFAO 97 (1997), 35–42.
4
See Beckerath, Chronologie, 147–152.
5
Baud, “Ménès”, 126–132.
dynasties 6 and 8 145

Reign of Teti

No dates are preserved on the annals (recto, first register),6 and the
absence of lines dividing year-compartments does not even allow an
estimate of the reign’s total length. The space allotted Teti seems much
too small to accommodate the available data; most likely, the monu-
ment displayed a summary of the reign with compartments of very
reduced size. Possibly what was initially considered the recto is rather
the verso,7 but arguments in favour of this are weak.8
The Abusir archive from the funerary temple of Neferirkare' pro-
vides a number of dates for the early part of Teti’s reign.9 All come
from pBerlin 10.474A–B recto and verso (HPBM V, 1969, pl. 92–95),
a narrow roll which belongs to a single reign, as is obvious from the
coherent time span of the recorded dates. Teti’s serekh (pl. 94) clearly
identifies the sovereign, at least for the left document of the presumed
verso preserving the earliest dates:
• [rnpt (m)-¢t zp 1], III prt sw 'r˚( y); rnpt (m)-¢t zp 1 [III prt] sw 10;
rnpt [(m)-¢t zp 1] III prt sw 9, in reverse order (pl. 94, left doc.),
to which the recto adds two:
• rnpt (m)-¢t zp 1, III “mw sw 3 (pl. 94, right doc.),
• rnpt (m)-¢t zp [1], IV “mw sw [x] and rnpt zp 2, II “mw sw 3, in
normal order (pl. 92).
pBerlin 15.729 verso (pl. 103) provides another date: [rnpt] (m)-¢t zp
[x], I #¢t sw 3, but the year is missing and the identity of the king
remains uncertain. It should not be Teti,10 since the recto displays a
basilophorous name citing a king Pepy11 (not necessarily Pepy II).
Whatever the numeral was, this document provides another example
of a post-census year.

6
Baud & Dobrev, “Annales”, 23–92.
7
V. Dobrev, “The South Saqqara Stone and the sarcophagus of Queen Mother
Ankhesenpepy”, in: M. Barta & J. Krejci, eds., Abusir and Saqqara in the Year 2000
(Prague: Archiv Orientální Supplementa 9, 2001), 382–384.
8
Baud & Dobrev, “Annales”, 54; Baud & Dobrev (n. 3; 1997), 38.
9
Posener-Kriéger, Archives II, 491.
10
Helck, Fs Goedicke.
11
Posener-Kriéger, Archives II, 491.
146 michel baud

Other dates for the reign of Teti are:


• rnpt (m)-¢t zp 6, III “mw sw [x]: a graffito at the alabaster quarry
of Hatnub (gr. no. 1);12
• rnpt zp 11, I #§Δ sw 20: an ink inscription in the tomb of Nykau-
Izezi (Teti Cemetery, Saqqara), added to a scene of fowling in the
marshes, just above the boat in which the owner stands.13 The
inscription dates the official’s burial “the 11th count, I #¢t sw 20:
burial in the necropolis of the prince, the treasurer of Lower Egypt,
Nykau-Izezi”. Although a king’s name is not mentioned, all argu-
ments favour the reign of Teti. By doubling the number of “occa-
sions” hitherto known, this new date raises several questions (see
below).

Reign of Userkare'

The South Saqqara royal annals,14 demonstrate conclusively the exis-


tence of this king, but almost nothing remains of the section in the
middle of the uppermost register devoted to his reign. The available
space between the titularies of Teti and Pepy I, when compared to the
size of an average year compartment of the latter, indicates that
Userkare'’s reign must have been brief, from two to four years. This
conclusion is consistent with the very few monuments of this king,
mostly seal impressions, so far recovered. The silence of contempora-
neous private biographies is disturbing. A funerary complex planned
but never erected, is a better explanation for this absence than a spec-
ulative damnatio memoriae.15 Regardless, the chronographic purpose of the
royal annals did not allow the omission of this king, whatever form his
titulary may have taken.

12
Eichler, E., Untersuchungen zum Expeditionswesen des ägyptischen Alten Reiches (Wiesbaden,
1993), 41, no. 36.
13
(a) N. Kanawati & M. Abder-Raziq, The Teti Cemetery at Saqqara VI. The Tomb of
Nikauisesi (Warminster: ACE Reports 14, 2000), pl. 50; (b) N. Kanawati, “A new ˙#t/
rnpt-zp for Teti and its implication for Old Kingdom chronology”, GM 177 (2000)
25–32.
14
Baud & Dobrev, “Annales”, 28, 53, 59–62.
15
R. Stadelmann, “König Teti und der Beginn der 6. Dynastie”, in: Fs Leclant I,
335.
dynasties 6 and 8 147

Reign of Pepy I

Again, the discussion must start with the data preserved in the annals
stone from South Saqqara. The reign extends from the last third of
the first register (A) to the very beginning of the fifth (E). Although
there is again no demarcation of year-compartments, traces of a num-
ber of memorial formulae (nswt bjt Ppy jrn.f m mnw.f ) provide clues for
reconstructing the original layout.16 The twelve surviving formulae
(M3–M14) are spaced at rather regular intervals (× 2 or × 3 where
one, or perhaps two formulae are lacking), which supports as estimate
of the original number at up to 25. Since both “occasion” and “after-
occasion” years are known for the reign, obviously each mnw-formula
was associated with a pair of years, a census year and a post-census
year, presuming a regular biennial system.
Contrary to the editio princeps of the monument,17 it is, however, by
no means certain that a single heading systematically covered two years.
Some compartments, especially at the beginning of a register, indeed
appear much larger than others (see especially M5, second reg., and
one formula before M10, fourth reg.). Therefore, it is tempting to con-
clude that they group two different years, by contrast to the others,
which represent the vast majority of the (theoretical) compartments. But
there are a number of objections to such a radical proposition. First,
the size of a compartment might vary slightly according to the num-
ber of available and/or relevant data that needed to be recorded for
posterity, as exemplified by the difference between the first and the sec-
ond register where the average distance between successive mnw is
respectively 30 and 40 cm. For years of crucial importance to the
monarchy, such as the royal jubilee, the compartments could have been
much larger than the average, although this is not an absolute neces-
sity (cf. the minimal size of the first year, dedicated to the coronation
rites). Secondly, there is at least one surviving example of two years
grouped in a (theoretical) compartment of average size, reg. D, for-
mula following M11 (hereafter M11 + 1). The date preserved here, a
census year, occupies the middle of the available space and not the
end, as anticipated for the case of a single year covered by its own
mnw-formula. Nonetheless, if it is possible that most of the memorial

16
Baud & Dobrev, “Annales”, 50–53.
17
Baud & Dobrev, “Annales”, 50–52, fig. 19.
148 michel baud

formulae grouped census and post-census years together, the present


condition of the stone leaves some doubt about the generalization of
such a layout. Only parallel evidence from other sources might help
to solve this problem. The dates preserved in the annals are indeed
very few for the reign of Pepy I, and not unproblematic as regards
their reading:
• rnpt zp 18 (reg. D, M11 + 1, text zone D4);
• rnpt (m)-§t zp 23 (reg. E, M14 + 1, text zone E7);
• rnpt [m-¢t ?] zp 25 (reg. E, M14 + 3, text zone E8).
These high counts are also attested in several expedition graffiti18 and
a royal decree:19
• rnpt (m)-¢t zp 18, III “mw sw 27: Wadi Hammamat graffito no. 107,
mentioning the first jubilee;
• rnpt m-¢t zp 18, IV “mw sw 5: Sinai graffito no. 16, mentioning the
first jubilee;
• rnpt zp 21, I prt sw 23: decree for the Pyramid complex of Snofru,
Dahshur;20
• rnpt zp 25, I #¢t sw [x]: Hatnub quarry graffito no. III, once more
associated with the first jubilee.
Major clearance work at the king’s pyramid, South Saqqara, 1987–88
and 1993–97, revealed a few dates among the great number of masons’
marks.21 Most did not include the year but, according to common prac-
tice,22 only a season, month and day.23 A block from the eastern end
of the south side is a notable exception.24 After the group rnpt(?)-zp
there is an hieratic sign, which at first sight reads 30, followed by two
vertical strokes. Such a high date, count 32 (or even 22), from an early

18
Eichler (n. 12), nos. 133, 16, 30.
19
Spalinger, “Texts”, 303–304.
20
Goedicke, Dokumente, 55–77, fig. 5.
21
A. Labrousse, L’architecture des pyramides à textes. II, Saqqara Sud, (Cairo: BdE 131,
2000), 1–2.
22
For the mastaba of Ptahshepses see M. Verner, Abusir II. Baugraffiti der Ptahschepses-
Mastaba (Prague, 1992), 176–198.
23
V. Dobrev, “Observations sur quelques marques de la pyramide Pépi Ier”, in: Fs
Leclant I, 150–151.
24
Dobrev, pers. comm. 1994; see also F. Raffaele in 2001, “Sixth Dynasty Annals.
The South Saqqara Stone”, http://members.xoom.it/francescoraf/hesyra/ssannals.htm.
The mark is visible in Labrousse (n. 21), fig. 10, but at a very reduced scale.
dynasties 6 and 8 149

stage of the construction (corresponding to the 5th course of revetment


blocks), leaves some doubt about this reading. Furthermore, since the
highest count of the reign is 25, as confirmed by both the royal annals
and the other contemporary data, the only satisfactory solution is to
suppose that a double system of counting, one annual, the other bien-
nial (whether regular or not) existed. The annual rnpt-zp 32 should be
therefore equated with a lesser number of biennial census counts, at
least 16 if regular. (Note that an anomalous group of 19 strokes in two
rows (10 + 9) just before the date is obviously a calculation of some
sort, not necessarily connected to the dating system). At present, and
until adequate publication, this intriguing problem defies solution.
Conceivably, it may eventually provide a key to explaining the con-
tradictory dates for the king’s first jubilee.
The date of Pepy’s first heb-sed is controversial, since it is associ-
ated with two different years, counts 18+ and 25.25 Taking the two
figures at face value, they would have been separated by at least 6
years (systematically excluding post-census dates in this period, which
is unlikely), and as many as 12 (with systematic post-census years).
Therefore, the significance of both or either might reasonably be chal-
lenged, and could attest the prevalence of the Wunsch-Idee in the men-
tion of the jubilee, for the benefit of the king’s longevity.26 A strictly
historical/chronological interpretation is, however, still possible. Spalinger
ingeniously envisaged the existence of two parallel dating systems at
this period, one annual, i.e., 25 counts, the other (irregularly) biennial,
i.e., 18 census counts, plus presumably 7 post-census.27 Although inter-
pretation of building graffiti may support this hypothesis, there remains
the very confusing consequences of such a theoretical double system,
both citing all years simply rnpt-zp. Furthermore, Spalinger has not con-
sidered one important factor: the context in which the dates are actu-
ally associated with the jubilee. As for the Sinai relief, there is no direct
equation between the first jubilee and the date of the expedition. And
though the scene depicts the royal ceremony in a format typical of an

25
These do not refer to two different jubilees as P. O’Mara (“Dating the Sed-
Festival: Was there a Single Model?”, GM 136 [1993], 57–70) thought, nor can the
second belong to Pepy II (as proposed by J.v. Beckerath, “Gedanken zu den Daten
der Sed-Feste”, MDAIK 47 [1991], 30; tentatively Eichler [n. 12], 39).
26
E. Hornung, “Sedfest und Geschichte”, MDAIK 47 (1991), 169–171; with earlier
literature.
27
Spalinger, “Texts”, 305–306.
150 michel baud

year compartment in the annals, it is an all too frequent royal tableau


to be taken as a true date.28
The same may possibly hold true for the other inscriptions, although
the same historical connection between the Sinai and the Hammamat
graffiti, both under count 18+, end of the shemu-season, may not be
fortuitous.29 However, a tendency to mention the jubilee repeatedly in
the years following its celebration apparently existed,30 for example, in
connection with intense building activity at the royal funerary complex,
down until the very end of the reign (i.e., count 25). Here again, the
royal annals furnish a new argument favouring this hypothesis. Between
the mention of count 18 and the next memorial formula which belongs
to count 19, end of register D, the available space for count 18+ is
the expected half of the average size of a theoretical compartment. It
is hard to believe that such a narrow space corresponds to the jubilee
celebration, which obviously had a considerable importance for this
(and every) king, as is documented, for example, by the number of
stone vessels celebrating the event.31 (Count 25, the very last compart-
ment of the annals, is of course excepted). By contrast, the longest com-
partment of the reign—more than half again the average length—is
M10–1 (i.e., one formula before the preserved M10) at the beginning
of register D. Fortuitously or not, this compartment corresponds pre-
cisely to year 30/31, if a strictly biennial system of numbering is pre-
sumed.32 This could also explain why the handful of documents dated
to the first jubilee did not cite any other date. For example, decree
Coptos A simply epitomized the rule for the renewal of the king’s pow-
ers after 30 years.33
Specialists, however, remain divided on whether this rule obtained
during the OK.34 There would therefore be no necessity to place the

28
See, too, two inscriptions recording Merenre'’s visit to the First Cataract area to
receive the hommage of Nubian chief(s). One displays a real date (count 5, see infra),
reign of Merenre'; the other only a pictorial zema-tawy (Urk. I, 111), which may be
indicative of the theoretical date—the coronation year—in which such an event would
have taken place.
29
J. Vercoutter, L’Égypte et la vallée du Nil, 1: Des origines à la fin de l’Ancien Empire
(Paris, 1992), 326.
30
See Hornung (n. 26), 170.
31
A. Minault-Gout, “Sur les vases jubilaires et leur diffusion”, in C. Berger &
B. Mathieu, eds., Fs Jean-Philippe Lauer (Montpellier, 1997), 305–14.
32
Raffaele (n. 24).
33
Goedicke, Dokumente, 41–54, fig. 4.
34
For the range of interpretations see Beckerath (n. 25), Hornung (n. 26), O’Mara
(n. 25), A. A. Krol, “The representation of the ‘Sed-Platform’ in the Early Dynastic
monuments”, GM 184 (2001), 27–37.
dynasties 6 and 8 151

jubilee as late as year 35/36,35 nor to equate rnpt zp 18+, presuming


Userkare' was a usurper, with a canonical year 30 of strictly personal
rule,36 an assumption invalidated by the royal annals.37

Reign of Merenre'

The last register (F) of the recto of the Dyn. 6 annals is dedicated to
the first years of this reign. The number of compartments is uncertain,
but five or six is a reasonable estimate.38 The dates preserved are:
• rnpt zm#-t#wy, associated with the first memorial formula (M15) of
the register (text zone F1);
• rnpt zp 1 (+ 1?), with the second formula (M16, zone F3);
• rnpt (m)-¢t zp 1 (+ 1?), probably with the next formula, not pre-
served (zone F5).
In our initial publication, we logically assumed that the two last dates
were to be read as counts 2 and 2+, since in these annals, the Unification
of the Two Lands was apparently considered a year of cattle census
(the expression Δnwt j˙ does follow the date). After this first census (count 1),
a compartment was tentatively delineated to account for a post-census
year after the Unification,39 considering that for such years also the sys-
tem remained biennial. Although this remains a possibility, there are
weaknesses in such a reconstruction. In the first place, this so-called
count 1+ would be confined to a very narrow space, when compared
to the other very broad compartments of the last register.40 Secondly,
the figure of the next date consists of a very deeply carved single
stroke,41 and it is unlikely that another stroke ever existed; there is also
no space available for an alleged second stroke under the zp sign. It
may not be mere chance that the next date also retains only a stroke.
These two dates should be read accordingly as counts 1 and 1+, even
if the first year of the reign was labelled census year. Should the suc-
ceeding years be read 1/1+ or 2/2+, it is nonetheless clear that this

35
Contra e.g., H. Goedicke, “Two Mining Records from the Wadi Hammamat”,
RdE 41 (1990), 65–93, at 67, and O’Mara (n. 25).
36
Contra Spalinger, “Texts”, 305–306.
37
Baud & Dobrev, “Annales”, 61–62.
38
Baud & Dobrev, “Annales”, 48–49, 54.
39
Baud & Dobrev, “Annales”, fig. 19.
40
Baud, “Ménès”, 123–124.
41
Baud & Dobrev, “Annales”, pl. VII c.
152 michel baud

period experienced a biennial census. Since a year 5+ is also known


(see below) and since it is likely that the document did not end abruptly
in mid-reign, it must be concluded that the reign continued on the
verso (see above, contra Dobrev),42 even if the titulary of this king prob-
ably featured in the introductory column of the recto, as recently pro-
posed, citing new and convincing arguments.43 Other sources are restricted
to two rock inscriptions:44
• rnpt zp 5, II “mw sw 28, First Cataract area, Urk. I, 110, 12;
• rnpt (m)-¢t 5, Hatnub graffito no. VI.45
Once again, it is clear that the latest years of the reign experienced a
biennial system.

Reign of Pepy II

Spalinger’s list46 remains relatively current and must be consulted for


further references:
• rnpt (m)-¢t zp tpy, IV #¢t sw 10 (in two parts) and rnpt zp 2, pro-
cession graffiti in Wadi Hîlal (El Kab), although the beginning of
the reign of Pepy II is a good possibility,47 the dates could belong
to his predecessor Merenre';
• rnpt zp 2, III #¢t sw 15, letter of the king to Harkhuf in his tomb,
Aswan;
• rnpt zp 2, Sinai graffito no. 17;48
• rnpt zp 11, I “mw sw 23, the famous letter found in the workshop
adjacent to Temple T in the Djoser complex, Saqqara; the reign
is inferred from other chronological data of the archives;49
• rnpt (m)-¢t zp 11, II “mw sw 26, decree Coptos B, temple of Min;

42
Dobrev (n. 7).
43
Dobrev (n. 7), 384–385, pl. 58.
44
Spalinger, “Texts”, 306–307.
45
Eichler (n. 12), 40, no. 33.
46
Spalinger, “Texts”, 307–308.
47
H. Vandekerckhove & R. Müller-Wollermann, Elkab VI. Die Felsinschriften des Wadi
Hilâl (Turnhout, 2001), 210–211, gr. O 144, and 183–186, gr. O 74; conclusions:
375–379.
48
Eichler (n. 12) 35, no. 17.
49
P. Posener-Kriéger, “Fragments de papyrus provenant de Saqqarah”, RdE 32
(1980), 83–93.
dynasties 6 and 8 153

• rnpt zp 12, graffito of Tômas, either of the two Pepys, but more
probably Pepy II;50
• rnpt zp 14, I #¢t sw 23(?), Hatnub graffito no. 3;51
• rnpt (m)-¢t 22, IV “mw sw 28 (date in two parts), decree Coptos C;
• rnpt zp 31, III #¢t sw 3 [+ 3], decree for the cult of Mycerinus,
Gîza;
• rnpt zp 31(?), IV prt sw [x], graffito from the king’s funerary temple,
Saqqara;
• rnpt (m)-¢t zp 31, I “mw sw 20, Hatnub graffito no. 7;52
• rnpt zp 33 (?) or 24 (?), IV sw [x], decree for the cult of Queen
Udjebten, Saqqara.
There are also dates without a king’s name which can be placed securely
in the second half of Dyn. 6. The first two are from Giza:53
• rnpt zp 2, III prt sw 27, two mason’s marks on the walls of mastaba
G 7803C, Giza Eastern Cemetery, and another citing the same
year, but month, season and day lost;
• rnpt (m)-¢t zp 5, III prt sw 29, two execration texts from Giza.
Another date comes from an expedition graffito at Tômas;54
• rnpt 6, III “mw, probably either Pepy I or II.

Biennial, Irregular or Annual Census? The Case of Dynasty 6

The regularity of the census, backbone of the Ancient Egyptian dating


system, is still a matter of controversy for the OK. The most recent
discussions of this crucial problem present the largest possible spectrum
of interpretations, ranging from a regular biennial census55 through an
annual census with post-census years at irregular intervals56 to a strictly
annual one.57 That an annual count was already established by Dyn.

50
According to Eichler (n. 12), 105 (no. 227A).
51
Eichler (n. 12), 43, no. 39.
52
Eichler (n. 12), 44–45, no. 43.
53
Spalinger, “Texts”, 308–309.
54
Eichler (n. 12), 109, no. 245.
55
E.g., Baud, “Ménès”.
56
E.g., Verner, “Archaeological Remarks on the 4th and 5th Dynasty Chronology”,
Archiv Orientálni 69 (2001), 410–412.
57
E.g., Kanawati (n. 13).
154 michel baud

6 (and not in the FIP, according to the traditional view),58 is disproved


by the number of attestations of m-¢t zp years at this period. Kanawati
believes, however, that they resulted from provisional numbering, sub-
sequently altered to “normal 2 years counts”. For example, “the ref-
erence to the ‘year after the sixth count’ may simply refer to the seventh
year, but before the seventh count was undertaken”.59 This hypothesis
ignores the existence of the South Saqqara Stone, with at least two
examples of post-census years (one in the reign of Pepy I, and the sec-
ond under Merenre', see above). Since annals are an official recapitu-
lation of events, there is no reason why the entries should reflect a
provisional numbering system.
Kanawati’s proposal is an attempt to reconcile apparently contra-
dictory data in the case of Nykau-Izezi (see above), viz., (a) a basilophorous
name suggesting that Nykau-Izezi was born under Izezi60 (b) his rep-
resentation in the reliefs of the causeway of Wenis, with the high rank-
ing title ‘sole friend’; (c) the dating of his burial to the 11th count,
presumably of Teti; (d) an estimate of his age at death, based on exam-
ination of his remains, as 40–45 years or even slightly younger (35).
Kanawati was influenced by the difficulty of reconciling the relative
brevity of Nykau-Izezi’s life with the time-span between Djedkare'’s
reign and the 11th census in Teti’s reign, presuming a regular bien-
nial census. The 11th census of Teti corresponds to year 22/23 of a
regular biennial census system, but at least 13/14, if the census was
irregular (since two intervening years are known, 1+ and 6+, see above).
Nykau-Izezi was therefore between about 17/18 and 26/27 years old
when Teti ascended the throne. Since, on the same premise, Wenis
reigned between 16 (rnpt zp 8 as highest census) and 9 years,61 the
official was either born at the very end of Djedkare'’s reign, or ten
years earlier. This would account for Izezi in his name, but this expla-
nation is superfluous, since kings were celebrated thus for various rea-
sons, if indeed such names were not simply passed from father to son.62
Thus the name does not prove that Nykau-Izezi’s career began in
Djedkare'’s reign. If he is the like-named official in the Wenis cause-

58
E.g., Gardiner, “Years”, 14–16.
59
Kanawati (n. 13; 2000a), 21, 23, b; see also Helck (n. 2), 110.
60
N. Kanawati & M. Abder-Raziq, The Teti Cemetery at Saqqara V. The Tomb of Hesi
(Warminster: ACE Reports 13, 1999), 37–38, pl. 33, 59.
61
Verner (n. 56), 410–412, 416.
62
Another Nykau-Izezi is mentioned, for example, on three graffiti at the pyramid
of Pepy I, see V. Dobrev, “Les marques sur pierres de construction de la nécropole
de Pépi Ier. Étude prosopographique”, BIFAO 96 (1996), 112, D.1.
dynasties 6 and 8 155

way reliefs, he was promoted to ‘sole friend’ between the ages of 17


and 27—quite young in either case. The first alternative may indeed
seem much too young, but high-ranking titles may not have been
confined to mature officials. All in all, there are simply too many uncer-
tainties in Kanawati’s argument. The same situation obtains for other
officials who started their careers under Teti and died under Merenre'.
Some cases may still be debatable, as Kanawati rightly pointed out.63
Weni, for example, already held a relatively high position under Teti
and must have been about 70 when Merenre' ascended the throne,
since Pepy I’s reign amounts to 50 years, presuming a strictly biennial
system. This seems very old for the onerous duties reported in his biog-
raphy, and for his journeys to obtain materials and monuments for the
king’s tomb. However, this may have been a conceit to celebrate the
official’s longevity and his capacity to remain active at an advanced age.
If the theory of an annual census be discarded for sound reasons, it
must nevertheless be admitted that no basis exists for deciding in favour
of either of the alternatives, the regular or irregular biennial system.
On the one hand, the number of attested census years is very well bal-
anced by post-census years during the reigns of Teti to Pepy I. When
the sources shed some light on a segment of these reigns, i.e. the ear-
liest counts of Teti (1 to 2), the latest counts of Pepy I (18 to 25) and
the earliest and latest counts of Merenre' (1 and 5), the number of
intervening years equals, or nearly equals, the number of census years.
The alleged total imbalance between the two64 results from the limited
number of sources and prejudice associated with the category of sources,
as exemplified by the masons’ marks of Dyn. 4 at Giza.65 On the other
hand, the extreme imbalance for the reign of Pepy II could favour an
irregular counting system in his particular case. The celebrated longevity
of the king in tradition66 as well as the fact that he was a child at his
accession,67 demand reconciliation with the contemporaneous record,
to which a biennial count does justice.68 The table below summarizes
the dates for the period from Teti to Pepy II:

63
Kanawati (n. 13, 2000a), 22–23.
64
E.g., Helck (n. 2), 106–110; Spalinger, “Texts”, 314–316.
65
Baud, “Ménès”, 119–121.
66
Baud, “Ménès”, 129, with caution.
67
K. Ryholt, “The Late Old Kingdom in the Turin King-list and the Identity of
Nitocris”, ZÄS 127 (2000), 87–100, at 94.
68
H. Goedicke, “The Death of Pepy II-Neferkare”, SAK 15 (1988), 111–121; Beckerath,
Chronologie, 151–152.
156 michel baud

King Highest Minimal Number Minimal Reign Minimal Reign


Census of Post-census Length A Length B
Years

Teti 11 2 13 years 22/23 years


Userkare' ? ? [2/4] [2/4]
Pepy I 25(+?) 2 or 3 27 50/51
Merenre' 5+ 2 7 11/12
Pepy II 31, ev. 33 3 or 4 34 62/63
Total min. 83 9/10 83/85 147/153

X+ indicates a post-census year, for which X is the number of counts;


“minimal reign length A” is the sum of the highest count (col. 2) and
the attested intervening years (col. 3);69 “minimal reign length B” pre-
supposes a regular biennial census; the estimate for Userkare' is based
on the royal annals (see above).

Dynasty 8

The identity, number and order of the Memphite rulers of Dyn. 8


remain uncertain and identifications rely heavily on the much later
Ramesside lists.70 The TC counts 8 rulers after Pepy II (col. iv, no. 5
to 13, this name and some others in lacuna).71 It includes, in second
position, Queen Nitocris, who turns out to be a male ruler, Neitiqerty
Siptah, according to Ryholt’s recent examination of the papyrus.72 The
Abydos list (nos. 39 to 56) adds 10 more rulers, all probably between
Neitiqerty (Abydos no. 40, if identified with Netjerkare') and Neferka
Khered-seneb (Abydos no. 51, called Neferkare' Pepy-seneb),73 a group
which may have been in lacuna in the TC Vorlage and therefore prob-
ably reported as wsf/lost.74

69
After Verner (n. 56), 415–416.
70
E.g., Beckerath, Chronologie, 151–152.
71
This column should be renumbered 5, since according to Ryholt’s recent study
(see n. 67) there is evidence of an intermediate column between col. I and II of
Gardiner’s edition.
72
Ryholt (n. 67), 87–100.
73
See Ryholt (n. 67), 87–94.
74
Beckerath, Chronologie, 148–9; Ryholt (n. 67), 96–98.
dynasties 6 and 8 157

According to the partially preserved figures in the TC, both for reign
lengths and summations of grouped dynasties, Dyn. 8 covered a very
short period of about one generation. However, not only is the read-
ing of some of the figures problematic (e.g., the total for Merenre'),75
but the very value of the figures remains largely questionable, as
exemplified by contradictory OK data for a number of reigns. Recent
analysis would double the duration of this dynasty, to at least 50 years,76
or even slightly more,77 but this is not a significant change in the image
of a relatively short and obscure period. In this particular case, the TC
data is probably not far from the truth with its low figures for indi-
vidual reigns: 1 year for the immediate successor of Pepy II (name lost,
no. 6) and between 1 1/2 to 4 years for the last four rulers (nos.
10–13). The six wsf-years reported in the subtotals (col. iv, 14–17) for
ten missing kings probably represent an artificial emendation of the
scribe, as exemplified by other occurrences of this figure.78 All in all,
these brief reigns accord with the few royal monuments recovered so
far, and the low figures of the preserved dates.79 Arranged in increas-
ing numeric order, they are:
• rnpt zm#-t#wy, II prt sw 20, Coptos decree P of [Netjeri-bau] (Horus
name of Neferkauhor; identification from parallel decrees),80 tem-
ple of Min;81
• rnpt zp zm#-t#wy, IV “mw 1 (wpty), decree of [Demedj-ib]-tawy (?),
(Horus name),82 funerary complex of Queen Neith, Saqqara;

75
For the old reading ‘44 years’, see Ryholt (n. 67), 90, 98.
76
Beckerath, Chronologie, 151–152.
77
S. Seidlmayer, “Zwei Anmerkungen zur Dynastie der Herakleopoliten”, GM 157
(1997), 84–85.
78
Ryholt (n. 67), 97–98.
79
Spalinger, “Texts”, 312–313.—Note also the inscription Cairo JE 43290 dated to
rnpt zp <1 ?>; IV #¢t 25. The numeral is omitted, but 1 is the most likely emenda-
tion, see H. Goedicke, “A Cult Inventory of the Eighth Dynasty from Coptos (Cairo
JE 43290)”, MDAIK 50 (1994), 72. This could refer to the first incomplete civil year—
year 0—usually designated zm#-t#wy. The inscription presumably originates from Coptos
or nearby Khozam. Goedicke (ibidem) ascribed it tentatively to Nefer-kau-hor, but
Fischer, in: Manuelian, ed., Studies Simpson, 267–270, argues for a date towards the end
of Herakleopolitan rule in the Coptite nome. See also below Krauss, chapter III. 8
for the Khozam lunar date.
80
See W. C. Hayes, “Royal Decrees from the Temple of Min at Coptos”, JEA 32
(1946), 3–23.
81
Goedicke, Dokumente, 195–196, with Hayes (n. 80), pl. V.
82
According to the restoration proposed by Schenkel, Memphis, 24–25.
158 michel baud

• rnpt zp tpy, IV #¢t sw 2, Wadi Hammamat inscription of King Ity


(mentioned in the name of his presumed pyramid),83 possibly Dyn.
8 (O.Ham no. 169);
• rnpt zp tpy (?), III “mw sw 2, Wadi Hammamat inscription of an
unknown king, date uncertain but possibly Dyn. 8 (O.Ham no. 152);84
• [rnpt] zp 4 [+ x?],85 season etc. lost, Coptos decree H of king
Kha[bau?] (Horus name).86
The absence of post-census years probably testifies to a change in the
dating system from a regular (?) biennial to an annual one.87

83
Tentatively equated with Neferirkare' II of the Abydos list and the contempo-
rary Horus Demedjibtawy by Spalinger, “Texts”, 313, and n. 104. Goedicke’s read-
ing (n. 35), 66–67, (rnpt zp tpy <jb-sd> taken to refer to Pepy I is not supported by
the parallel evidence.
84
See Schenkel, Memphis, 32–33; Goedicke (n. 79), 83.
85
The stela is lost below the four aligned strokes. While 3 or even 4 more strokes
could possibly have figured in a lower line, it is rather unlikely in such a period of
ephemeral kings.
86
Goedicke, Dokumente, 163–164, fig. 16 and 23; for the date: Hayes (n. 80), 13,
n. 7, and pl. iiia, top, before col. 1.
87
Gardiner, “Years”, 14–16; Hayes (n. 80), 13; Spalinger, “Texts”, 312.
II. 6 THE RELATIVE CHRONOLOGY OF THE
FIRST INTERMEDIATE PERIOD

Stephan J. Seidlmayer

In dynastic history,1 the FIP spans the era of Herakleopolitan rule


(Dyns. 9 and 10) and the earlier part of the Theban Dyn. 11 up to
the re-unification of the country which occurred at some point in the
reign of Nebhepetre' Mentuhotpe II. When we attempt to clarify the
dynastic structure of this period and to estimate its chronological length,
our argument will be more straightforward if we first deal with Dyn.
11 and only then turn our attention to the problems of the Herakleo-
politan dynasties.

Dyn. 11

We are comparatively well informed about Dyn. 11.2 While the king
lists of Abydos and Saqqara omitted all FIP rulers and listed only
Nebhepetre' Mentuhotpe II and S'ankhkare' Mentuhotpe III of Dyn.
11,3 the TC gives a full account of its rulers, omitting only (as do the
lists of Abydos and Saqqara) its last ruler Nebtawyre' Mentuhotpe IV,
whose reign was relegated to a group of “missing” (wsf ) years.4 While
the names of most of the kings are destroyed in the TC, a number of
reign length data are preserved. Most valuable, however, is the fact
that this document also provides a figure for the total length of the
dynasty which allows us to determine the combined length of the first
two reigns, for which individual length data are not preserved. The
royal names which are destroyed in the TC can easily be reconstructed.

1
To define the FIP as a distinctive phase in the history of pharaonic culture or its
political structure, one would envisage a more extensive period including at least the
end of the OK after the demise of Pepy II.
2
For accounts of the chronology of Dyn. 11 see Schenkel, Studien, 145–149; Gestermann,
Kontinuität, 22–31; Beckerath, Chronologie, 139–142.
3
Abydos list nos. 57–58, Saqqara list nos. 37–38.
4
TC V, 11–18.
160 stephan j. seidlmayer

The list of royal ancestors inscribed by Tuthmosis III in the temple of


Karnak5 and a relief block from the temple of Tod which lists the pre-
decessors of king Mentuhotpe II are particularly relevant.6 Biographical
inscriptions of officials also attest the names and sequence of the more
important rulers,7 while dated monuments and inscriptions concur to
confirm the accuracy of the reign length data of the TC.8 Confusion
about the number of rulers named Mentuhotpe, caused by the fact
that Nebhepetre' Mentuhotpe II changed his protocol twice, were set-
tled by Gardiner.9 On this basis, the data relating to the sequence and
length of reigns of Dyn. 11 can be summarized with some confidence
as follows:

TC Years Karnak Urk. IV Tod Modern

jrj-p't [˙#tj-'] Jnj-[ jtj.f ] 608.15 count Inyotef


[MnΔw-˙tp] 16 Órw tpj-' Mn[Δw-˙tp] 608.14 Mentuhotpe I
[ Jnj-jtj.f ] Órw [S]h[r-t #.wj] Jnj-[ jtj.f ] 608.13 Órw Shr-t#wj
Jnj-jtj.f Inyotef I
[ J]n[ j-jtj.f ] 49 Órw [W#˙-'n¢] Jnj-jtj.f 608.12 Órw W#˙-'n¢
Jnj-jtj.f Inyotef II
[ Jnj-jtj.f ] 8 [Órw N¢t-nb-tp-nfr Jnj-jtj.f ] 608.11 [Órw N¢t-nb-tp
nfr] Jnj-jtj.f Inyotef III
Nb-˙pt-R' 51 Nb-˙pt-R' 609.14 Mentuhotpe II
S'n¢-k #-R' 12 S'n¢-k #-R' 609.15 Mentuhotpe III
wsf 7 [Nb-]t #[.wj]-R' 609.16 Mentuhotpe IV
Total 143

5
Urk. IV, 608–609.
6
J. Vandier, “Un nouvel roi Antef de la XIe dynastie”, BIFAO 36 (1936), 101–116.
7
The sequence Inyotef II, Inyotef III, Mentuhotpe II is attested in three biographical
inscriptions from Thebes, see Clère & Vandier, Textes, 15–16 § 20, and 19–20 § 23–24.
8
For Inyotef II a regnal year 50, probably the year of his burial, is attested on a
stela from his tomb, Clère & Vandier, Textes, 11 § 16; no dated monuments are known
for Inyotef III; for Mentuhotpe II, stela Turin 1447 (Schenkel, Memphis, 240) attests
year 46; for Mentuhotpe III a rock inscription in Wadi Hammamat attests year 8
(Schenkel, Memphis, 253 no. 426), and for Mentuhotpe IV a series of graffiti at the
same place, year 2 (Schenkel, Memphis, 263–268, nos. 441–444).
9
A. H. Gardiner, “The First King Menthotpe of the Eleventh Dynasty”, MDAIK
14 (1956), 42–51; see also L. Habachi, “King Nebhepetre Menthuhotep: his monu-
ments, place in history, deification and unusual representations in the form of gods”,
MDAIK 19 (1963), 16–52.
first intermediate period 161

A number of comments are in order. “Count” Inyotef, who was prefixed


as a non-royal ancestor to the line of Dyn. 11 kings in the Karnak
inscriptions, is in all probability identical with the jrj-p'.t ˙ #.tj-' Jnj-jtj.f
ms Jkwj to whom Senwosret I dedicated a votive statue in this temple.10
Whether he can also be identified with one of the attested pre-11th
dynasty nomarchs of this name,11 cannot be strictly demonstrated. The
lacunary state of the evidence available would seem to recommend a
careful stance in such matters. In any case, this question has no direct
consequences for the chronology of the period.
The Horus name tpj-' “the ancestor” accorded the first king of the
dynasty is evidently a later fiction intended to prolong the royal line
into the past. No contemporary monuments are attested for this person;
however, a statue erected by Inyotef II at Elephantine calls him “father
of the gods”, i.e. the father of the first two kings of Dyn. 11.12 Nor is
his direct successor, Sehertawy Inyotef I ever attested in contemporary
inscriptions.13 However, the sequence of tombs in the royal necropolis
at el-Tarif suggests that the Saff Dawaba belonged to this ruler. The
truly extraordinary size and layout of this tomb substantiate his claim
to royal status.
All other kings of the dynasty are well documented in contemporary
sources. Of some interest is the fact that the last ruler, Nebtawyre'
Mentuhotpe IV, attested in rock inscriptions from his second regnal
year in Wadi Hammamat,14 was omitted from all of the NK kings lists.
The reason is speculative, but it does not seem very far fetched to sup-
pose that his absence was motivated in some way by the circumstances
of the transfer of power to a new royal house. In view of the lack of
sufficient evidence it remains uncertain whether all 7 “missing”-years
in the TC belonged to him or whether there was a period of disputed
rule at the end of the dynasty,15 for which, however, there is no pos-
itive evidence.

10
Cairo CG 42005.
11
See the discussion in Gestermann, Kontinuität, 24–26, and F. Gomaa, Ägypten während
der Ersten Zwischenzeit (Wiesbaden: TAVO B27, 1980), 138–144.
12
On this king see L. Habachi, “God’s fathers and the role they played in the his-
tory of the First Intermediate Period”, ASAE 55 (1958), 167–190.
13
W. Schenkel in: D. Arnold, Gräber des Alten und Mittleren Reiches in El-Târif (Mainz:
AV 17, 1976), 50.
14
Schenkel, Memphis, 263–268 nos. 441–444.
15
J.v. Beckerath, “Zur Begründung der 12. Dynastie durch Ammenemes I.”, ZÄS
92 (1965), 8–9.
162 stephan j. seidlmayer

While none of the Manethonian sources lists individual kings, all of


them claim that Dyn. 11 comprised 16 kings who ruled for 43 years.
Evidently, the number of kings was inflated by 10 from 6 to 16, while
the number of years was reduced by 100 from 143 to 43. Nevertheless
the basic similarity of the figures again shows how close Manetho’s
data are to earlier pharaonic tradition.
Unfortunately, the date of the re-unification of Egypt cannot be deter-
mined precisely within the reign of Nebhepetre' Mentuhotpe II.16 A
stela of one Inyotef from Thebes is dated to year 14 of Mentuhotpe II,
“the year of the rebellion of Thinis”.17 This event is in all probability
to be seen in the context of the war between Thebans and Herakleo-
politans, which was fought, at least in its earlier phases, in the Abydos-
Asyut region. An inscription of an official who governed the Heliopolitan
nome, dated to year 41 of Mentuhotpe II, provides proof that by then
Mentuhotpe II controlled all of Egypt.18
Information which could clarify the sequence of historical events in
the period between these two dates is lacking. The phases and modal-
ities of Mentuhotpe II’s conquest of the Herakleopolitan kingdom, as
well as the development of his political aspirations and his propaganda,
remain unknown. Nevertheless it seems likely that both the actual course
of political events and the ideological dimension of the war between
Thebans and Herakleopolitans did have some influence on how the
end of the Herakleopolitan dynasty was chronologically fixed in later
annalistic tradition. Therefore one should be well aware that not only
the date of the historical event of the re-unification of Egypt is lacking
from our documentation but that the historical process itself in its sub-
stance eludes us.
It is even more difficult to use indirect criteria in an attempt to fix
the date of the re-unification of Egypt. The two alterations of Mentuhotpe
II’s royal protocol19 may be linked to the stages of his rule over Egypt.
In particular his latest Horus name zm #-t #.wj “uniter of the two lands”,
attested for the first time in regnal year 39,20 invited direct historical

16
On this issue see Franke, “Chronologie I”, 133, and Gestermann, Kontinuität, 35–42,
with further literature.
17
Clère & Vandier, Textes, 19 § 23; Schenkel, Memphis, 227.
18
Gestermann, Kontinuität, 42 n. 5.
19
Beckerath, Handbuch, 78–79.
20
Inscriptions in Wadi Schatt el-Rigâl, Schenkel, Memphis, 207–208 no. 318 and
320. Beckerath’s doubts that the date belongs to the rock inscription of Mentuhotpe II
(Chronologie, 141 with n. 632) are in no way convincing.
first intermediate period 163

interpretation. However, as Gestermann rightly pointed out, this is not


at all certain and in fact all his Horus names imply a claim of domi-
nance over the whole of Egypt.21 Rather more significant seems to be
the development of the structure of the royal protocol. While in the
first stage, Mentuhotpe, like his Theban FIP predecessors, used only a
Horus name and the titles nswt-bjt and z #-R' with personal name, he
introduced in the second stage both a nb.tj-name and the throne name
Nebhepetre', thus claiming full royal status.22 Whether this move reflected
only his aspirations or his actual taking over of power remains unknown,
however. In addition, from a strictly chronological point of view, this
discussion is bound to remain more or less fruitless, since we do not
know when the changes in the royal protocol occurred. Speculating
that the earlier, rather than the later change of the protocol might be
linked to the end of Herakleopolitan rule over northern Egypt would
only intuitively make it more likely that Mentuhotpe’s victory had
occurred in the earlier part of his reign. Strictly speaking, however, an
uncertainty of about 25 years in fixing the date of the re-unification
of Egypt remains until new sources become available.

The Herakleopolitan Dyns. 9 and 10

Determining the identity and length of Herakleopolitan rule is much


more difficult.23 Manetho’s account listed two dynasties of rulers from
Herakleopolis,24 Dyn. 9 with four (Eusebius) or 19 kings (Africanus)
who ruled for 100 (Eusebius) or 409 years (Africanus), and Dyn. 10
with 19 kings who ruled for 185 years according to all sources. The
TC, in contrast, listed only a single dynasty of 18 rulers;25 unfortunately
nearly all of the royal names and all reign length data as well as the
total for this dynasty, which originally was given in line V.10, are lost.

21
Gestermann, Kontinuität, 35–39.
22
This view was envisaged already by Hayes and Arnold, see Gestermann, Kontinuität,
37, n. 2–3.
23
See J.v. Beckerath, “The Date of the End of the Old Kingdom in Egypt”, JNES
21 (1962), 140–147; idem, “Die Dynastie der Herakleopoliten (9./10.)”, ZÄS 93 (1969),
13–20; H. Goedicke, “Probleme der Herakleopolitenzeit”, MDAIK 24 (1969), 136–143;
Beckerath, Chronologie, 143–145; S. J. Seidlmayer, “Zwei Anmerkungen zur Dynastie
der Herakleopoliten”, GM 157 (1997), 81–90, with additional literature.
24
Waddell, Manetho, 60–63.
25
TC IV.18–V.10.
164 stephan j. seidlmayer

The king lists of Abydos and Saqqara omitted the period of Herakleo-
politan rule completely.
While following the authority of the more ancient king list the unity
of the Herakleopolitan dynasty was already assumed by Schenkel and
others,26 Malek was able to account convincingly for the discrepancy
between the TC and Manetho. He demonstrated that the first four
kings of the dynasty came to be separated from the main group of
kings as a result of a series of misunderstandings which occurred when
the original text was copied.27 The number of 19 kings which Manetho
gives for Dyn. 10 and which Africanus duplicated also for Dyn. 9 accu-
rately reflects the total number of kings as given for the Herakleopolitan
dynasty in TC V.10. The difference of one king between TC and
Manetho can very likely be explained by suggesting that the TC omitted
the last Herakleopolitan ruler, just as the last king of Dyn. 11 was
omitted, probably because he was not considered legitimate, having
been removed from power under shameful circumstances as a result of
the victory of the Thebans over the Herakleopolitan kingdom.28 According
to this analysis which interprets the separation of Dyns. 9 and 10 merely
as a result of textual corruption in post NK tradition, the terms “Dyn.
9” and “Dyn. 10” should, for the sake of terminological precision, no
longer be used in historical interpretation to designate an earlier and
a later phase of Herakleopolitan rule.29 This argument has, of course,
no implications whatsoever for the length and historical structure of
the period.
Since the relevant entries are destroyed in TC, and since only very
few contemporary monuments of the Herakleopolitan kings are pre-
served, the names and the sequence of the 19 Herakleopolitan kings
cannot be reconstructed coherently.30 Manetho names a king Khety as
the founder of the dynasty, and the fact that the Herakleopolitan king-
dom was referred to as pr flty “the house of Khety” in contemporary

26
Schenkel, Studien, 149–150; Beckerath, “Herakleopoliten” (n. 23).
27
J. Malek, “The Original Version of the Royal Canon of Turin”, JEA 68 (1982),
105; Redford, King-Lists, 238–239, arrives at an equivalent conclusion, though with
different arguments.
28
For another solution see Malek, “Version” (n. 27), 105.
29
Seidlmayer, “Anmerkungen” (n. 23), 85–86.
30
For the available data see Beckerath, Handbuch, 72–75; to the material listed there
add M. Abd el-Gelil, A. Saadani & D. Raue, “Some Inscriptions and Reliefs from
Matariya”, MDAIK 52 (1996), 146–147.
first intermediate period 165

sources31 lends credibility to this tradition. For the third ruler, the name
Nfr-k #-R' is preserved in TC. He could be identical with King Nfr-k #-R'
whose name appears (with a graphic transposition k #-nfr-R', typical for
the Herakleopolitan period)32 on a label in the tomb of 'Ankhtifi at
Moalla.33 The last ruler of the dynasty was probably Merykare', whose
name appears in texts relating to the final phase of the Theban-
Herakleopolitan war in Siut tomb IV34 and whose pyramid complex at
Saqqara is attested in the titles of funerary priests.35 In fact, the gen-
eral tenor of the “Teachings for king Merykare'” seems to suggest that
this king suffered defeat by the Thebans.
Determining the length of the Herakleopolitan period is fraught with
all but unsurmountable difficulties. Data for the length of individual
reigns or of the entire dynasty are not preserved. Since it is clear, how-
ever, that the Herakleopolitan dynasty did not start after Dyn. 11 but
ran parallel to it up to the re-unification of Egypt under Nebhepetre'
Mentuhotpe II, it must have lasted for between 87 and 114 years at
least, depending on the exact date of the re-unification. Dismissing the
data given by Manetho as unreliable, Beckerath developed the hypoth-
esis of an ultra-short Herakleopolitan period making the beginning of
Dyn. 11 more or less coeval with the establishment of the new Herak-
leopolitan line of rulers in the north.36 Apart from speculations about
the historical development, this hypothesis rests mainly on the fact that
so very few monuments are attested for the Herakleopolitan period,
implying a short duration. This argument, however, is clearly not valid,
since the 100 years or so when the Herakleopolitan dynasty existed
parallel to the Theban kingdom did not leave many traces in the mon-
umental record either. Evidently, this situation is due to the bias of

31
E.g. the stela of Djari, Clère & Vandier, Textes, 14 § 18, l. 3; see also O. D. Berlev,
“The Eleventh Dynasty in the Dynastic History of Egypt”, in: D. W. Young, ed.,
Studies presented to Hans Jakob Polotsky (East Gloucester/Mass., 1981), 361–377.
32
Schenkel, Studien, 150; D. B. Spanel, “The Date of Ankhtifi of Mo'alla”, GM 78
(1984), 89.
33
J. Vandier, Mo'alla (Cairo: BdE 18, 1950), 36.
34
Schenkel, Memphis, 86 No. 64.
35
PM (2) III, 562–563; J. Malek, “King Merykare and his Pyramid”, in: C. Berger,
G. Clerc & N. Grimal, eds., Hommages à Jean Leclant IV (Cairo: BdE 106.4, 1994),
203–214; Kh. A. Daoud, “The Herakleopolitan Stelae from the Memphite Necropolis”,
in: Eyre, Proceedings, 303–308.
36
Beckerath, “Herakleopoliten” (n. 23); for the details of the discussion see Seidlmayer,
“Anmerkungen” (n. 23), 82–83.
166 stephan j. seidlmayer

our sources, which tend to favour UE, and to the special character of
FIP kingship which was no longer able to muster the resources of the
country to construct monumental buildings. The argument that the
Herakleopolitan kings were dropped from the Abydos king list and
therefore never ruled UE does not carry much weight either. Even if
the Herakleopolitan dynasty never wielded effective rule over UE, it could
nevertheless have held nominal supremacy over the whole country for
a period of time. The fact that a Herakleopolitan ruler was mentioned
in the tomb of 'Ankhtifi and that a certain Setka in his still unpub-
lished biographical inscription in his tomb on Qubbet el-Hawa claimes
to have entertained peaceful trade relations with the “House of Khety”
seem to provide direct proof that there was indeed a period of
Herakleopolitan sole rule before the advent of Dyn. 11 at Thebes.
The only source which could provide data on the length of this
period is the account of Manetho which has to be used with extreme
caution, however. As was argued elsewhere,37 Manetho’s data for the
length of Dyn. 9 can be dismissed since these numbers clearly could
have been created only after the artificial division of the dynasties had
occurred in post NK textual tradition. The information on the length
of Dyn. 10, however, derives from a year total which was present
already in earlier annalistic tradition and therefore could be of real his-
torical value. Taking Manetho’s figure at face value, the period of
Herakleopolitan rule before the start of Dyn. 11 would amount to some-
thing between 71 and 98 years, again depending on the date of the
re-unification of the country. Of course, there is no way to be sure
about the correctness of Manetho’s figure; if one chooses to disregard
Manetho’s data, however, the length of the Herakleopolitan dynasty
becomes entirely a matter of speculation, since there are no other
sources available for fixing the length of Herakleopolitan rule before
Dyn. 11. Nevertheless a number of excellent scholars have tried, hypo-
thetically linking the prosopographical and historical data which can
be derived from tomb inscriptions from UE, to reach a reconstruction
of the historical events of the earlier part of the FIP which would pro-
vide the basis for a tentative estimate of its length.38 It should be not
overlooked, however, how very problematic this type of reasoning is

37
For details see Seidlmayer, “Bemerkungen” (n. 23), 86–88.
38
An intermediate estimate between Beckerath’s minimum solution and the Mane-
thonian maximum model was proposed e.g. by Fischer, Dendera, 131 n. 579.
first intermediate period 167

bound to remain and how very unlikely it is that the dispersed pieces
of fragmentary evidence which were preserved by mere chance could
indeed link up to form a historically coherent picture.
The customary “short model” and the Manethonian “long model”
result in two profoundly different views about the FIP. Therefore a
look at the documentation for the period beyond the material relating
to dynastic history in the strict sense could provide indications whether
a short or a long model is more likely. Basing the argument solely on
material from the royal sphere, as is often done in discussions of dynas-
tic chronology, leads inevitably to a foreshortened perspective for peri-
ods of weak kingship. The extensive prosopographic data from the FIP
led Brovarski and Spanel to conclude that a succession of several gen-
erations of local administrators held office in many UE towns between
the end of the OK and the beginning of Dyn. 11,39 thus clearly favour-
ing a long model for the period in perfect accord with the data of
Manetho. Also, the archaeological record can be made to bear on the
problem. As was argued by Ward and Seidlmayer, the large number
of burials in Upper Egyptian cemeteries which are to be dated to the
earlier part of the FIP, as well as the fundamental morphological change
which can be discerned in the archaeological material exactly in this
phase, argue for a period of several generations.40 Therefore, substan-
tial evidence seems to support Manetho’s figure for the length of the
Herakleopolitan period. Nevertheless, the chronology remains on shaky
ground. Since the discovery of new historical sources cannot be pre-
dicted (although the excavations at the site of Herakleopolis Magna
could certainly turn up new evidence), and since the potential of radio-
carbon dating for this period seems to be limited, current efforts to
establish a dendrochronological series which covers the second millen-
nium BC might offer the prospect of substantial progress.41

39
E. Brovarski, “The Inscribed Material of the First Intermediate Period from Naga-
ed-Dêr”, AJA 89 (1985), 581–583; idem, The Inscribed Material of the First Intermediate
Period from Naga-ed-Dêr (Diss. Chicago, 1989) I, 15–53; D. B. Spanel, “The Date of
Ankhtifi of Mo"alla”, in GM 78 (1984), 87–94; idem, Beni Hasan in the Herakleopolitan
Period (Diss. Toronto, 1984).
40
W. A. Ward, Egypt and the East Mediterranean World 2200–1900 BC (Beirut, 1971),
10–11; S. J. Seidlmayer, Gräberfelder aus dem Übergang vom Alten zum Mittleren Reich
(Heidelberg: SAGA 1, 1990), 378; idem, “Anmerkungen” (n. 23), 84.
41
See below Cichocki, chapter III. 3, with additional literature.
II. 7 THE RELATIVE CHRONOLOGY OF THE MIDDLE
KINGDOM AND THE HYKSOS PERIOD (DYNS. 12–17)

Thomas Schneider

1. Terminology and Methodology

Substantial progress has been made in the study of the relative Chronology
of the MK and Hyksos Period (MK/SIP)1 since the presentation of the
state of our knowledge fifteen years ago by D. Franke2 thanks to new
discoveries and the reappraisal of core issues. Moreover, there have
been impulses to improve the terminological precision—particularly due
to K. S. B. Ryholt’s reassignment of dynasty designations—that are
basically desirable.
Ryholt3 has reassigned the term “Dyn. 16”, which had traditionally
been used to designate vassals of the Dyn. 15 Hyksos, although it was
known that this was based on an erroneous reading of the Manethonian
tradition,4 to a sequence of Theban rulers which would have been listed
in TC X, 31–XI,14 and which he identifies as a “First Theban Dynasty”.5
While the term “Dyn. 17” is generally understood as referring to all
of the Theban rulers between Dyns. 13 and 18 who were believed to
be listed in TC X, 31–XI, 14, this term is used by Ryholt only for a

1
The designation of the period is a historiographical, rather than a chronological,
concern. Cf. Franke, “Chronologie II”, 245–274, esp. 245f.; idem, Heqaib, 77–78;
Schneider, Ausländer, 155–156; differently, Ryholt, Situation, 311. For the present chap-
ter, I prefer the title “MK and Hyksos Period”, as Ryholt’s use of “Intermediate Period”
is based on the alleged political division of Egypt since the end of Dyn. 12, which is
not necessarily valid (cf. below) whereas our term corresponds to the period. I would
like to thank M. Bietak, D. Franke, E. Hornung, D. Polz and A. Spalinger who read
and commented upon earlier drafts of the text. Bibliography has been considered until
spring 2002 when the manuscript was delivered.
2
Franke, “Chronologie I. II.”
3
Ryholt, Situation. Reviews: D. Ben-Tor, S. Allen, J. P. Allen, “Seals and Kings”,
BASOR 315 (1999), 47–74; R. Holton Pierce, Acta Orientalia 60 (1999), 207–213; J. v.
Beckerath, AfO 46/47 (1999/2000), 433–435; W. Grajetzki, OLZ 95(2000), 149–156;
A. Dodson, BiOr 57(2000), 48–52; A.S. Spalinger, JNES 60 (2001), 296–300.
4
Beckerath, Untersuchungen, 17–20, Schneider, Ausländer, 123.
5
Ryholt, Situation, 151.
middle kingdom and the second intermediate period 169

chronologically later group of Theban kings, his “The Second Theban


Dynasty”. The rulers of Near Eastern origin, but not members of Dyn.
15, are then designated as “Dyn. 14”.6
In a response, Beckerath has suggested that the entire sequence of
Theban kings between Dyn. 13 and 18 should henceforth be termed
“Dyn. 16”, and to use the term “Dyn. 17” for the rival dynasty at
Abydos postulated by Ryholt.7 Depending upon the course and out-
come of the discussion (for details, cf. below section 4) it might be use-
ful to maintain “Dyn. 17” for the (larger number of ) Theban kings.
In order to avoid the menace of conceptual confusion, this contri-
bution will distinguish the different terminologies by adding the expo-
nents “T” (for traditional) and “R” (for Ryholt) to positions of kings where
these vary in the chronological reconstruction. Furthermore, alternative
dynasties differing from the customary usage will be put in quotation
marks, e.g. Dyn. 16 signifies the conventional Dyn. 16, whereas “Dyn.
16” refers to the terminological reassignment by Ryholt or later. The
numbering of the columns in the TC follows Gardiner’s standard pub-
lications whereas different systems of numbering are explicitly stated
(“Ryholt’s ninth column”).
Methodologically, it is important to note that a chronological frame-
work for the period can be established that extends from Dyn. 12 to
Dyn. 13 where it can be roughly fitted into Dyns. 15 and 17 (or “Dyns.
16/17” according to Ryholt). In its original state, the TC recorded at
least 50 kings for Dyn. 14, for whom, however, the length of the reign
is preserved for only a few, and only two ('#-s˙-r' N˙sj; Mrj-≈f #-r") are
otherwise attested through monuments. The rulers known from scarabs
are not among those listed in the entries for Dyn. 14 in the TC. Ryholt’s
attempt to create a typological sequence of scarabs upon which to
build chronologically acceptable successions has encountered consider-
able criticism.8 As other hypotheses are likewise difficult to support (an
overlap of early Dyn. 14 and the end of Dyn. 12;9 the relationship be-
tween Dyns. 13 and 14 in the sense of a “trade agreement”),10 Dyn. 14

6
Ryholt, Situation, 94ff.
7
Cf. below, section 4. A decision depends largely upon the interpretation of TC XI.
8
Ben-Tor et al. (n. 3), 53–65.
9
Ben-Tor et al. (n. 3), 55, 59, 66.
10
Ben-Tor et al. (n. 3), 59f.; Spalinger (n. 3), 299; more positive, Beckerath (n. 3),
434; Grajetzki (n. 3), 153–154 (who does point out that of 600 seals from Dyn. 14,
only one was found in Ryholt’s postulated capital of Dyn. 14, Avaris/Tell el-Daba).
170 thomas schneider

disappears as an alternative to Dyn. 13 as a chronological link between


the MK and Dyns. 15 and 17 (“Dyns. 16/17” according to Ryholt).
Ryholt weeds out a number of fragments of TC col. X which allegedly
record fictitious royal names, and places them in a postulated addi-
tional column of gods and demigods at the start of the papyrus (new
second column) with the result that the numbering of the columns
would increase by one for all following columns.11

2. Dynastie 12

The more recent chronological discussion on Dyn. 12 has been dom-


inated by the issue of coregencies, whereas the succession of rulers
and their reign-lengths are in principle resolved.12 After Franke in his
survey in 1988 assumed the validity of the coregencies of Dyn. 12,13
C. Obsomer has not only rejected the alleged 10-year coregency of
Amenemhet I and Senwosret I, but all other coregencies of Dyn. 12
as well, in the extensive discussion of the problem in his study of the
reign of Senwosret I (as R. D. Delia and W. Helck before him).14 Delia
remarked about this: “Obsomer’s reconstruction largely rests upon an
all-or-nothing foundation. If one is unconvinced that Obsomer has sat-
isfactorily explained away all of the coregency evidence, then much of

11
Ryholt, Situation, 24–25. In some responses to Ryholt’s work, the elimination of
the allegedly fictitious names of the TC have been met with enthusiastic agreement:
Ben-Tor et al. (n. 3), 315, 49; Beckerath (n. 3), 433; idem, Handbuch, 282; Dodson
(n. 3), 49. Note however, that Ryholt’s proposal (Situation, 24 n. 59) takes part of its
legitimation from considering Redford’s efforts to recognize West Semitic names in the
notations a failure. A modified version of the approach has been offered by Schneider,
Ausländer, 99–122; the doubts mentioned there (100) on the reorganization of the frag-
ments have not been eliminated. There is also the question whether names such as
“Protector of noble women” etc. would be more plausible among the mythical demigods.
Note also that Ryholt moves fr. 42 with the divine name “Apis” whereas fr. 123 with
the same name “Apis”, followed by two entries written in group writing, is left in the
section of the SIP. Fr. 22 which Ryholt also shifts might preserve the name of a ruler
known from scarabs, 'nt-hr (Schneider, Ausländer, 134), and fr. 152 another king known
from scarabs, Pns (Schneider, Ausländer, 106–109, 140).
12
An absolute chronological date in the early 12th dynasty could perhaps be pro-
vided by the astronomical reference of Sesostris’s temple at Karnak, cf. L. Gabolde,
Le „Grand Chateau d’Amon de Sesostris Ier à Karnak“ (Paris, 1998), 123–134; but see below
Belmonte, Chapter III. 5.
13
D. Franke, Chronologie II”, 114–125. In idem, Heqaib, XII, he shares the view
rejecting the coregency of Amenemhet I and Senwosret I.
14
C. Obsomer, Sésostris I er. Étude chronologique et historique du règne (Bruxelles, 1995).
middle kingdom and the second intermediate period 171

the structure of his history and chronology crumbles although many


gems remain in the details.”15 After a detailed investigation of Obsomer’s
arguments and the entire situation, K. Jansen-Winkeln came down clearly
in favour of coregencies in 1997.16
In fact since Franke’s summary, a number of new documents and
archaeological evidence, particularly from the sites of the pyramid com-
plexes of Dyn. 12, has appeared which objectively speaking can hardly
be interpreted in a different manner. In the fundamental case, of the
first royal transition from Amenemhet I to Senwosret I, the 10-year
coregency of these first two kings is favoured not merely by the two
main witnesses to which appeal is usually made (Stela Cairo CG 20516
naming both rulers in the field at the top with the regnal years “30”
and “10” and Louvre C1 with a date naming both kings).17 An archi-
trave from Matariya published in 1990 names both kings symmetri-
cally with their titularies and apparently as co-reigning builders; both
are designated as nsw bjt and living Horus (i.e. as reigning king).18 Finally,
the control marks from Lisht published by F. Arnold reveal that it was
only in regnal year 10 of Senwosret I that the construction of his pyra-
mid began, i.e., apparently after the death and burial of Amenemhet
I in his pyramid complex.19 The description of the attack on Amenemhet
I in the Teaching of Amenemhat I may have served, as Jansen-Winkeln
suspected,20 to legitimize the introduction of the institution; making, in
any case, “a decision in favour of a coregency practically unavoid-
able”.21 The length of the reign of Senwosret I is somewhat more than

15
R. D. Delia, JARCE 34 (1997), 267–268; sceptical is also C. Gallorini, DE 39
(1997), 135–137.
16
K. Jansen-Winkeln, “Zu den Koregenzen der 12. Dynastie”, SAK 24 (1997),
115–135. Beckerath, Chronologie, 133, also comes out in favour of the coregencies.
17
Jansen-Winkeln (n. 16), 122–125.
18
Both are also wished “eternal life”, which clearly underscores that Amenemhet I
could not have been dead at the time (cf. Ryholt, Situation, 273); A. Awadalla, “Un
document prouvant la corégence d’Amenemhat et de Sesostris I”, GM 115 (1990),
7–14; Jansen-Winkeln (n. 16), 125.
19
F. Arnold, The Control Notes and Team Marks (New York: PMMA 23, 1991) 19ff.,
30ff.; Jansen-Winkeln (n. 16), 125–126.
20
K. Jansen-Winkeln, “Das Attentat auf Amenemhat I. und die erste ägyptische
Koregentschaft”, SAK 18 (1991), 241–264; idem (n. 16), 128–135. Cf., differently,
N. Grimal, “Corégence et association au trône: l’Enseignement d’Amenemhet Ier”,
BIFAO 94 (1994), 143–172.
21
L. M. Berman, Amenemhet I (New Haven: Ph.D. Yale, 1985), 173–213. R. Leprohon,
“The Programmatic Use of the Royal Titulary in the Twelfth Dynasty”, JARCE 33
(1996), 167 now supports the coregency.
172 thomas schneider

45 years (as deduced from the highest attested date in a graffito south
of Amada and the entry in the TC, whereas Manetho has 46 years).
A coregency of roughly 3 years with his successor can be plausibly
deduced from the stela Leiden V.4, where apparently, regnal year 44
of Senwosret I is made equivalent to regnal year 2 of his son Amenemhet
II.22 A still more explicit equivalency of regnal years using the prepo-
sition ¢ft comes from the coregency of Amenemhet II and Senwosret
II, whose regnal year 3 is identified as the regnal year 35 of his pre-
decessor in a stela near Konosso.23 This date is also the highest recorded
regnal year for Amenemhet II.
A regnal year 8 (9?) is preserved for Senwosret II on a stela from
Toshka; and a “year 8” is also now known from the stela Cairo JE
59485.24 It is therefore necessary to emend accordingly the Manethonian
tradition and the figure of “19 years” in TC VI, 23 (see below). There
would not appear to have been a coregency with his successor Sen-
wosret III.
The debate over a short (19 year) or a long (39 year) reign for
Senwosret III can now be viewed as resolved. A regnal year “39” found
in the valley temple of the cenotaph of Senwosret III at Abydos in
1994 confirms a long reign,25 supported by a control mark of year “30”
from the royal pyramid complex at Dahshur,26 and a reference to a
first sed-festival of this king27—as had already been argued by W. Helck,
W. Barta and J. v. Beckerath; TC VI, 24 indicates 30+ regnal years.
In terms of relative chronology, this is not relevant as there had been
a coregency with Amenemhet III since year 20,28 and thus regnal year
20 of Senwosret III is identical to regnal year 1 of his son. It is in this

22
Jansen-Winkeln (n. 16), 117–118.
23
Jansen-Winkeln (n. 16), 118–119. See also W. K. Simpson, “Studies in the Twelfth
Egyptian Dynasty III: Year 25 in the Era of the Oryx Nome and the Famine Years
in Early Dynasty 12”, JARCE 38 (2001), 7f.: Regnal year 43 of Senwosret I corre-
sponds to (¢ft) year 25 of the nomarch Amenemhet.
24
M. C. Stone, “Reading the Highest Attested Regnal Year Date for Senwosret II:
Stela Cairo JE 59485”, GM 159 (1997), 91–99.
25
J. M. Wegner, “The Nature and Chronology of the Senwosret III—Amenemhet
III Regnal Succession: Some Considerations Based on New Evidence from the Mortuary
Temple of Senwosret III at Abydos”, JNES 55 (1996), 249–279; cf. idem, The Mortuary
Complex of Senwosret II (Philadelphia: PhD Thesis, 1996), 416.
26
F. Arnold, “New Evidence for the Length of Reign of Senwosret III?”, GM 129
(1992), 27–31.
27
D. Arnold & A. Oppenheim, “Reexcavating the Pyramid Complex of Senwosret
III at Dahshur”, KMT 6/2 (1995).
28
Jansen-Winkeln (n. 16), 119–120.
middle kingdom and the second intermediate period 173

fashion that one must understand the change of dates in pBerlin 10055
from Illahun where a “year 19” is followed by a “year 1”, which has
previously been used by those favouring a short reign (M. Bietak,
D. Franke) and would entail an emendation in Manetho and the TC.29
Chronologically, it is these 19 years that must be accounted for in a
relative chronology.
Until recently, the alternatives for Amenemhet III did not affect
chronology: whether proposing a short reign for Senwosret III without
a coregency (Franke) or a long reign with a 20-year coregency. With
Senwosret’s long reign confirmed, a chronological consequence would
result if one rejected the coregency (as do Delia30 and Obsomer).31
However, the coregency is also confirmed by the presence of the names
of both kings on scarabs and cylinder seals as well as fragments of
inscriptions of the coronation ritual of Amenemhet III which was appar-
ently performed by Senwosret III.32
The highest explicit regnal year attested for Amenemhet III is the
45th, but a “year 46” in the Illahun letters should probably also be
assigned to him. A short coregency with Amenemhet IV is assured by
a double date in a rock inscription at Semna (RIS 7), where the first
regnal year of Amenemhet IV is made equivalent to regnal year 44
(or 46 or 48?) of Amenemhet III. The coregency is also supported by
representations of two kings from the pyramid complex of Amenemhet
III in Hawara.33
The TC records a reign of 9 years, 3 months and 27 days for
Amenemhet IV. The highest date known from inscriptions is that of
the 9th year (Sinai 122), but a “year 10 (?)” from the Illahun papyri
may belong to him as well. His successor Nofrusobek reigned for 3 years,
10 months and 24 days according to the TC.34 Her highest date doc-
umented epigraphically is regnal year 3 on a Nile level mark at Kumma.

29
See also Ryholt, Situation, 212 n. 728, who still assumes a short reign for Senwosret
III and a one-year coregency with Amenemhet III.
30
R. D. Delia, A Study of the Reign of Senwosret III (New York: PhD Thesis, 1980),
239–253.
31
Obsomer (n. 14).
32
Jansen-Winkeln (n. 16), 120; Ryholt, Situation, 212 n. 728.
33
Cf. Franke, “Chronologie,” 120; Ryholt, Situation, 209–210 with n. 716. This core-
gency is also supported by Leprohon (n. 21), 170.
34
For her, cf. also V. G. Callender, “Materials for the Reign of Sebekneferu”, in:
Eyre, Proceedings, 227–236; S. Roth, Die Königsmütter des Alten Ägypten von der Frühzeit bis
zum Ende der 12. Dynastie (Wiesbaden: ÄAT 46, 2001), 242–245 (against Ryholt’s hypoth-
esis of a kinship and its implications).
174 thomas schneider

The original text of the TC seems to have corresponded to the dates


known from epigraphical sources in almost every instance. The reigns
preserved in the TC are:

Amenemhet I (2)9 years (x) months (x) days


Senwosret I 45 (x) (x)
Amenemhet IV 9 3 27
Nofrusobek 3 10 24

Rests of the entries (without names) of the four reigns between Senswosret
I and Amenemhet IV are probably preserved on fr. 67 and have been
discussed most recently by Krauss,35 Franke,36 and Ryholt.37 The four
lines give these numbers of regnal years:

TC fr. 67,1 10 (or 20/30?) + x


67,2 19
67,3 30 (+ x)
67,4 40 (+ x)

Of these, only TC fr. 67,2 has to be emended (19 instead of 9). The
entire picture would thus be:

King Highest Year Coregency with Successor Turin Canon

Amenemhet I 30 10 years 29 years


Senwosret I 45 2–3 years 45 years
Amenemhet II 35 3 years 10/20/30 + x years
Senwosret II 8/9 – 19 years
Senwosret III 39 20 years 30 + x years
Amenemhet III 46 1 year (?) 40 + x years
Amenemhet IV 10 9 y., 3 m., 27 d.
Nofrusobek 3 3 y., 10 m., 24 d.

35
Krauss, Sothis, 194–195.
36
Franke, “Chronologie I,” 122, 126.
37
Ryholt, Situation, 14–15.
middle kingdom and the second intermediate period 175

The total length of Dyn. 12 would be the sum of the reigns minus the
coregencies, which amounts to ca. 181 years according to this table.
This would confirm the year sum recorded by Eusebius of “182”. By
contrast, the sum in TC VI, 3 of “213 years, 1 month; 15/17/19 days”
was simply the total of the individual reigns without subtracting the
coregencies.38

3. Dynastie 13

Ryholt’s monograph provides a comprehensive new discussion of Dyn.


13 which necessarily entails numerous shifts and reassignments of chrono-
logical positions. TC VII, 5—VIII, 27 listed a total of 51 kings, which
he raises to 57, based on the restoration of lacunae which he postu-
lates for the three wsf-notes in VII, 6, VII, 7 and VII, 17.39 Manetho
gives a total of 60 kings for Dyn. 13. In order to make the changes
clear, I provide a table with both the traditional sequence (according
to Franke) and the new one proposed by Ryholt in Fig. II. 7.1.
According to TC VII, 5, a king Wegaf founded Dyn. 13. As earlier
H. Stock and K. A. Kitchen, Ryholt shifts this king 20 places so that
he follows Amenemhet VII, since the names of Wegaf appear on both
sides of the titulary of Amenemhet VII on the back of a statue base
dedicated to Month in Medamud.40 However, the fact that the name
Wegaf was later added to the statue does not necessarily demand the
assumption of a later reign (after Amenemhet VII). As only about 30
years separate the reign of Amenemhet VII and the foundation of the
dynasty, it is also conceivable that a later reference to the founder of
the dynasty by a king who may have been related to him was desired.
In Ryholt’s reconstruction, Sekhemre'-khutawy Sobekhotep I who is
traditionally identified as Sobekhotep II and listed as the 16th king of
the dynasty, opens Dyn. 13 which is justified by assuming an erroneous
exchange of names in the TC. A definite judgement on the matter is
not possible at present.41

38
Ryholt, Situation, 16 (partially based on other figures; differently also Franke,
“Chronologie I,” 122, 126f.; Krauss, Sothis, 198; Beckerath, Chronologie, 134.
39
Ryholt, Situation, 72.
40
Ryholt, Situation, 317–318.
41
Positive is Allen in: Ben-Tor et al. (n. 3), 50; cf., however, Franke, “Chronologie II”,
249.
176 thomas schneider

Fig. II. 7.1


middle kingdom and the second intermediate period 177

Ryholt postulates a crisis of legitimacy at the start of Dyn. 13, on


the assumption that Dyn. 14 had recently come to power. The dou-
ble names of early Dyn. 13 would indicate a filiation that was to legit-
imize their rule. In this vein, a name such as Ameni Qemau would
mean “Ameni’s (Amenemhet’s) son Qemau”. However, the proposed
early date for Dyn. 14 rests on very uncertain foundations and has
probably to be rejected.42 By contrast, the interpretation of the double
names as filiations (without any further political implications) appears
very plausible.43 In the 2nd and 4th positions of the TC are “Amenem-
hetsenbef ” and “Amenemhet” (V). Whereas Ryholt proposes reading
“Amenemhet (IV)’s son Senbef ”, earlier scholars had considered that
the two were identical.44 The assumption of a name indicating filiation
has implications for other relative positions at the start of Dyn. 13.
King Ameni Qemau, who had not been assigned a firm place until
now, probably also occupied a position in early Dyn. 13. If his dou-
ble name is to be understood as a filiation, Qemau could, following
Ryholt, be understood as the son of Ameni = Amenemhet V, and
assigned the 5th position. A more convincing solution is that proposed
by J. P. Allen where the fragmentary entry TC VI, 7 “Amenemhet” is
believed not to denote Amenemhet V, but Amenemhet Qemaw, and
thus the lacuna of TC VI, 6 would once have held the name of
Amenemhet V.45 The 5th position (Sehetepibre') is shifted to place 10
by Ryholt.
The 3rd position in Dyn. 13, where only a figure of 6 regnal years
is preserved in the TC, remains unclear. Ryholt fills this with a Nerikare,
known from a Theban stela, on the basis of a Nile level mark at Semna,
but the Semna record should be read in a different fashion.46 Beckerath
proposes the hypothetical Pentjini here, who is a ruler in the parallel
provincial line in Abydos according to Ryholt.47
The following five rulers (places 6–9) are identical in both the tra-
ditional arrangement and in the new reconstruction, with only the last
two Har-nedjheritef and Sehetepibre' having been switched. The TC

42
Cf. the references above in note 10.
43
Beckerath (n. 3), 434; Allen in: Ben-Tor et al. (n. 3), 50; Dodson (n. 3), 50.
44
Ryholt, Situation, 208, 212.
45
Allen in: Ben-Tor et al. (n. 3), 50.
46
Beckerath (n. 3), 434. The existence of this king had earlier been thrown into
doubt by L. Gabolde, “Nerkarê, a-t-il existé?”, BIFAO 90 (1990), 213–222.
47
Beckerath, Chronologie, 137.
178 thomas schneider

records the throne name of Sehetepibre' for both so that an error must
be in place in the case of Har-nedjheritef (with a throne name Hetepibre').
The reason for Ryholt’s proposed switch of Har-nedjheritef (read as
Sihar-nedjheritef by him) is the assumption of a filiation, as m¡w z¡
precedes the name in the cartouche. Ryholt takes the word “son” as
the first part of the personal name, and interprets the sequence of the
two names as a filiation: Qemaw’s son Sihar-nedjheritef. Accordingly,
he places the king immediately after his hypothetical father Qemaw.
However, as the inverted filiation can be written with or without the
word for “son”, the proper name can still be read Har-nedjheritef; even
as the son of Qemaw he could have come to power a few years after
his father’s reign.48
The next five positions are once again identical (Franke 10–14 =
Ryholt 11–15). One should merely note that Sobekhotep bears the ordi-
nal number “II” in Ryholt’s reconstruction because the traditional
Sobekhotep II has been shifted from the 16th place to the first. The
last of these five kings is Hor (I) Awibre', whose Horus name is attested
on a reused block in Tanis together with the Horus name Khabaw.
Ryholt therefore concludes that Khabaw Sekhemre'-khutawy was the
successor of Hor (I) Awibre' (places 15 and 16), whereas the lost 3rd
position of Dyn. 13 in the TC was customarily reserved for him.49
Accounting for the fact that neither does Khabaw Sekhemre'-khutawy
follow Awibre' in the TC, nor does a certain Djedkheperu (Horus name)
attested in seal impressions from Uronarti alongside Khabaw’s, but
immediately Amenemhet VII Sedjefakare' (TC VII, 18), Ryholt postu-
lates a lacuna of four places in the Vorlage of the TC, to be filled by
Khabaw, Djedkheperu, Seb and Kay.50 To Djedkheperu he assigns the
Osiris bed from Abydos,51 and suggests that the effaced titulary con-
sisted in the name of Hor (I Awibre) as the beginning of his proper
name and simultaneously filiation, but J. P. Allen has plausibly argued
that it was Pentjini who was responsible for the inscription.52

48
N. Swelim & A. Dodson, “On the Pyramid of Ameny-Qemau and Its Canopic
Equipment”, MDAIK 54 (1998), 330, suggest that Har-nedjheritef was only the fourth
successor of his father.
49
Critically, and with an alternative Ben-Tor et al. (n. 3), 50.
50
Ryholt interprets the note wsf of the TC as an indication of a lacuna in the Vorlage,
not as a reference to a king whose name had been deliberately omitted. For this ques-
tion, cf. C. Bennett, “King Qemau: a Reconsideration”, GM 159 (1997), 11–17.
51
This was assigned to Khendjer by A. Leahy and to Nebiriraw II by Beckerath.
52
Allen in: Ben-Tor et al. (n. 3), 50–51.
middle kingdom and the second intermediate period 179

Ryholt postulates Seb and Kay as new kings of Dyn. 13, as he dissolves
the proper name Sebkay (Sbk #y) on the magic ivory CG 9433 ( JdE
34988) into the filiation, “Seb’s son Kay”. Since Amenemhet VII names
himself Kay-Amenemhet, three generations of kings would thus have
reigned within the span of a few years. This “daring construction”
( J. von Beckerath)53 which in addition requires the postulation of a
lacuna in the Vorlage of the TC does not seem necessary. If we do not
move Sobekhotep II from his traditional place 16 to the top of the
dynasty (against Ryholt), the proper name of the magic ivory can be
assigned to him. In this case we would have a pet form of a divine name
formed by means of the suffix -li. In the NK this suffix was written
<nr>, <nrj< (Hali for Hathor, Wurel for Weret),54 for which in Dyn. 13
a notation with <#> would have been mandatory. The sounding of the
name would thus have been something like Sabkuli. The possible
patronymic for Amenemhet (K #y = Kuli) would be the abbreviated form,
so that Sobekhotep II (and not an otherwise unknown K #y) should be
viewed as the father of Amenemhet. This would require a switch between
TC VII, 18 and TC VII, 19 at the most. Amenemhet VII is followed
in the traditional chronology by Sobekhotep (II) Sekhemre-khutawy,
and by Wegaf who has changed places with him according to Ryholt.
This takes us into uncontested territory: the following 16 kings of
Dyn. 13 are arranged in the same succession in the two competing
reconstructions, with a single minor variation: Sobekhotep Merhotepre'
would be inserted as Sobekhotep V between Sobekhotep IV Khaneferre'
and Sobekhotep VI (previously labelled ‘Sobekhotep V’), and his
identification with Ini Merhotepre' completely abandoned.55 A relief
from the reign of Neferhotep I depicting the enthroned prince of Byblos
Antin serves as the traditional basis for the synchronism between Egypt
and Babylonia for the first half of the second millennium BC (Neferhotep:
Jantin'ammu of Byblos: Zimrilim of Mari: Hammurabi),56 which is,
however, probably not reliable.57

53
Beckerath (n. 3), 434; a similar appreciation by Grajetzki (n. 3), 155; cf. Allen in:
Ben-Tor et al. (n. 3), 51.
54
T. Schneider, Asiatische Personennamen in ägyptischen Quellen des Neuen Reiches (Freiburg:
OBO 114, 1992), 276–277.
55
Ryholt, Situation, 231–232.
56
E. Hornung, “Lang oder kurz?—Das Mittlere und Neue Reich Ägyptens als
Prüfstein”, in: High, Middle or Low? 1, 27–36; K. A. Kitchen, “The Basics of Egyptian
Chronology in Relation to the Bronze Age”, in: High, Middle or Low? 1, 37–55; Franke,
“Chronologie II”, 273–274; Ryholt, Situation, 87–88.
57
C. Eder, Die ägyptischen Motive in der Glyptik des östlichen Mittelmeerraumes zu Anfang des
180 thomas schneider

Aya Merneferre' is the last ruler of the dynasty attested in both the
north and the south, whereas his successor Ini Merhotepre' is the first
to be only attested in southern Egypt. The lower part of TC’s column
VII (Ryholt’s eighth column)—after Sobekhotep VII Merikawre'; TC
VII, 8—is very badly preserved after a long gap. At the end of the
column (now lost) was probably the total for Dyn. 13 and the start of
Dyn. 14, which continued on column VIII (Ryholt’s ninth column),
with Nehesi in TC VIII,1.58 That Nehesi should have had five prede-
cessors which would justify the postulation of an additional lacuna in
the Vorlage of the TC 59 is hardly plausible.60
Of the lengths of the reigns of Dyn. 13 kings in the TC, the fol-
lowing entries display preserved year dates:61

TC VI, 5 Wegaf 2 years, 3 months, 24 days


TC VI, 6 wsf 6 years
TC VI, 7 Amenemhet V 3 or 4 years
TC VI, 8 Sehetepibre' 1–4 years
TC VI, 14 Nedjemibre' 0 years, 7 months
TC VI, 16 Reniseneb 0 years, 4 months
TC VI, 24 Sobekhotep III 4 years, 2 months
TC VI, 25 Neferhotep I 11 years, 1–4 months
TC VI, 26 Sihathor 0 years, 1 + x months, 3 days
TC VII, 1 Sobekhotep V T/VIR 4 years, 8 months, 29 days
TC VII, 2 Ibiau 10 years, 8 months, 28 days
TC VII, 3 Aya 23 years, 8 months, 18 days
TC VII, 4 Ini 2 years, 2–4 months,62 9 days
TC VII, 5 Sewadjtu 3 years, 2–4 months
TC VII, 6 Ined 3 years, 1 months, 1 day
TC VII, 7 Hori 5 years, ? months, 8 days
TC VII, 8 Sobekhotep VII 2 years, ? months, 463 days

2 Jts. v.Chr. (Leuven: OLA 71, 1996), 13; T. Schneider, ZDPV 114 (1998), 184–188.
Jantin itself would have been a hypocoristic abbreviation of Jantin'ammu, but what
we have is only "Antin.
58
For Nehesi, see now A. Loprieno, “N˙sj, ‘der Südländer’?”, in: Stationen, 211–217.
59
Ryholt, Situation, 94.
60
See above, notes 8–10. Spalinger (n. 3), 297 assumes (following Ryholt) that it is
demonstrated that Sheshi belongs in early Dyn. 14, whereas this assignment is dis-
puted by Ben-Tor et al. (n. 3), 61.
61
Cf. Ryholt, Situation, 192, and Fig. 10 with the sheet joins as opposed to Franke,
“Chronologie II”, 267ff.
62
Franke, “Chronologie II”, 267ff. has “2” months, Ryholt “3–4”; cf. however TC
VIII, 5, and thus “2–4”.
63
Ryholt, Situation, 192 has erroneously “3” days.
middle kingdom and the second intermediate period 181

Down to the reign of Aya, a minimum of 100 years must be accounted


for since the beginning of Dyn. 13. The last ca. 25 kings of Dyn. 13
are very badly documented in the epigraphic material, and their mon-
uments are restricted to a few places (Abydos, Thebes, Deir el-Bahri,
Gebelein, Edfu). “By the end of Aya’s 24-year reign, the administra-
tion seems to have collapsed [. . .] This period further witnessed no less
than 24 kings, whose reigns on the average lasted little more than one
year.”64 In place of a precise chronology, we can merely estimate a
total of some two decades, after which Dyn. 13 would be immediately
followed by the new Theban dynasty (Dyn. 17 according to the tra-
ditional version;65 “Dyn. 16” according to Ryholt).66 There are some
dated monuments for a few kings data for whose reigns are not pre-
served in the TC: Sobekhotep II T/I R (year 4), Amenemhet-Senbef (year
5), Khendjer (year 5), Sobekhotep IV (year 9) and a few additional of
contested attribution.67 The only indication of the total length of Dyn.
13 remains the total given by Manetho of 453 years, which is usually
emended to *153 or 133 (Beckerath) years and represents the basis for
approximate estimates of 133 years (Beckerath),68 some 130 years
(Franke),69 152 years (Kitchen),70 and 154 years (Ryholt);71 cf. our final
remarks in section 6.

4. Dyn. 17 (Ryholt: “Dyn. 16” & “Dyn. 17”)

4.1. Generalities Arranging the relative chronology of the kings attested


in the Thebaid between Dyns. 13 and 18 is one of the most difficult
challenges of the SIP. The traditional reconstruction of Dyn. 17 assumes
15 rulers, based on the sum in TC XI, 15 (which must apparently be
emended to “15”); these kings would have been named in TC X, 31–XI,
14.72 Against this, Ryholt (following H. E. Winlock) postulates two

64
Ryholt, Situation, 298.
65
Franke, “Chronologie II”, 259.
66
Ryholt, Situation, 302.
67
Ryholt, Situation, 193f. For the dates in pBoulaq 18 and pBrooklyn 35.1446 see
also Franke, “Chronologie II”, 254f. and esp. S. Quirke, The Administration of Egypt in
the Late Middle Kingdom. The Hieratic Documents (New Malden, 1990).
68
Beckerath, Untersuchungen, 220f.
69
Franke, “Chronologie II”, 265; table, 267ff.
70
Kitchen (n. 56), 45.
71
Ryholt, Situation, 195–196.
72
Beckerath, Untersuchungen, 194f.; Ryholt, Situation, 151.
182 thomas schneider

Theban dynasties, “one prior and one subsequent to the Hyksos con-
quest of the south”. He assignes all the aforementioned 15 places of
the TC to his first Theban dynasty, which he terms “Dyn. 16”, cor-
recting an error in the terminology of Africanus and traditional research
(where Dyn. 16 were considered princes and vassals of Dyn. 15), whereas
the designation “Dyn. 17” applies exclusively to his second Theban
dynasty.73 This “Dyn. 17” would have been listed in a now lost col-
umn of the TC, being separated from “Dyn. 16” by an alleged local
dynasty in Abydos (TC XI, 16–31).74
That allowance must be made for more space after the end of Dyn.
13 for kings of the Thebaid than is permitted in the traditional 15
places is an indisputable fact which means that Ryholt’s revision is a
welcome attempt. It is, however, reasonable to suppose that the cor-
rect reconstruction could have a different appearance. According to
Beckerath,75 a column might have been lost after TC X (with the first
entry being TC X, 31), and this column would have included all of
the Theban kings as a single dynasty; only then would one have to
add the rulers from Abydos. Beckerath proposes that all of the Theban
kings between Dyns. 13 and 18 be termed “Dyn. 16”, and that the
local dynasty at Abydos be termed “Dyn. 17”. A. Spalinger76 is also
opposed to assuming two Theban dynasties, divided by the postulated
Hyksos conquest of Thebes.
Ryholt concludes that the number of Theban rulers was 15 (“Dynasty
16”, listed in the TC ) + about 10 (“Dynasty 17”, known from the
monuments and lasting a maximum of ca. 40 years). Taking Beckerath’s
version of Ryholt’s reconstruction, the succession of Theban rulers would
be from TC X, 31 through Beckerath’s new column “Xa” to TC XI, 14
(with the total in TC XI, 15), which would thus include 1+31+14 =
46 places, nearly twice as many as in Ryholt’s approach. This is not
conceivable for chronological reasons and with regard to the extant
epigraphical evidence.

73
King Sekhemre'-wahkhaw Ra'hotep, traditionally viewed as second king of Dyn.
17, occupies the first place in Ryholt’s Dyn. 17 (Dyn. 17R/1 vs. Dyn. 17T/2) and is
thus moved 14 places away. Sobekemsaf (traditionally I, Ryholt: II) has been shifted
18 places.
74
Ryholt, Situation, 164.
75
Beckerath (n. 3), 434.
76
Spalinger (n. 3), 298: “There is little evidence for a Theban dynasty à la Ryholt
unless we reinterprete the data to suit the hypothesis.”
middle kingdom and the second intermediate period 183

Regardless of the judgment of the order in the TC, we must con-


cur with Ryholt that the total number of Theban rulers between Dyns.
13 and 18 cannot be higher than 25. However, a division of the
sequence into two dynasties based upon Ryholt’s postulated Dyn. 15
Hyksos conquest of Thebes is just as difficult to accept as Ryholt’s pro-
posed scorched earth policy during the withdrawal of Apopis from the
south.77 It is entirely possible that there was originally a division of two
groups of kings in the TC, but this may have resulted from other con-
siderations (cf. below). These 25 kings at the most seem not to have
ruled longer than 90–115 years.

4.2. Dyn. 17 (Ryholt: “Dyn. 16” & “Dyn. 17”)—Details The issue of
ordering these kings is extremely complex due to the state of preser-
vation of the TC and the epigraphic situation. Even for the more
important rulers and their families, the data is fragmentary and open
to diverse interpretations. Three problems of the interconnected fami-
lies of Inyotef/Sobekemsaf should suffice as exemplary in prohibiting
a definitive resolution of the issue.
1) Inyotef Nebukheperre'’s consort Sobekemsaf (N.B. “Sobek is his
(!) protection”), presumably the daughter (or granddaughter?) of a king,
seems to have been named after a king Sobekemsaf; and the name
Sobekemsaf was also borne by Ra'hotep’s consort (or mother?).78 Both
the identity of the Queen(s) Sobekemsaf and the identity of the king
Sobekemsaf are subject to debate.
2) Whether this king was the father of Inyotef Nebukheperre' or
merely an indirect predecessor depends upon the interpretation of a
newly found fragmentary inscription on the Luxor-Farshut road.79
3) The precise placement of the Inyotef kings, the Sobekemsaf kings
and the king Ra'hotep is subject to controversy. That a prince with
the basilophorous names Inyotefmose was praised by a king Sobekemsaf
for his actions during a Sokar festival could demand at the most the
placing of a king Sobekemsaf after the Inyotef kings.80
The following brief presentation attempts to provide a solution along
this line of arguments: A detailed discussion must take account of the

77
Ryholt, Situation, 143–148.
78
Ryholt, Situation, 265–266, 268ff.
79
J. C. Darnell & D. Darnell, “The Luxor-Farshût Desert Road Survey”, Oriental
Institute Annual Report (1992–93), 50, fig. 4; Ryholt, Situation, 270.
80
Ryholt, Situation, 170.
184 thomas schneider

aforementioned differences between Ryholt’s interpretation of the TC


and Beckerath’s response to it (cf. above, with Figure). One possible
solution is to reject the proposed additional column TC Xa (with Ryholt,
against Beckerath), while simultaneously assuming (with J. P. Allen,
against Ryholt)81 that all of the Theban rulers between Dyns. 13 and
18 were originally listed between TC X, 31 and TC XI, 25. A first
group (TC X, 31–XI, 14) would close with the total at TC XI, 15; and
then 9 additional rulers would follow in TC XI, 16–XI, 24 whose total
would once have been noted in TC XI, 25 (before the inception of fr.
163).82 The division into two groups could easily have been motivated
by their lineage or residence (e.g., Thebes vs. Dendera).83
The five kings mentioned on fr. 163 would only be inserted then,
and possibly others on a possible column “XII”. That these will have
been Abydenian rulers—Ryholt suggests that the remainder of his
Abydos dynasty was listed here—is not demonstrable since the traces
of the names do not match any epigraphically attested names.
Despite diverging on the reconstruction of TC XI, I agree with Ryholt
in the number of rulers who can be viewed as Theban kings between
Dyns. 13 and 18. In addition to the 15 places given in the TC (i.e.,
the traditional Dyn. 17), Ryholt names 9 kings of a “second Theban
dynasty” (i.e. his “Dyn. 17”). According to the placement of these rulers
in TC XI such as proposed above, there would in fact be exactly that
many places.
The first three places of TC XI are damaged and begin with S¢m-
r'-, S¢m-r'- and S¢m-r'-s-. The traditional restoration (since H. Stock) to
the throne names S¢m-r'-w # ˙-¢'w (= Ra'hotep), S¢m-r'-w # ≈-¢'w
(= Sobekemsaf I) and S¢m-r'-smn-t #wj (= Djehuti) has recently been
thrown into doubt by Beckerath himself, who considers Ra'hotep as
doubtful, the third place as completely uncertain, and views only
Sobekemsaf I as certain.84 On account of their building activity in
Abydos and their restorations in Medamud and Coptos, Ryholt places
Ra'hotep and Sobekemsaf Sekhemre'-wadjkhaw in his chronologically

81
Allen in: Ben-Tor et al. (n. 3), 49.
82
An alternative placing of the fragments is not possible; cf. W. Helck, “Anmerkungen
zum Turiner Königspapyrus”, SAK 19 (1992), 151–216.
83
Cf. below, text referring to n. 111.
84
J. v. Beckerath, “Theban Seventeenth Dynasty”, in: E. Teeter & J. A. Larson,
eds., Gold of Praise. Studies on Ancient Egypt in Honor of Edward F. Wente (Chicago: SAOC
58, 1999), 23–24. The earlier sequence can still be found in Beckerath, Chronologie, 139.
middle kingdom and the second intermediate period 185

Fig. II. 7.2


186 thomas schneider

later “Dyn. 17”85 and restores TC XI, 1–3 with the throne names S¢m-
r'-smn-t #wj (= Djehuti), S¢m-r'-swsr-t #wj (= Sobekhotep VIII) und S¢m-r'-
s'n¢-t #wj (= Neferhotep III), of whom the last two were traditionally
acknowledged as contemporary but had not been placed.86 Djehuti is
thus shifted only two places in contrast to the traditional sequence (Dyn.
17 T/4 and “Dyn. 16”R/2).87 Placing Neferhotep III here is favoured by
the position of Se'ankhenre' Mentuhotpi (Mentuhotep VI) in TC XI,
4, as two similarly executed stelae of the two kings could possibly have
come from the same workshop.88 The next five places TC XI, 5–9 are
preserved whereby TC XI, 6 (Nebiriaw II) could be a copying error89
and in TC XI, 9 Sekhemre'-shedwaset (as traditionally maintained)
might be a deliberate change for the correct Sekhemre'-shedtawy =
Sobekemsaf [I].90 The proper name of TC XI, 7, Semenenre', is un-
known; as the successor is named “Bebiankh” and “Bebi” is a possi-
ble abbreviation of a name containing the element “Sobek”,91 one could
speculate on a Sobek-name. Of the following five places (TC XI, 10–14),
only the insignificant beginning is preserved, before the total (TC XI,
15). TC XI, 15–17 apparently introduces a new group of kings.
The main difficulty here is the correct restoration of the five places
before the sum. If one places the traditional Dyn. 17 in TC X, 31–XI,
14, the last three places are occupied by Senakhtenre', Seqenenre' Tao
and Wadjkheperre' Kamose.92 Ryholt assigns to these positions the

85
Ryholt, Situation, 152, 170. Caution is, however, advised in basing the restoration
of Sobekemsaf Sekhemre'-wadjkhaw on the alleged destruction of the Thebaid by the
Hyksos (145–146; 170).
86
N. Dautzenberg, “Neferhotep III. und Sebekhotep VIII.”, GM 140 (1994), 19–25,
also views the second as being the immediate successor, but places them in Dyn. 13.
The same restoration for TC XI, 2 was accomplished by C. Bennett, “The First Three
Sekhemre Kings of the Seventeenth Dynasty”, GM 143 (1994), 21–28.
87
A connection to Dyn. 13 follows via the vizier Ibiaw, for which link one should note
the critical remarks by Grajetzki (n. 3), 151–152. Franke, Heqaib, 79 is more positive.
88
P. Vernus, “La stèle du roi Sekhemsankhtaouyrê Neferhotep Iykhernofret (Stèle
Caire JE 59635) et la domination Hyksos”, ASAE 68 (1982), 129–135 with pl. I; idem,
“La stèle du pharaon MnΔw-˙tpô à Karnak: Un nouveau témoignage sur la situation
politique et militaire au début de la D.P.I.”, RdE 40 (1989), 145–161, pls.6–7; Ryholt,
Situation, 154. Beckerath, Chronologie, 139, places Senwosret IV Seneferibre' at TC XI, 4.
89
Franke, “Chronologie II”, 263.
90
Following Beckerath, Untersuchungen, 168, 290, contra Ryholt, Situation, 156.
91
P. Vernus, Le surnom au Moyen Empire (Rome: Studia Pohl 13, 1986), 111 n. 122;
cf. the two sons of Sobekhotep VII Merkawre' with the names Bebi and Sobekhotep
(Ryholt, Situation, 235–236).
92
Franke, “Chronologie II”, 271.
middle kingdom and the second intermediate period 187

rulers Dedumose I, Dedumose II, Mentuemsaf, Mentuhotep (VIR)


Mer'ankhre' and Senwosret IV Seneferibre', who are documented epi-
graphically at Thebes but had not been positioned chronologically.
However, the Dedumose kings certainly belong in Dyn. 13.93 If one
follows the scheme outlined above, with the assumption that further
Theban kings down to Kamose were named in TC XI, 16–XI, 25, we
face the question of the placement of the kings Inyotef/Sobekemsaf/
Ra'hotep and thus of the exact relationship between the late Dyn. 13
and the kings named between TC X, 31 and TC XI, 25.
As Ryholt correctly notes, the traditional placing of Nebukheperre'
Inyotef (V) at the beginning of Dyn. 17 (his “Dyn. 16”), in the lost
entry at TC X, 31, cannot be justified. D. Polz, who succeeded in
finding the tomb of the king in Dra Abu’l Naga in 2001, argues like-
wise for situating him before the final kings of Dyn. 17.94 Such a place-
ment is likewise favoured by the evidence of the box of Minemhat,
who was mayor of Coptos under Inyotef V, which was part of the
funerary equipment of an Aqher who lived under Seqenenre'.95
From the legend on the coffin Louvre E 3019, it follows that Inyotef
Nebukheperre' (Dyn. 17T/1: Inyotef V; Dyn. 17R/4: Inyotef VII) arranged
the burial of his brother Inyotef Sekhemre'-upimaat (Dyn. 17T/11:
Inyotef VI; Dyn. 17R/3) and must therefore have followed him on the
throne.96 In his Untersuchungen, Beckerath had viewed Inyotef Sekhemre'-
upimaat (VI) and Inyotef Sekhemre'-herhermaat (VII) as brothers,
whereas he had separated Inyotef Nebukheperre' (VI; coffin BM 6652)
from them as a king he considered not necessarily related to them,
placing him at the beginning of the dynasty. Ryholt equally bases his
arguments upon a consistent palaeographic peculiarity (the Pleneschrei-
bung of “j”) in the case of the coffin of Inyotef Sekhemre'-herhermaat

93
Franke, Heqaib, 77–78.
94
D. Polz & A. Seiler, Die Pyramidenanlage des Königs Nub-Cheper-Re Intef in Dra' Abu
el-Naga. Ein Vorbericht (Maniz: DAIKS 24, 2003). J. C. Darnell views Inyotef Nebukheperre'
as the addressee of a newly discovered royal hymn at Wadi el-Hôl; however, the king
is not named (“A New Middle Egyptian Literary Text from the Wadi el-Hôl”, JARCE
34 [1997], 85–100).
95
H. Winlock, JEA 10 (1924), 258 with n. 1 (taking up an idea of P. Newberry);
W. Helck, “Der Aufstand des Tetian”, SAK 13 (1986), 126 (who erroneously refers to
a spoon; the spoon from the burial of Aqhor came from a mayor Sobekwer). In favour
of a placement in late Dyn. 17 (as traditionnally conceived) is N. Dautzenberg, “Die
Wahl des Königsnamens in der Hyksoszeit”, GM 159 (1997), 43–52, based on the
titulary.
96
Ryholt, Situation, 270.
188 thomas schneider

(Dyn. 17T/12: Inyotef VII; Dyn. 17R/5: Inyotef VIII; Louvre E 3020)
so that he becomes an ephemeral coregent of Inyotef Nebukheperre'.97
However, this “stock coffin” was certainly not originally intended for
Inyotef Nebukheperre', and to infer from it any hypothetical coregency
is doubtful in the extreme.98
Placing the three Inyotef kings together appears to be plausible, not
the least on account of their canopic equipment.99 It seems possible
that the legend on Louvre E 3020 can be used to argue that Inyotef
Sekhemre'-herhermaat (VIII) was the son of Inyotef (VI) Sekhemre'-
upimaat,100 which would produce the sequence: Inyotef (VI) Sekhemre'-
upimaat (“the elder”)—Inyotef (VII) Nebukheperre'—Inyotef (VIII)
Sekhemre'-herhermaat.
If we can identify Sobekemsaf, the consort of Nebukheperre' with
the mother of Ra'hotep (presumably) of the same name,101 then Ra'hotep
would have to be placed chronologically after Nebukheperre'.102 The
evidence is of similar ambiguity with regard to the mutual position of
Ra'hotep and Sobekemsaf Sekhemre'-wadjkhaw. If a prince Ameni
whom Ra'hotep grants a bow, arrows and the right to participate in
the ceremonies for Min of Coptos, is identical to the son-in-law of
Sobekemsaf Sekhemre'-wadjkhaw, then Ra'hotep would plausibly be “a
close predecessor of Sechemre'-wadjkhau Sobekemsaf ”.103 If the two are

97
Cf. Ryholt, Situation, 267–268.; the earlier discussions will be found in Beckerath,
Untersuchungen, 267f.
98
Dodson (n. 3), 50–51; Spalinger (n. 3), 300.
99
A. Dodson, The Canopic Equipment of the Kings of Egypt (London, 1994), 42; C. Bennett,
“The Date of Nubkheperre Inyotef ”, GM 147 (1995), 19–27 (22f.: all three Inyotefs
closely associated).
100
The coffin was not originally manufactured for Inyotef Nebukheperre' (A. Dodson).
When the coffin was reworked, the <y> of the name of Inyotef was replaced with
<' #>. An explanation can be found in assuming that the correction was an attempt at
specification: the proper name Inyotef, “The one who brings back his father” is an
Ersatzname (a newborn child is believed to replace a recently deceased relative, in this
case the father). The insertion of <' #> and the addition of the throne name would
have adapted the meaning of the name to the contemporary situation: “he who brings
back the elder, his father, Sekhemre'-herhermaat”. As the epithet <' #> “the elder” is
only attested for Sekhemre'-upimaat, it would follow that Inyotef Sekhemre-herher-
maat was the son of Inyotef Sekhemre'-upimaat.
101
For the discussion, cf. Ryholt, Situation, 265–268 (who rejects the identification).
102
Ryholt, Situation, 170, opposes this order: “Likewise, it may be argued that
Ra'hotep, whose restoration of the temple of Min at Coptos is described on a stela,
was a predecessor of the Inyotef kings, since such a claim could hardly have been
made in the years subsequent to the reign of Inyotef N[ebucheperre'] who built exten-
sively at this temple.”
103
C. Bennett, “The First Three Sekhemre Kings of the Seventeenth Dynasty”, GM
143 (1994), 21–28.
middle kingdom and the second intermediate period 189

not the same,104 then the chronological place of Sobekemsaf Sekhemre'-


wadjkhaw is completely uncertain; rather than placing him, with Ryholt,
after the Inyotef kings,105 he could also precede them, perhaps after
Sekhemre'-shedtawy Sobekemsaf (I) or even in the lost entry at TC X,
31.106 A possible candidate for TC X, 31 or one of the other open
places could, however, also be seen in Sekhemre'-neferkhaw Wepwaw-
temsaf (previously unplaced, assigned by Ryholt to the Abydos dynasty).
In principle, the block of the Inyotef kings and Ra'hotep could have
stood in TC XI, 10–14 or else in TC XI, 18–21 (with the four fol-
lowing entries to be restored as Senakhtenre/Seqenenre'/Kamose/total).
In view of these possibilities, of particular interest are the two entries
where the beginning is still preserved in TC XI, 16 and TC XI, 17,
which Ryholt views as undocumented throne names of kings of his pos-
tulated Abydos dynasty. In TC XI, 16 Weser-. . . r' is preserved, and in
TC XI, 17 probably only Wsr without the solar disk,107 i.e., there is a
proper name as in TC VIII (Ryholt: IX), 1 (N˙sj), TC XI, 5/6 Nb-jrj-
#w, and elsewhere.108 For both places, which have not hitherto been
explained, a restoration can be suggested:

• In the tablet of kings from Karnak, the fourth row (no. 27) has a
royal name Wsr-n-r' along with Seqenenre', Senakhtenre' and Inyotef
Nebukheperre', who has usually been emended to the Swsr-n-r' named
in TC XI, 8.109 However, the kings to be placed in TC X, 31–XI,
14 (including Swsr-n-r' ) are certainly listed in other rows of the tablet
of kings if the assignment of kings to the TC passage as proposed
above can be accepted.
• A king who fits well with the notation beginning Wsr- at TC XI, 17
is the hitherto unsituated Senwosret (IV) Seneferibre', known from
a colossal statue and a stela in Karnak as well as blocks from Tod
and Edfu.110

104
Ryholt, Situation, 266, 272 (the names display different endings: Amanja, Amana).
105
Decisive for Ryholt is the praise of a king Sobekemsaf for a Inyotefmose, whom
Ryholt places after the Inyotef kings due to his basilophorous name.
106
Franke, Heqaib, 84: the dyad of king Sobekemsaf Sekhemre'-wadjkhaw with Satet
is characteristic for the beginning of Dyn. 17.
107
The weser-sign follows on the cartouche more closely than in TC XI, 16.
108
In detail: Ryholt, Situation, 27–28.
109
Beckerath, Untersuchungen, 27; idem, Handbuch, 126. It is less probable that the Beni
Hasan graffito (cf. Beckerath, Untersuchungen, 69) should be read as Weserre'[-nefer-
khaw].
110
Ryholt, Situation, 391 (16/e), 157, 306; Beckerath, Untersuchungen, 255 (13 F), 62.
190 thomas schneider

In the lacuna between TC XI, 17 and XI, 26, I would propose plac-
ing the kings Senakhtenre', Seqenenre' Tao, Wadjkheperre' Kamose
(and possibly others) as well as the total. The rulers named in the lower
part of TC XI would thus be the kings of Ryholt’s “Dyn. 17” (whom
he believes to have been listed on a lost * twelfth column). Their sep-
aration from those named in the upper part of the column could have
been motivated by this family’s suggested provenance from Dendara
and its struggle with the supplanted Inyotef family that continued until
early Dyn. 18.111
Frag. 163 placed at the end of TC XI has a throne name ending
in—˙ #b in line 2 6, and another in line 31 might be constructed with
wbn. A parallel can be found in TC VIII, 4 (S˙ #b-r' ) and TC VIII, 11
(Wbn-r' ), whereas there are absolutely no corresponding throne names
in Dyns. 17–19.112
What follows is a listing of TC X, 31–XI, 26 with the hypothetical
restorations proposed above in the right column:

TC X, 31 --- Sekhemre'-neferkhaw
Upuautemsaf ??
TC XI, 1 Sekhemre'- Sementawy Djehuti
TC XI, 2 Sekhemre'- Seusertawy Sobekhotep VIII
TC XI, 3 Sekhemre'-Se- -'anchtawy Neferhotep III
TC XI, 4 Se'ankhenre' Montuhotpi
TC XI, 5 Nebiriaw (I.) Sewadjenre'
TC XI, 6 Nebiriaw (II.)
TC XI, 7 Semenenre' (proper name unknown;
Sobek- ?)
TC XI, 8 Seweserenre' Bebiankh
TC XI, 9 Sekhemre'-shedwaset (< tawy ?) Sobekemsaf (I)
TC XI, 13 Sekhemre'-wadjkhaw
Sobekemsaf II
TC XI, 10 Sekhemre'-upimaat (“the
elder”) Inyotef (VI)

(continued on next page)

Cf. also Beckerath’s “Sesostris V”. Improbable seems Weser[monthu] (?) (Beckerath,
Untersuchungen, XIII, L).
111
Helck (n. 95), 125–133; cf. the Coptos decree of Inyotef (VIII, according to the
sequence presented here): E. Martin-Pardey, “Zum Koptosdekret Antefs V.”, in: FS
Jürgen von Beckerath zum 70. Geburtstag am 19. Februar 1990 (Hildesheim: HÄB 30, 1990),
185–197.
112
From late Dyn. 17 there is only a titulary prince Sbk-m-˙b attested in Esna; and
the consort Sbk-m-˙b of a prince Ameni. The next ruler with a name (but not the
throne name) ending with -m-˙b is Horemhab.
middle kingdom and the second intermediate period 191

table (cont.)

TC XI, 11 Sekhemre'-Herhermaat
Inyotef (VII)
TC XI, 12 Nebukheperre' Inyotef (VIII)
TC XI, 14 Sekhemre'-wahkhaw
Ra'hotep
TC XI, 15 [1]5 kings
TC XI, 16 Weser . . . re' Weserenre'
TC XI, 17 Weser- Senwosret (IV) Seneferibre'
TC XI, 18
TC XL, 19 [placing of the following
kings uncertain]
TC XL, 20 Senakhtenre'
TC XL, 21 Seqenenre' Tao
TC XI, 22 Wadjkheperre' Kamose
TC XI, 23 (total)
TC XI, 24
TC XI, 25
TC XI, 26 . . . hab

The lengths of the reigns are listed in the TC as follows: TC XI, 1: 3


years (rest lost); TC XI, 2: 16 years (rest lost); TC XI, 3: 1 year (rest
lost); TC XI, 4: 1 year (rest lost); TC XI, 5: 26 years (rest lost); TC
XI, 8: 12 years, x months, 12 days.113 Of those rulers who are hypo-
thetically arranged here among their successors, a regnal year 7 is
attested for Sobekemsaf II, and for Inyotef VIII (Ryholt: Inyotef VII)
Nebukheperre' a year 3. These eight kings produce a total of 69 years;
for all 15 entries TC X, 31–XI,14 we might estimate 75–100 years.114
For the kings placed from TC XI, 16 onward, there are good reasons
for assigning them a significantly shorter length of rule. Kamose will
have died not long after his regnal year 3 (attested on the Kamose
Stela), at a time when Ahmose was still quite young.115 As his mummy
testifies, Seqenenre' suffered an early death on the battlefield. Ryholt
postulates 4 regnal years for Seqenenre', and one year for Senakhtenre',116
who is not recorded in contemporary documents. If Senwosret IV is

113
Cf. Ryholt, Situation, 202.
114
Cf.—with a different succession—the estimates by Ryholt, Situation, 204.
115
For the coregency, cf. Ryholt, Situation, 273; otherwise, 172ff., 309.
116
Ryholt, Situation, 206.
192 thomas schneider

to be placed here, a first year is attested in a biographical inscription.117


It follows that the rulers listed here were in power only very briefly;
with a total of perhaps 15 years.
Despite various differences in the internal sequence of kings between
Ryholt’s version and that presented here, both reach a relatively reli-
able estimate of the length of time to be assigned to the rulers between
Dyn. 13 and 18: roughly a century. Ryholt has 67 years (“Dyn. 16”)
+ 31 years (“Dyn. 17”). According to our reading TC X, 31–XI, 14
covers 75–100 years, and TC XI, 16 to Kamose amounts to 15 years
(cf. below, final remarks, section 6).

5. Dynastie 15

The kings of Dyn. 15, or the “Hyksos” (Ó˚ #.w-¢ #s.wt, “Rulers of Foreign
Countries”) were rulers of (at least remote) Syro-Palestinian descent
whose power was based in the eastern Nile Delta where a high pro-
portion of the population were of Palestinian origin. Establishing the
relative chronology of this dynasty suffered from the inadequate docu-
mentation and the widely assumed discrepancy between the royal names
in the epigraphic sources and the Manethonian tradition.118 Previously,
moving beyond the Khamudi attested in the TC and the epigraphi-
cally recorded kings involved deciding which of the many scarabs named
kings of Dyn. 15 and which did not. Hitherto, there were no gener-
ally recognized criteria for the attributions. The distribution and fre-
quency of their scarabs often seemed to favour including Sheshi and
Ja'qubhaddu (“Jaqobher”) whereas Helck believed that “Semqen” and
'Anathaddu (“'Anather”) could be assigned a place among the “great Hyksos”
(i.e. Dyn. 15 of the traditional approach). Ryholt proposed to include
the Skr-hr now attested in Tell el-Daba and to consider, from the three
rulers whose scarabs include the title ˙q#-¢#swt, Sm˚n and 'pr-'nt as the
first two kings of Dyn. 15.119 None of these names can be equated with
the Hyksos names such as preserved by Manetho, and this alleged dis-

117
W. Helck, Historisch-biographische Texte der 2. Zwischenzeit und neue Texte der 18. Dynastie
(Wiesbaden: KÄT, 19832), 41 [no. 56].
118
For the history of the debate from 1936–1997, see in detail, Schneider, Ausländer,
58–70.
119
Ryholt, Situation, 118–125. Moving the third—'nt-hr —to Dyn. 12 is apparently
not correct, cf. Ben-Tor et al. (n. 3), 63.
middle kingdom and the second intermediate period 193

crepancy seemed to indicate to most scholars that the Manethonian


evidence has to be discarded despite of the close to complete lack of
contemporary sources that might fill the gap. The situation was that
summarized by W. A. Ward: “It is impossible to equate the names preserved
in the various recensions of Manetho with these actually known from the monu-
ments (. . .) It is clear that most of the names preserved in this tradition are too
corrupted to have any value (. . .) The present discussion will therefore ignore
Manetho as being unreliable.”120
This traditional standpoint relies on the improper assumption that
even if most of the documentary evidence on the Hyksos is lost, we
nevertheless possess at least all their names, and does not accurately
reflect on how their names were handed down and copied in the later
chronographical tradition. Instead of rejecting Manetho, the author has
tried121 to solve the issue by reckoning with kings absent from our con-
temporary documentation122 and by accounting for textual mistakes in
the process of the copying of the king lists by late scribes who could
solely rely on the written form of the names but were not aware of
their original sounding. Our correlation of the traditions does not leave
any lacunae which must be filled having recourse to scarabs. The nearly
complete loss of the sequence of Dy. 15 kings in TC (where only the
last one, Khamudi is preserved) gives priority to Manetho of whose
epitomists those can be shown to be correct that place Apophis at the
end of the dynasty (Africanus, Eusebius’s Armenian version, the scho-
lion to Plato’s Timaios). The names of this sequence—1. Salitis, 2.
Bnon, 3. Apachnan, 4. Iannas, 5. Archles/Assis, 6. Apophis—can all
be equated with names attested epigraphically with one exception for
which a postulated original name can be supplied.123 It has to be noted,

120
W. A. Ward, “Royal-Name Scarabs”, in: Studies on Scarab Seals. Vol. II. Scarab Seals
and their Contribution to History in the Early Second Millennium BC (Warminster, 1984), 162.
Similarly, M. Bietak, Historische und archäologische Einführung, in: Pharaonen und
Fremde. Dynastien im Dunkel. Sonderausstellung des Historischen Museums der Stadt Wien in
Zusammenarbeit mit dem Ägyptologischen Institut der Universität Wien und dem Österreichischen
Archäologischen Institut Kairo, Rathaus Wien, Volkshalle, 8. Sept.–23. Okt. 1994, 17–57: 24.
121
Schneider, Ausländer, 33–56.70–75.
122
Striking examples are the two Hyksos •kr-Hr (architrave found in Tell el-Daba
and first published in 1994) and Khamudi (attested only in TC) who could not possi-
bly have been postulated by modern research. A late tradition of •kr-Hr seems to be
preserved in pCarlsberg 642 where an impious ruler Saker is mentioned (cf. J. F. Quack,
“Zwischen Sonne und Mond—Zeitrechnung im Alten Ägypten”, in: H. Falk, ed., Vom
Herrscher zur Dynastie. Zum Wesen kontinuierlicher Zeitrechnung in Antike und Gegenwart, Bremen
2002, 27–67: 47f ), equated by Quack with Salitis, but •kr-Hr is certainly preferable.
123
For what follows see in detail Schneider, Ausländer, 50–56.
194 thomas schneider

that the TC apparently once listed a total of 6 rulers of this dynasty,


including one named Khamudi, the copyists of Manetho also name 6
rulers, but without Khamudi.124

Hieroglyphic: Copyists of Manetho:

a. ”arà-Dagan (” #rk[n]) 1. Salitis > 36 years, 7 months


b. *Bin-'Anu 2. Bnòn > 44 years
c. ("Apaq-) ›ajran 3. Apakhnan > 19 years
d. Jina≤≤i"-Ad 4. Iannas > 50 years, 1 month
e. Sikru-Haddu125 5. Archlès/Assis > 49 years 2 months
f. Apapi 6. Apophis > 61 years
6. ݈lmu"di Р(not in Manetho)

Total TC: 108 years Manetho: 260 years

The total given at TC X, 21 for the entire Dyn. 15 is 108 years.126 By


contrast, according to Africanus’s version of Manetho, the dynasty would
have lasted 284 years. Assigning reign lengths to individual rulers is at
the present possible only in some cases. Beckerath proposed a hypo-
thetical reign for all individual rulers,127 but this remains speculative
since it is not clear which of the reigns have been lengthened by decades
by Manetho and which not. On an unnumbered fragment of the TC
(Ryholt’s “b”)128 are the remains of the totals of two reigns: “10 (or
20, or 30)” + x years and “40” + x years. The latter could hypo-
thetically be assigned to Apapi, who according to the colophon on the
Rhind Mathematical Papyrus must have reigned at least 33 years. The
first reign might then be assigned to •kr-Hr.129 Another reign length
can be inferred from the note on the verso of the Rhind Mathematical

124
Cf. Schneider, Ausländer, 56 (if not in the name Chamois of the book of Sothis).
125
The interpretation of major elements of these names such as offered by Ryholt,
Situation, 126ff. (particularly 'pr and hr) reflects older positions (Albright) which are no
longer supported.
126
See Ryholt, Situation, 118.—It must be noted at the outset, however, that in an
unpublished communication made while this book was in its final stages of prepara-
tion, Ryholt announced that the traces on the actual papyrus did not confirm Gardiner’s
reading and that another date, possibly of more than 140 years should instead be con-
sidered possible (editors’ note).
127
Beckerath, Chronologie, 137.
128
Ryholt, Situation, 118–119.
129
Contra Ryholt, Situation, 119.
middle kingdom and the second intermediate period 195

Papyrus whereby in the 11th regnal year of the ruling king, Heliopolis
has been conquered, and “he of the South” has attacked and taken
Sile. Since “he of the South” must denote the Theban ruler Ahmose,
the regnal year 11 can only be assigned to the successor of the Hyksos
king Apapi: Khamudi.130 The Hyksos capital Avaris will have fallen to
Ahmose not much later. An inscribed spear point from the booty of
Avaris which includes a dating criterion (orientation of the lunar hiero-
glyph)131 enables us to place the conquest of Avaris in the regnal year
18/19 of Ahmose. A possible check is provided by a graffito in the
quarry at Tura whereby “oxen from Palestine” were used at the open-
ing of the quarry in Ahmose’s regnal year 22.132 The cattle could have
been brought to Egypt after the three-year siege of south Palestinian
Sharuhen, which followed the conquest of Avaris. This link between
the reigns of Khamudi and Ahmose means that the beginning of Dyn.
15 should be 90 years before the death of Kamose (108 years—pro-
vided this number from the TC is correct—minus the 18 years of the
reign of Ahmose parallel with the last two Hyksos).

6. Total Length of the Period of Dyns. 13–17

Providing any exact figure for the total duration of the period of Dyn.
13–17 is impossible. The difficulties are not merely due to the inade-
quacy of the documentation of the various dynasties, but rather that
these are exacerbated by our inability to establish their mutual rela-
tionships. In particular, it is far from certain that Dyn. 15 and the
Theban dynasty (Ryholt’s “Dyns. 16/17”) followed immediately on Dyn.
13 as Ryholt proposes, with both starting in 1649 BC. It seems plau-
sible that the transition was both spatially and temporally more com-
plex in the final decades of the 17th century. The fact that the numerous
ephemeral rulers of the last 25 years or so of Dyn. 13 were probably
incapable of maintaining administrative control of the entire country

130
D. Franke, “Chronologie II”, 263; M. Bietak, “Historische und archäologische
Einführung”, in: Bietak, SCIEM 1996/1998, 29; Beckerath, Chronologie NR, 115; Spalinger
(n. 3), 299. Ryholt is opposed, but procedes a priori from a short reign for the last
Hyksos king; likewise, Kitchen, “Regnal and Genealogical Data of Ancient Egypt,” in:
Bietak, SCIEM 1996/1998, 39–52, esp. 45–46 (who also considers assigning the date
to the last Hyksos).
131
Kitchen (n. 130), 46 correctly stresses that this criterion is not well founded.
132
Franke, “Chronologie II”, 264.
196 thomas schneider

suggests that there may have been a number of competing local dynas-
ties at the time.133 It would be plausible to suppose that the origins of
the power of these local dynasts lie in the period before the ephemeral
end of Dyn. 13, and therefore that an approximate chronological fixpoint
might be established at the end of the reign of Aya, at ca. 1650 BC.134
This linkage between Dyns. 13 and 17 finds some support in bio-
graphical information of the time.135
Calculating back from the start of the reign of Ahmose (around 1539
BC) and suggesting that the Theban kings reigned for 90–115 years
before Ahmose leads to a beginning of the Theban dynasty at ca.
1654/1629; based on the TC, the dates for Dyn. 15 would be 1639–1521
BC (cf. above). Given the uncertainties in the length of individual reigns
here and in Dyn. 13, a conclusive judgment about the length of the
SIP136 still needs further evidence.

133
Spalinger (n. 3), 297–298 assumes that an independent kingdom (= Dyn. 17)
already appeared a few decades before the end of Dyn. 13, and poses the question of
the possible coexistence of a truncated state in the North with the Hyksos.
134
Cf. still the remark of the Greek author Artapanos (2nd century BC), that under
a king named “Chanephres” (Khaneferre' Sobekhotep IV), Egypt was divided into var-
ious kingdoms (Schneider, Ausländer, 158–159). Whether the pyramidion of Aya found
at Faqus was removed there from Saqqara by the Hyksos, as Ryholt (Situation, 147)
and Dodson (n. 3; 50) assume is not certain; Beckerath (Untersuchungen, 73) assumed
the Ramesside era. Of the various possible synchronisations between the dynasties,
Manetho’s remark about a supposed conquest of Egypt by the Hyksos under a
“Toutimaios”, in whom one of the Dedumose kings of Dyn. 13 had earlier been sus-
pected, is no longer tenable as the reference is clearly due to a misunderstanding of
the Vorlage, or a textual error: cf. Schneider, Ausländer, 159; for a different interpreta-
tion, see A. Bülow-Jacobsen in: Ryholt, Situation, 327ff.
135
The stèle juridique records that an Aya received the governorship of el-Kab in reg-
nal year 1 of Merhotepre' Ini (successor of the king Aya) and that his grandson Kebsi
sold it in regnal year 1 of Nebiriaw. In TC XI, 1–5, the lengths of the reigns up to
Nebiriaw I are preserved (21 years, with a few months missing), so that including the
reign of the founder of the dynasty in TC X, 31, some 25 years will have passed
before regnal year 1 of Nebiriaw (cf. above, section 4). If the dynasty began between
1654 and 1629 BC, Merhotepre' Ini (cf. above, section 3) should be placed around a
century after the start of the dynasty, i.e., ca. 1656 BC. This would result in 27–52
years for the period during which the father and grandfather of Kebsi exercised the
office, which appears plausible. Ryholt (Situation, 197, 202) has a far larger margin of
more than 70 years (1 Merhotepre' Ini = 1677 BC; 1 Nebiriaw I = 1627 BC), assum-
ing that the father and grandfather of Kebsi reached a very ripe old age of more than
70 years each.
136
Ryholt, Situation, 191 opts for a long, of 254 years (as R. Parker proposed based
on his absolute dates for the MK), and thus proposes the dates of 1803–1549 (Dyn.
14 from 1805; Dyn. 15 to 1540; similarly Kitchen (n. 130), 46. For a short chronol-
ogy cf. E. Hornung (“Lang oder Kurz?,” in: High, Middle, Low 1, 36; 1756–1539 BC,
with Dyn. 15 down to 1521. Dodson (n. 3, 50–51) also concludes that a maximum
chronology is not necessary, but is willing lower the beginning of the NK to 1500 BC.
II. 8 THE NEW KINGDOM

Erik Hornung

For some time there has been a consensus about the relative chronol-
ogy of the New Kingdom. Since most reign lengths are well docu-
mented, they are not problematic. Nor does the sequence of kings pose
difficulties—except for Akhenaten’s immediate successors (who are not
mentioned in the cultic lists). However, the issue of coregencies for
Thutmose III/Amenhotep II and Amenhotep III/Akhenaten continues
to stimulate debate, as does the exact relationship between Amenmesses
and Sety II, although otherwise dates important for relative chronol-
ogy are particularly abundant for the Ramesside Period, leaving very
little “empty space”.
This positive situation is enhanced by the fact that regnal years were
counted from a king’s accession and appear in the dates of documents
which allows precise calendar dates to be suggested for some reigns,
the necessary first step leading to a “day-exact chronology” (Depuydt),
as is in fact possible in the Late Period. The identification of the exact
day of accession or the establishment of very precise possible limits
allows additional checks through the months offered by Manetho via
Josephus. A few problematic issues remain, especially the length of
Haremhab’s reign which has been estimated at between 13 and 27
years. In this particular case, additional criteria, such as the number
of monuments or the sequence of officials cannot solve the problem.
Nor do we have Haremhab’s mummy which could have provided a
potential check based on his estimated age at death. But in fact, age
estimates for royal mummies have not generally yielded satisfactory
results. As Kitchen noted in his review of the basic work by Harris &
Wente, their analyses offers an over-abundance of comparatively very
young kings.1 The cases of Thutmose III and Ramesses II are partic-
ularly revealing. The estimates provided by Wente and Harris for the
ages of the unequivocally identified mummies of these kings (40 and
55 years, respectively) are glaringly at odds with the well-documented

1
Harris & Wente, Atlas.
198 erik hornung

long reigns of both rulers (53 and 66 years). “Something somewhere


is badly wrong”, as Kitchen remarked.2 In other cases, as with Amenhotep
III, the identification of the mummy itself is not beyond doubt, which
adds to the uncertainty.
The totals which the Manethonian sources cite include part of Dyn.
19 and thus provide only a limited control. For the period from 'Ahmose
to Merneptah, Eusebius and the Old Chronicle give 348 years while
Africanus has 263 years The tally resulting from the following list
amounts to a minimum of 307 years and a maximum of 335 years for
the same time span, effectively excluding Africanus’ total.

Dyn. 18

'Ahmose No data relating to the accession date is currently available.


Most royal monuments are undated. Year 17 is cited on a stela from
the foundations of the third pylon at Karnak (Abdul Qader Muhammad,
ASAE 59, 1966, 148–149, pls. IV–V); year 18 is recorded on the stela
Hanover 1935. 200. 209 (Im Zeichen des Mondes, Exhibition catalogue,
Munich 1999, No. 4; A. Klug, Königliche Stelen in der Zeit von Ahmose bis
Amenophis III, Brussels 2002, 49–51, reviving doubts about its authen-
ticity), and year 22 in the Turah quarries (Urk. IV 25,7).
Josephus gives 'Ahmose a reign of 25 years, 4 months; Eusebius
rounds this off to 25 years. (The figure is missing in Africanus.) The
autobiography of an anonymous Viceroy of Nubia covers the period
from 'Ahmose to Thutmose II (Urk. IV 39–41), and thus a maximum
of 33 (21 + 12) years between the two rulers. For 'Ahmose, the min-
imum reign length should be 21 years and the maximum 25 full years.
His mummy (CG 61057) was estimated to be that of a man 25–30
years old (Harris & Wente, Atlas, 202), but a general uncertainty pro-
hibits drawing any conclusions.
The precise date of the defeat of the Hyksos and thus the end of
Dyn. 15 must lie in 'Ahmose’s second decade. “Year 11” in the Rhind
Mathematical Papyrus should probably be assigned to the last Hyksos
ruler Khamudi (A.-F. el-Sabbahy, GM 133, 1993, 97–99, cf. above
Schneider, Chapter II. 7), but Kitchen still maintains that the date

2
K. A. Kitchen, JNES 44 (1985), 235–237.
the new kingdom 199

belongs to 'Ahmose (in SCIEM, 2000, 45). Since P. Rhind refers to


the conquests of Heliopolis and Sile, the capture of Avaris could fol-
low somewhat later.

Amenhotep I For the accession, W. Helck (in: Fs S. Schott, Wiesbaden


1968, 71–72) assumed that the festival dates and the months in Josephus
indicated 29–30/I/ Akhet, as opposed to D. B. Redford ( JNES 25,
1966, 115–116) who used the same festival dates to argue for 11/III/
Shemu while Krauss (Sothis, 115) considers the actual date of P. Ebers,
9/III/ Shemu, to be the accession date, as do U. Luft (GM 92, 1986,
69–77) and Beckerath (Chronologie NR, 110).3 G. Vittmann believes the
title “Royal Mother” borne by 'Ahmes Nefertari in the Turah inscrip-
tion of year 22 (Urk. IV 25,4) supports a coregency; but he also stresses
its chronological irrelevance, since Amenhotep I counted his regnal
years from the death of his father (“Was there a coregency of Ahmose
with Amenophis I?”, JEA 60, 1974, 250–251).
Dated monuments belong to year 7 (graffito of the Viceroy Tury in
Semna: J. H. Breasted, AJSL 25, 1908, 108), year 8 (Uronarti: Urk. IV
78,8; stela from Qasr Ibrim: J. Plumley, JEA 50, 1964, 4 with pl. I,3),
year 9 (rock inscription in Semna: F. Hintze, ZÄS 111, 1984, 137–138
and Hintze & Reineke, Felsinschriften, No. 512), and 1/I/Shemu of year
10 (Kares Stela, CG 34003: Urk. IV 45,9). Thereafter there is only a
graffito at Saqqara dated 19/IV/ Akhet of year 20 (Step Pyramid I, 79).
Josephus assigns 'Ahmose 20 years 7 months; the other Manethonian
sources round this up to 21 years, which accords well with the 21 years
that the “astronomer” Amenemhat spent ( jrj) under Amenhotep I
(L. Borchardt, Altägyptische Zeitmessung, Berlin & Leipzig 1920, pl. 18).
Wente & Van Siclen argue that this refers only to sole rule and add
6 years coregency with 'Ahmose, but even this would not produce the
30 years necessary for a “real” sed-festival which they presume (“Chron-
ology”, 225). The accession date of Thutmose I means fixes the death
date of 'Ahmose on 20/III/ Peret.

Thutmose I The accession on 21/III/ Peret is certain (Urk. IV 81,4).


Further dates are 15/II/ Akhet year 2 (Tombos: Urk. IV 82,9), 22/I/
Shemu year 3 return from Nubia (Urk. IV 88,11; 89,6/16), year 4 on

3
On the problems of the Ebers date, see below Chapter III. 10.
200 erik hornung

a naos from Giza (Urk. IV 91,9; perhaps posthumous or modern:


R. Krauss, Ä & L 3, 1992, 86), and year 8/9 on a block from Karnak
(A. Mariette, Karnak, Leipzig 1875, pl. 32.f ), with doubtful attribution
(see Wente & Van Siclen, “Chronology”, 225–226, and R. Krauss, Ä
& L 3, 1992, 86–87 with fig. 3). Perhaps the “11 years” on the stela
of Nebwawy (Urk. IV 208,16) also refer to him, although D. B. Redford
( JNES 25, 1966, 118–119) prefers Thutmose II.
With an estimated age of 18–22 years, the mummy (CG 61065) seems
to be extremely young (Harris & Wente, Atlas, 202) for a reign of 12
years 9 months, according to Josephus.

Thutmose II Following Gardiner, Redford accepts 8/II/ Akhet ( JNES


25, 1966, 117) as the accession date, whereas Beckerath (Chronologie NR,
117) assumes that it took place in III/IV Akhet, on account of the
months assigned to his predecessor by Josephus. The only certain date
is 9/II/ Akhet of year 1 (Aswan: Urk. IV 137, 9). A short reign is sup-
ported by the small number of scarabs (105 examples collected by
B. Jaeger for Thutmoses II as opposed to 292 for Thutmoses I)4 and
gaps in the succession of officials, e.g. the stewards of Karnak, as well
as the lack of a mortuary temple (signalled by L. Gabolde, “La chronolo-
gie du règne de Thoutmosis II”, SAK 14, 1987, 61–81). However,
Beckerath (SAK 17, 1990, 65–74; Chronologie NR, 121), W. Barta ( JEOL
26, 1980, 33–34) and Wente & Van Siclen, “Chronologie”, 226–227)
have argued for a 13-year reign on the bases of the age of his chil-
dren at the end of the reign,5 the Ebers date, and the sed-festival of
Hatshepsut, which is capable of different interpretations.
The “Year 18” which often appears in this context does not belong
to Thutmose II but rather to Amenhotep II, according to L. Gabolde
(SAK 14, 1987, 61–81), or to Maatkare Hatshepsut, according to Beckerath
(SAK 17, 1990, 66, cf. R. Krauss, Ä & L 3, 1992, 86 with n. 3). The
identification of the mummy (CG 61066) is uncertain and thus its age
estimate of 25–30 years is not relevant. Manetho assigned “Chebron”,
who is generally identified as Thutmose II, 13 years, but this could
easily be an extension of 10 years; 3 years fit the sources, but they are

4
Cf. Ä&L 3 (1992), 88.
5
On the problem of the expression “in the nest”, cf. R. Krauss, OLZ 90 (1995),
239–240.
the new kingdom 201

no more than a possibility. At present, we can assume perhaps 2 to 4


years for the reign of this king.

Hatschepsut When Hatshepsut seized power continues to be debated.


The oracle of 29/II/ Peret year 2, initially postulated by Schott, is not
sufficiently explicit (cf. Ch. Cannuyer, in: Studies Lichtheim I, 109–115).
The queen counted her own years from her coregent’s accession, so
that Thutmose III’s accession date is taken for hers, and confirmed
through Urk. IV 367,3–5 on the creation of her obelisks in 7 months,
from 1/II/ Peret year 15 to 30/IV/ Shemu year 16.
The first certain date is year 9 for the Punt expedition (Urk. IV
349,10). 12/III/ Peret year 12 in West Tangur (Hintze & Reineke,
Felsinschriften, 172, No. 562) is linked to both kings, as is year 16 in
Wadi Maghara (Urk. IV 393,15), as well as an inscription of the over-
seer of works Nakht in year 206 and a graffito at Saqqara of 2/III/
Peret year 20 (Step Pyramid I, 80, F). An inscription of 30/I/ Akhet
year 17 in Karnak (Urk. IV 376,13) names Hatshepsut alone, and
another of year 20 in Nubia (Urk. IV 1375,3) only Thutmose III. An
ostracon of IV Peret year 20 from Deir el-Bahri should probably be
assigned to Hatshepsut (W. C. Hayes, JEA 46, 1960, 38).

Thutmose III His accession occurred with certainty on 4/I/ Shemu


(Urk. IV 180, 15–16). The earliest document thereafter is a graffito at
Saqqara (Step Pyramid I, 80, D) which mentions a royal sojourn at
Thebes for 5/IV/ Akhet of year 1. His sole rule could have begun on
10/II/ Peret year 22, recorded on a stela from Armant (Urk. IV 1244,14).
Josephus gives Hatshepsut a reign of 21 years 9 months, which would
cover her joint reign with Thutmose III. According to Urk. IV 895,16–17,
Thutmose III died on 30/III/ Peret of his year 54. Before his sole
rule, we have precise dates for a renewal of offerings in the Semna
Temple on 8/II/ Shemu year 2 (Urk. IV 193,13), for the inauguration
of the Vezier Useramun on 1/I/ Akhet year 5 (Urk. IV 1384,3), a
mention of 26/II/ Akhet year 7 in the Akhmenu at Karnak (Urk. IV
1256,8), a renewal of offerings in Karnak on 27/I/ Shemu year 15
(Urk. IV 172,15); year 15 (without a royal name) is mentioned in the
autobiography of Iamunedjeh (Urk. 940,5), year 16 in a rock inscrip-

6
Urk. IV 1377,3.
202 erik hornung

tion at Abusir (Hintze & Reineke, Felsinschriften, 38 No. 64), and 16/IV/
Shemu year 18 at Shalfak (ibidem, 90 No. 365).
The stela of Senimes bears the date 25/III/ Peret year 21 (Urk. IV
1066,10). 25/IV/ Peret year 22 appears in the annals as the begin-
ning of the first campaign (Urk. IV 647,12); the same year 22 also
occurs in a reference to the restoration of a statue (Urk. IV 605,17).
Dates from the Megiddo campaign include year 23, 4/I Shemu for
arrival at Gaza (Urk. IV 648,9) and 16/I/ Shemu for the council of
war at Yehem (IV 649,3ff.). Year 23 is also documented from the tem-
ple in Wadi Halfa (IV 806,11), and the Akhmenu at Karnak records
2/I/ Shemu of a “year after 23” (IV 1252, 11, cf. A. H. Gardiner,
JEA 38, 1952, 9). On 30/II/ Peret year 24 a foundation ceremony in
Karnak is mentioned (Urk. IV 836,2), and for year 24 also a list of
tributes (IV 671,6). Year 25 occurs in the Botanical Garden in Karnak
(IV 777,2) and on a stela at Serabit el-Khadim, Sinai (IV 886,5), year
27 on the statue of Sabastet (IV 1369,4), year 28 in the tomb TT 82
(IV 1043,15), year 29 in the annals for the fifth campaign (IV 685,3),
and year 30 for the sixth campaign (IV 689,3).
3/I/ Shemu year 31 is mentioned in the list of booty for the sev-
enth campaign (Urk. IV 690,14). An inscription of Sennefer at el-Bersheh
(Urk. IV 597,12) cites 12/IV/ Shemu year 33 in relation to the sed-
festival (a parallel inscription in Krakow gives 4/IV/ Shemu); year 33
is documented from the annals for the eighth campaign (Urk. IV 696,15),
year 34 for the ninth campaign (Urk. IV 703,16), and year 35 for the
tenth campaign (Urk. IV 709,15); year [38] can be restored for the
thirteenth campaign (Urk. IV 716,12); year 39 for the fourteenth (Urk.
IV 721,9) is also known from a graffito at Saqqara (13/III/?: Step
Pyramid I, 80, E), and year 40 from a list of tribute (Urk. IV 668,4).
22?/I/ Akhet year 42 is documented with the restoration of a statue
in Karnak (Urk. IV 606,6), year 42 at the end of the annals (IV 734,14);
2/II/ Peret year 43 is known from an ostracon from the west bank of
Thebes (IV 1374,8), and year 45 from the dedication for a Mnevis-
bull (IV 1373,2); 10/III/ Akhet year 47 is the date of the Gebel Barkal
stela (IV 1228,6), year 47 of the Mnevis-bull stela Cairo JdE 65830
(M. Moursi, SAK 14, 1987, 233–235), and of the stela Berlin 1634 con-
cerning an enclosure wall in the temple of Heliopolis (Urk. IV 832,12);
22/I/ Shemu year 50 is from a rock inscription at Sehel (Urk. IV
814,10; cf. also J. Leclant & G. Clerc, Or 61, 1992, 299). 14/II/Shemu
year 51 is the date of a rock inscription at Ellesia (Urk. IV 811,10),
and 23/III/ Peret year 53 is found on a scribal palette in Hanover
(W. Helck, MDAIK 48, 1992, 41–44).
the new kingdom 203

The reign length which Josephus associated with Amenhotep II is


30 years 10 months (reduced by ten years by Theophilus), and this
could be identified as the sole rule of Thutmose III, if one assumes a
coregency with his son, but might actually be Amenhotep III, if his
reign length be reduced by ten years.

Amenhotep II The accession date (or the date when he was named core-
gent) was 1/IV/ Akhet (Urk. 1343,10). According to Bierbrier (Or 49,
1980, 108), who argues against a coregency, this is in error for 1/IV/
Peret. Many authors side with Redford ( JEA 51, 1965, 107–122) and
Parker (in Studies in Honor of John A. Wilson, Chicago 1969, 75–82), in
favor of a coregency lasting 2 years 4 months. Despite P. Der Manuelian’s
lengthy discussion in his Studies in the Reign of Amenophis II (Hildesheim
1987), 19–40, the matter remains unresolved, cf. R. Krauss, OLZ 90,
1995, 241–242.
The earliest date is 15/III/ Shemu year 3 (Amada stela: Urk. IV
1289, 1) which, in the event of a coregency, might mark the begin-
ning of sole rule. To year 4 belong an inscription from the Turah
quarries (IV 1448,4), and an addition to the duplicate of the Amada
stela in Cairo (IV 1299,2), as well as the 10/I/ Peret graffito at Saqqara
(Step Pyramid I, 80, G). The years 5, 6, 7 and 9 are documented by
several dates on stelae from Memphis and Karnak (Urk. IV 1301–1314),
but there follows a gap until at least year 17 (uncertain: L. Habachi,
Kemi 18, 1968, 55 with fig. 5 on the graffito of Nakht in Aswan) or
even until 1/IV/ Akhet year 23, the date of the stela of the Viceroy
Usersatet from Semna, in Boston (Urk. IV 1343,10).
A jar label from the mortuary temple names year 26 (Urk. IV
1365,18–20), which suits the 25 years 10 months of Josephus, but unfor-
tunately, it is not certain that this figure belongs to Amenhotep II,
although it makes no sense for Thutmose III either. No available infor-
mation suggests a longer reign. Only the upper limit for the estimated
age of the mummy (CG 61069; Harris & Wente, Atlas, 202) at 35–45
years can be reconciled with the historical data.

Thutmose IV There is no basis for determining the date of his acces-


sion, other than Josephus’ figure of 8 months which suggests the end
of II or beginning of III Akhet. On this assumption the date of the
sphinx stela, 19/III/ Akhet year 1 (Urk. 1540,2) is plausibly the earli-
est of the reign. Somewhat later in the first year would be 7/II/ Shemu
from a stela in Luxor (El Sayed Higazy & B. M. Bryan, “A New Stela
of Thutmose IV from the Luxor Temple”, VA 2, 1986, 93–100). Bryan
204 erik hornung

provides a list of documented years in the reign (The Reign of Thutmose


IV, Baltimore & London 1991, 5–9.
An inscription on Sinai (Urk IV 1634,9) is dated to year 4, others
to years 5 and 7 (IV 1564). A stela on the art market, probably from
Medamud, names 10/I/ Shemu year 5 (R. G. Bigler & B. Geiger,
“Eine Schenkungsstele Thutmosis’ IV.”, ZÄS 121, 1994, 11–17), two
Theban tomb inscriptions (LD Text III, 273 and Urk. IV 1618,5) year
6; year 7 (to be corrected to 8) appears on a stela from the island of
Konosso (IV 1555–1556). The highest date is 2/III/ Peret year 8 on
the Konosso stela (IV 1545,6), when the king is informed about a
Nubian rebellion.
Josephus gives “Thmosis” 9 years 8 months. The other Manethonian
sources round down to 9 years and the Book of Sothis has an inflated
39 years. Years 9–10, like years 2 and 3, are not documented in con-
temporaneous sources. This seems acceptable, whereas a longer reign
results in difficulties, and certainly the extreme lengthening proposed
by Wente & Van Siclen on the basis of their ideas about sed-festivals
is precluded. The age at death of his mummy (CG 61073) was esti-
mated at 30–40 years (Harris & Wente, Atlas, 202; cf. Bryan, Reign,
9–13). According to the inscription on the Lateran obelisk (Urk. IV
1550,5), it lay “on its side for 35 years” before being erected, which
would include the entire reign of Amenhotep II and an unknown num-
ber of the years of Thutmose III and IV.

Amenhotep III Based on the date of the first sed-festival, Helck (Manetho,
67) argued for an accession on 3/III/ Shemu. An inscription from III
Shemu year 1 was found in el-Bersheh (Urk. IV 1677–1678); in year
2 the quarries at Turah were opened (Urk. IV 1681,2) and the com-
memorative scarab of the wild bull hunt was issued (Urk. IV 1739,1).
A graffito in the royal tomb WV 22, at the passage into the antecham-
ber, is dated to 7/III/ Akhet year 3 ( J. Kondo, in: R. H. Wilkinson
(ed.), Valley of the Sun Kings, Tucson 1995, 29–30, fig. 3); a wine jar
label from Deir el-Medina also mentions year 3 (Valbelle, Ouvriers, 23).
Several dates in year 5 are linked to the Nubian campaign: 24/II/
Akhet on the island of Sai (Urk. IV 1959,11), 2/III/ Akhet near Aswan
(Urk. IV 1665,15), I Shemu at Buhen (Urk. IV 1758,17) and a stela at
Konosso on the return from Kush (Urk. IV 1662,7). The lion-hunt
commemorative scarab dates to year 10 (Urk. IV 1740,12), as does the
scarab commemorating the wedding with Gilukhepa (Urk. IV 1738,6),
and a graffito of 13/III/ Shemu at Saqqara (Step Pyramid I, 81, H).
the new kingdom 205

The second decade of the reign is poorly documented. The series of


commemorative scarabs ends with that relating to the excavation of a
lake for Queen Teye, dated 1–16/III/ Akhet in year 11 (Urk. IV
1737,8). Labels on wine jars from Deir el-Medina document years 14,
17 and 19 (Valbelle, Ouvriers, 23); a papyrus fragment in Turin men-
tions IV Peret year 14 (GLR II 310, XI). For II Akhet year 20 the
statue of Nebunefer in Brussels records a royal visit to Memphis (Urk.
IV 1885,10), and 2/I/ Peret (without royal name) is documented in
an historical inscription from the mortuary temple of Amenhotep-son-
of-Hapu (A. Varille, Inscriptions concernant l’architecte Amenhotep fils de Hapou,
Cairo 1968, 96–97).
The third decade is likewise sparsely documented. A jar-label from
Amarna of year 21 should belong to this reign ( JEA 67, 1981, 2), as
should the dates of 20/III/ Shemu year 27 of P. Berlin 9784 (A. H.
Gardiner, ZÄS 43, 1906, 28–35) and O. Cairo CG 25242 of 20/IV/
Akhet year 29. Many dates are associated with the first sed-festival in
year 30, beginning on 27/II/ Shemu (Kheruef: Urk. IV 1867,2) and
extending through III Shemu (Urk. IV 1869,2; 1837,9); these are sup-
plemented with numerous labels from Malqata, some from year 29,
but generally mentioning year 30 and the first sed-festival (W. C. Hayes,
JNES 10, 1951, 35–56, 82–112, 156–183, 231–242, and additional
labels in M. A. Leahy, Excavations at Malkata and the Birket Habu 1971–1974.
The Inscriptions, Warminster 1978).
Further deliveries at Malqata are dated to year 31, along with the
hieratic stela BM 138 with the decree for Amenhotep-son-of-Hapu
(6/IV/ Akhet: Varille, Inscriptions, 67–85; D. Wildung, Imhotep und
Amenhotep, Munich & Berlin 1977, § 182). A contract from Gurob is
dated 5/I/ Akhet year 33 (P. Gurob II,1: A. H. Gardiner, ZÄS 43,
1906, 35–37), and the second sed-festival of year 34 is documented with
numerous labels from Malqata (Hayes, in JNES 10, 1951). 1/I/ Shemu
year 35 is the date of two stelae at Silsileh (Urk. IV 1920,3; 1678,8);
the third sed-festival of year 36 is documented in the tomb of Kheruef
(Urk. IV 1860,2–7), and 9/II/ Peret of year 36 on a Sinai stela (Urk.
IV 1891,4). Labels from Molkata mention year 37 for the third sed-
festival (Urk. IV 1954,12), stretching through 1/III/ Shemu of year 38
(Hayes, JNES 10, 1951, fig. 11, no. 142). So at least 37, and perhaps
38 full years should be assumed for Amenhotep III; Manetho’s copyists
give 38 years 7 months, as well as 36 and 37 years (Helck, Manetho, 67).
A long coregency with Akhenaten has been posited time and again
since 1899 (Petrie, A History of Egypt II, London, 208), but it has been
206 erik hornung

challenged by W. Helck (MIO 2, 1954, 196–209), and many others,


most recently by Gabolde, Akhenaton, 62–98, who discusses most of the
arguments. Gardiner’s remark years ago (Egypt of the Pharaohs, Oxford
1961, 213) still stands: “the much canvassed co-regency must be an
illusion”.

Amenhotep IV/Akhenaten Analysis of the sequence of dates on the later


boundary stelae at Amarna, Akhenaten’s newly founded capital, places
the accession in the time span 1/I Peret—8/I Peret (see W. J. Murnane,
“On the Accession Date of Akhenaten”, in Studies in Honor of George R.
Hughes, Chicago 1976, 163–167; Murnane & Van Siclen, Stelae, 155;
Beckerath, Or 63, 1994, 123; Gabolde, Akhenaton, 14–16).
One of the earliest dates of the reign is the hieratic docket on the
Amarna letter EA 27, from 5?/I/ Peret year 2 (Urk. IV 1995,16, cf.
W. Fritz, SAK 18, 1991, 207–214). 27/?/Peret year 2 and epagome-
nal day 4 of year 3 are mentioned in P. Berlin 9784 (Murnane, Texts,
44), 7/II/ Akhet year 4 in P. Berlin 9785 (Murnane, Texts, 46), 11/III/
Akhet year 4 with the expedition of the High Priest May in the Wadi
Hammamat (Murnane, Texts, 68), 19/III/ Peret year 5 in the letter of
Ipi to the king (Murnane, Texts, 51). The earlier series of boundary
stelae is dated 13/IV/ Peret year 5, the later series exactly a year later,
with repetition of the oath on 8/I/ Peret year 8 and a colophon of
30/IV/ Akhet year 8 (Murnane & Van Siclen, Stelae, 73–86). There is
a reference to “Year 9 of the heretic” in a letter from the reign of
Ramesses II (P. Berlin 3040: KRI III 5).
The later years are less well documented in monumental inscriptions.
8/II/ Peret year 12 is mentioned in the tribute scenes in the tombs of
Meryre and Huya (Urk. 2003,5; 2006,11); 20/III?/ Akhet year 12 is
the date of the Nubian campaign on a stela in Buhen (Murnane, Texts,
101–102); 2?/II/ Peret year 14 of a graffito at Saqqara ( J. Malek, DE
32, 1995, 105–106). Wine jar labels from the site of Amarna docu-
ment Akhenaten’s regnal years 4 through 17 (Hornung, Untersuchungen,
90–91; Krauss, MDOG 129, 1997, 227–229), so that he could have
died at the earliest during the sealing of the wine jars in year 17 (II
Akhet, corresponding to August 22 to September 20 [ Julian] in 1336
BC) or just before the sealing in year 18 at the latest. Akhenaten does
not seem to have been preserved in the Manethonian tradition. The
conventional identification of VROS (Horos) with Akhenaten is problematic.
the new kingdom 207

Smenkhkare' and 'Ankh(et)kheperure' It is now certain that not only a man


'Ankhkheprure', but also a woman 'Ankhetkheprure' ruled between
Akhenaten and Tut'ankhamun. The king is first known as 'Ankhkhe-
prure' (throne name) Smenkhkare' ≈sr-¢prw (personal name), later as
'Ankhkheprure' mrjj /Neferkheprure'/Waenre'/and Nefernefruaton mrjj
Waenre'. The two names of the queen, i.e. 'Ankhetkheprure' mrjj/
Neferkheprure'/Waenre'/ and Nefernefruaton #¢t n hj.s, are nearly the
same as the king’s later set of names and epitheta. The “funerary” epi-
theton #¢t n hj.s (beneficial for her husband) is hers alone and indicates
that she succeeded her husband 'Ankhkheprure'. Her identity remains
problematic; Kiya, Nefertiti, and Merytaton have been proposed. Items
of her funerary equipment were adapted for Tut'ankhamun (M. Gabolde,
Égypte Afrique & Orient 33, 2004, 19–26). Josephus lists three rulers
named AXENXERRSS < 'Ankh(et)kheprure', i.e. two male rulers (one of
which might be due to a corruption in the text) and a female, described
as a king’s daughter. In its transmitted form the Manethonian tradi-
tion ascribes 12 years and some months to either of the kings named
Akhenkherres. Possibly the figures reflect an original 2 years and some
months. The mummy of King 'Ankhkheprure' is identifiable as the
occupant of KV 55; his age at death was 18–22 years (W. J. Murnane,
OLZ 96, 2001, 22).
The wine jar labels from Amarna attest sixteen successive wine vin-
tages during the occupation of the site, 13 corresponding to years 5
through 17 of Akhenaten, whereas 3 vintages correspond to regnal
years of his successors. The vintage of Akhenaten’s year 4 occurred in
the year before the foundation of the city; wine of year 4 was con-
sumed at the site before the vintage of year 5 became available. Thus
altogether seventeen successive vintages are attested at Amarna.
Up to year 13 of Akhenaten the chief vintners held the title ˙rj k #mw.
The title ˙rj b#˙ is attested from year 13 through 17 and its use con-
tinued in year 1 and 2 of King 'Ankhkheprure'. The last vintage that
is documented at Amarna dates to a regnal year 1; in that year the
vintner’s title ˙rj k #mw was reintroduced and continued to be used as
wine jar labels in the tomb of Tut'ankhamun show (see below). Regnal
year 1 of the last vintage at Amarna could belong either to the queen
'Ankhetkheprure' or to Tut'ankhamun. The only inscription from this
period that is dated by a regnal year with a royal name is the graffito
in the Theban tomb of Pairy: 10/III/Akhet year 3 of 'Ankhkheprure'
mrjj /// Nefernefruaton mrjj /// (Urk. IV 2024, 14). The date seems
to relate to the king, but the queen is not excluded (Gabolde, Akhenaton,
208 erik hornung

161–162, 184). If the graffito relates to her, then she continued the
year count of her husband and the last vintage that is attested at
Amarna dates to year 1 of Tut'ankhamun. If the graffito relates to the
king, then the queen started a regnal year count of her own; year 1
of the last vintage belongs to her and there would be no dated mate-
rial of Tut'ankhamun at Amarna.
A regnal year 3 is also attested at Amarna in the labels on vessels
for various commodities. Year 3 continues year 1 and 2 of King
'Ankhkheprure' as labels of year 2 and 3 belonging to a single deliv-
ery of olive oil prove (Hornung, Untersuchungen, 88–89). There are only
3 wine jar labels of year 3 which cannot represent a complete vintage,
because the yearly mean number of wine jar labels is 50 to 60. The
disproportion is explicable if the change from regnal year 2 to 3 occurred
during the sealing of the wine jars. Thus King 'Ankhkheprure' would
have counted his reign from a day in ca. II Akhet (Krauss, MDOG
129, 1997, 238), which may have coincided with the occurrence of
Akhenaten’s death.

Tut'ankhamun There are no plausible proposals for his accession date,


nor by dated inscriptions for his first three years. A graffito at Saqqara
is dated 2/IV/ Shemu year 4 (Step Pyramid I, 78); year 4 is also attested
on a donation stela (W. Kaiser et al., Ägyptisches Museum Berlin [1967]
no. 776); a label on a textile from his tomb cites year 6 (Urk. IV 2062,4);
the decree for the Overseer of the Treasury Maya is dated 22/III/
Peret year 8 (Stela Liverpool E. 583: A. A. M. A. Amer, “Tutankhamun’s
Decree for the Chief Treasurer Maya”, RdE 36, 1985, 17–20). Wine
jar labels from the tomb document years 4, 5, 9 and 10 ( J. Cernÿ,
Hieratic Inscriptions from the Tomb of Tut'ankhamun, Oxford 1965, 1–3), but
the last could also have been Akhenaten’s (R. Krauss, OLZ 90, 1995,
245–246, and P. Tallet, BIFAO 96, 1996, 369–383).
Analysis of botanical remains from the funeral show that the burial
took place in March/April, and thus the death in January, III or IV
Peret (R. Krauss, “Nochmals die Bestattungszeit Tutanchamuns”, SAK
23, 1996, 227–254). The age of the mummy was estimated at ca. 18
years (F. Filce Leek, The Human Remains from the Tomb of Tut'ankhamun,
Oxford 1972).

Aya The death of Tut'ankhamun in III or IV Peret should date the


accession of Aya. A wine jar label from Deir el-Medina mentions Aya’s
mortuary temple and a year 2 (KRI VII 65,9), but the date could belong
the new kingdom 209

to Haremhab’s reign. A donation stela of the king is dated to 1/III/


Shemu year 3 (Urk. IV 2109,8), and two others bear the date 1/IV/
Akhet year 4 (Urk. IV 2110,13), the highest date known, which accords
well with Josephus’ 4 years 1 month.

Haremhab The accession may have been in II Akhet, provided it was


related to the king’s participation in the Opet Festival (Hornung,
Untersuchungen, 38–39). But the change of year from 26 to 27 in O.
IFAO 1254 leads Krauss to conclude that the accession may be placed
in the interval between 28/IV/ Peret and 13/I/ Shemu (DE 30, 1994,
73–85). However, it is not certain that these dates should be assigned
to Haremhab.
Festival foundations of 22/IV/ Akhet year 1 are known from Karnak
(Urk. IV 2132,4); the rewards of Neferhotep in TT 50 (Urk. IV 2177,6)
date to year 3; 1/I /Akhet year 6 is attested on a stela from the mor-
tuary temple of Amenhotep III (G. Haeny, BÄBA 11, 1981, 65–70);
year 7 is mentioned in O. BM 5624 (IV 2162,10); and Maya’s graffito in
the tomb of Thutmose IV is dated in III Akhet year 8 (Urk. IV 2170,15).
Helck compiled a number of additional dates from years 1 to 13 (Ä
& L 3, 1992, 64), and the highest certain date at present is III Akhet
year 13, in a wine jar label from Saqqara (Martin, JEA 65, 1979, 15;
cf. Murnane, Texts, 107 C 2). Hari (Horemhab, 392) wanted to assign a
year 20(?) in El Kab (Capart, ASAE 37, 1937, 10) to Haremhab. Krauss
argues that O. IFAO 1254 supports a longer reign (DE 30, 1994,
73–85), and a longer reign is also favored by Beckerath (SAK 6, 1978,
43–49) and van Dijk (GM 148, 1995, 29–34: 15–17 years).
The interpretation of the graffito in the royal mortuary temple refer-
ring to the “day of entering of King Haremhab”, dated to 9/I/ Shemu
year 27, is contentious. Beckerath (Tanis und Theben, Glückstadt 1951,
104) argued that this was the date of death (better: burial); the use of
the expression jrjt '˚w n with the meaning “burial” is occurs in the con-
temporaneous TT 50 (Hari, La tombe thébaine du père divin Neferhotep,
Geneva 1985, pl. X), and '˚w is also used for the burial of the Apis-
mothers (Smith, in Lloyd, ed., Studies J. G. Griffiths, London 1992,
203–204). Thus it seems legitimate to understand this date as a refer-
ence to the burial of Haremhab, which suggests that he died at the
end of II Peret. O. IFAO 1254 also points to year 27, and this in turn
is compatible, without any emendation, with the year 58 or 59 of the
Mes inscription, including the reigns of all the proscribed kings of the
Amarna Period.
210 erik hornung

Dyn. 19

With the exception of the relationship between Sety II and Amenmesses,


the royal succession of this dynasty is not a matter of debate, and,
excluding some minor details, the lengths of the various reigns are like-
wise relatively certain. Helck provided a list of the confirmed regnal
years from Ramesses I to year 10 of Ramesses II (Ä & L 3, 1992, 64),
and Wente & Van Siclen compiled another covering the time span
from year 32 of Ramesses II through Ramesses X (“Chronology”,
251–261). Kitchen’s index (KRI VIII 70–84), supplemented by Helck,
Ostraka, is more comprehensive for Ramesses I through Ramesses XI.
However, as many dates are not linked to a royal name, assignment
to particular reigns are subject to change. Demarée has supplied an
overview of accession dates from Sety I through Ramesses XI (GM
137, 1993, 52).
The surviving portion of Manetho’s list for the later phase of the
dynasty is problematic since only Twosre’s (“Thoeris”) reign is correct
at 7 years, including the coregency with Siptah and her sole rule.
Africanus provides a total of 209 years, and Eusebius 194, but both
are too high as the actual sum cannot be more than 104 years.

Ramesses I A possible date for his accession can be deduced using Sety
I’s accession date in combination with the months indicated by Josephus
(neither of which is certain, however), resulting in III, or possibly IV
Peret. On this basis, the only certain date of the reign, 20/II/ Peret
year 2 on the stela Louvre C 57 from Buhen (KRI I 2,9), would lie at
the end of the second year of the reign. The king may have died at
the beginning of his third year, which would allow for two full years.
(For dates from his reign see also Krauss, DE 30, 1994, 83, n. 13.)
Josephus assigns him 1 year 4 months which Africanus rounds down
to one year, allowing for an estimate of one or, alternatively, two years.

Sety I Utilizing the date of a later processional festival celebrating Sety


I (KRI VI 249,7), Helck (CdE 41, 1966, 233–234; SAK 17, 1990, 207–208)
and Krauss (OLZ 90, 1995, 246; SAK 24, 1997, 168) propose 24/III/
Shemu (cf. KRI VIII 70) for the accession. Basing his analysis on palace
accounts in Memphis, Murnane identified a period between 18/III to
23/IV/ Shemu, and favored the festival date of 2/IV/ Shemu (Serapis
3, 1975–1976, 23–33). Using pRollin 1889 Redford (King-Lists, 113 with
n. 46) chooses 23/IV/ Shemu, or the interval between 30/II/ Shemu
the new kingdom 211

and 2/I/ Akhet, which is compatible with the festival date. For Brand
(The Monuments of Seti I, Leiden 2000, 302), 24/III/ Shemu is “the most
likely solution”, but the date cannot be considered certain.
Analyzing the inscription of Bakenkhons, Bierbrier reckons the length
of the reign at 15 years ( JEA 58, 1972, 303), and Kitchen ( JNES 39,
1980, 170–171; High, Middle or Low I, 3, and III, 153–154) concurs.
However, Jansen-Winkeln argues (JNES 52, 1993, 221–225) that the
autobiographical details of the high priest’s career do not allow for such
extrapolations. Spalinger suspects that 10 years and a fraction is more
appropriate, given the data from the campaigns ( JARCE 16, 1979, 41
n. 106). The abundance of sources for years 1–11 suggests that Sety I
died in year 11; only year 10 is missing (KRI VIII 70), while 13/IV/
Shemu year 11 is known from Gebel Barkal (KRI I 75,8). The limited
production of the Aswan quarries which were opened in year 9 (Brand,
JARCE 34, 1997, 101–114) favors 11 years, as does the extended sum
of 51 regnal years in Africanus ( Josephus gives 59). After discussing
the sources, Brand favors 11 full years, or perhaps 10 years (Monuments
of Seti I, 305–309), and Kitchen now suggests 11–15 years (in SCIEM,
2000, 42–43).
Arguments for a coregency between Sety I and Ramesses II (Murnane,
Ancient Egyptian Coregencies, Chicago 1977, 57–87, 183–185) have been
criticized by Kitchen ( JNES 39, 1980, 169–171), and Brand also comes
to a negative conclusion (Monuments of Seti I, 312–332. Regardless,
Ramesses II’s regnal year count did not begin under Sety I. Clearly,
the kings of the Ramesside period found a way for the designated suc-
cessor to share in sovereignty other than using a titulary and separate
dating system.

Ramesses II Helck suggested 27/III/ Shemu for the accession of Ramesses


II (AnBib 12, 1959, 118–120); Krauss (SAK 5, 1977, 146–148), Valbelle
(Ouvriers, 167) and Demarée (GM 137, 1993, 52) concur with this date
which is that of a regularly documented festival. An alternative sugges-
tion made by Larson and Wente & Van Siclen was countered by Helck
(SAK 17, 1990, 205–207). After exhaustive discussion, both Beckerath
(Chronologie NR, 68–70) and Brand (Monuments of Sety I, 302–305) agree
with Helck and Krauss.
The king certainly ruled for 66 full years (66 years 2 months accord-
ing to Josephus), and every year of his reign is documented (KRI VIII
70–73), with the highest date being 18/I/ Akhet of year 67 (RAD
30,10), soon followed by 19/II/ Akhet of a year 1 (RAD 30,14). There
212 erik hornung

is an indication that he died after 29/I/ Akhet (Valbelle, Ouvriers, 176


with n. 4).

Merenptah O. Cairo CG 25504 shows that the accession occurred


between 19/I and 13/II/ Akhet. Peden (“A Note on the Accession
Date of Merenptah”, GM 140, 1994, 69) narrows this interval to
3–13/II/ Akhet, on the basis of a Theban graffito from 2/II/ Akhet
year 1 (but cf. Krauss, OLZ 90, 1995, 246–247). Other options are
19/I/ Akhet (Demarée, GM 137, 1993, 52, opposed by Kitchen, BiOr
60, 2003, 586) and 5–7/II/ Akhet (Beckerath, GM 191, 2002, 5–6).
The highest date is IV/ Shemu year 9 (O. Gardiner 197: KRI IV
159,5). The report of his death reached Deir el Medina on 16/?/ Peret
(HO 64,1,1); presumably he died during year 10. 7/IV/ Akhet year
10 is mentioned in P. Sallier I, 3,4 (LEM 79), and KRI IV 160 also
assigns two Theban graffiti of 7 and 13/II/ Akhet year 10, announc-
ing the inundation, to the reign of Merenptah. The 19 years 6 months
given by Josephus can thus be reduced by a decade. Sety II was his im-
mediate successor, without any intervening rule by Amenmesses (Krauss,
SAK 24, 1997, 174–177).

Sety II Helck calculated that the accession took place between the end
of I Peret and the beginning of III Peret (AnBib 12, 1959, 123). He is
followed by Janssen (Varia, 101 with n. 21), whereas Krauss posits an
interval of 29/II/ Peret through 3/III/ Peret (extended up to 6/III/
Peret by Beckerath, Chronologie NR, 71). For Demarée (GM 137, 1993,
52), it is 29/II/ Peret, announced on 16/?/ Peret in HO 64,1,1.
According to O. Cairo CG 25560 (KRI IV 302), the king was in Thebes
on 10/II/ Akhet year 1.
The king’s death was reported in Thebes on 19/I/ Peret year 6 (O.
Cairo CG 25515: KRI IV 322); O. CG 25516 from 17/I/ Peret is still
dated to his reign (KRI IV 328). A graffito above the tomb of Twosre
dates his burial to 11/III/ Peret in year 1 (of Siptah; Altenmüller, in
Reeves, ed., After Tut'ankhamun, London & New York 1992, 148, fig. 19).
The mummy (CG 61037) suggests that Sety II died young.

Amenmesses Janssen (Varia, 99–109), Krauss (SAK 24, 1997, 161–181)


and Beckerath (Chronologie NR, 70–73) have recently discussed the rela-
tionship between the reigns of Amenmesses and Sety I. We know from
HO 64,1 that the Foreman Neferhotep was absent from work after the
accession of Sety II; thus his murder by “the enemy” can only have
the new kingdom 213

taken place later if the enemy be identified with Amenmesses or, rather,
his followers. The date of accession can be placed between 27/I/Shemu
and 18/III/ Shemu (O. Cairo CG 25783 & 25784: KRI IV 227,6). He
was in power in Thebes during years 3 and 4 (perhaps earlier in Nubia),
which are unaccounted for in the dates of Sety II, but are followed by
a great “clean-up” in Deir el-Medina by Sety II. Treating Amenmesses
as a rival king provides the best explanation for the various phases in
the decoration of Sety II’s tomb, as pointed out by Dodson (“The
Decorative Phases of the Tomb of Sethos II and their Historical
Implications”, JEA 85, 1999, 131–142), and also for the interruption
of work in the tomb of Twosre (Altenmüller, in After Tut'ankhamun,
141–164, esp. 149, 159). The last date of the rival king is 29/III/
Shemu year 4 (O. Cairo CG 25784, 15—the only ostracon from this
year!). Assigning ostraca from Deir el-Medina dated in years 1 and 2
to Amenmesses (so Helck, Ostraka, 97) cannot be justified.

Siptah Helck’s suggestion that O. Cairo CG 25521 allows the acces-


sion to be placed between 28/IV/ Akhet and 3/I/ Peret (AnBib 12,
1959, 123–124) presumes a scribal error, but is supported by the year
change in P. Greg between 28/IV/ Akhet and 11/I/ Peret ( Janssen,
Varia, 116, and Beckerath, Chronologie NR, 74). Following Helck (Studies
Kakosy, 270), Demarée sets the date as 2/I/ Peret (GM 137, 1993, 52).
Given the report of the death of Sety II on 19/I/ Peret (see supra),
the accession must have taken place early in I Peret.
The execution of the “Chancellor” Bay was announced on 27/III/
Shemu year 5 (Grandet, BIFAO 100, 2000, 339–342); therefore he can-
not have been buried on 22/IV/ Akhet in year 3, as Altenmüller sug-
gested (SAK 23, 1996, 1–9, and GM 171, 1999, 13–18). This date
(without year) is better linked to the burial of Siptah by the Vizier Hori
(O. Cairo CG 25792: KRI IV 414–415). To year 6 belongs a graffito
found in Buhen (KRI IV 365 [2]). Siptah died between 9 and 12/II/
Akhet and was buried on 22/IV/ Akhet year 7, according to Beckerath
(Chronologie NR, 74, following Helck, in Studies Kakosy, 270). Alternatively,
Krauss (OLZ 90, 1995, 247–248) and Schneider (ZÄS 130, 2003, 144,
146) suggest year 6.

Twosre Work in her tomb began in year 2 of Sety II (W. Helck, SAK
17, 1990, 208–210), but was interrupted for a time (see Amenmesses).
After her debut as sole ruler, the queen counted her years as a con-
tinuation of the deceased Siptah’s reign.
214 erik hornung

A graffito in Deir el-Bahri mentions a “visit” of Amun to the mor-


tuary temple of the reigning pharaoh Twosre on 28/II/ Shemu year
7 (Marciniak, Inscriptions, 59–60, No. 3). A date 5/III/ Peret year 8
(O. Deir el-Medina 594: KRI IV 407,16) is assigned by Krauss (SAK
4, 1976, 191, and OLZ 90, 1995, 248 n. 29) to Merenptah’s reign
instead of hers. For IV [Peret?] of year 8 on O. Cairo 25293 see
Altenmüller, JEA 68, 1982, 114 who suggests that she died in I Shemu
year 8, which would correspond to the reign of 7 years recorded by
the Manethonian copyists.

Dyn. 20

The ancient compilers who excerpted Manetho’s history did not record
any names for this dynasty, simply citing “12” (actually 10) kings who
reigned for 135 (Africanus), 172 (Eusebius) or 178 years. Only Africanus’
total could be correct, provided it follows directly on the sum down to
Merneptah and so includes the last rulers of Dyn. 19. The actual length
of the various reigns is both certain and precise, thanks to an abun-
dance of dated documents. The only significant change resulting from
recent scholarship is a shortening of the reign of Ramesses X. Parker
assigned him 9 full years based on a presumed lunar date, whereas
contemporaneous sources do not justify more than 3 years. Problems
were also created by a hypothetical “interregnum” at the inception of
the dynasty, but again, our sources preclude such a proposal; there is
simply no space available.
Demarée (GM 137, 1993, 49–52) and Helck (Ostraka) have discussed
the accession dates, which are in general clearly restricted. Documented
regnal years are listed in KRI VIII.

Sethnakhte His accession date is not known, but Beckerath proposes the
beginning of II Shemu (Studies Kakosy, 63–67; Chronologie NR, 75–76).
Dates are only known from year 2, on 10/II and 24/III/ Shemu, but
possibly also (without month) from year 3 on Sinai (Beckerath, Studies
Kakosy, 63–67). He probably died on 25/I/ Shemu, at the start of his
year 3 (Altenmüller, GM 145, 1995, 29–36). If his year count subsumes
the sole rule of Twosre, he will have ruled for only a year; however,
the graffiti Deir el-Bahri 3 and 10 are linked to the Festival of the
Valley, and imply an interval of 10 years between year 7 of Twosre
and year 7 of Ramesses III (Krauss, Sothis, 139), so that Sethnakhte
must be assigned some 3 years.
the new kingdom 215

Ramesses III His accession was on 26/I/ Shemu (Beckerath, Chronologie


NR, 77, with n. 431). Every year of the reign is documented, except
for the first, down through year 32. His death occurred on 14?/III/
Shemu year 32. The burial equipment was transported into the Valley
of the Kings on 4/I/ Akhet, and the burial will have taken place on
24/I/ Akhet (O. Deir el-Medina 40: KRI VI 106, see Cerny, ZÄS 72,
1936, 113, and Helck, Studies Kakosy, 269). According to information
provided by O. Chicago 12073 (Allam, Ostraka, No. 40, 73–76) 18 years
elapsed between year 17 of Ramesses III and year 3 of Ramesses IV.

Ramesses IV His accession on 15/III/ Shemu is certain (O. Deir el-


Medina 44); word reached the workers at Deir el-Medina on the fol-
lowing day (O. Deir el-Medina 39 and P. Turin 1949 + 1946), as the
accession took place in Thebes. For the accession dates of Ramesses
IV, V and VI see Janssen, Varia, ch. VIII. Dates extend to 23/III/
Akhet year 7 (O. Deir el-Medina 207: KRI VI 149), and the Turin
Indictment Papyrus (RAD 80,8–9) indicates that Ramesses IV must have
died before the harvest of his year 7, i.e., before May (Gregorian),
probably in IV Peret. The estimated age of the mummy (CG 61041)
was at least 50 years.

Ramesses V Beckerath calculated the accession at around 1/IV/ Peret


(ZÄS 122, 1995, 98; GM 157, 1997, 7–10); Gutgesell (Die Datierung der
Ostraka und Papyri, Hildesheim 1983, 227–229), followed by Helck (Ostraka,
411), prefers 29/III/ Peret, and Janssen (GM 155, 1996, 62) tries to
set the date between 29/IV/ Peret and 7/I/ Shemu, but this remains
uncertain. Recently, Beckerath put the accession before I Peret (GM
188, 2002, 16–17), supported by Demarée’s new interpretation of P.
Turin 2044.
O. Cairo CG 25247 (without royal name) documents 25/II to 17/IV/
Shemu of year 4; a year 5 is not attested. The accession date of his
successor indicates that he will have died between the middle of I Peret
and the start of II Peret of his year 4. His mummy (CG 61042) shows
him to be the victim of a smallpox epidemic, who was probably just
over 30 years of age (Harris & Weeks, X-raying the Pharaohs, London
1973, 166–167). Krauss assigns him a reign of 3 years 10 months (OLZ
90, 1995, 249).

Ramesses VI His accession can be restricted to between 28/I and 11/II


Peret (Beckerath, GM 79, 1984, 7, based on O. IFAO 1425); argu-
ments for 8/II/ Peret are given by Janssen (Varia, 131–138). The
216 erik hornung

announcement followed on ?/II/ Peret (KRI VI 36 4). The highest date


of the reign is 11/II/ Peret year 8 ( Janssen, “Year 8 of Ramesses VI
Attested”, GM 29, 1978, 45–46), and a reign of 7 years 9 months can
be deduced from Theban graffito 1860a (Krauss, OLZ 90, 1995, 249).
The earlier literature reflects confusion about the sequence (and thus
names, numbers, dates and chronology) of the next two kings, but it
is now clear that Ramesses VII (Itiamun) preceded Ramesses VIII
(Sethherkhepeshef ).

Ramesses VII. (Itiamun) Suggestions for the accession date vary widely:
from between 20/II/ Peret and 5/II/ Shemu ( Janssen, JEA 52, 1966,
92), to between 30/III and 26/IV/ Peret (Beckerath, ZÄS 122, 1995,
99; but idem, Chronologie NR, 85: on 30/III or 1/IV/ Peret), between
10/III and 26/IV/ Peret ( Janssen, GM 155, 1996, 61), and “shortly
after” 14/IV/ Peret (Demarée, GM 137, 1993, 52). Utilizing P. Amiens
and other documents, Beckerath recently proposed 30/III/ Peret (GM
188, 2002, 17). Year 7 is documented in P. Turin 1907/1908 (KRI VI
403–409) and O. Strasbourg H 84 with 16/II/ Shemu ( Janssen, JEA
52, 1966, 91 n. 2). While Eyre argues for a 25/IV/ Shemu year 8
(P. Turin 1883 + 2095: “The reign-length of Ramesses VII”, JEA 66,
1980, 168–170), Valbelle suggests that the king in question may actu-
ally be Ramesses IX (Ouvriers, 39 with n. 8 and 204).

Ramesses VIII The accession date has been fixed between 2/I/ Peret
and 13/I/ Akhet (Amer, GM 49, 1981, 9–12) and, according to Beckerath
(in: Deir el-Medina in the third millenium AD, Leiden 2000, 4–5), must lie
before 13/I/ Akhet. Only year 1 is documented, but the calendar pro-
vides a basis for arguing a year 2 if he reigned a full year, as his death
must have occurred on 20/I/ Akhet because of the accession date of
his successor.

Ramesses IX His accession probably occured on 21/I/ Akhet (Beckerath,


GM 79, 1984, 7–8), but in any case on or after 18/I/ Akhet (Beckerath,
ZÄS 127, 2000, 112; Janssen, GM 191, 2002, 59–65). He died in his
year 19 with III and IV Akhet documented (Botti, JEA 14, 1928, 48–51,
without royal name), and Beckerath suggests more precisely the end of
I Peret (ZÄS 127, 2000, 112).

Ramesses X The accession was between 25 and 27/I/ Peret (Beckerath,


GM 79, 1984, 8–9, based on Turin Cat. 2075 + fragm. = KRI VI 653,
the new kingdom 217

and ZÄS 127, 2000, 112, followed by Valbelle, Ouvriers, 43, n. 3, and
Demarée, GM 137, 1993, 50). IV Akhet year 3 is documented in the
Giornale, followed by 24/I/ Akhet year 1 (Helck, GM 70, 1984, 32).
Krauss allows for a year 4 (GM 70, 1984, 42–43), but a year 8 pro-
posed by Bierbrier ( JEA 61, 1975, 251) is quite uncertain, and Parker’s
astronomical arguments for 9 full regnal years can be discarded (Krauss,
Sothis, 151–154, and Jansen-Winkeln, ZÄS 119, 1992, 32–33).

Ramesses XI According to Helck (SAK 17, 1990, 211–212), his acces-


sion date was 27 or 28/IV/ Shemu, but 20/III/ Shemu according to
Gardiner ( JEA 26, 1940, 23–29) and others (Cannuyer, Studies Lichtheim
I, 98–105; Ohlhafer, GM 135, 1993, 59–72; Demarée, GM 137, 1993,
50), but not between 26/III/ Shemu and the 4th epagomenal day
(Beckerath, OLZ 85, 1990, 657 n. 1). His year 27 is documented (8?/IV/
Shemu: KRI VI 701,15). The Renaissance (w˙m-mswt-) era counted 10
years and thus lasted into his 28th year (25/I/ Shemu year 10 in the
letter BM 10326 = LRL no. 9). This means that a 29–year reign as
given in the Book of Sôthis is possible. Dates from the Renaissance
era are listed by Thijs, GM 173, 1999, 190–191. Years of this era are
to be correlated with the regnal years of Ramesses XI and do not
accord with calendar years (Ohlhafer, GM 135, 1993, 59–72). Thijs’
proposal that Ramesses IX–XI were contemporary (GM 167, 170, 173;
SAK 31, 2003, 289–306) has been countered by Beckerath (ZÄS 127,
2000, 114–116, and GM 181, 2001, 17).
II. 9 RELATIVE CHRONOLOGY OF DYN. 21

Karl Jansen-Winkeln

At the beginning of Dyn. 21 Egypt was split in two, with two centres
of power, each ruled individually. UE, whose northern frontier was
located in the region of Herakleopolis, was governed by a military com-
mander who, at the same time was HPA of Thebes.1 In texts and
depictions some of these UE regents (Herihor, Pinudjem I and
Menkheperre) assume in varying degrees attributes which are reserved
for a king. Kings reigned in LE, but at least two of them (Psusennes
and Amenemope) occasionally bear the title of “HPA”. Contemporaneous
documents of which only a small number survived do not give any
direct indication as to the reason for this partition of Egypt.2 The only
large group of finds are the graves of the kings in Tanis and the col-
lective interments in the Theban necropolis (including replacements and
re-interments of older mummies). Among these Theban funeral sites
various dated objects can be found, but unfortunately most dates are
anonymous and not ascribed to any explicit regent. Of this twofold
line of regents, Manetho lists only the kings of LE, namely (1) Smendes,
(2) Psusennes [I], (3) Nepherkheres, (4) Amenophthis, (5) Osochor, (6)
Psinaches, (7) Psusennes [II]. Contemporary documents contain ample
reference of the kings Psusennes (P #-sb #-¢'j-m-nwt; only in LE), Amenemope
( Jmn-m-Jpt) and Siamun (Z #-Jmn) (both in LE and UE). The first two
kings can be straightforwardly identified as Manetho’s Psusennes (I) and
Amenophthis. A king named Smendes (Ns-b #-nb-ddt) is attested by only
a few, undated inscriptions, but the history of Wenamun shows clearly
that he was a contemporary of Herihor and thus the first king of Dyn.
21. The identification of the remaining four kings, on the other hand,
has caused some problems.
The Nepherkheres of Manetho is not attested as the personal name
of any king. Two bow caps from the grave goods of Psusennes I dis-

1
The first two rulers also called themselves Viceroys of Nubia; the first three had
the title Vizier.
2
For an attempted explanation, see K. Jansen-Winkeln, Orientalia 70 (2001), 153–182.
dynasty 21 219

play the throne-name and the personal name of Psusennes opposite the
throne-name Nfr-k #-R' (˙q # W #st) and the personal name Mrjj-Jmn Jmn-
m-njswt.3 Obviously the throne-name Nfr-k #-R' has been handed down
as Nepherkheres by Manetho. The proper name Amenemnisut ( Jmn-
m-njswt) is attested only a second time on the relief Berlin 23673 from
the reign of Shoshenq V, on which a long line of ancestors of the
owner is named, sometimes together with the reigning king. On this
relief, Amenemnisut is the predecessor of Psusennes I, whereas Manetho
names him as the successor. His true position has not yet been identified.
The Berlin genealogy was compiled only about 250–300 years after
the reign of Amenemnisut and should, therefore, be given greater con-
sideration than Manetho’s frequently garbled tradition. But the fact that
Psusennes and Amenemnisut appear together on one funeral object,
strengthens the idea that Amenemnisut was the successor of Psusennes
and that he donated the object.4 Nevertheless, Amenemnisut (Nepher-
kheres) was without question an ephemeral king.
The Osochor of Manetho is attested contemporarily only by one
inscription from Karnak, which registers the inauguration of a priest
in year 2 of a king with the throne-name '#-¢pr-R' Stp.n-R' (the personal
name is missing in a lacuna).5 E. Young has demonstrated6 that this
king cannot be Psusennes I, as believed in the past, because he always
bears the epithet Stp.n-Jmn. Furthermore, a few lines further down, the
text refers to the inauguration of the priest’s son in the year 17 of
Siamun. If ' #-¢pr-'# Stp.n-R' really were identical with Psusennes I, then
the inaugurations of father and son had to have been almost three gen-
erations apart. Therefore this otherwise unknown throne-name from
Dyn. 21 may well be that of Manetho’s Osochor. In this case the sec-
ond inauguration would have taken place only 21 years, or about one
generation, later, if Manetho’s 6 years for Osochor be accepted.
Actually, the personal name of the king is mentioned once, but not
contemporarily.7 An inscription (no longer traceable) from the roof of

3
Montet, Tanis II, 105; 108, Fig. 44; pl. 72 (No. 413/414).
4
See also Kitchen, TIP, 70–71. An alternative to this could be that the bow was
made during a co-regency of the two kings, cf. TIP, 70–71 and Beckerath, Chronologie,
101. However the reign of Nepherkheres only lasted for a few years, and a co-regency
is more likely at the end of a long reign. For the Berlin genealogy see Bochardt, Mittel,
96–112; Bl. 2/2a.
5
No. 3B of the “Annals of the Priests”, see G. Legrain, RT 22 (1900), 53; Kruchten,
Annales, pl. 2; 17.
6
JARCE 2 (1963), 100–101.
7
Concerning the following see J. Yoyotte, BSFE 77–78 (1976/77), 39–54; cf. also
Kitchen, TIP, § 437.
220 karl jansen-winkeln

the temple of Khonsu from year 9 of Takelot III8 mentions, among


the author’s ancestors, a king Osorkon and his mother M˙jt-m-ws¢t.
This Osorkon cannot be identical with one of the kings named Osorkon
from Dyns. 22–23, because their mothers had different names.9 A king’s
mother called M˙jt-m-ws¢t is known from Dyn. 21; on the stela of P #-
sn-Ór from the Serapeum the grandmother of Shoshenq I is named
likewise.10 The two texts complement each other optimally and indi-
cate the existence of a king Osorkon in Dyn. 21, the uncle of the later
Shoshenq I, who can be identified as Manetho’s Osochor. They also
match in time: Psusennes II, father-in-law of Osorkon I, is assumed to
have been a contemporary of Shoshenq I. Because “Osochor”11 was
the older brother of Shoshenq I’s father, he might well have been the
second predecessor of Psusennes II.12 The identification of Manetho’s
Osochor by Young and Yoyotte has gained general acceptance.
Manetho’s last king but one, Psin(n)aches, cannot be found in any
Egyptian sources whatever. The only name that could be considered
(with some modifications), would be P #-sb #-¢'j-m-nwt,13 but that name
has already, and rightly so, been identified as Psusennes. On the other
hand, contemporary documents reveal a King Siamun (Z#-Jmn) bear-
ing the throne-name NΔrj-¢pr-R' as the last but one king of Dyn. 21,
who does not appear in Manetho’s history. It is tempting, therefore,
to identify Manetho’s Psinaches with Siamun,14 even though the lengths
of their reigns do not match: Manetho’s Psinaches is supposed to have
reigned for nine years, Siamun, by contrast, for at least 17 years. A
solution would be to amend the number 9 to <1>9. This identification
and emendation have become traditional, as the most obvious. Anyway,
we should always bear in mind that this identification originates only

8
LD, III, 258c; G. Daressy, RT 18 (1896), 51–52.
9
Nor can Osorkon III and IV be considered, for chronological reasons.
10
See CSSM, 30–31; Kitchen, TIP, § 85.
11
This form of the Egyptian-Libyan word Ws(j)rkn (“Osorkon”) is attested elsewhere
in Manetho.
12
F. Payraudeau, “Remarques sur l’identité du premier et du dernier Osorkon”,
GM 178 (2000), 75–80, is of the opinion that two objects of a king ' #-¢pr-R' stp.n-Jmn
Mrjj-Jmn Wsjrkn, which until now have been ascribed to Osorkon IV (whose throne-
name is unknown), originally belonged to Osochor. If this is correct, Osochor would
have taken turns using the epithets stp.n-R' and stp.n-Jmn in his throne-name.
13
Cf. M. Römer, GM 114 (1990), 94.
14
Cf., most recently, J.v. Beckerath, GM 130 (1992), 17–19 and (concerning a pos-
sible explanation for the varying information in Manetho’s work) GM 131 (1992), 11.
dynasty 21 221

from the fact that we can neither find a king from the end of Dyn.
21 who is named in contemporaneous documents in Manetho’s work,
nor can we find Manetho’s last but one king Psinaches on Egyptian
monuments. The remaining two criteria for the identification have not
been met: neither name nor length of reign being the same.
The identification of Manetho’s second king called Psusennes with a
(Ór-) P #-sb #-¢'j-m-nwt is, on the one hand, unequivocal and undisputed.
On the other hand, however, there is the question as to whether the
last king of Dyn. 21 is identical with the last HP of Thebes of that
dynasty who has the same name.15 Actually, the evidence weighs heav-
ily in favour of his being one and the same man, who was first HP
and then successor to King Siamun in Tanis, without giving up his
Theban office.
The only reference for the HP Psusennes can be found on shrouds
and mummy-braces (etc.) from the priests’ mummies in the so-called
second Cachette (Bab el-Gusus).16 From 10 references, 8 name him
HP, whereas on the other 217 his name appears in a cartouche. No
other titles are mentioned, which for H. Kees meant that he—in con-
trast to his predecessors—no longer possessed military power.18 But this
conclusion was perhaps overly hasty, because the HP Menkheperre,
who held the highest offices,19 is referred to on mummy wrappings from
the second Cachette as only a HP,20 his name otherwise appearing in
a cartouche;21 his military titles are not mentioned at all, and in the
filiations of his descendants his name is often cited without any titles.22
It can be established that the HP Psusennes’ name is sometimes written
in a cartouche like the names of Herihor, Pinudjem (I) and Menkheperre,
whilst his father and predecessor Pinudjem II never used any royal

15
In Kitchen, TIP they are distinguished from one another as Psusennes II (= the
king) and III (= the HP).
16
Burials A.17; 43; 48; 58; 65; 66; 125; 132; 133; 148, see G. Daressy, ASAE 8
(1907), 23–37.
17
A.58 and 66, see Daressy (n. 16).
18
H. Kees, Die Hohenpriester des Amun von Karnak von Herihor bis zum Ende der Äthiopen-
zeit (Leiden: PÄ 4, 1964), 79: “In contrast to all of his predecessors in Thebes he did
not style himself supreme commander of the UE army.”
19
Cf. M. Römer, Gottes- und Priesterherrschaft am Ende des Neuen Reiches (Wiesbaden:
ÄUAT 21, 1994), 66–73.
20
Burials A.2; 13; 96; 105; 109; 113, see Daresssy (n. 16), 22–31.
21
A.11 and 64, Daressy (n. 16), 22; 27.
22
A.12; 26; 32; 38; 81 (Daressy [n. 16], 22–28) and elsewhere.
222 karl jansen-winkeln

attributes. A graffito from the Temple of Abydos23 reveals the com-


plete titles of a king Tjt-¢pr-R' Stp.n-R' P #-sb #-¢'j-(m-)nwt Mrjj-Jmn, who
is simultaneously HPA and supreme military commander. Actually, the
elements of the titles of his kingship, his duties as HP and his military
titles blend into each other in a peculiar manner, not to be found else-
where. He is called
njswt-bjt nb t #wj Tjt-¢pr-R' Stp.n-R' <mrjj>(?) Jmn-R' njswt nΔrw24
˙m-nΔr tpj n Jmn-R' njswt nΔrw
z # R' nb ¢'w
˙ #wtj P #-sb #-¢'j-(m-)nwt Mrjj-Jmn ntj (r-)¢ #t n # m“ '[w n Kmt ≈rw]
˙m-nΔr tpj n Jmn-R' njswt nΔrw
jrj hpw nfrw n Kmt
˙ #wtj pr-' # P #-sb#-¢'j-(m-)nwt Mrjj-Jmn.
The military title, ˙#wtj pr-'# P #-sb #-¢'j-(m-)nwt ntj (r-)˙ #t n # m“'[w n Kmt
≈rw], is very informative. It reveals distinctly that this is the HP Psusennes,
the successor of Pinudjem II, and not a king who has adopted the
additional title of HP (as Psusennes I and Amenemope did). The rea-
son is that this title is only to be found in connection with Theban
HP and military commanders,25 but never in connection with a Tanite
king. The throne-name of Psusennes in this graffito also appears with
slight variation (Tjt-¢prw-R' ) on a vessel fragment from Abydos.26 A king
bearing almost the same name, Tjt-¢pr-R' Stp.n-R' Mrjj-Jmn Ór-P #-sb #-
¢'j-m-nwt, can be found outside Abydos on two Theban statues: (1)
Cairo CG 42192, on which he is named as an ancestor of his grand-
son M#'-¢pr-R' Stp.n-R' Mrjj-Jmn ”“nq (Shoshenq II);27 (2) the Nile-statue
London BM 8 of that particular grandson which also mentions the

23
M. A. Murray, The Osireion at Abydos (London, 1989), 36; pl. XXI; G. Daressy,
RT 21 (1899), 9–10.
24
Concerning this epithet, cf. M.-A. Bonhême, Les noms royaux dans l’Egypte de la
Troisième Période Intermédiaire (Cairo: BdE 98, 1987), 61.
25
See GM 99 (1987), 19. No. 8 is to be crossed out of this list, see JEA 81 (1995),
130; instead, the HP Jwlt is attested a second time on an altar-stand in Moscow, see
S. Hodjash & O. Berlev, The Egyptian Reliefs and Stelae in the Pushkin Museum of Fine Arts,
Moscow (Leningrad, 1982), 157/161 (No. 105).
26
E. Amélineau, Les nouvelles fouilles d’Abydos 1897–1898 (Paris, 1904), 146 (24).
27
Shoshenq II donated the statue, and not Schoschenq I, as often reported; cf. J.v.
Beckerath, Orientalia 63 (1994), 84–87 and K. Jansen-Winkeln, JEA 81 (1995), 145–148,
who both render the text.
dynasty 21 223

daughter of Psusennes and mother of Shoshenq II, M#'t-k #-R'.28 The


additional Ór(-P #-sb#-¢'j-m-nwt)29 is not a distinctive feature, but appears
with reference to one and the same person30 as demonstrated by the
Decree for Maatkare.31 Here the very same Psusennes and father of
Maatkare is solely called Mrjj-Jmn P #-sb#-¢'j-m-nwt. The obvious con-
clusion is that all these cartouches refer to the same person.32 The
graffito from Abydos also demonstrates that he was king and at the
same time HP in Thebes; he had clearly not resigned this office.33 He
was probably buried in Tanis (and later re-buried in the Antechamber
of the tomb of Psusennes I).34 A limestone-fragment with his name has
been found near Tell el-Daba.35
Dodson drew the conclusion that Psusennes did not have a reign of
his own at all, but was only an UE ephemeral King next to Sho-
shenq I,36 from the fact that many of his attestations are posthumous
and that he is often mentioned together with Shoshenq I.37 This is not
at all convincing: on CG 42194 and BM 8 he is only mentioned in
his grandson’s genealogy, and together with Shoshenq I he only appears
in the tomb TT A.18.38 By contrast the latter inscription provides

28
C. R. Lepsius, Auswahl der wichtigsten Urkunden des Aegypischen Alterthums (Leipzig,
1842), pl. XV.
29
Another reference is an inscription on a bead of unknown provenance, see GLR
III, 300 (IV). The ivory stick-handle, which in Gauthier’s opinion also belonged to
Psusennes II (GLR III, 302[IV]), more probably belonged to Psusennes I.
30
In contrast to Bonhême (n. 24), 60, who inexplicably would like to recognize up
to four different persons in Tjt-¢pr-R' (Ór-)P#-sb#-¢'j-m-nwt; cf. also J. Yoyotte, BSFFT
1 (1988), 46(1).
31
J. Winand, Cahiers de Karnak XI (2003), 672ff.; 707 (Fig.4), l.3, 5.
32
Beckerath’s distinction (GM 130 [1992], 18) between a Tanite king (Hor-) Psusennes
with the throne-name Tjt-¢prw-R' Stp.n-R' and a HP who, in the role of a (mock-) king
(on the graffito in Abydos) bore the throne-name Tjt-¢prw-R' Stp.n-Jmn is not correct.
The epithet is Stp.n-R' in this graffito, too. Furthermore he is also called Tjt-¢prw-R'
Stp.n-R' on a vessel fragment from Abydos (n. 26, above) which cannot be connected
to any other Psusennes than the one from the graffito.
33
It is inexplicable why Beckerath (GM 130, 1992, 18) writes that if the HP Psusennes
had inherited the crown from Siamun he would have had to appoint a new HP. A
HP’s and a king’s office do not exclude each other in dynasty XXI.
34
Cf. Yoyotte (n. 30), 41–53; idem, Tanis, L’or des pharaons. Exhibition-catalogue
(Paris, 1987), 136–137.
35
M. Bietak, Avaris and Piramesse: Archaeological Exploration in the Eastern Nile Delta
(Oxford, 1981), 271; but cf. LÄ V, 131.
36
RdE 38 (1987), 49–54; BES 14 (2000), 9–12.
37
The statues Cairo CG 42192; CG 42194 (name destroyed), London BM 8 and
the Decree for Maatkare.
38
A. Dodson, JEA 79 (1993, 267–268; pl. 28. On Cairo CG 42192, on the other
hand, Psusennes does not appear together with Shoshenq I, but with another king
called Shoshenq, see above, footnote 27.
224 karl jansen-winkeln

weighty evidence that Shoshenq I was Psusennes’s successor: someone


is promoted by Psusennes (s¢ntj.f ), and is promoted once again during
the reign of Shoshenq (w˙m ¢ntj.f ). There is no reference that the two
kings reigned in parallel. Considering the fact that Psusennes II was
buried in Tanis and appears in Manetho’s list of kings, we cannot say
that he was only an UE ephemeral king.39
There is thus evidence of the following LE kings in Dyn. 21: (1)
Smendes, (2/3) Psusennes/Amenemnisut (Nepherkheres), (4) Amenemope,
(5) Osorkon (“the Elder”, Osochor), (6) Siamun (“Psinaches” in Manetho’s
work) and (7) Psusennes II (at the same time HP in Thebes). We do
not know much about the familial relationships of these kings. There
is nothing to be said about the origin of Amenemnisut, Amenemope
and Siamun. Smendes I’s wife, Tentamun,40 is mother of Henuttawy,
wife of Pinudjem I and mother of Psusennes I.41 Consequently Smendes
I would be father-in-law of Pinudjem I and grandfather of his succes-
sor(?) Psusennes I. Osochor, being son of Nimlot I and of M˙jt-m-ws¢t,
is uncle of Shoshenq I. Psusennes II is son of HP Pinudjem II and
father-in-law of Osorkon I; the latter already concerns the relationship
to the kings of Dyn. 22.
Nine UE rulers are known as belonging to Dyn. 21. Eight of them
are part of a lineage of fathers and sons:

Payankh

Pinudjem I

Masaharta Djedkhonsiuefankh Menkheperre

Smendes II Pinudjem II

Psusennes (II =) III

39
Cf. also Beckerath, GM 130 (1992), 17f; Kitchen, TIP 3 1995, XIX–XXI.
40
Herself being the daughter of a man without any important titles, called Nebseni.
41
Under the probable circumstance that the King’s Mother who is mentioned on
some funerary objects, Henuttawi is identical with Pinudjem’s wife. Kitchen’s postu-
dynasty 21 225

Only Herihor does not belong to this lineage;42 his position as a pre-
decessor or successor of Payankh is the only one disputed (see below).
The order of the others is clear, even though some overlap. There is
evidence of the HP (and king) Pinudjem I until a year 15, and of his
son Masaharta in the years 16 and 18 following. His son Menkheperre
who is clearly younger takes up the duties of the HP in a year 25 and
from then holds office for almost five decades. A third son of Pinudjem
I, called Djedkhonsiuefankh, is recorded only once as a HP on a coffin
which is at present missing;43 he most probably held office for a very
short time between Masaharta and Menkheperre.44 However, Pinudjem
survived his son’s term of office and died in that of Menkheperre (see
below). Evidence of Smendes II is, admittedly, somewhat better than
that of Djedkhonsiuefankh, but his term of office can only have been
very short, either as Menkheperre’s successor or as his “co-regent” (see
below). His brother Pinudjem II came next in office, followed by his
own son Psusennes, who is probably identical with King Psusennes II
(see above). Consequently we have the following order: (1/2) Payankh and
Herihor (see below), (3) Pinudjem I, (4) Masaharta, (5) Djedkhonsiue-
fankh, (6) Menkheperre, (7) Smendes II, (8) Pinudjem II, (9) Psusennes III.
Some of the UE regents are related by blood or marriage to those
of LE: Smendes I seems to be the father-in-law of Pinudjem I, Pinudjem
himself is Psusennes I’s father (see above). HP Psusennes himself becomes
king in Tanis.
Concerning the succession of the first two HP, Herihor was on
account of a copying error believed for a long time to be father and
predecessor of Payankh. Since this error has been corrected, 45

late of a second (older) Henuttawi “Q” as a hypothetical second wife of Smendes and
Psusennes’ mother is only rooted in his wish for a genealogical bridge to the Ramessides
for Psusennes on account of his occasionally being called “Ramses-Psusennes”. This
has only confused matters unnecessarily.
42
Depending on whether he was predecessor or successor of Payankh, he might
have been his father-in-law (Kitchen, TIP 2 § 438) or son-in-law (K. Jansen-Winkeln,
ZÄS 119 [1992], 25) or he might have married his wife after Payankh died ( J. Taylor,
in: Eyre, Proceedings, 1143–1155).
43
Kitchen, TIP § 392.
44
We cannot totally exclude the possibility that he was a predecessor of Masaharta’s
who was in office only for a short period. According to A. Niwinski (BES 6, 1984,
83–6) he was a son of Pinudjem II; Torr’s filiation data would in consequence not
name his father but his great-grandfather (!) Pinudjem I.
45
Cf. E. F. Wente, (Fs Korostovtsev) Drevny Vostok (Moscow, 1975), 36–38; The Temple
of Khonsu. I. OIP 100 (1979), p. 13(d); pl. 26, l. 4.
226 karl jansen-winkeln

the succession has had to be explained by other means. The term of


office of both HP or at least part of it can be said to have taken place
in the later years of the reign of Ramses XI. Records mention Herihor’s
years 5 and 6 (without any explicit relation), and Payankh’s year 7 of
the w˙m-mswt-era and a year 10. At first sight it would be logical if
Herihor had held office in the first half of the w˙m-mswt-era and Payankh
in the second. Even so, a series of arguments favour a reverse order:46
(1) The form of the titles: We can recognise Payankh’s origin from the
rank of officers much more clearly than that of Herihor. He is mostly
referred to simply as “The General”, his military titles being much
more prominent and detailed than those of Herihor. His titles are in
general similar to those of Pinhasi, who was in charge of UE from the
beginning of the w˙m-mswt-era. The titles of Herihor on the other hand
are more related to those of the later HP. Furthermore, Payankh’s titles
almost always refer to the king (. . . n pr-' #), as was usual in the Ramesside
period, whereas those of Herihor no longer do so. (2) Payankh never
assumes any royal titles or attributes, whereas Herihor and the later
HP do. (3) Herihor and Pinudjem I are both recorded as builders in
Thebes, and Pinudjem directly succeeds Herihor with regard to the
decoration of the temple of Khonsu. Payankh on the other hand is not
recorded as a builder. A similar situation is to be found regarding the
(re-)burials in the Theban necropolis. On shrouds, bandages etc. of
these mummies, every single HP of Dyn. 21 is recorded, except Payankh.
Thus these burials must have taken place after his term of office. (4)
The genealogical information corresponds more to a Payankh-Herihor
succession. The order of these HP is still being discussed,47 but in my
opinion the order Payankh-Herihor is the more probable solution. At
any rate, this problem has a direct influence on the chronology of the
whole dynasty.
Most of the dates preserved from Dyn. 21 are from Thebes, and
most do not refer to a specific ruler. Breasted presupposed that all

46
ZÄS 119 (1992), 22–25.
47
The following authors do not agree with the thesis published in ZÄS 119 naming
Herihor as Payankh’s predecessor: A. Niwinski, BIFAO 95 (1995), 346–47; J.v. Beckerath,
in: D. Kessler & R. Schulz, eds., Gedenkschrift für Winfried Barta (Frankfurt: MÄU 4,
1995), 49–53; A. Gnirs, Militär und Gesellschaft (Heidelberg: SAGA 17, 1996), 199–201;
Kitchen, TIP 3 1995, XIV–XIX (A–N). For a response see K. Jansen-Winkeln, GM
157 (1997), 49–74. In favour of the succession Payankh—Herihor are A. Egberts (GM
160 [1997], 23–25; ZÄS 125 [1998], 93–108) and J. Taylor (see above, footnote 42).
dynasty 21 227

those dates were related to the LE kings48 and in recent times this
opinion has found general acceptance. The opinion is supported by
some explicit dates which almost always mention the name of a Tanite
King: There is a date which is explicitly related to Amenemope,49
another one is related to Osochor,50 and six are related to Siamun.51
Furthermore Amenemope and Siamun are quite well documented in
Thebes. On the other hand there is only one date which is explicitly
related to a HP.52 Under that condition, the following years would be
recorded:53
W˙m-mswt-era: 4; 5; 6; 7; 10
Smendes I: 1; 4(?); 6; 9; 10; 11; 12; 13; 15; 16; 18; 19; 20; 21; 25
Amenemnisut: –
Psusennes I: 6; 7; 8; 19; 27;54 30; 40; 48; 49
Amenemope: 1; 3; 5; 10(?)55
Osochor: 2
Siamun: 1; 2; 3; 5; 6; 7; 8; 9; 10; 12; 14; 16; 17
Psusennes II: 5; 13(?)56
The currently recorded dates can be made compatible in this order
with Manetho’s lengths of reign as preserved by Africanus. He gives
Smendes 26 years, Nepherkheres (Amenemnisut) 4, Psusennes 46 (41
according to Eusebius), Amenemope 9, Osochor 6, Psinaches (that
means Siamun [?]) 9 and Psusennes (II) 14 (35 according to Eusebius).
At the end of Psusennes I’s reign there was supposedly a co-regency
with Amenemope. On condition that this dating system was used, the
famous linen-bandage with the inscription “King Amenemope; year 49”
can be restored beyond doubt to “[year X under] King Amenemope;
year 49 [under King Psusennes; linen made by HP NN . . .]”,

48
BAR IV, § 604–607.
49
Kitchen, TIP § 388, no. 54.
50
Kitchen, TIP § 388, no. 56.
51
Kitchen, TIP § 389, no. 73; 74; 77; 82 (from LE); 83; 84.
52
Kitchen, TIP § 387, no. 46. This latter record must be interpreted differently if
we adhere to a continuing dating by LE kings, cf. e.g., E. Young, JARCE 2 (1963),
102–103, n. 21; Kitchen, TIP § 377.
53
With reference to the evidence listed in Kitchen, TIP § 379–381. Only the under-
lined dates are connected explicitly with the king.
54
A. Dodson & J. J. Janssen, JEA 75 (1989), 128, 134.
55
This date may also refer to Siamun, cf. Kitchen, TIP § 388, no. 55.
56
Possibly referring to Shoshenq I, cf. Kitchen, TIP § 391, no. 86; 87.
228 karl jansen-winkeln

the juxtaposition of the years being evidence of a co-regency.57 As a


result, Amenemnisut must have been the predecessor of Psusennes, and
the only contemporary record of this king would indicate a co-regency
Amenemnisut—Psusennes58 at the beginning of Psusennes I’s reign.
Various suppositions have been made concerning the length of these
two (hypothetical) co-regencies,59 almost all of them deriving from
Manetho’s information: Only 46 of the 49 recorded years were to be
taken into consideration.
On the other hand, in the case of Psinaches/Siamun, Manetho has
to be emended. Siamun’s attested 17 years mandates the emendation
9 > 19 (y > iy). Altogether Dyn. 21 would have lasted 124 years which
is the result of adding the lengths of reign according to Africanus and
this emendation. The difference between these 124 years and Manetho’s
sum of 130 years (indicated in all versions, regardless of the actual,
correct total) might be explained by suggesting that Manetho calcu-
lated those years in which there was a co-regency for both rulers.60
According to this hypothesis, the lengths of reign for the UE rulers
would be as follows:
Herihor until year 6 (or 7) of the w˙m-mswt-era;
Payankh from year 6 (or 7) until year 1 of Smendes I at most;
Pinudjem at the earliest from year 10 of the w˙m-mswt-era onward,
until year 15 (year 16 at most) of Smendes I in his position as HP,
after that at least until year 8 of Psusennes I as king;
Masaharta from year 16 (15 at the earliest) until year 25 of Smendes
I as a HP at the latest;
Djedkhonsiufankh only for a very brief period between Masaharta
and Menkheperre;
Menkheperre from year 25 of Smendes I until (at least) year 48 of
Psusennes I.
Smendes II for a brief period between Menkheperre and Pinudjem II;
Pinudjem II from year 1 of Amenemope or shortly thereafter;61
Psusennes “III” from year 10 of Siamun on.
If, however—which seems probable—Payankh is not the successor but
the predecessor of Herihor, this system cannot easily be maintained.

57
Kitchen, TIP § 29.
58
See above, footnote 4 and Kitchen, TIP § 56.
59
Cf. for example Kitchen, TIP § 29, 465 (table I) or Beckerath, Chronologie, 101–102.
60
Cf. Beckerath, Chronologie, 101–102; idem (n. 47), 54–55.
61
Kitchen, TIP § 388 (51).
dynasty 21 229

The highest recorded date for Herihor is a year 6, 15/III/Peret.62 This


date could only refer to Smendes if Herihor followed Payankh and if
the dates refer exclusively to the LE kings. But Pinudjem was already
recorded in year 6, 7/III/Peret (of Smendes after this system). The above
given dating-system could only be retained if Herihor’s date was to be
read 7/III/Akhet 63 instead of 7/III/Peret (or emended accordingly), but
that would be an unhappy solution.
Even so, there is some information for Theban dates of the UE
kings. There is a record of a year 48 of HP Menkheperre;64 moreover,
a closer look reveals a complementary distribution of the records con-
cerning the rulers of that time. In the first half of Dyn. 21, HP Herihor,
Pinudjem I65 and Menkheperre have royal attributes and titles to differing
extents. On the other hand, the LE kings of that time are virtually not
recorded at all in UE: there is a graffito mentioning Smendes66 and a
rock-stela,67 and nothing for Amenemnisut and Psusennes I, even though
the latter reigned for a long time. Subsequently, however, Amenemope
and Siamun are well documented in Thebes, and Osochor at least
once, whereas HP Pinudjem II (who held office parallel to them), does
not adopt any royal attributes or titles. It is, therefore, likely that the
HP who called themselves kings counted their own years of reign
whereas during the second half of the dynasty the dates refer to the
LE kings. This would mean that the beginning of Amenemope’s reign
might have implied a change in the dating-system and concurrently a
change in the political structures.68
A possible, but very hypothetical explanation would be that a new
family or a new branch of the same family gained power in Tanis and

62
Kitchen, TIP § 379, no. 3
63
Cf. ZÄS 119, 26; Beckerath (n. 47), 51.
64
Kitchen, TIP § 387, no. 46.
65
This does not, of course, apply to Masaharta and Djedkhonsiuefankh since their
period is equal to that of Pinudjem I.
66
A. Varille, Karnak(-Nord) I (Cairo 1943), 36, Fig. 26, pl. 98 (71); L. A. Christophe.
Karnak-Nord III (Cairo 1951), 77.
67
G. Daressy, RT 10 (1888), 135f. Already in Daressy’s time part of the text was
gone; in the meantime everything has been destroyed. The genre of the text (Königsnovelle)
normally requires a date, but the structure of the text does not require a date in that
part which was already missing in Daressy’s time.
68
P. Brooklyn 16.205 might contain some information with regard to a critical sit-
uation in UE, referring to a year 49 of Dyn. 21 as a “bad time” (h#w bjn); concern-
ing the dating of the papyrus to Dyn. 21, see J.v. Beckerath, GM 140 (1994), 15–17;
Kitchen, TIP 3, XXVI (Y).
230 karl jansen-winkeln

then successfully laid claim to supremacy over the whole of Egypt. We


know that Smendes and Psusennes I were closely related to the UE
family of HP (see above). No family relationships whatsoever are known
for Amenemope and Siamun, but Osochor, who held office between
them, was a son of the Libyan great chief of the Meshwesh, Shoshenq
A, and the uncle of the later Shoshenq I. We do not know if this fam-
ily was in any way related to the descendants of Payankh, although it
is possible that Amenemope, Osochor and Siamun all belonged to this
family, or to a branch of it. It is also striking that HP Pinudjem II,
son of Menkheperre is not only called his son (z # Mn-¢pr-R' ), but also,
sometimes even on the same object, the son (= descendant) of King
Psusennes (I).69 Thus it seems to have been important to stress his being
part of this half of the royal family. A change of royal family with
Amenemope could explain a change within the dating-system.
Assuming that the UE regents Herihor, Pinudjem I and Menkheperre
counted their own regnal years, we can draw some conclusions. For
the period of the LE kings Smendes I, Amenemnisut70 and Psusennes
I just one single date would have been recorded in Egyptian sources,
and even that from later times: Year 19 of a king Psusennes is men-
tioned in retrospect on a stela from the Dakhla Oasis dated to year
five of Shoshenq (I).71 Under these conditions only the regnal years as
given by Manetho could be used as evidence for the dates of these
kings—which is precisely what scholars have done.
Uncertainty prevents us from precisely calculating regnal years for
the first three UE rulers. Herihor reigned for at least 5 whole years
(year 6 is recorded), possibly slightly longer (up to 8 years). In year 25
of Pinudjem, his son Menkheperre was installed as HP,72 and soon after
that a new count of years begins.73 As a result we have to calculate at

69
Daressy (n. 16), 23 (no. 24); 27 (no. 61); 28 (no. 81, no. 82); 31 (no. 113); 32
(no. 119, 120); 36 (no. 139).
70
If the linen-band with the regnal year 49 (cf. above) is not to be associated with
Psusennes but rather with Menkheperre, there is no need to propose a co-regency for
Psusennes and Amenemope. As a result, the question of whether Amenemnisut was
predecessor or successor of Psusennes reappears (see above, footnote 4.
71
A. H. Gardiner, JEA 19 (1933), 32; pl. VI, l.11. Concerning the dating see
H. Jacquet-Gordon, in: Hommages à la mémoire de Serge Sauneron I (Cairo: BdE 81/1,
1979), 180–182; O. Kaper, BACE 12 (2001), 77, n. 6; R. Krauss, DE 62 (2005), 43–48.
72
“Banishment Stela,” l. 1–8, see J.v. Beckerath, RdE 20 (1968), 10.
73
In line 7/8 of the Banishment Stela a lower date follows (RdE 20, 10–11; 33).
The two events described in the text should not lie too far apart from each other.
dynasty 21 231

least 24 years for Pinudjem, at most 25 years. The highest date recorded
for Menkheperre is the year 49 and in that year (his last?) Amenemope
may have already reigned in Tanis.74
At first sight this seems to be contradicted by the fact that Smendes
II, son of Menkheperre would have had to be HP at the latest when
Psusennes I died,75 because he donated goods for the burial. For this
reason he cannot have been Menkheperre’s successor if the reign of
Menkheperre overlaps with that of Amenemope and even less so if Ame-
nemnisut was Psusennes’s successor. Niwinski presumed that Smendes
II was only HPA in Tanis at that time, later becoming Menkheperre’s
successor for a short time.76 This is possible, but in my opinion it is
more probable that Smendes—like Masaharta previously—held office
parallel to his father at the end of his father’s reign, while the count-
ing of regnal years continued to follow Menkheperre’s reign. However
that may be, 48 years is the most likely calculation for Menkheperre.77
Consequently, the first three UE rulers could be reckoned to have
held office for at least roughly as long as the LE kings, namely 77
years (5 + 24 + 48), possibly 1 or 2 years less, if the overlap between
Menkheperre and Amenemope is greater. A slightly longer period seems
to be more probable, including some leeway for Herihor, altogether
perhaps 80 years, hardly significantly longer. In other words, the dates
we have from Manetho’s tradition, 124 years (the sum of the lengths
of reign according to Africanus with emendation 9 to 19 for Psina-
ches) and 130 years (sum total in all versions), set the limits of what
is possible. Most likely is a total of about 126–8 years. If there is a
difference in the lengths of the reigns of the HP Herihor, Pinudjem I

The lower date is probably the first year of Menkheperre, who grants an amnesty on
New Year, which is an appropriate act at the beginning of a reign. Thus the event
would mark the transition from Pinudjem to Menkheperre, not only the inauguration
of a new HP under the royal authority of Pinudjem which would have been implied
by the other dating-system.
74
The linen-band with the inscription “[Year X under] King Amenemope; Year
49 [under NN]” (see above) under this circumstance would contain a common date
of Amenemope and the HP Menkheperre.
75
Cf. Kitchen, TIP § 25.
76
A. Niwinski, JARCE 16 (1979), 59–60; idem, 21st dynasty Coffins from Thebes (Mainz:
Theben 5, 1988), 50–51 (§ 43).
77
Less only if the overlap with the era of Amenemope lasted longer. But if the
“bad time” for year 49 (footnote 68) is connected with the change of regency to
Amenemope, year 49 of Menkheperre could be the same as year 1 of Amenemope.
232 karl jansen-winkeln

and Menkheperre and the parallel reigning LE kings (from the reign
of Amenemope onwards there is no difference in the two dating-sys-
tems anyway) it would only amount to a few years. And we do not
know if Smendes and Herihor started their reign at the same time or
whether the Manethonian numbers are all correct.78
At the beginning of Dyn. 22 there is a certain fixed point which
links Dyn. 21 to absolute chronology, i.e. Shoshenq I’s campaign in
Palestine. According to the OT,79 the Egyptian King Shishak besieged
Jerusalem in year 5 of Rehabeam, king of Judah. On the Egyptian
side, the campaign is attested by a victory scene in Karnak. Year 5 of
Rehabeam can be pinned down to about 926/925 BC with the aid of
the known lengths of reign of the kings of Israel and Judah and their
synchronisms—although there are some inconsistencies—as well as by
means of two synchronisms with the Assyrian chronology.80
From Egyptian sources we do not know when Sheshonq’s campaign
took place. Construction work on the pylon and the court, on whose
exterior walls the scene of triumph is depicted, began in his year 21
(possibly his last year but one), as recorded on a rock-stela.81 The major-
ity opinion is that the construction work and the campaign were con-
nected to each other and that the campaign did not take place very
long before construction work started, in year 20 at the earliest. An
essential point for the temporal connection between the campaign and
the construction work could be that of the whole decoration which was
planned in Karnak only this triumphal scene has been completed. So
if this campaign really took place in year 20 or 21, Shoshenq’s reign
would have begun in 946 or 945 BC. However, we can in no way
be certain that the campaign took place immediately before the
construction work started.82 There is no reason why it could not have

78
When dating according to the High Priests’ years of office, we nevertheless have
to consider the necessity of adding a few (possibly 2–3) years to Herihor’s term of
office under Ramesses XI subsequent to Payankh’s term of office. Anyway, Ramesses
XI’s absolute length of reign (or the length of the w˙m-mswt-era) is uncertain.
79
Kings I 14,25; II Chronicles 12,2.
80
Cf. Hornung, Untersuchungen, 24–29; Kitchen, TIP § 59; Beckerath, Chronologie
68–70. This fixed point is only valid if we work on the assumption that the information
concerning the kings’ lengths of reign in the OT has been taken from reliable sources.
81
R. A. Caminos, JEA 38 (1952), 46–61.
82
Almost unanimous in the literature: presumably supported by the wish for at least
one fixed point.
dynasty 21 233

taken place several years earlier.83 In that case, the beginning of


Shoshenq’s reign would have to be set slightly later, and thus the entire
Dyn. 21.

83
Even if we could establish that there was a causal relationship between the cam-
paign and the construction work, the work in Thebes could still have been begun long
after the campaign. One could argue that the first priority was the enlargement and
decoration of the LE temples and that simultaneous work in LE and UE was beyond
the capacity of both the labour force and the architects. At least the temple of El-
Hibeh in Middle Egypt had a depiction of the triumph, cf. ASAE 2 (1901), 85–87;
154–156; H. Ranke, Koptische Friedhöfe bei Karâra und der Amontempel Scheschonks I. bei el
Hibe (Berlin & Leipzig, 1926), 50–52; pls. 19–21; E. Feucht, SAK 9 (1981), 105–117;
pl. 2.
II. 10 THE CHRONOLOGY OF THE THIRD
INTERMEDIATE PERIOD: DYNS. 22–24

Karl Jansen-Winkeln

The prevailing conditions and patterns of rule during Dyns. 22–23 were
basically similar to the state of Egypt during Dyn. 21. UE—with the
important centers Thebes and Herakleopolis—was administered by a
military governor who was simultaneously the High Priest of Amun;
LE was directly governed by the king with residences in Memphis and
Tanis (and in Bubastis as well, since Osorkon I). With the aid of their
sons, the first kings of Dyn. 22 maintained their rule over the entire
country. However, since the reign of Osorkon II at the latest, they
gradually lost out to the powers of decentralisation, when (due to a
divided inheritance?) clearly defined and separate spheres of power and
local potentates appeared, particularly in LE.1 In the same fashion, the
separation of UE and LE remains tangible under Libyan rule.2
The most important chronological sources for UE are the records
of the Nile levels,3 the annals of the priests at Karnak,4 the “Chronicle
of Prince Osorkon”,5 and the statues (and other objects) belonging to
dignitaries from certain families which permit detailed and extensive
genealogies;6 for LE, we only have the donation stelae7 and the stelae
from the Serapeum.8 Altogether, there are relatively few actual dates

1
It is not clear whether this regionalisation only came into existence at this time,
or whether it existed earlier, i.e., already perhaps in Dyn. 21, but only became clear
in the sources at this time (the most important sources are the donation stelae, and
these only become abundant from later Dyn. 22, being totally absent in Dyn. 21). It
is probable that there were at least incipient developments in this direction, which
became more strongly expressed later.
2
In fact, this division led to different cursive scripts used in the administration: the
“anormal” hieratic in UE, and “Demotic” in LE.
3
PM II2, 21–22; J.v. Beckerath, JARCE 5 (1966), 43–55; G. Broekman, JEA 88
(2002), 163–178.
4
PM II2, 108; G. Legrain, RT 22 (1900), 51–63; Kruchten, Annales.
5
PM II2, 35–36; Reliefs III, pl. 16–22; Caminos, Chronicle.
6
Cf. TIP, §§ 157–205; Bierbrier, LNK, passim.
7
See Meeks, Donations.
8
See CSSM; PM III2, 780ff.
the third intermediate period 235

surviving from this period. As a rule—in contrast to the NK9—we lack


a continuous series (or even relatively complete chain) of dates for any
given sovereign, and thus by no means can we confidently suggest that
the highest known date for any reign reflects its actual length. Given
this paucity of dates, the chronology of this era is imprecise and uncer-
tain in many respects.
The actual means of dating was presumably the same as that of the
NK, 10 as is suggested by the dates from one Serapeum stela.11 These
affirm that an apis bull, born in year 28 of Shoshenq III, was intro-
duced on 1/II/Akhet of the same year: if the year began on 1/I/Akhet,
the Apis would have been a month old at the most—and this is highly
unlikely.12 Furthermore, his predecessor was buried in the same year,13
and there are generally several months between the burial of the pre-
vious Apis and the introduction of the new one.14 It follows that the
regnal year still began with the accession of the king; unfortunately,
there are no surviving accession dates for the TIP.

1. The Rulers of Unified Egypt of Early Dyn. 22

According to Manetho, following Africanus, Dyn. 22 consisted of 9


kings from Bubastis who ruled for 120 years: Sesonchis (21 years),
Osorthon (15), three others (25), Takelotis (13) and three more (42).15
The family tree in the Serapeum stela of Pasenhor from year 37 of
Shoshenq V (' #-¢pr-R' )16 includes a reference to a King Osorkon who
ruled six generations earlier, whose father, grandfather and great-
grandfather were kings named Takelot, Osorkon und Shoshenq, while
their forefathers were not kings, but rather Libyan princes. The non-
royal origins of the earliest named king, Shoshenq, the exact corre-
spondence of the names of the kings with those listed by Manetho for

9
Cf. KRI, VIII, 70–84.
10
Thus also Beckerath, Chronologie, 10. It is a priori probable that the MK concept
of “predating” was among the anachronisms introduced during Dyns. 25–26.
11
Louvre SIM 3697, cf. CSSM, 21–22; pl. VIII (no. 22). R. Krauss drew my atten-
tion to the importance of these dates.
12
Cf. E. Winter, Der Apiskult im Alten Ägypten (Mainz, 1983), 18.
13
Stela Louvre SIM 3749, CSSM 19–20; pl. VII (no. 21).
14
See L. Depuydt, JNES 54 (1995), 123; Kienitz, Geschichte, 155.
15
According to Eusebius only 3 kings in 49 years, namely Sesonchosis (21), Osorthon
(15) und Takelothis (13).
16
Louvre SIM 2846, cf. CSSM 30–31; pl. X (no. 31).
236 karl jansen-winkeln

this dynasty, and the period of time separating Pasenhor (nine generations
to year 37 of Shoshenq V) clearly reveal that these were the first kings
of Dyn. 22. In addition, the grandparents of this oldest Shoshenq link
him to Dyn. 21, as he is the nephew of the third to the last king of
that dynasty, Osorkon (Osochor).17 This gives a sequence of 4 kings,
each pair being father and son, for the start of Dyn. 22: Shoshenq I
(Ó≈-¢pr-R'), Osorkon I (S¢m-¢pr-R' ), Takelot I (Ó≈-¢pr-R' ) and Osorkon
II (z# B #stt Wsr-m #'t-R' ).18
Although each king is the son of a former king, this does not nec-
essarily mean that each son immediately followed his father in office.
It is entirely possible that other sovereigns can be fitted into the sequence.
According to Africanus, Manetho inserts three other kings, and the fol-
lowing are candidates for this:
a) On his own documents, and in the patronymic of his son (a priest
of Amun named Osorkon), the HPA, Shoshenq, son of Osorkon I and
grandson of Psusennes II is designated as HP and Generalissimo and
not as king.19 Only on the London statue BM 8 does he enclose his
name (in the titulary of HP) in a cartouche, adding the epithet mrjj-
Imn.20
b) The statue Cairo CG 41292 from Karnak21 was re-inscribed by
a king Shoshenq with the throne-name M#'-¢pr-R' Stp-n-R', and to the
benefit of his “begetter” (msj sw) Psusennes II. It is entirely possible that
this is an otherwise completely unknown son of Psusennes II,22 but it
seems more reasonable to assume that this is the (earlier?) high priest
and son of Osorkon I,23 who could easily have designated himself as
“begotten” by Psusennes, his grandfather.24

17
J. Yoyotte, BSFE 77–78 (1977), 39–54; cf. above Jansen-Winkeln, Chapter II. 9.
18
The throne-names were not listed on the stela of Pasenhor. Assigning the kings
with these throne-names to the first kings of the dynasty results from a (i.a.) compar-
ison with the kings appearing in the family tree of the Theban Nakhtefmut family, cf.
TIP, § 88. For the throne-name of Takelot I, cf. VA 3 (1987), 253–258; TIP 3, XXII–XXIII.
19
The catalogue of these monuments (all from UE): K. Jansen-Winkeln, “Historische
Probleme der 3. Zwischenzeit”, JEA 81 (1995), 145–146.
20
S. PM II2, 289.
21
G. Legrain, Statues et statuettes de rois et de particuliers, III (Cairo, 1914), 1–2; pl. 1;
J.v. Beckerath, Orientalia 63 (1994), 84–87; K. Jansen-Winkeln (n. 19), 147–148; pl. XIII.
22
G. Broekman, GM 176 (2000), 39–46, considers Shoshenq M#'-¢pr-R' to be a son
of Psusennes II who was able to assert his claims to be the royal successor of his father
in Thebes at least, while Shoshenq I was recognized in LE (and dates in Thebes fol-
lowed his reign).
23
Thus also Beckerath (n. 21), 86; N. Dautzenberg, GM 144 (1995), 21.
24
As jtj and z# can mean “grandfather” and “grandson”.
the third intermediate period 237

c) A number of kings were subsequently interred in the antecham-


ber of the tomb of Psusennes I at Tanis, including two anonymous
mummies25 and a Shoshenq Óq#-¢pr-R' Stp-n-R',26 who was presumably
already more than 50 years of age27 and whose throne-name bore a
form reminiscent of early Dyn. 22 (before Osorkon II),28 and the same
applies to the iconographic details of his shabtis.29 The interment also
included a pectoral of the great chief of the Ma, Shoshenq A, and a
bracelet of Shoshenq I30—and thus the same person before and after
the accession. As the individuals interred in the royal tombs often bore
objects belonging to their parents,31 this king is probably a son of
Shoshenq I.32 The commonly assumed identification of this king with
the (earlier) HP and son of Osorkon I33 does not appear to be very
probable.
d) A king Shoshenq with the throne-name Twt-¢pr-R' is known from
the sherd Louvre E.31886 from Abydos,34 and apparently also from a
fragmentary relief from Tell Basta.35 This is evidently a king of the
entire country and not a minor UE king or a local ruler. The form
of the throne-name implies that he too belongs near the start of Dyn. 22.

25
Possibly Siamun and Psusennes II, cf. J. Yoyotte in Tanis: L’or des pharaons (Paris,
1987), 48.
26
Montet, Tanis II, 36–51
27
D. E. Derry, ASAE 39 (1939), 549–551.
28
TIP, § 93.
29
G. Broekman, GM 181 (2001), 29–31.
30
Montet (n. 26), 43–45 (219; 226/227); fig.13.
31
TIP, § 93; K. Jansen-Winkeln, VA 3 (1987), 256–257; D. Aston, “Takeloth II—
A King of the ‘Theban Twenty-Third Dynasty’?”, JEA 75 (1989), 139–153, esp.
143–144.
32
In addition he also bore the ring of a ˛d-Pt˙-jw.f-'n¢ (Montet [n. 26], 44, fig. 13;
46 [228]), perhaps his brother: a prince und 2nd/3rd Prophet of Amun of this name
was interred in the cachette of Deir el-Bahri in year 11 of Shoshenq I (G. Maspero,
Les momies royales de Déir el-Baharî (Paris 1889), 572–574; GLR, III, 284, n.2). He was
presumably a son of Shoshenq I.
33
TIP, §§ 93–94; 452; most recently with new arguments Broekman (n. 29), 27–37.
Rather than identifying Shoshenq Óq#-¢pr-R' with the son of Osorkon I and grandson
of Psusennes II, and thus being obliged to reckon with yet another new and hitherto
unknown son of Psusennes, it appears more reasonable to identify the grandson of
Psusennes II with the donor of CG 42192 and to identify Schoschenk Óq#-¢pr-R' as a
son of Shoshenq I, based upon his grave goods.
34
According to the reading of J. Yoyotte, cf. M.-A. Bonhême, BSFE 134 (1995), 53.
35
E. Lange, GM 203 (2004), 65–72. The arrangement of the cartouches does not
allow one to deduce a coregency of Twt-¢pr-R' (= Psusennes II) and Shoshenq (I) as
Dodson does (BES 14 [2000], 9–10). Aside from this, Osorkon I is thus far consid-
ered to be the first sovereign of the TIP documented in Bubastis.
238 karl jansen-winkeln

The HP Shoshenq (“II”) is presumably identical with Shoshenq M#'-


¢pr-R', but most certainly did not have an independent reign, but rather
was responsible for UE during the reign of his father. Shoshenq Óq#-
¢pr-R' may have ruled briefly after his father, if Shoshenq I was his
father, or perhaps after his brother Osorkon I. He could thus have
been one of the “three other kings” Manetho places between Osorkon
(I) and Takelot (I).36 The same applies to Shoshenq Twt-¢pr-R' who
should most probably be put between Osorkon I and Takelot I. In
contrast to his father and his son, not one single royal monument is
known for Takelot I;37 his brothers in UE probably dated according to
his reign (cf. below), but they do not name him. This could indicate
that his rule was undisputed.
For the first part of Dyn. 22 we would thus have the following kings,
and dates:
1. Shoshenq I; documented years 2, 5, 6, 10, 11, 13, 2138
2. Osorkon I: regnal years [1]-4, 6, 10, 11, 12, 23, 3339
3. Shoshenq Óq#-¢pr-R': no dates
4. Shoshenq Twt-¢pr-R': no dates
5. Takelot I: years: 9,40 dubious 5, 8, 13/14, 14 (cf. below)
6. Osorkon II: years 12, 16, 21, 22, 23,41 29(?)42
For Shoshenq I, Manetho’s 21 years appear to be possible, and a reign
of 35 years is quite probable for Osorkon I.43 Only a year 9 is certain
for Takelot I. The Nile level records nos. 16–21 are generally assigned
to his reign: nos. 16 (year 5) and 20–21 (years lost) belong to the HP

36
Takelot II cannot be implied, as he was an UE sovereign, cf. below.
37
Cf. also TIP, §§ 95; 270.
38
The highest regnal year on the rock stela of Gebel Silsila, cf. JEA 38 (1952),
pl. XIII.
39
Of these, only the year 10 in lines 2–3 of the “stèle de l’apanage” (ZÄS 35 [1897],
14) and year 12 of the Nile level record no. 2 (Beckerath [n. 3], 49) are explicitly related
to Osorkon. Regnal year 33 is on the mummy wrappings of a burial, which also had
a “counterweight” bearing the name and throne-name of Osorkon I, cf. J. E. Quibell,
The Ramesseum (London 1998), 10–11; pl. XVIII.
40
G. Daressy, RT 18 (1896), 52–53, earlier ascribed to Takelot II, cf. now Aston
(n. 31), 144; TIP 3, XXIII.
41
Serapeum stela Louvre SIM 3090, s. CSSM, 17; pl. VI (no. 18).
42
Nile level record no. 14, cf. below.
43
Were one to follow Manetho here, we would still be obliged to emend 15 years
to 35. Aside from the 33 which should in all probability to assigned to Osorkon I (cf.
above), there are further indications of a long reign, cf. TIP § 89.
the third intermediate period 239

Iuwelot who was still a youth in year 10 of the reign of his father,
Osorkon I.44 The year 5 must therefore relate to a successor of Osor-
kon I.45 The records nos. 17–19 are from the HP Smendes III, doubtless
the brother and successor of Iuwelot;46 no. 17 is from year 8, no. 18
from year 13 or 14. A block, presumably from the Serapeum, bears
the names of Takelot I and the HP of Memphis, Merenptah;47 Mariette
noted that this was found together with a stela from a year 14.48 This
might be a stela in Alexandria dated to a year 14 (without a royal
name), and originally came from the Serapeum, as the inscription sug-
gests.49 This would thus support Manetho’s 13 (full) years for Takelot.
His possible predecessors (see above) have not left many traces and
assuredly did not reign for a long period.50 Thus for Takelot and the
others, 15 years is a reasonable suggestion.51 One can therefore adopt
Kitchen’s suggestion of 21 + 35 + 15 years for the first 3 to 5 kings
of Dyn. 22. However, these dates should be viewed as the minimum
to which a few more years might be added.
The length of the reign of Osorkon II is a matter of debate, and
Manetho cannot aid here. The highest date which can with certainty
be assigned to his reign is year 23 (see above), linked to an Apis bur-
ial, where his son, the Crown Prince and HP of Memphis, Shoshenq
D apparently also took part.52 Shoshenq D will thus have died after
that time, but apparently before his father,53 and thus Kitchen assigned
Osorkon II 24–25 full years, to allow a margin for these events.54

44
Lines 2–3 of the “stèle de l’apanage”, cf. ZÄS 35 (1897), 14.
45
But certainly not to Osorkon II, whose Nile level records take a different form,
cf. Broekman (n. 3), 171.
46
These records have exactly the same form as those of Iuwelot and differ from all
others, cf. most recently Broekman (n. 3), 164; 170–171.
47
CSSM, 18; pl. VII (no. 19).
48
Mariette’s remarks are, however, rather doubtful, cf. n. 47.
49
G. Daressy, ASAE 5 (1904), 121 [XXIV]. The stela Louvre SIM 2810 (CSSM,
18–19; pl. VII [no. 20]) of a ˛d-Pt˙-jw.f-'n¢ of a year 10 [+ X] (without royal name)
dates to a later epoque, cf. A. Leahy, SAK 7 (1979), 149.
50
If there was a conflict over the throne, it is conceivable, that some of them ruled
parallel to Takelot.
51
If there really was an Apis burial in year 14 of Takelot, and the Apis buried in
year 23 of Osorkon II was the successor of this bull (which is, of course, uncertain) it
would favour placing year 14 towards the end of the reign of Takelot, as 26 years are
the longest documented life of an Apis bull.
52
Can no longer be verified, cf. TIP, § 81, with n. 77; GM 207 (2005), 76, n. 16.
53
This was generally assumed because he is also designated as Crown Prince (rp't
wr tpj n ˙m.f ) in his tomb.
54
TIP, § 87.
240 karl jansen-winkeln

This logic is no longer tenable since Shoshenq D did in fact outlive


his father. In his undisturbed burial was a chain of 8 W≈ #t-amulettes
(Cairo JE 86786), and one of them bore the name of Shoshenq III
(Wsr-m#'t-R' Stp-n-Jmn Mrjj-Jmn z# B #stt ”“nq).55
On the other hand, Aston has produced arguments that Osorkon
II’s reign was clearly longer than previously assumed, and perhaps even
40–45 years.56 Aston’s argument is based on the family trees of two
Theban families which reveal that several generations lived in the reign
of Osorkon II; other genealogical data likewise allegedly favours a longer
reign; furthermore, there would be a whole series of HPA belong-
ing to the reign of Osorkon II, and his three known sons would all
have predeceased him. Of these arguments, only the family tree of the
Nakhtefmut family57 is really reliable, but this actually supports a
relatively long reign for Osorkon II. Whether the genealogy of the
Nebneteru-family58 must also be understood in this sense is more debat-
able: the statue Cairo CG 42225 was erected after the death of its
owner, so that the name of the king and the high priest there could
relate to the date of erection and not necessarily hint at the lifetime
or term of office of the statue’s owner. The other genealogical data
which Aston introduces does favour a long life, but not necessarily a
long reign for Osorkon II. As HPA under Osorkon II only his son
Nimlot C, his grandson Takelot F59 and Harsiese B are documented.60
Of the sons of Osorkon II, Harnakht C died as a child, Shoshenq D
probably did outlive his father (see above), and thus effectively only
Nimlot C predeceased him.61 Nevertheless, I consider the basic sense of
Aston’s arguments to be correct. There is a Nile level record (no. 14)

55
K. Jansen-Winkeln, “Der Prinz und Hohepriester Schoschenk (D)”, GM 207 (2005),
77–78. It is conceivable (although rather improbable) that Osorkon II died immedi-
ately after his son, and thus his successor may have been able to arrange for a gift
for the burial. In this case, it would be certain that Shoshenq III was the immediate
successor of Osorkon II (cf. below).
56
Aston (n. 31), 145–148.
57
Ibidem, 145.
58
Ibidem, 146.
59
Presumably the later Takelot II, cf. K. Jansen-Winkeln (n. 19), 138–139; Dautzenberg
(n. 23), 24.
60
Jansen-Winkeln (n. 19), 135–139.
61
It is interesting to note incidentally that this HPA did leave hardly any traces in
Thebes, being almost exclusively recorded in the genealogical records of his descen-
dents there, and even there he is consistently designated as HPA and General of
Herakleopolis.
the third intermediate period 241

from the year 29 of an Wsr-m #'t-R', who is most probably Osorkon II


and not Shoshenq III or Osorkon III.62 At the very least, the family
tree of the Nakhtefmut-family clearly supports a reign for Osorkon II
of more than the 24 or 25 years Kitchen allows him. In addition, it
must be recalled that it is precisely from the reign of Osorkon II that
we have comparatively numerous monuments, both royal and private:
far more than from the eras of Shoshenq I, Osorkon I and Shoshenq
III who are otherwise the best documented of the TIP. It is therefore
not too bold to suggest a reign of at least 30–40 years for Osorkon
II. In this era, it is hardly surprising that we do not have any dates
from the final decade of the reign.
The king Harsiese (A) also belongs to the period of Osorkon II: on
the stelaphoric statue Cairo CG 42208 we see the complete titulary of
Osorkon II, but the statue was dedicated “by the grace” of Harsiese.63
This Harsiese is known only from UE,64 and was buried in Thebes.
There are no known regnal years relating to him, and dating in his
era presumably followed Osorkon II.65 His reign should probably be
assigned to the beginning of the reign of Osorkon II;66 in any case, it
is not chronologically relevant.
The period from Shoshenq I to Osorkon II should have lasted about
100–111 years (21 + 35 + 15 + 30–40), and would be ca. 945/40–
844/29.

62
Cf. Broekman (n. 3), 174–5.
63
K. Jansen-Winkeln, Ägyptische Biographien der 22. und 23. Dynastie (Wiesbaden: ÄUAT
8, 1985), 453.
64
Jansen-Winkeln (n. 19), 133–5. He is only documented as king, contrary to the
common view, he is not documented as HPA even one single time. Earlier, he was
viewed as the son of the HPA Shoshenq (II); but since it became evident that this
was based on a mistaken reading (ibidem, 129–132), he has become an orphan. In
the necropolis of the TIP at Herakleopolis was the burial of a woman named T#-nt-
Jmn, in Tomb 4. According to the inscriptions of the tomb and grave goods (M. Perez-
Die/P. Vernus, Excavaciones en Ehnasya el Medina (Madrid 1992), 50–59; 128–132;
156–159; Docs. 21–26), she was wrt ¢nrt n Órj-“.f, her father was the ˙m-nΔr tpj (n) Jmn
mr m“ ' ˙#wtj Ns-b#-nb-˛dt, and her mother was [#st-]#¢bjt or J¢y (shortened version) and
she is designated as mwt nΔr. Represented together with T#-nt-Jmn was a man named
Osorkon, who was wr '# n <pr>-S¢m-¢pr-R'. It necessarily follows that the HP Smendes,
the father of T#-nt-Jmn, cannot have been Smendes II of Dyn. 21. If this is not an
HP Smendes unknown from other sources, the only candidate is Smendes III of Dyn.
22. As his wife is designated a “king’s mother” (mwt nΔr cannot be a sacerdotal title
here), Smendes III must have had a son who became king, and who belongs to the
generation of Osorkon II. Harsiese is the obvious candidate.
65
However, the lack of dates could simply be the result of the type of documents
which are preserved.
66
Jansen-Winkeln (n. 19), 135.
242 karl jansen-winkeln

2. Takelot II

Related to the length of the reign of Osorkon II and equally contro-


versial is the question of the identity of his successor; the stela of
Pasenhor has nothing to say on the matter. The HP Osorkon (B) who
left a long inscription (“The Chronicle of Prince Osorkon”) was a son
of Takelot II (throne-name Ó≈-¢pr-R' as with Takelot I), his mother was
a daughter of the HP Nimlot (C) and a granddaughter of Osorkon II.
In the inscription, the donations are at first dated according to the
reign of Takelot II (until year 24), and then according to the reign of
Shoshenq III (years 22–29), and thus a sequence of Osorkon II—Takelot
II—Shoshenq III was deduced.67
D. Aston has dismissed this long established chronology for several
reasons:68 (1) Takelot II is only known in UE; (2) he has the epithet
nΔr ˙q # W #st in his throne-name; (3) his consort and children do not
reveal any known links to LE either; (4) the genealogical details of his
dependents hint that he belonged to the generation of the grandchil-
dren of Osorkon II; (5) in the “Chronicle of Prince Osorkon”, the years
22–29 of Shoshenq III follow years 11–24 of Takelot II: were Takelot
the predecessor of Shoshenq III, we would face a lacuna of more than
two decades. Aston thus assumes that Takelot was a “Theban” ruler
whose realm was restricted to UE, and thus that he ruled parallel to
a LE sovereign. This would have major chronological consequences.
K. A. Kitchen has strongly rejected this approach by attempting to
disprove or disarm Aston’s arguments:69 Takelot II left relatively few
traces in Thebes; other kings who definitely lived in a Delta residence
had relations with Thebes; the epithet nΔr ˙q# W #st was also borne by
Shoshenq V (in Tanis); other kings of the TIP, such as Osochor,
Psusennes II or Osorkon IV were rarely or not attested in LE, although
they actually resided there. On the other hand, he suggests that the
scenario leading to this “Theban” Takelot II is historically excluded:
the Thebans would hardly have accepted a king in Thebes but rejected

67
TIP, § 86; as noted already in principle by R. Lepsius (Über die XXII. ägyptische
Königsdynastie, Berlin 1856, 271–274), who inserted yet another Shoshenq (“II”, our
Shoshenq D) between Osorkon II and Takelot II.
68
Aston (n. 31), 140–144.
69
TIP 3, XXIII–XXIV; JEA 85 (1999), 247; BiOr 58 (2001), 383.
the third intermediate period 243

and opposed his son as HP, and they would never have tolerated this
HP as the later king Osorkon III. This argumentation is not convinc-
ing. Takelot II and his son definitely belonged to a common “party”
in the civil war; had Osorkon B been expelled from Thebes, the same
would be true of his father. And it is rather doubtful that the opinion
of the people (the “Thebans”) would have had any role to play. Kitchen’s
replique does not dispose of the really decisive point: Takelot II and
his entire family are attested only in UE and not at all in the Delta,
and this point cannot be dismissed by references to such ephemeral
rulers as Osochor, Psusennes II or Osorkon IV. The period from
Osorkon II to Shoshenq III is the best documented of the TIP and
both kings are demonstrably present in LE. That anyone else reigned
in the same place for a quarter of a century, of whom (and whose
dependents) no trace can be found, must be excluded. The genealog-
ical connections of Takelot II and the sequence of years in the “Chronicle
of Prince Osorkon” are likewise very clear. In addition, the HP Osorkon
B disappears at the very moment (year 39 of Shoshenq III) when an
otherwise unknown Osorkon appears as a new king; this is the only
sovereign of Dyn. 22 who occasionally uses the title of HP in his royal
name,70 and his mother has the same name as the mother of the HP.
It therefore follows that Osorkon B and Osorkon III are the same per-
son, and that also demands that Takelot II must be placed parallel
with Shoshenq III. There is thus a whole set of reasons supporting
Aston’s assumption, and nothing which contradicts it. Therefore, I con-
sider the point to be certain.

3. The LE Sovereigns of Dyn. 22 to Shoshenq V

This would thus mean that Shoshenq III was the immediate successor
of Osorkon II, and there is not the slightest hint of any other hitherto
unknown king between them.71 With Shoshenq III and his successors
until Shoshenq V, we stand on firmer ground chronologically. For

70
The Paleological Association of Japan, Akoris. Report of the Excavations at Akoris
in Middle Egypt 1981–92 (Kyoto, 1995), 301–305; pl. 116; idem, Preliminary Report.
Second Season of the Excavations at the Site of Akoris, Egypt 1982 (Kyoto, 1983),
14–15; pl. 11. No other HPA is known from the period before Osorkon III with this
name, aside from Osorkon B.
71
Cf. also Aston (n. 31), 144.
244 karl jansen-winkeln

Shoshenq III, recorded years include: 3, 5(?), 6, 12, 14, 15, 18(?), 22,
23, 24, 26, 28, 29, 30, 31, 32, 33, 38, and 39.72 An Apis-bull was
buried in his year 28, and a stela commemorating the event73 was
erected for the great chief of the Ma and HP of Memphis74 P #-dj-#st,
who was the grandson (through his mother Ôz-B #stt-prt) and at he same
time the great-grandson (through his father Vater Ôkrjt) of Osorkon II.
The successor of this Apis bull (introduced in the same year year,
1/II/Akhet) in turn died in year 2 (II/Peret) of Pami, after reaching the
age of 26 years.75 Year 2 of Pami thus lies 26 years after year 28 of
Shoshenq III. Were Pami the successor of Shoshenq III, the latter
would have had a reign of no less than 52 years. In fact, however, it
would appear highly probable that another king Shoshenq with the
throne name Ó≈-¢pr-R' should be inserted here,76 who was buried in
the tomb of his predecessor. The most important piece of evidence
here is a donation stela of year 10 from a King Shoshenq Ó≈-¢pr-R',77
mentioning a Great Prince of the Libu named Niumateped, and a man
apparently bearing the same name and title is documented from year
8 of Shoshenq V.78 If, as would appear reasonable, this is the same
person, then a king Shoshenq Ó≈-¢pr-R' should be placed here, who
reigned not long before Shoshenq V, but after Shoshenq III. As Shoshenq
V probably reigned immediately after or following a very short inter-
val after his father Pami, yet 13 years lay between year 39 of Shoshenq
III and year 2 of Pami, for which we have no dates for Shoshenq III,
then everything favours placing a 10–13 year reign of this Shoshenq
Ó≈-¢pr-R' into this period.79 The precise length of his reign is chrono-
logically not very important since the total for the period between year

72
Nile level record no. 22, cf. Beckerath (n. 3), 51; Annals of the priests, no.7, cf.
Kruchten, Annales, pl. 4; 19.
73
Louvre SIM 3749, cf. CSSM, 19–20; pl. VII (no. 21).
74
His son P#j.f-Δ #w-(m-)'wj-B #stt likewise bears the title of HP of Memphis on this
stela.
75
Louvre SIM 3697, CSSM 21–22; pl. VIII (no. 22); cf. also the Stelae Louvre
SIM 3736 and 4205, ibidem, 22–24; pl. VIII–IX (nos. 23/24).
76
A. Dodson, GM 137 (1993), 53–58; TIP 3, XXV–XXVI.
77
Meeks, Donations, 666 (22.1.10).
78
While in year 31 of Shoshenq III, yet another great chief by the name Jnj-Jmn-
n #j.f-nbw is documented, cf. J. Yoyotte, in: Mélanges Maspero I, Orient ancien 4 (Cairo
1961), 143 (§ 31).
79
Numbered variously in the literature: Ib, IIIa, IV or “quartus”; IIIa would be
preferable, as this would eliminate all possible sources of misunderstanding.
the third intermediate period 245

28 of Shoshenq III and year 2 of Pami is certain. For this king Pami,80
the years 2, 4, 5, and 6 are documented; from the structure of the text
on his “annals” in Heliopolis, the presence of the years 3 and 7 can
be deduced.81 Were these “annals” to have covered the entire reign of
Pami,82 this would confirm Kitchen’s assessment of 6 full years for the
reign.83 This assumption of a rather short reign for Pami is supported
by the paucity of monuments he has left, and further by the fact that
the reign of his son was quite long. However, the assumption of a mere
6–7 years is not really certain. His son Shoshenq V followed Pami,
probably as his immediate successor: a stela from the Serapeum from
year 37 of Shoshenq V bears the name of the same (still living) donor
as in year 2 of Pami.84 It is thus improbable that this long period can
be stretched any further. But, it cannot be excluded that another king
(e.g., an older son of Pami) may have ruled between Pami and Shoshenq
V, but then if at all, only very briefly.85
For Shoshenq V, the years 7, 8, 11, 15, 17, 19, 22, 36, 37, and 38
are documented,86 and the interval between year 28 of Shoshenq III
and year 2 of Pami is 26 years long. If 6 full years are assigned to
Pami, and Shoshenq V was his immediate successor, the period from
Shoshenq III to year 38 of Shoshenq V would be 27/28 + 26 + 4/5 +
37 years, and thus 94–96 years depending upon exactly when that Apis
which died under Pami was introduced under Shoshenq III and when
it died under Pami. The interval is probably 95 years.

4. The Successors of Shoshenq V

Shoshenq V is not among the rulers named on Piye’s victory stela. He


was probably already dead at the time. Appearing on a dedicatory stela

80
See J. Yoyotte, RdE 39 (1988), 160–169.
81
S. Bickel, L. Gabolde & P. Tallet, BIFAO 98 (1998), 31–56, esp. 41.
82
Cf. ibidem, 42.
83
TIP, § 83.
84
Louvre SIM 3441 and 3091, cf. CSSM, 24–25; pl. IX (no. 25); 41; pl. XIII (no.
42), cf. TIP, § 84, n. 97.
85
However, the documented lifetime of the Apis-bulls do allow a somewhat longer
period between Pami and Shoshenq V. A bull was buried in year 11 of Shoshenq V:
between this one and the last known predecessor, buried in year 2 of Pami, are only
15–16 years if Shoshenq V immediately followed Pami.
86
PM III2, 787–789.
246 karl jansen-winkeln

of his year 3687 is Tefnakhte, the Great Chief of the Ma, commander
and prince of the Libu, and again on another of year 38, that same
Tefnakhte is called “Great Prince of the entire land”.88 The extension
of this prince’s power, which later obliged Piye to intervene, was thus
already apparent at this time. It thus follows that the interval between
the last years of Shoshenq V and Piye’s campaign was not long. Shoshenq
V is documented in Memphis and in diverse areas of the Delta, includ-
ing Tanis, Bubastis, Buto and Kom Firin. On Piye’s stela, Tefnakhte is
lord of Memphis, Buto and Kom Firin; Iuput II rules in Leontopolis,89
Osorkon IV in Bubastis and the region of Tanis.90 Osorkon IV would
thus be spatially and temporally the successor of Shoshenq V, and the
contemporary documents do not provide any reason to assign him to
another dynasty.91
On the issue of the identity of Shoshenq’s immediate successors, the
temporal and spatial position of Manetho’s Dyn. 23 could play a role.
If Petubaste I and Osorkon III were UE rulers (cf. below, section 5),
then Manetho certainly did not take them into consideration. Thus
they could not be those kings whom he assigned to his Dyn. 23 of
Tanis (consisting of Petubaste, Osorkon, “Psammus” and “Zet”). Priese92
thus suggested that Osorkon IV (rather than III) be assigned to Manetho’s
Dyn. 23, A. Leahy has further elaborated on this idea.93 Thus, Osorkon
IV would be the successor of the ephemeral Petubaste, S˙tp-jb(-n)-R',

87
From Buto (former collection Farouk), cf. Yoyotte (n. 78), 153, § 48; Meeks,
Donations, 670 (22.10.36).
88
From Tell Farain (in the storeroom?), cf. S. Sauneron, BSFE 24 (1957), 51; 53–54,
figs.1–2. The cartouches were left blank, but it unquestionably concerns the year 38
of Shoshenq V, cf. TIP, § 84. The king is omitted on the other stela as well, which
only has the year.
89
A regnal year 21 is documented ( J.-L. Chappaz, Genava 30 [1982], 71–81), but
neither precedessors nor successors are known and thus he cannot be linked to any
dynasty.
90
If R'-nfr is to be understood in this way, cf. Yoyotte (n. 78), 129, n. 2; F. Gomàa,
Die libyschen Fürstentümer des Deltas (Wiesbaden: Beihefte TAVO B, 6, 1974), 132–134.
If not, then it means that, astonishingly, Tanis—one of the most prominent cities of
the TIP—was not mentioned on the stela of Piye. This could only be explained if the
rulers of Tanis declined to submit to the Nubian king.
91
Leahy’s interpretation, that “there is nothing to warrant his inclusion in . . .
Manetho’s Twenty-second Dynasty” (Libya, 189) is thus not entirely convincing. Inci-
dentally, Osorkon IV is only documented with certainty on the stela of Piye: the other
references could also be assigned to Osochor of Dyn. 21, cf. Leahy, Libya, 189; F.
Payraudeau, GM 178 (2000), 75–80.
92
ZÄS 98 (1972), 20, n. 23.
93
Libya, 186ff.
the third intermediate period 247

who is known from Memphis and Tanis (among other places),94 and
otherwise identified with Putubi“ti of the annals of Assurbanipal.95 Aston96
and Beckerath97 have both followed him. At the very least, this would
be a means of integrating Manetho’s Dyn. 23 into the previously known,
although identifying Petubaste S˙tp-jb-R' with the Putubi“ti of the Assyrians
is at least equally plausible. In any case, the result would be that
Manetho’s Dyn. 23 would be nothing but a continuation of Dyn. 22.98
As regnal years have not been preserved from the reign of either
Osorkon IV, nor of his supposed predecessor, Petubaste S˙tp-jb-R', and
the transition from Shoshenq V (—Petubaste)—Osorkon IV is to be
dated to before Piye’s campaign, this possible insertion of a Petubaste
(Manethonis gratia) is not of chronological significance. Osorkon IV is
only dated through the campaign of Piye. Were he the king Shilkanni
who paid tribute to Sargon II (cf. below), then he will still have been
in office around 715/716.

5. UE Kings and Dynasties from Takelot II to Dyn. 25

Along with two Lower Egyptian rulers, the stela of Piye names two
Upper Egyptians: Nimlot D of Hermopolis and Peftjau'awybast of
Herakleopolis. At this time, Thebes itself will have already been under
Nubian control, but before this time we find Harsiese A and Takelot
II (cf. above, section 2) as UE kings who ruled Thebes. Of Kitchen’s
Dyn. 23 (Petubaste I, Iuput I, Shoshenq IV, Osorkon III, Takelot III,
Rudamun und Iuput II, as well as perhaps also Shoshenq VI;99 Residence:

94
Cf. Habachi, ZÄS 93 (1966), 69–74; pls. V–VI; P. Montet, Le lac sacré de Tanis
(Paris, 1966), 63–5; pl. XXX.
95
Cf. TIP, § 357.
96
Aston (n. 31), 140.
97
Chronologie, 99.
98
This could have been another branch of the family, with deeper roots in Tanis
than Bubastis. In any case, according to our present knowledge, Manetho’s king list
of Dyn. 23 is more or less useless for the historical (and chronological) reconstruction:
the last two of his four kings are virtual phantoms, the first two cannot be identified
with certainty, and the note that the first Olympiad took place during the reign of
Petubaste is generally dismissed as a later invention, calculated by the Christian chrono-
graphers who used Manetho, cf. TIP, § 419, n. 134; Redford, King-lists, 311–312;
Beckerath, GM 147 (1995), 9.
99
Shoshenq VI (W#s-nΔr-R', cf. TIP, §§ 67; 110; 146; 336; M.-A. Bonhême, Les noms
royaux de l’Egypte de la Troisième Période Intermédiaire [Cairo: BdE 98, 1987], 140–141) is
not considered in the following, since his very existence is debatable, and there is in
any case no indication of where he should be placed chronologically.
248 karl jansen-winkeln

Leontopolis)100 Iuput II is only documented in LE, Petubaste I mainly


in UE, but a few times in LE; the others are known exclusively from
Upper and Middle Egypt. Osorkon III is the father of Takelot III and
Rudamun, and the later is the father-in-law of Peftjau'awybast. All of
the members of this family are known exclusively from UE sources.101
They are doubtless UE rulers in the tradition of Harsiese A and Takelot
II, and thus are not Manetho’s Dyn. 23. The issue is thus the temporal
relationship of those kings known from UE sources to one another and
to the kings of Dyn. 22. The sources allow for the following synchronisms:
a) In the “Chronicle of Prince Osorkon”, years 22–29 of Shoshenq III
follow year 24 of Takelot II.102 This suggests that Takelot II became
king in UE during the reign of Osorkon II (as Harsiese A before him,
but with his own count of regnal years) and that in his year 4,
Shoshenq III became the successor of Osorkon II (in LE).
b) The year 12 of a king who can only be Shoshenq III corresponds
to the year 5 of Petubaste I, with Harsiese (B) as HPA.103 Petubaste I
thus began his reign in year 8 of Shoshenq III (= year 11 of Takelot II)
and HP Harsiese (B) is linked to this regency. Harsiese (B) is subse-
quently documented in the years 18 and 19 of Petubaste (= years 25
and 26 of Shoshenq III),104 and previously in year 6 of Shoshenq III,105
and already under Osorkon II.106 A Takelot (E) was HP at the latest
from year 23 of Petubaste,107 who then assumes Harsiese’s post.
It is therefore highly probable that the “rebellion” of year 11 of
Takelot II mentioned in the “Chronicle of Prince Osorkon” was the
accession to the throne of Petubastis,108 which was understood as a
usurpation, as he thus became a kind of rival king to Takelot II. The

100
TIP, §§ 102; 297; 519; p. 588.
101
And the same applies, as described above, to Takelot II, the father of Osorkon III.
102
Reliefs, III, pl. 22, Z.7–22.
103
Nile level record no. 24, cf. Beckerath (n. 3), 51. On the identification of the
unnamed king as Shoshenq III, cf. TIP, §§ 106–107. On purely technical grounds, fol-
lowing the chronology proposed by Aston, Takelot II could also be considered, but
historically, he is out of the question, as an opponent of Petubaste and Harsiese B.
104
Nile level record nos. 28 and 27, cf. Beckerath (n. 3), 52.
105
Nile level record no. 23, cf. Beckerath (n. 3), 51.
106
On the statue, Cairo CG 42225, for this, cf. Jansen-Winkeln (n. 19), 135–6.
107
Nile level record no. 29, cf. Beckerath (n. 3), 52.
108
Chronicle of Prince Osorkon, A, 22ff., cf. Reliefs, III, pl. 16; 18.
the third intermediate period 249

HP Osorkon B is documented for years 11 and 12 in Thebes,109 whereas


Petubaste I and Harsiese B are not, but another revolt erupts in year
15 of Takelot,110 and exactly in this year, Petubaste and Harsiese B
reappear in the Theban sources.111 In year 24 and 25 of Takelot (= 14
and 15 of Petubaste I), Osorkon B donated offerings in Thebes, and
at this time Petubaste and Harsiese are not documented here. Evidently,
there were two parties in this civil war: Osorkon B and his father
Takelot II on the one hand, and Petubaste I and the HP Harsiese B,
later Takelot E, on the other. 112 This Takelot is also mentioned in the
year 6 of a king Shoshenq Wsr-m#'t-R' Mrjj-Jmn,113 who cannot be
Shoshenq III,114 but must rather be an another (certainly UE) King
Wsr-m#'t-R' Shoshenq (IV).115
c) The highest documented regnal year for Takelot II is year 25,116
and as in the donation lists of the “Chronicle of Prince Osorkon”, year
24 of Takelot II is followed by year 22 of Shoshenq III, it was appar-
ently his last.117 Despite publicly announced claims,118 the successor of
Takelot II was not his son Osorkon B: the latter is still General and
HP in year 39 of Shoshenq III.119 It was presumably Iuput I who was

109
The son of Takelot II who commissioned the Chronicle of Prince Osorkon.
110
Chronicle of Prince Osorkon, B,7, s. Reliefs, III, pl. 21.
111
Nile level record no. 24, cf. Beckerath (n. 3), 51.
112
The role played by Shoshenq III in these events is not evident.
113
Nile level record no. 25, cf. Beckerath (n. 3), 52.
114
Cf. Aston (n. 31), 151: Shoshenq III does not use the epithet Mrjj-Jmn in his
throne-name, and there is already a Nile level record (no. 23) for his year 6, naming
HPA Harsiese.
115
The latest documented date for him is year 6, cf. Beckerath (n. 3), 52. (Nile level
record no. 25); Jacquet-Gordon, Graffiti, 40–41 (no. 100).
116
Donation stela Cairo JE 36159, cf. ASAE 4 (1903), 183.
117
The years 24 and 26 (without the king’s name, cf. Capart, BMRAH, 3. série,
13, 1941, 26), are recorded on the mummy wrapping Brussels E.7047b/c of a mrjj-nΔr
named Ns-p#-nΔr-n-R' var. Ns-nΔr-p#-R'. As the father of this man is Ns-r-Jmn (Cartonage
Berlin 30, cf. ÄIB II, 381–382), Kitchen (TIP § 86, n.115; 294) and Bierbrier (LNK,
71) have both identified him as Ns-p#-R', son of Ns-r-Jmn (I), the donor of the statue
Cairo CG 42221, whose family tree (TIP, § 166) suggests that he belongs roughly in
the period of Takelot II, and they have thus deduced a year 26 of Takelot II. Since,
however, both the name (Ns-p#-nΔr-n-p#-R' vs. Ns-p#-R', cf. M. Thirion, RdE 46 [1995],
181–182) and the title (mrjj-nΔr vs. ˙m-nΔr n Jmn-R' njswt nΔrw jmj-#bd.f n pr Jmn ˙r z# tpj)
of these individuals differ, this identification (and thus a possible source for a year 26
of Takelot II) cannot be maintained.
118
Chronicle of Prince Osorkon, A, 53, cf. Reliefs, III, pl. 16; 17; CPO, §§ 101–102.
119
Nile level record Karnak no. 22, cf. Beckerath (n. 3), 51; Annals of the Priests
at Karnak, no. 7, cf. Legrain (n. 4), 55–56; Kruchten, Annales, pl. 4; 19.
250 karl jansen-winkeln

the successor, for year 16 of Petubaste I corresponds to year 2 of a


king Iuput (I),120 and thus his year 1 (corresponding to year 15 of
Petubaste and year 22 of Shoshenq III) follows immediately on the last
full year of Takelot II. As these dates match, it is more probable that
Iuput I was the successor of Takelot, and not a “short-lived coregent”
of Petubaste.121 On the other hand, Shoshenq IV may have been the
successor of Petubaste as Petubaste appears initially together more fre-
quently with the HP Harsiese B, and then with Takelot E, who him-
self is then named likewise together with Shoshenq IV (cf. above). These
synchronisms produce the relations presented in Fig. II. 10.1.122
King Petubaste is documented in Thebes with the throne-name Wsr-
m #'t-R' Stp-n-Jmn and with the unique epithet z# #st. 123 A king with the
same prenomen and throne-name, but with the epithet z# B#stt is known
from a donation stela from Memphis (year 6),124 from Herakleopolis or
the eastern Delta (?),125 and Bubastis (year 23),126 as well as on a statue
of uncertain provenance.127 This has been interpreted as being two
different kings with the same prenomen and throne-name, 128 but this
is hardly plausible.129 The idea that both the UE and LE Petubaste
would have the same highest known date of 23 years appears rather
suspicious. In addition, one of the Theban retainers of Petubaste, the

120
Nile level record no. 26, cf. Beckerath (n. 3), 52.
121
In TIP, § 448; cf. also Aston (n. 31), 151. Against this, one could argue that all
of the other synchronisms in the Nile level records give only the links between the
rulers of one “party” to the LE king (Shoshenq III). If Iuput I was the successor of
Takelot II, he should have belonged to the foes of Petubaste. However, from the
Chronicle of Prince Osorkon (B,7ff.) it is evident that at this time, there was a tem-
porary unity among the various rivals in the civil war (cf. Jansen-Winkelen [n. 19],
140–141 on this).
122
Abbreviations: NLR, Nile Level Records, cf. Beckerath (n. 3), 43–55; OC =
Caminos, Chronicle; OC, A = Reliefs, III, pl. 16–19; B = ibidem, pl. 21; C = ibidem,
pl. 22; AP = Annals of the Priests at Karnak, cf. Legrain (n. 4), 51–63; Kruchten,
Annales; Stela 22.8.26 = Meeks, Donations, 669 [22.8.26]. Years in brackets are postulated.
123
Nile level record no. 24; Beckerath (n. 3), 51.
124
Cairo JE 45530, cf. Schulman, JARCE 5 (1966), 33–41; pl. 13.
125
Copenhagen Ny Carlsberg AEIN 917, cf. O. Koefoed-Petersen, Recueil des inscrip-
tions hiéroglyphiques, pl. 5; J. Yoyotte, BIFAO 58 (1959), 97 (2); Meeks, Donations, 671
(23.1.00).
126
Florence 7207, cf. R. A. Caminos, Centaurus 14 (1969), 42–46; pl. 1–2.
127
Gulbenkian Museum Lisbon, cf. M. Hill, Royal Bronze Statuary from Ancient Egypt
(Leiden/Boston, 2004), 155–156; pl. 18 (12).
128
E.g., A. S. Schulman, JARCE 5 (1966), 37–39; Beckerath, GM 147 (1995), 9–13.
129
Cf. B. Muhs, JEA 84 (1998), 223; J.v. Beckerath in: Es werde niedergelegt als Schriftstück:
Festschrift für Hartwig Altenmüller zum 65. Geburtstag (Hamburg 2003), 31–36.
the third intermediate period 251

Fig. II. 10.1

Dyn. 22 Dyn. 23 (UE) Rival Kings High Priests Sources

Osorkon II Harsiese B CG 42225,a


Takelot II
Shoshenq III
year (1) (4)

6 (9) Harsiese B NLR, no. 23


Petubaste I
(8) 11 (1) Osorkon B OC, A 18–53: 1st
“rebellion”
(9) 12 (2) Osorkon B OC, B 1–6
12 15 5 Osorkon B/ NLR, no. 24;
Harsiese B OC, B 7:
2nd “rebellion”
(14) (17) 7 AP, no. 1, 1.1
(15) (18) 8 Harsiese B AP, no. 1, 1.2;
no. 2, 11.1/3
(21) 24 (14) Osorkon B OC, C 7 (cf. B
7–C1)
22 25 (15) Osorkon B OC, C 12 (year
Iuput I 22); ASAE 4,
(1) 183
23 2 16 Osorkon B NLR, no. 26
(y. 2/16); OC,
C 12
24 (3) (17) Osorkon B OC, C 13–16
25 (4) 18 Osorkon B/ NLR no. 28;
Harsiese B OC, C 1–2; 17:
3rd “rebellion”
26 (5) 19 Harsiese B NLR, no. 27;
Stela 22.8.26
28 (7) (21) Osorkon B OC, C 12; 17
29 (8) (22) Osorkon B OC, C 22
(30) (9) (23) Takelot E NLR, no. 29
Shoshenq IV Jacquet-Gordon,
Graffiti, 85
(33) 12 (?)
(?) (?) 6 Takelot E NLR, no. 25
39 (18) Osorkon B NLR, no. 22; PA,
no. 7, 11.1–3
(39?) Osorkon III
(1)
252 karl jansen-winkeln

prophet of Amun and royal scribe Ór (IX), is unexpectedly documented


at Memphis,130 and perhaps also in Tell el-Balamun.131 There can only
be one single king Petubaste, who used the epithet z# B#stt in LE. He
may have been a rival king who attempted to re-establish a unified
kingship over the entire land, a situation which had ceased to exist at
the very latest by Takelot II. Regardless, the “dynasty” of Petubaste is
not chronologically relevant. It is not known when Shoshenq IV suc-
ceeded Petubaste, the length of whose reign is likewise unknown. This
dynasty presumably ended in year 39 of Shoshenq III, at the latest.132
By contrast, the dynasty of Takelot II can be followed: a year 12 is
documented for his presumed successor, Iuput I (cf. above),133 and his
successor can only have been Osorkon B/III. He appears for the last
time in year 39 of Shoshenq III, as High Priest. As he had this office
since year 11 of Takelot II (= year 8 of Shoshenq III), and then reigned
for 28 years as king, he must have become king in or shortly after
year 39 of Shoshenq III. If he, as is probable, followed immediately
after Iuput I, the latter must have reigned for at least 17 years.
For Osorkon III, the regnal years 1(?), 3, 5, 6, 14(?), 15, x + 6, 23(?)
and 28 are documented, with his regnal year 28 being equal to year
5 of his son Takelot III,134 the only completely unambiguous coregency
in the TIP. 135 For Osorkon III, 23 full years can be accounted for,
and for Takelot III, years 5, 6, and 7 are clearly attested.136 If Osorkon

130
K. Jansen-Winkeln, SAK 27 (1999), 123–139; pls. 1–4.
131
A. J. Spencer, Excavations at Tell el-Balamun 1995–1998 (London 1999), 13–15;
83–86; 90–91.
132
In this year, the HPA Osorkon B claimed that he and his brother defeated all
of those with whom they fought, cf. Legrain (n. 4), 55–56; Kruchten, Annales, pl. 4;
19. It would still be conceivable that the later “dynasty” of Hermopolis (Nimlot D and
Thotemhat) continued that of Petubaste, as Hermopolis could have been a major cen-
ter in Petubaste’s “rebellion”, cf. Jansen-Winkeln (n. 19), 142. However, there does not
appear to be any trace of a temporal link between these regents.
133
Aside from the graffito of year 9 of Iuput, the same priest also left graffiti from
years 9 and 12 (without the name of a king), cf. Jacquet-Gordon, Graffiti, 84–85 (nos.
244–245).
134
Nile level record no. 13; Beckerath (n. 3), 50. For the uncertain numbers, cf.
Jacquet-Gordon, Graffiti, 41 (nos. 101: year 1); 68–9 (no. 190: year 14); 69 (no. 191:
year 23).
135
This conregency is also confirmed by the statue Cairo CG 42211, dated by the
cartouches of njswt-bjt Mrjj-Jmn z# #st Ôkrt and z # R' Mrjj-Jmn z# #st Wsrkn, cf. Jansen-
Winkeln (n. 63), 470.
136
Daressy (n. 40), 51–52.
the third intermediate period 253

III ascended the throne in year 39 of Shoshenq III, then the tempo-
ral relationship between Dyn. 22 and the UE rulers of the line of
Takelot II will have been that presented in Fig. III. 10.2. If he became
king at a later date (year 40 or 41), then the dynasty must be pressed
down a bit, but this can hardly be a matter of more than a few years.137
Of his successors, only his well documented son Takelot III spatially
and temporally anchored in Thebes. The length of the reign remains
unclear: he is occasionally assigned a reign of more than 6 full years,
and not least because several of his children were still alive shortly
before 700 as the family trees of their descendents and the style of
their tombs reveal.138 F. Payraudeau has recently attempted to link a
year 14 of a Takelot z# #st in P. Berlin 3048 to Takelot III rather than
Takelot II.139 This is possible but by no means certain.140 However long
he reigned, the problem of the “generation shift” does not disappear:
perhaps Takelot III and/or Osorkon III only became fathers late in
their lives.
At the very latest, after the reign of Takelot III the situation in UE
becomes quite obscure. At the time of the Piye campaign, the Nubians
ruled the Thebaid, while other UE kings were in Hermopolis and
Herakleopolis. The later successors of Osorkon III were thus driven
out of Thebes. There is no clear indication of when this happened,
but at the very latest the inauguration of Amenirdis I as the adoptive
daughter and heir of the Divine Wife Shepenupet I marks that Thebes
was definitely governed by the Nubians. According to Kitchen, it was
Piye, the brother of Amenirdis, who ordered the adoption,141 but Morkot

137
On the condition that the HPA Osorkon B and Osorkon III were in fact one
and the same person, cf. above, section 3.
138
Cf. Aston & Taylor, in: Leahy, Libya, 138–143.
139
GM 198 (2004), 82–85. Palaeographically it appears probable that on the same
papyrus (debt note) one should read year 23 (rather than 13), cf. S. Vleeming, OMRO
61 (1980), 3, n. 14; B. Menu, CRIPEL 1 (1973), 89–90; K. Donker van Heel, in:
K. Ryholt, ed., Acts of the Seventh International Conference of Demotic Studies (Copenhagen:
CNI Publications 27, 2002), 142.
140
The palaeography can hardly aid with the date as there are already very cur-
sive texts in Dyn. 21, cf. M. Malinine, in: Textes et langages de l’Égypte pharaonique I (Cairo,
1973), 31. The identification of some ancestors of the scribe with individuals from
dated contexts is unreliable in the extreme, or would even favour a date under Takelot
II (as with the vizier Órj ). Nor can an argument be made using the Overseers of the
Treasury, as four of them appear in this one Papyrus (cf. Donker van Heel [n. 139],
143).
141
TIP, § 122.
254 karl jansen-winkeln

Fig. III. 10.2 (Abbreviations: see Fig. III. 10.1)

Shoshenq III Takelot II

Year 1 Year 4

21 24
22 25
Iuput I
1
33 12
39 (18)
Osorkon III
1
Shoshenq IIIa 2
1
13 14
Pami 15
1
6 20
7
Shoshenq V
1 21
3 23
4 24 = 1 Takelot III
8 28 = 5
10 7 (sole rule)

has convincingly shown that it was probably her father Kashta who
installed her.142 This would mean that the successors of Osorkon III
were swiftly removed from Thebes. If Piye’s campaign (in his year 20)
took place within five years of the death of Shoshenq V (see below,
section 7), then his reign must have begun at the latest in year 25 of
Shoshenq V, and probably somewhat earlier. The inauguration of
Amenirdis could thus have taken place in years 20–24 of Shoshenq V.
As year 28 was probably the final year of Osorkon III, and corre-
sponds to year 8 of Shoshenq V, at the earliest (cf. above), his suc-
cessors have a mere 10–15 year in Thebes, before they had to withdraw
to the North. All of their dated sources from Thebes must be assigned
to this short period.

142
In: S. Wenig, “Studien zum antiken Sudan”, Meroitica 15 (1999), 194–196.
the third intermediate period 255

Aside from Takelot III, the following UE kings are known from the
period after Osorkon III.
• Rudamun, the brother of Takelot;143 no known regnal years.
• Peftjau'awybast, the son-in-law of Rudamun, king of Herakleopolis
at the time of Piye’s campaign;144 regnal year 10 is documented.145
• G. Broekman has recently shown that it is highly probable that
there was an UE king Shoshenq (“VII”) with the epithet z # #st and
the throne-name Ó≈-¢pr-R' Stp.n-R',146 who was recognized as king
in Thebes in his regnal year 5,147 and who is to be inserted after
Shoshenq III and thus also after Takelot III.
• Another candidate would be the king Iny who is documented sev-
eral times in Thebes (including a regnal year 5) and perhaps also
in Abydos.148
• In addition, there is a dynasty residing in Hermopolis, whose most
prominent member, Nimlot D, is chronologically anchored in the
stela of Piye. His predecessor or (more probably) successor could
have been Thotemhat,149 and a later successor may have been
Padinemti(?).150

143
Cf. O. Perdu, RdE 53 (2002), 157–178, for this person.
144
Even if his power was restricted to the Herakleopolis region, during this period
when the Nubians controlled the Thebaid and there appeared yet another UE king-
dom, he could still have been the heir of an UE dynasty with a much larger realm.
In Herakleopolis and the surrounding area at least, the dynasty of Takelot II is well
documented, e.g, the HPA Osorkon B (cf. Caminos, Chronicle, §§ 28–30) and the later
Takelot III (ASAE 37 [1937), 16–24). Payraudeau’s ([n. 139], 79–81) attempt to dis-
tinguish the general of Herakleopolis from the son of Osorkon III, who bears the same
name, cannot be accepted in view of the fact that both are HPA and had a mother
with the same rather uncommon name.
145
Donation stelae Cairo JE 45948 and 11/9/21/14, cf. G. Daressy, ASAE 17 (1917),
43–45; ASAE 21 (1921), 138–139.
146
Broekman (n. 3), 163–78, esp. 176–177.
147
The only certain document is the Nile level record no. 3; Beckerath (n. 3), 49,
hitherto assigned to Shoshenq I. However, one cannot exclude a possible reference to
Shoshenq IIIa; his predecessor Shoshenq III is in fact mentioned in his last (or next
to last) year in the Nile level records, cf. Broekman (n. 3), 176. It is conceivable that
there was still resistance after Osorkon III ascended the throne, and that one of his
enemies was able to establish himself briefly in Thebes, and dated according to the
LE king. It is highly probable that the Nile level record no. 45 does not belong to
Shoshenq VII (cf. Broekman [n. 3], 177); there does not remain any time for a year
17/19/25 of a sovereign in Thebes before the Nubians after Osorkon III (cf. above).
148
J. Yoyotte, CRIPEL 11 (1989), 113–131; pl. 14.
149
For him, cf. H. Wild, RdE 24 (1972), 209–215; P. Spencer & A. Spencer, JEA
72 (1986), 198–199; pl. 21; Bierbrier, LNK, 84.
150
For him most recently, cf. A. Leahy, JEA 85 (1999), 230–232.
256 karl jansen-winkeln

As the brother of Takelot III, Rudamun was most probably his suc-
cessor, as is generally assumed. It is, however, remarkable, that he is
better documented in Hermopolis than in Thebes.151 It is thus also con-
ceivable that Rudamun became king in Hermopolis after the death of
his father, alongside his brother Takelot in Thebes (and Herakleopolis?).
The Libyan period does reveal a tendency to multiply both rulers and
principalities. The line of Takelot would then have been reduced to
Herakleopolis after the Nubian intervention. Shoshenq “VII” is only
documented in Thebes, with a year 5. He too can belong only to the
dynasty of Osorkon III (as a son of Takelot III?). If Rudamun was the
successor of Takelot III (in Thebes), Shoshenq VII would most prob-
ably have been a successor of Rudamun, although a sequence of
Takelot—Shoshenq—Rudamun cannot be excluded.152 If Rudamun was
a local ruler in Hermopolis, then Shoshenq VII would have followed
immediately after Takelot. The year 5 of king Iny should be situated
roughly two generations after year 4 of a king Shoshenq;153 and this
may have been Shoshenq III, IV or VII. Were it Shoshenq III, the
reign of Iny would fall under the reign of Osorkon III, and that is
improbable. Otherwise, he should be assigned either to the period after
Takelot III (successor of Shoshenq VII?), or indeed placed in Dyn. 25.
Unusually his name was effaced, and thus he might have been a pre-
tender (during the reign of Osorkon III or Dyn. 25),154 in which case
the reign would be of no chronological relevance.
In any case, the rulers of the house of Osorkon III were swiftly
evicted from Thebes. The Peftjau'awybast of Herakleopolis named on
the stela of Piye is the last of this line. The “dynasty” of Hermopolis
(whether from Rudamun or by another line) may have been founded
by descendents of Osorkon III, but it could equally easily have been
the late revival of the rival dynasty of Petubaste.

151
Cf. Perdu (n. 143), 169–170.
152
Cf. G. Broekman, “The Chronological Position of King Shoshenq Mentioned in
Nile Level Record No. 3 on the Quay Wall of the Great Temple of Amun at Karnak”,
SAK 33 (2005), 75–89.
153
Graffito no. 11 from the roof of the temple of Khonsu, cf. H. Jacquet-Gordon,
in: Hommages à la mémoire de Serge Sauneron I (Cairo, 1979), 174–183; pl. 27–28; Yoyotte
(n. 148), 115.
154
Cf. Yoyotte (n. 148), 131. A “reign” of at least 4 years for a rival king could be
possible, but it would be quite unusual, and particularly so in Dyn. 25.
the third intermediate period 257

The chronology of the UE kings after Osorkon III thus remains quite
uncertain: there are only a few (low) dates, and it is unclear which
kings reigned in parallel and which in succession. Of the kings attested
on the stela of Piye, Nimlot D cannot be linked to either a predeces-
sor or a successor,155 and Peftjau'awybast can only be identified genealog-
ically as the son-in-law of Rudamun.
The familiar “graffito” from Wadi Gasus could offer a chronologi-
cal connection with the following Dyn. 25.156 To the right is the car-
touche of the Divine Adoratrice Amenirdis (I), above which is regnal
year 12, to the left the cartouche of the Divine Wife Shepenupet (I),
above which is regnal year 19, both names have the epithet “living”
('n¢.tj ). It is today agreed that the year 12 of Amenirdis can only be
related to Piye,157 and thus the year 19 should be assigned to one of
the UE rulers recognized by Piye. It is thus immediately clear that
Takelot III cannot possibly be the king designated by the year 19 of
the graffito158 since his year 19 must have corresponded to year 22(–25,
or so) of Shoshenq V (cf. above, Fig. III. 10. 2), and thus clearly before
year 12 of Piye in whose year 20 the campaign to the North took
place, Shoshenq V, who reigned at least 37 years, no longer in office.
Rudamun would only be a candidate for the year 19 if he was not
the predecessor of Shoshenq VII, as they were certainly not dating in
Thebes according to the dynasty of Osorkon III 19 + 5 years after
Takelot III (cf. above). The year 19 can also be linked to Shoshenq
VII or Peftjau'awybast. In any case, it should certainly be someone rel-
atively close to Piye’s house: an ally. Nimlot D of Hermopolis would
thus also be a candidate. He appears in an ambivalent fashion on the
stela of Piye: on the one hand the Nubian king expresses his particu-
lar irritation over the alliance with Tefnakhte of Sais, while on the
other, he is given preferential treatment.159 This can be most easily
explained by the fact that he was an ally of the Nubians who then

155
No regnal year is preserved, and the same is true of several other members of
this dynasty: Rudamun, Thotemhat, and Padinemti; for the latter two, not even the
exact position in the sequence of the “dynasty” is known.
156
L.-A. Christophe, BIE 35 (1952/53; 1954), 141–152.
157
TIP, §§ 143–145.
158
Thus Payraudeau (n. 139), 85–86.
159
He is the only prince admitted into the Palace to Piye, cf. the great stela of vic-
tory, ll. 148–53 (Urk. III, 54) and is the only one pictured standing, but actually like
a woman, with a sistrum in his hand.
258 karl jansen-winkeln

switched sides.160 As an ally before these events, he would have been


a suitable candidate for the double dating, and in fact he does appear
a second time with the Divine Wives Shepenupet I and Amenirdis I.161
Nimlot D thus appears to me to be a particularly suitable candidate
for the year 19 in this graffito. Chronologically, however, this does not
aid at all: in temporal terms, neither Nimlot D nor the other possible
candidates can be pinned down to sufficiently narrows slots in time so
as to allow a direct link with between the house of Osorkon III and
Dyn. 25.
A somewhat more precise knot making a temporal link between the
Libyan and Nubian periods is possible only via Dyns. 22 and 24, and
possible fixpoints can only be gained for Dyn. 25.

6. The Chronological Framework for Dyn. 25

The beginning of the reign of Taharqa lies in year 690 BC, and this
is not disputed.162 For a long time, his predecessor Shebitku (highest
date is year 3163 was assigned a reign of 8–12 years, and at the most
13 regnal years.164 However, the inscription of Sargon II at Tang-i Var
reveals that Shebitku was already (at the latest) king in 706,165 and thus
reigned for at least 16 years. As his predecessor Shabaka ruled for at
least 14 full years (cf. below), the beginning of his reign would be at
the latest in 720 BC. Since one had once assumed that there were
good reasons for believing that the Nubian rule in Egypt could not
have begun before 716 or indeed 712 (cf. below), it was suggested a
number of times that Shebitku was only (co)regent in Nubia while his
senior partner, Shabaka (with dates according to his reign) ruled in
Egypt.166 This is historically quite improbable, aside from the fact that
there has never been the slightest hint at any form of coregency of the

160
Opposing D. Kessler, SAK 9 (1981), 238.
161
On the fragment of a vessel in the Museo Barracco in Rome, cf. L. Bongrani
Fanfoni, OrAnt 26 (1987), 65–71; pls. 2–3.
162
Cf. TIP, §§ 130–131; Beckerath, Chronologie, 91.
163
Nile level record no. 33, cf. most recently J.v. Beckerath, GM 136 (1993), 7–9.
164
TIP, §§ 126; 468; Beckerath, Chronologie, 92.
165
G. Frame, Orientalia 68 (1999), 31–57; cf. D. Kahn, Orientalia 70 (2001), 1–3. Cf.
also N. Na’aman, N.A.B.U. 1999, Nr. 3, 63 (65).
166
So Redford, Orientalia 68 (1999), 58–60; Beckerath, SAK 29 (2001), 3–6; Kitchen,
in: Bietak, ed., SCIEM Haindorf 1996/98, 50–51.
the third intermediate period 259

Nubian kings of Dyn. 25. Had Shabaka been ruler of Egypt in the
year 707/706 and Shebitku his “viceroy” in Nubia, one would definitely
expect that the opening of diplomatic relations with Assur as well as
the capture and extradition of Yamani would have been part of Shabaka’s
responsibility. Sargon can also be expected to have named the regent
of Egypt and senior king, rather than the distant viceroy Shebitku. If,
on the other hand, Shebitku was already Shabaka’s successor in 707/706,
the reports of the Yamani affair become clearer and make more sense.
It had hitherto been assumed that the Nubian king (Shabaka) handed
over Yamani more of less immediately after his flight to Egypt.167 Now
it appears to be certain that Yamani was only turned over to the
Assyrians a couple of years later.168 It then becomes much more prob-
able that Shabaka awarded him asylum, but that Shebitku did not feel
bound by his predecessor’s word and that he desired to make a ges-
ture of good will towards the Assyrians at the start of his reign, and
that he extradited Yamani.169 This interpretation also matches with the
peculiar insertion into Sargon’s large “ceremonial inscription” in
Khorsabad where the king of Nubia is described as residing in a very
distant, inaccessible land.170 The formulation of his Nile level record
(no. 33) also supports the idea that Shebitku only came to Egypt in
his year 3.171
Shabaka must, therefore, have already been dead in 707/706. The
“international” reasons which have hitherto been used to justify plac-
ing his reign in Egypt after 716 or even 712 cannot therefore be cor-
rect, and in fact they are wrong. The events of the years (around) 725
(when Hosea of Israel addresses an appeal for aid to a “So, King of
Egypt”)172 and around 720 (when an unknown Egyptian sovereign sends
a general named Re’e leading an army into Palestine to support a
revolt against the Assyrians only to be defeated at Raphia),173 are not

167
Cf. e.g., TIP, § 341.
168
Cf. A. Fuchs, “Die Annalen des Jahres 711 v. Chr,” State Archives of Assyria Studies
VIII, 127–31. The actual text of the Assyrian reports recording that Yamani lived in
Egypt “like a thief ” likewise indicate a prolonged stay.
169
Had Shabaka himself extradicted Yamani after having granted him asylum for
years, that would have been an inconvertible sign of weakness.
170
A. Fuchs, Die Inschriften Sargons II. aus Khorsabad (Göttingen 1994), 221–222; 348–349;
Frame (n. 165), 53.
171
Cf. Beckerath (n. 163), 7–9.
172
2 Kgs. 17,4.
173
Annals of Sargon II from Khorsabad, ll. 53–5, cf. Fuchs, Inschriften, 90; 315; cf.
also the threshold inscriptions from Khorsabad, ll. 38–41, ibidem, 262; 360; and a
clay cylinder from Khorsabad, l. 19, ibidem, 34; 290.
260 karl jansen-winkeln

relevant for the dating of the Nubian rule in Egypt.174 In the year 716,
Sargon II extends his sphere of control further south, and receives trib-
ute (or the like) “from Pharaoh, the king of the Land of Egypt”.175
Another source is more precise, recording that Shilkanni, the king of
Egypt, sent 12 large steeds as a greeting present.176 Shilkanni could be
Osorkon IV,177 but he is in any case a LE and not a Nubian king. In
the case of the Yamani-affair (711–706)178 the city of Ashdod asks
“Pharaoh, the king of Egypt, a prince, who could not rescue it” for
an alliance, apparently in vain. As the Assyrians attack, Yamani flees
“to the border of Egypt in the area of Nubia”,179 where he lives “(secretly)
like a thief ”, until extradited by Shebitku. Neither the events of 716
nor 711 can possibly serve as a terminus post quem for the beginning of
Nubian rule. The pharaoh whose alliance was requested in 712/711
can only be either Shabaka or a Delta Prince, but even in the latter
case, it would not imply that Shabaka had not yet been recognized in
Memphis. Shilkanni apparently had good reasons for trying to reassure
the Assyrians; but this does not solve the issue of who had the upper
hand in Egypt. In the account of his third campaign, Sennacherib
reports that at the battle at Eltekeh (701), Hezekiah made appeals to
“the kings of Egypt” and the troops of the King of Nubia (Shebitku).
On this occasion, the Assyrian king captured “the charioteers and the
sons of the kings of Egypt” and “the charioteers of the king of Nubia”.180
From the Assyrian point of view, the enemies are perceived primarily
as a kind of coalition, and this may have correspond to the facts, for

174
There is one hint that Nubian soldiers took part in the battle at Raphia (cf.
Kahn, Orientalia 70, 11–12), but these could have been mercenaries.
175
Annals of Sargons II from Khorsabad, ll. 123–4, cf. Fuchs, Inschriften 110; 320.
176
Assur Prisma, cf. Fuchs, Annalen, 28–29; 57.
177
This is, however, by no means certain, a name such as ”rkn or the like would
be more reasonable, cf. J. Yoyotte, Kêmi 21 (1971), 51–52.
178
Mentioned in the following inscriptions: Annals of Sargon II from Khorsabad,
ll. 241–254, cf. Fuchs, Inschriften, 132–5; 326; small ceremonial inscription from Khorsabad,
ll. 11–14, cf. Fuchs, Inschriften, 76; 308; Orientalia 68 (1999), 52–53; large ceremonial
inscription from Khorsabad, ll. 90–112, cf. Fuchs, Inschriften, 219–222; 348–9; Orientalia
68, 53; Niniveh Prisma VIIB, cf. Fuchs, Annalen, 44 –6; 73–4; inscription from
Tang-i Var, cf. Frame (n. 165), 31–57.
179
This frequently discussed phrase (cf. most recently L. Depuydt, JEA 79 [1993],
272, n. 24; Fuchs, Inschriften, 220; 348; 452; Frame (n. 165), 52, n. 24) seems to mean
something like “to that part of Egypt, which was under the direct control of the
Nubians”.
180
Cf. E. Frahm, “Einleitung in die Sanherib-Inschriften”, AfO Beiheft 26 (1997), 54; 59.
On the 3rd campaign of Sennacheribs as a whole, cf. ibidem 10–11.
the third intermediate period 261

even under Asarhaddon and Assurbanipal the princes of the Delta are
represented as acting independently on the international stage. It is thus
inadmissible to use evidence of such activities as a base for defining
the beginning of Dyn. 25.
There are no obstacles to ending the reign of Shabaka in 706 at the
latest; on the contrary, everything suggests that Shebitku ruled alone
from 707/706 to 690. Year 2, 3, 4, 6, 10, 12, 13, 14, 15181 are doc-
umented for Shabaka, and he is generally assigned 14 full years. An
indirect confirmation of this can be found in Manetho, if one allows
for a slip,182 by assigning the 14 years Africanus gives to Shebitku to
Shabaka. However, in view of the unreliability of the Manetho tradi-
tion concerning Dyn. 25183 this does not mean much. A possibility for
calculating can also be deduced from two stelae from Kawa where
Taharqa states that he was 20 years old when Shebitku called upon
him to go from Nubia to Egypt.184 As this will doubtless have taken
place in the course of the preparations for the campaign which led to
the battle at Eltekeh where Taharqa saw action,185 he must have been
born ca. 722/721. If he was a son of Piye’s (as is generally assumed),186
the latter must have lived until at least 723 and perhaps a bit longer.187
However, it is by no means certain that Taharqa was really the bio-
logical brother of Shepenupet II and thus the son of Piye.188 Nevertheless,
a reign of 14–15 years for Shabaka remains highly probable. Favouring
this is also the fact that there is a relatively complete coverage of dates
from the second decade of his reign (10, 12, 13, 14, 15), and a large
hole would be improbable. He must thus have come to the throne at
the latest in 720, or more probably 721 or 722. His second year would
thus be ca. 720 (721–719), and also year 6 of Bocchoris.189 It is calculating

181
Block statue BM 24429, cf. Leclant, Enquêtes, 15–27; pl. 5–6.
182
Thus Beckerath, Chronologie, 92; TIP, § 421.
183
Cf. TIP, § 468.
184
Stela IV, ll. 7–9; V, ll. 16–7, cf. Macadam, Kawa I, 15; 28; pl. 7–10.
185
TIP, § 127–9; 133.
186
According to the stela of Nitokris, ll. 3–4 (cf. JEA 50 [1964], 74; pl. VIII) the
Divine Wife Shepenupet II, a daughter of Piye, was his sister, cf. TIP, §§ 120–121.
187
Cf. D. Kahn, Orientalia 70 (2001), 7.
188
Cf. A. Leahy, GM 83 (1984), 43–45.
189
An inscription from year 2 of Shabaka was found in the Serapeum, and this—
despite some inconsistencies in the secondary literature—should be related to the same
Apis burial as the stelae from the beginning of year 6 of Bocchoris, cf. TIP, § 114;
J. Vercoutter, Kush 8 (1960), 62–67; PM III2, 789. That year 2 of Shabaka was either
the same as, or close to, year 6 of Bocchoris is clear from the sources: Manetho assigns
262 karl jansen-winkeln

the regnal years of Piye, the predecessor of Shabaka, which is uncer-


tain, and thus likewise the link to the major campaign of year 20.190
In Egypt, the years 20(?), 21, 22, and 24 are documented,191 but he is
generally assigned a reign of 31 years as a few years must be inserted
for Tefnakhte before the reign of Bocchoris his successor. This rests
on the correct assumption that the various rulers of Egypt listed on
the stela of Piye are actually identified by their rightful titles—includ-
ing the foes of the Nubian king. If Tefnakhte is not designated a king
there,192 he will thus have become such only after the campaign of
Piye. As a year 8 is recorded for Tefnakhte as king,193 at least an addi-
tional 7 years must have passed between Bocchoris’s accession to the
throne (ca. 725, cf. above) and the campaign of Piye,194 and thus the
campaign will have taken place shortly before ca. 732, perhaps 733/734.
This is possible, but not compelling. Tefnakhte’s predecessors were not
kings,195 and on two donation stelae from years 36 and 38 of Shoshenq
V—certainly only a few years before the campaign of Piye196—he him-
self does not yet bear the royal title,197 and dates himself according to

Bocchoris 6 years (following Africanus), and reports that Shabaka burnt him alive.
Shabaka himself is only known in Egypt (even LE) in his regnal year 2.
190
Only the erection of the stela with the record of this campaign is dated, in the
first month of year 21. It is generally agreed that the campaign must have taken place
in the previous year.
191
JEA 54 (1968), 165–172; pl. XXV; for the alleged year 30 on the mummy wrap-
ping London BM 6640 cf. D. B. Redford, JARCE 22 (1985), 9–12; figs. 1–2, accord-
ing to which it can be read as either 20 or 40.
192
In ll. 19–20 he is named “Great Prince of the West”, along with a few of his
other titles; in general, however, he is merely the “Chief of the Ma” (ll. 28; 80; 126).
193
A hieratic donation stela in Athens, cf. R. el-Sayed, Documents relatifs à Saïs et ses
divinités, BdE 69 (1975), 37–53; pl. 7. K.-H. Priese (ZÄS 98 [1972], 19–21) and
K. Baer ( JNES 32 [1973], 23–24) have disputed that the king Tefnakhte with the
throne-name ”pss-R' is the same as the Prince Tefnakhte on the stela of Piye. They
assume instead that this is the first king of Dyn. 26 (before the predecessor of Neco
I) mentioned by Manetho (“Stephinates”), and thus a local prince of Sais. Opposing
this stance is the fact that one of the stelae of ”pss-R' Tefnakhte actually probably
comes from the eastern Delta (cf. Yoyotte [n. 177], 37–40), which was most assuredly
not under the control of the local princes ruling in Sais during Dyn. 25. Furthermore,
Diodor (I, 43) specifies that the king Tefnakhte, predecessor of the sage Bokchoris,
undertook an expedition to “Arabia”, and this would only have been possible from
the eastern Delta.
194
The possibility that Tefnakhte only became king after the campaign, but that his
regnal years were then post facto extended back to a point in time before the cam-
paign is rejected by Kitchen (TIP, § 112).
195
J. Yoyotte, BSFE 31 (1960), 13–22; TIP, § 113; 468.
196
Cf. above, section 4.
197
The unusual designation, “Great Prince of the Entire Land” reveals that his
ambition (and certainly also his power) extended far beyond that of the ordinary Libyan
local princes.
the third intermediate period 263

Shoshenq V. If he became king shortly thereafter, e.g., after the death


of Shoshenq V, this could only have taken place on the basis of his
own power. As one very conscious of legitimacy, Piye would thus not
have had the slightest reason to have designated someone as a king if
that person had only just shortly before proclaimed himself king, and
even less so if this person was his major opponent.198 It would thus be
possible to set the campaign of Piye somewhat closer to the accession
of Bocchoris, perhaps, between 734 and 726 BC; his accession to the
throne would thus be ca. 753–745 BC.

7. Connecting Dyns. 22 and 25

Shoshenq V died before the campaign of Piye, but not long before,
since Tefnakhte claims the title “Great Prince of the Entire Land” in
year 38. On the other hand, however, Shoshenq’s rule was apparently
uncontested in Memphis in his year 37, and thus Tefnakhte’s expan-
sion was not as advanced as at the beginning of the campaign of Piye.
In addition, there may be another king Petubaste (cf. above) to insert
before Osorkon IV who reigned in Bubastis and Tanis during the cam-
paign. A period of about 5 years between the death of Shoshenq V
and year 20 of Piye would appear reasonable.
For the kings from Shoshenq I to Takelot I we can reckon at least
21 + 35 + 15 years, for Osorkon II at least 30 years, as a year 29 is
very probably documented and the genealogical data favours a long
reign (cf. above, section 1). For the kings Shoshenq III, Shoshenq IIIa,
Pami and Shoshenq V, we have made a minimal period of 95 years
(cf. above, section 3). If we start with year 945 as the beginning of
Dyn. 22, the year 38 of Shoshenq V is to be set in 749 at the earliest.
The campaign of Piye can be placed in the years 734–726 (cf. above,
section 6); 10–18 years would remain to bridge the period between the
campaign and year 38 (+ 5) of Shoshenq V. This result is also realis-
tic as those reigns the duration of which is not certain were assigned
minimal values here. Where these missing years must be placed is a

198
On his “small stela” (Khartoum 1851, G. A. Reisner, ZÄS 66 [1931) 89–100;
pls. V–VI) Piye clearly enunciates that only that person is king whom he makes king,
and not those whom he forbids. This could apply to Tefnakhte, nor does he call
Tefnakhte “Great Prince of the Entire Land”, but rather “Chief of the Ma”.
264 karl jansen-winkeln

matter of speculation, but there are several possibilities.199 First of all,


the year 945 is not certain. If the campaign of Shoshenq in Palestine
(926/925 in year 5 of Rehabeam) did not take place in his year 20,
but rather a few years earlier—as is entirely possible200—the beginning
of the reign must accordingly be placed somewhat later. Candidates
for a somewhat longer reign are Shoshenq V himself and Osorkon II.
Likewise, Pami may have reigned for more than 6 years,201 if his “annals”
were not written posthumously (cf. above, section 3), and Osorkon I
and his successors could have ruled longer than we have assumed
above.202 In any case, a very slight extension of a few reigns is just as
unproblematic as setting the beginning of Dyn. 22 marginally later in
history.

8. Conclusion

For the chronology of the TIP, Egyptian sources only supply the year
690 as a certain point of departure. Additionally, the date of the cam-
paign of Shoshenq I, presumably towards the end of his reign, can be
placed with the aid of Near Eastern chronology in 925/926.203 Between
these two there is not one single firm date, but the sequence of kings
and the highest known dates for these kings does not leave significant
gaps. The general framework of the chronology of this age is certain.
Additional finds of dated monuments from this period will hopefully
add to the previous discoveries, and lead to an even higher degree of
resolution, leaving still less uncertainty.

199
Assigning the entire sum of years to the reign of Osorkon II, as Aston (n. 31;
145–148) does, is not necessarily the most logical possibility.
200
Cf. above, Jansen-Winkeln, Chapter II. 9.
201
Beckerath, Chronologie, 98, assigns him 11 years.
202
The usual numbers still depend to a great extent upon the very doubtful figures
for this period provided by the copyists of Manetho.
203
Cf. above, n. 200. Following alternative and acceptable calculations in OT stud-
ies, the year 5 of Rehabeam would not have been 926/5, but rather 922/1 (H. Donner,
Geschichte des Volkes Israel und seiner Nachbarn in Grundzügen, 2 [1995], 274); and this would
correspond to the Egyptian dates quite well. A “chronological problem” noted by
Donner ibidem, 321, n. 14) does not exist in this fashion: the Egyptian chronology is
absolutely dependent upon Near Eastern chronology. If one follows Begrich/Jepsen
and not Thiele, one simply shifts the accession of Shoshenq I by the same margin.
II. 11 SAITE AND PERSIAN EGYPT, 664 BC–332 BC
(DYNS. 26–31, PSAMMETICHUS I TO ALEXANDER’S
CONQUEST OF EGYPT)

Leo Depuydt

In the period at hand, events can dated exactly in absolute terms, that
is, in relation to the present moment in time. Thus, the death of
Psammetichus II in Month 1 Day 23 of his Year 7, that is, on 9 Feb
589 BC, is separated from the same time of day (whatever it was) on
1 Jan AD 2003 by 946311 full days or 24-hour periods. Chapter III.
11 outlines the general principles of day-exact chronology, with refer-
ences to contributions that provide more detail.1 In terms of chrono-
logical structure, the period has three natural subdivisions, each with
its own anatomy: (1) Dyn. 26, the Saite dynasty; (2) Dyn. 27, consist-
ing of Persian rulers; and (3) the fourth century BC up to Alexander’s
conquest of 332 BC, when Egypt was independent yet in Persia’s sphere
of influence. Accordingly, the following three equations have been
obtained in fundamentally different ways.
1. Year 7 Month 1 Day 23 (I #¢t 23) = 9 Feb 589 BC
of Psammetichus II
2. Year 15 Month 4 Day 16 (16 Hathyr) = 4 Mar 471 BC
of Xerxes I
3. Year 16 Month 8 Day 21 (21 Pharmouthi) = 5 Jul 343 BC
of Nectanebo II
The structural hierarchy of the chronology of Dyns. 26–31 is as fol-
lows. Dyn. 27 is most secure, ultimately owing to synchronies with
Mesopotamia established through Babylonian astronomical texts. As the
anchor, Dyn. 27 is treated here first. Dyn. 26 depends on Dyn. 27 and
is reasonably secure. It is treated next, along with Taharqa’s reign, the
last of Dyn. 25, which depends on Dyn. 26. The fourth century BC
is best seen as a gap to be bridged between Dyn. 27 and Alexander’s
conquest of 332 BC. It is only partly day-exact.

1
For a brief outline, see Beckerath, Chronologie, 79–88, with supplements in L. Depuydt,
WdO 30 (1999), 143–51, a review. See also H.-J. Thissen, “Chronologie der frühde-
motischen Papyri”, Enchoria 10 (1980), 105–25.
266 leo depuydt

1. Dynasty 27

The four pioneers who in the more recent past have done the most
to cement the chronology of Egypt in 664 BC–332 BC by consolidat-
ing Dyn. 27 as its anchor are E. Meyer, R. A. Parker, A. Sachs, and
above all F. X. Kugler.2 By manipulating (1) Ptolemy’s Canon, (2) pre-
dating of postdating, and (3) the cuneiform record,3 the following approx-
imate dates for the beginnings of the reigns are obtained. Details follow
in the notes to the table in section 4. Day dates derived from ancient
lunar dates can be one to two days off.

Cambyses Aug 530 BC


Darius I 29 Sep–22 Dec 522 BC
Xerxes I late Nov 486 BC
Artaxerxes I 5 Aug 465 BC–2 Jan 464 BC
Darius II 25 Dec 424 BC–13 Feb 423 BC
Artaxerxes II 18 Sept 405 BC–9 Apr 404 BC

Babylonian regnal Year 1 began on the first Babylonian new year fol-
lowing these dates (see 1.2 in Chapter III. 11), that is, around the
spring equinoxes of 529, 521, 485, 464, 423, and 404 BC. The num-
berless period lasting from the beginning of the reign to the first Baby-
lonian new year may be called the accession year. The above dates
can be translated as follows into reign lengths detailing years only.

Dyn. 27 527–525 – ca. 400


Cambyses (second half of reign) 527–525 – 522
Darius I 522 – 486
Xerxes I 486 – 465
Artaxerxes I 465 – 424/3
Darius II 424/3 – 405/4
Artaxerxes II (beginning of reign) 405/4 – ca. 400
En-dashes (–) denote time-periods. Slashes (/) mark ancient calendar-years straddling
two Julian years.

2
Pioneering contributions are: E. Meyer, Forschungen zur Alten Geschichte, II: Zur
Geschichte des fünften Jahrhunderts v. Chr. (Halle, 1899); F. X. Kugler, Sternkunde und Sterndienst
in Babel (Münster, 1907–24); R. A. Parker, “Persian and Egyptian Chronology”, AJSL
58 (1941), 285–301; id., “Darius and His Egyptian Campaign”, AJSL 58 (1941), 373–377;
Parker & Dubberstein, Chronology; A. Sachs’s work on the Babylonian diaries.
3
See sections 2.1, 1.2, and 2.3 in Chapter III. 11.
saite and persian egypt, 664 bc‒332 bc 267

2. Dynasty 26

The main structural feature of the day-exact chronology of Dyn. 26 is


its dependence on the day-exact chronology of Dyn. 27. Already in the
nineteenth century AD, data from Greek historians and Serapeum
stelae (see 2.2 in Chapter III. 11) had fixed the reigns of Dyn. 26
absolutely to within one or two years. In the 1950s, two contributions
by R. A. Parker pushed the limit of day-exact chronology back from
525 BC in two moves: (1) from 525 BC back to 664 BC; (2) from 664
BC back to 690 BC. Each move rests on a single piece of evidence.4
(1) A civil-lunar double date deciphered jointly by M. Malinine and
R. A. Parker at Brown University in a photograph of the abnormal
hieratic papyrus Louvre 7848 equates civil Month 10 Day 13 (II Shemu
13) with lunar Month 9 Day 15 (I Shemu 15) in Year 12 of Amasis.
Lunar Day 15 ought to fall around full moon. Before this date sur-
faced, Amasis’s Year 12 had mostly been equated with the 365-day
year 10 Jan 558 BC–9 Jan 557 BC, in which civil Month 10 Day 13
equals 19 October. However, 19 Oct 558 BC is not close to full moon.
In the preceding Egyptian year, 10 Jan 559 BC–9 Jan 558 BC, civil
Month 10 Day 13 also equals 19 Oct. and 19 Oct 559 BC does occur
near the full moon of 21 Oct 12:09PM.5 The earlier Egyptian year is
therefore in all probability Amasis’s Year 12.6
This backward shift of Amasis’s regnal years affects the dating of
Cambyses’s conquest. Traditionally, the conquest had been dated to
spring 525 BC and Amasis’s Year 44 had been considered his last. But
now, Day 1 of Year 44 was moved back from 2 Jan 526 BC to 2 Jan
527 BC. To keep the end of Amasis’s reign close to the conquest,
R. A. Parker postulated an unattested Year 45 for Amasis beginning
on 2 Jan 526 BC. However, postulating a Year 45 was rendered unnec-
essary when the arguments dating the conquest to spring 525 BC.

4
A passage in Demotic papyrus Berlin 13588 that has been interpreted variously
as a solar eclipse (E. Hornung, “Die Sonnenfinsternis nach dem Tode Psammetichs
I”, ZÄS 92 (1966), 38–9) and as a lunar eclipse (M. Smith, “Did Psammetichus I Die
Abroad?”, OLP 22 (1991), 101–9) has possible chronological relevance (see also L. Depuydt,
“On the Consistency of the Wandering Year as Backbone of Egyptian Chronology”,
JARCE 32 (1995), 43–58, at 53, note 50); for a different view, see Chapter III. 4.
5
Goldstine’s time, Moons, 37, for Babylon, minus 47 minutes for Thebes.
6
R. A. Parker, “The Length of Reign of Amasis and the Beginning of the Twenty-
sixth Dynasty”, MDAIK 15 (1957), 208–12. pLouvre 7848 has now been published by
K. Donker van Heel, Abnormal Hieratic and Early Demotic Texts Collected by the Theban
Choachytes in the Reign of Amasis (Leiden, 1995), 93–99.
268 leo depuydt

became invalid. The conquest is now better dated more imprecisely to


527–525 BC.7
(2) Serapeum stela Louvre IM 3733, the official epitaph of an Apis
bull born in Taharqa’s Year 26 and deceased in Psammetichus I’s Year
20 pushes day-exact chronology back further from 664 BC to 690 BC.8
Accordingly, Taharqa would have counted his reign from some day in
the Egyptian year 12 Feb 690 BC–11 Feb 689 BC. The year 12 Feb
691 BC–11 Feb 690 BC is also a possibility. This window of doubt
exists because IM 3733 lacks certain information: (1) the months and
days of the bull’s life-span, given as 21 years; (2) the month and day
of his birth in Taharqa’s Year 26; and (3) the year of his day of instal-
lation, which was Month 8 Day 9.9 The lengths of the reigns of Dyn.
26 are as follows. Details follow in the notes to the table in section 4.

Dyn. 26 664/3 – ca. 527–5


Psammetichus I 664/3 – 610
Necho II 610 – 595
Psammetichus II 595 – 589
Apries 589 – 570
Amasis 570 – 527/6
Psammetichus III? some (6?) months in 527–525?
En-dashes (–) denote time-periods. Slashes (/) mark ancient calendar-years straddling
two Julian years.

3. Dynasties 28 to 31

The main structural feature of the chronology of Egypt in the fourth


century BC up to Alexander’s conquest of 332 BC is that it is partly
day-exact. This feature deserves attention first. The evidence is in
Nectanebo’s Dream, which prophesies the demise of Egypt’s last native
ruler. The king of the Greek version had been identified with Nectanebo II,

7
For a detailed discussion, see L. Depuydt, “Egyptian Regnal Dating under Cambyses
and the Date of the Persian Conquest”, in: Studies in Honor of William Kelly Simpson
(Boston, 1996), 179–190, at 184, note 23. See now also J.v. Beckerath, “Nochmals die
Eroberung Ägyptens durch Kambyses”, ZÄS 129 (2002), 1–5; and E. Cruz-Uribe, “The
Invasion of Egypt by Cambyses”, Transeuphratène 25 (2003), 9–60, at 54–57.
8
R. A. Parker, “The Length of Reign of Taharqa”, Kush 8 (1960), 267–269.
9
For details, see Depuydt, JARCE 32, 52–53.
saite and persian egypt, 664 bc‒332 bc 269

not Nectanebo I, but it took a Demotic fragment to confirm this


definitively.10 In the Greek version in papyrus Leiden I 396, the Dream
is dated to the night from 21 to 22 Pharmouthi (Day 21 to Day 22
of Month 8) of Nectanebo II’s Year 16.11 It was about full moon (kata
yeon dia dexomenian). The lunar date can be identified with the help
of Manetho’s king-lists. In Manetho,12 about 72 to 74 years separate
the end of the reign of Darius II in 405/4 BC from Alexander’s con-
quest of 332 BC. Nectanebo II’s Year 16 is therefore about 62 years
away from 405/4 BC, or falls about halfway between 345 BC and 340
BC. By far the closest match for the combination of full moon with
21/2 Pharmouthi falls in 343 BC. In this year, 21/2 Pharmouthi equals
5/6 Jul 343 BC. Full moon occurred about noon of 6 July.13 The fol-
lowing equations result. Details follow in the table in section 4.

Nectanebo II Year 1 21 Nov 359–20 Nov 358


Nectanebo II Year 2 21 Nov 358–19 Nov 357
And so on
Nectanebo II Year 19 16 Nov 341–15 Nov 340
Nectanebo II Year 20 16 Nov 340–15 Nov 339

It is not known when in the Egyptian year 21 Nov 359 BC–20 Nov
358 BC Nectanebo II assumed power. According to Manetho, he reigned
20 years. The highest regnal date in hieroglyphic sources is Year 18.14
According to Manetho, or an addition to Manetho, Artaxerxes III,
also called Ochos, conquered Egypt in his Year 20.15 His Babylonian
Year 20 ran from spring 339 BC to spring 338 BC. According to
predating of postdating (see 1.2 in Chapter III. 11), his Egyptian
year 20 would begin on the Egyptian new year of 16 Nov 340 BC.

10
K. Ryholt, “A Demotic Version of Nectanebos’ Dream (P. Carlsberg 562)”, ZPE
122 (1998), 197–200, with bibliography. Cf. also A. Spalinger, “The Date of the Dream
of Nectanebo”, SAK 19 (1992), 295–304.
11
The Demotic version’s Year 18 must be an error. Indeed, three other Demotic
versions, Carlsberg 424, 499, and 559, date a sequel to the Dream to Year 16 (Ryholt,
ZPE 122, 198).
12
Waddell, Manetho.
13
Goldstine’s time (Moons, 55) for Babylon, minus 53 minutes for Memphis.
14
Kienitz, Geschichte, 216 (part of a useful and extensive listing of documents dating
to the fourth century BC).
15
Waddell, Manetho, 184–187.
270 leo depuydt

Remarkably, the conquest is now mostly dated earlier, to 343/2 BC.16


But on closer inspection, the sole argument ever adduced in favor of
343/2 BC appears non-binding. Persian envoys visited Athens when
Lukiskos was archon, from July 344 BC to July 343 BC. It has been
assumed that the army of Greek mercenaries hired on the occasion
could not have been maintained long without action. The invasion of
Egypt must therefore have followed soon, in the fall of 343 BC.17 This
argument is reasonable, but hardly conclusive. It contradicts other evi-
dence that seems firmer. 340/39 BC remains preferable as the date of
the re-conquest.
For the rest of Dyns. 28 to 30, Manetho serves as the basis, fine-
tuned by data from the monuments. Current assignments of regnal
years to Egyptian wandering years are not definitive, but in all prob-
ability correct to within one or two years. Africanus, generally consid-
ered the best source for Manetho, who is himself not preserved, gives
73 years and 4 months for that period: 6 years for Dyn. 28; 20 years
4 months for Dyn. 29; 38 years for Dyn. 30; and 9 years for Dyn. 31.
Provisional dates for the reigns are as follows.18

Dyn. 28
Amyrtaios about 404/3–398/7
Dyn. 29
Nepherites I about 398/7–392/1
Achoris about 392/1–379/8
Psammuthis brief reign
Nepherites II brief reign
Dyn. 30
Nectanebo I (Nectanebes) about 379/8–361/0
Teos/Tachos about 361/0–359/8
Nectanebo II (Nectanebos) about 359/8–342/1

16
Kienitz, Geschichte, 172; A. B. Lloyd, “Egypt, 404–332 BC”, in: CAH 2, vol. 6
(Cambridge, 1994), 337–360, 981–987, at 359, where the date is deemed a ‘fact’.
17
E. J. Bickerman, “Notes sur la chronologie de la XXXe dynastie”, in: Mélanges
Maspero I: Orient Ancien (Cairo: MIFAO 56, 1934), 77–84, at 81.
18
From Lloyd, CAH 2, vol. 6, 358. Lloyd’s survey of Egypt in 404–332 BC includes
an appendix on chronology. See also J. D. Ray, “Psammuthis and Hakoris”, JEA 72
(1986), 149–58; C. Traunecker, “Essai sur l’histoire de la XXIXe Dynastie”, 79 (1979),
395–436. M. Chauveau (“Les archives d’un temple des oasis au temps des Perses”,
BSFE 137 (1996), 32–47) suggests that Amyrtaios, sole ruler of Dyn. 28, sometimes
used the name Psammetichus. That would him make the fifth ruler of that name, in
addition to Psammetichus I and II of Dyn. 26 and two postulated ephemeral rulers
dating to about 526 BC and to about the mid 480s BC.
saite and persian egypt, 664 bc‒332 bc 271

The evidence for Dyn. 31 is sparse.19 The following dates for the lengths
of the Egyptian reigns of its three Persian kings are derived from the
date of Artaxerxes III’s conquest (see section 3 above) and from Ptolemy’s
Canon (see 2.1 in Chapter III. 11). The first Babylonian regnal years
of Arses and Darius III began in the spring of 337 BC and of 335 BC
respectively, that is, on the first Babylonian new year following the
Canon’s beginnings for their reigns, namely 16 Nov 338 BC and 15
Nov 336 BC. Their actual reigns may have begun up to a year earlier
and therefore have either preceded or followed the Canon’s beginning.

Dyn. 31 ca. 340 – 332


Artaxerxes III (end of reign) ca. 340 – 338/7
Arses 338/7 – 336/5
Darius III 336/5 – 332

4. Regnal Years in 690 BC–332 BC

All the years in the following table are exactly 365 days long. The
years printed in italics include a Julian 29 February. The Era of
Nabonassar in the first column is a year-count from the first king of
Ptolemy’s Canon (see 2.1 in Chapter III. 11). The table joins the Canon
in 525 BC. Details on the transitions of the historical reigns appear in
the notes. The years from 525 BC to the end of the fifth century BC
are almost certainly also the actual historical Egyptian regnal years; the
years back to 664 BC, with high probability; those back to 690 BC,
quite possibly. For the fourth century BC, only regnal years placed
with reasonable certainty are listed. To convert, say, Month 5 Day 29
of Year 5 of Cambyses, one may proceed as follows.

19
See D. Devauchelle, “Réflexions sur les documents égyptiens datés de la Deuxième
Domination perse”, Transeuphratène 10 (1995), 35–43; A. B. Lloyd, “Manetho and the
Thirty-first Dynasty”, in: Pyramid Studies and Other Essays Presented to I. E. S. Edwards
(London: EES Occasional Publications 7, 1988), 154–160; A. Spalinger, “The Reign
of King Chababash: An Interpretation”, ZÄS 105 (1978), 142–54. On the Demotic
Chronicle, see J. H. Johnson, “The Demotic Chronicle as an Historical Source”, Enchoria
4 (1974), 1–17.
272 leo depuydt

2 Jan 525 = Month 1 Day 1


+ 29 + 29
31 Jan 525 = Month 1 Day 3
+ 1 + 1
1 Feb 525 = Month 2 Day 1
+ 28 + 28
29 Feb 525 = Month 2 Day 29
+ 1 + 1
and so on

Era of Ruler+ Length of 365-day


Nabonassar Regnal Year Egyptian Year

58 Shabataka? Year ?
+ Taharqa 1 12 Feb 690–11 Feb 689i
59 Taharqa 2 12 Feb 689–10 Feb 688
60 Taharqa 3 11 Feb 688–10 Feb 687
61 Taharqa 4 11 Feb 687–10 Feb 686
62 Taharqa 5 11 Feb 686–10 Feb 685
63 Taharqa 6 11 Feb 685–9 Feb 684
64 Taharqa 7 10 Feb 684–9 Feb 683
65 Taharqa 8 10 Feb 683–9 Feb 682
66 Taharqa 9 10 Feb 682–9 Feb 681
67 Taharqa 10 10 Feb 681–8 Feb 680
68 Taharqa 11 9 Feb 680–8 Feb 679
69 Taharqa 12 9 Feb 679–8 Feb 678
70 Taharqa 13 9 Feb 678–8 Feb 677
71 Taharqa 14 9 Feb 677–7 Feb 676
72 Taharqa 15 8 Feb 676–7 Feb 675
73 Taharqa 16 8 Feb 675–7 Feb 674
74 Taharqa 17 8 Feb 674–7 Feb 673
75 Taharqa 18 8 Feb 673–6 Feb 672
76 Taharqa 19 7 Feb 672–6 Feb 671
77 Taharqa 20 7 Feb 671–6 Feb 670
78 Taharqa 21 7 Feb 670–6 Feb 669
79 Taharqa 22 7 Feb 669–5 Feb 668
80 Taharqa 23 6 Feb 668–5 Feb 667
81 Taharqa 24 6 Feb 667–5 Feb 666
82 Taharqa 25 6 Feb 666–5 Feb 665
83 Taharqa 26 6 Feb 665–4 Feb 664
84 Taharqa 27 5 Feb 664–
+ Psammetichus I 1 –4 Feb 663ii
85 Psammetichus I 2 5 Feb 663–4 Feb 662
86 Psammetichus I 3 5 Feb 662–4 Feb 661
87 Psammetichus I 4 5 Feb 661–3 Feb 660
(continued on next page)
saite and persian egypt, 664 bc‒332 bc 273

Table (cont.)
Era of Ruler+ Length of 365-day
Nabonassar Regnal Year Egyptian Year

88 Psammetichus I 5 4 Feb 660–3 Feb 659


89 Psammetichus I 6 4 Feb 659–3 Feb 658
90 Psammetichus I 7 4 Feb 658–3 Feb 657
91 Psammetichus I 8 4 Feb 657–2 Feb 656
92 Psammetichus I 9 3 Feb 656–2 Feb 655
93 Psammetichus I 10 3 Feb 655–2 Feb 654
94 Psammetichus I 11 3 Feb 654–2 Feb 653
95 Psammetichus I 12 3 Feb 653–1 Feb 652
96 Psammetichus I 13 2 Feb 652–1 Feb 651
97 Psammetichus I 14 2 Feb 651–1 Feb 650
98 Psammetichus I 15 2 Feb 650–1 Feb 649
99 Psammetichus I 16 2 Feb 649–31 Jan 648
100 Psammetichus I 17 1 Feb 648–31 Jan 647
101 Psammetichus I 18 1 Feb 647–31 Jan 646
102 Psammetichus I 19 1 Feb 646–31 Jan 645
103 Psammetichus I 20 1 Feb 645–30 Jan 644
104 Psammetichus I 21 31 Jan 644–30 Jan 643
105 Psammetichus I 22 31 Jan 643–30 Jan 642
106 Psammetichus I 23 31 Jan 642–30 Jan 641
107 Psammetichus I 24 31 Jan 641–29 Jan 640
108 Psammetichus I 25 30 Jan 640–29 Jan 639
109 Psammetichus I 26 30 Jan 639–29 Jan 638
110 Psammetichus I 27 30 Jan 638–29 Jan 637
111 Psammetichus I 28 30 Jan 637–28 Jan 636
112 Psammetichus I 29 29 Jan 636–28 Jan 635
113 Psammetichus I 30 29 Jan 635–28 Jan 634
114 Psammetichus I 31 29 Jan 634–28 Jan 633
115 Psammetichus I 32 29 Jan 633–27 Jan 632
116 Psammetichus I 33 28 Jan 632–27 Jan 631
117 Psammetichus I 34 28 Jan 631–27 Jan 630
118 Psammetichus I 35 28 Jan 630–27 Jan 629
119 Psammetichus I 36 28 Jan 629–26 Jan 628
120 Psammetichus I 37 27 Jan 628–26 Jan 627
121 Psammetichus I 38 27 Jan 627–26 Jan 626
122 Psammetichus I 39 27 Jan 626–26 Jan 625
123 Psammetichus I 40 27 Jan 625–25 Jan 624
124 Psammetichus I 41 26 Jan 624–25 Jan 623
125 Psammetichus I 42 26 Jan 623–25 Jan 622
126 Psammetichus I 43 26 Jan 622–25 Jan 621
127 Psammetichus I 44 26 Jan 621–24 Jan 620
128 Psammetichus I 45 25 Jan 620–24 Jan 619
129 Psammetichus I 46 25 Jan 619–24 Jan 618
(continued on next page)
274 leo depuydt

Table (cont.)
Era of Ruler+ Length of 365-day
Nabonassar Regnal Year Egyptian Year

130 Psammetichus I 47 25 Jan 618–24 Jan 617


131 Psammetichus I 48 25 Jan 617–23 Jan 616
132 Psammetichus I 49 24 Jan 616–23 Jan 615
133 Psammetichus I 50 24 Jan 615–23 Jan 614
134 Psammetichus I 51 24 Jan 614–23 Jan 613
135 Psammetichus I 52 24 Jan 613–22 Jan 612
136 Psammetichus I 53 23 Jan 612–22 Jan 611
137 Psammetichus I 54 23 Jan 611–22 Jan 610
138 Psammetichus I 55 23 Jan 610–
–no later than 18 Nov 610iii
+ Necho II 1 by 19 Nov 610 at the latestiii–
–22 Jan 609
139 Necho II 2 23 Jan 609–21 Jan 608
140 Necho II 3 22 Jan 608–21 Jan 607
141 Necho II 4 22 Jan 607–21 Jan 606
142 Necho II 5 22 Jan 606–21 Jan 605
143 Necho II 6 22 Jan 605–20 Jan 604
144 Necho II 7 21 Jan 604–20 Jan 603
145 Necho II 8 21 Jan 603–20 Jan 602
146 Necho II 9 21 Jan 602–20 Jan 601
147 Necho II 10 21 Jan 601–19 Jan 600
148 Necho II 11 20 Jan 600–19 Jan 599
149 Necho II 12 20 Jan 599–19 Jan 598
150 Necho II 13 20 Jan 598–19 Jan 597
151 Necho II 14 20 Jan 597–18 Jan 596
152 Necho II 15 19 Jan 596–18 Jan 595
153 Necho II 16 19 Jan 595–
–at least until 4 May 595iv
+ Psammetichus II 1 by 23 Nov 595 at the latest–
–18 Jan 594iv
154 Psammetichus II 2 19 Jan 594–18 Jan 593
155 Psammetichus II 3 19 Jan 593–17 Jan 592
156 Psammetichus II 4 18 Jan 592–17 Jan 591
157 Psammetichus II 5 18 Jan 591–17 Jan 590
158 Psammetichus II 6 18 Jan 590–17 Jan 589
159 Psammetichus II 7 18 Jan 589–9 Feb 589
+ Apries 1 10 Feb 589–16 Jan 588v
160 Apries 2 17 Jan 588–16 Jan 587
161 Apries 3 17 Jan 587–16 Jan 586
162 Apries 4 17 Jan 586–16 Jan 585
163 Apries 5 17 Jan 585–15 Jan 584
164 Apries 6 16 Jan 584–15 Jan 583
(continued on next page)
saite and persian egypt, 664 bc‒332 bc 275

Table (cont.)
Era of Ruler+ Length of 365-day
Nabonassar Regnal Year Egyptian Year

165 Apries 7 16 Jan 583–15 Jan 582


166 Apries 8 16 Jan 582–15 Jan 581
167 Apries 9 16 Jan 581–14 Jan 580
168 Apries 10 15 Jan 580–14 Jan 579
169 Apries 11 15 Jan 579–14 Jan 578
170 Apries 12 15 Jan 578–14 Jan 577
171 Apries 13 15 Jan 577–13 Jan 576
172 Apries 14 14 Jan 576–13 Jan 575
173 Apries 15 14 Jan 575–13 Jan 574
174 Apries 16 14 Jan 574–13 Jan 573
175 Apries 17 14 Jan 573–12 Jan 572
176 Apries 18 13 Jan 572–12 Jan 571
177 Apries 19 13 Jan 571–12 Jan 570
178 Apries 20 13 Jan 570–
–at least until 19 Oct 570vi
+ Amasis 1 by Day x in 11 Jul–9 Aug 570
at the latestvi–12 Jan 569
179 Amasis 2 13 Jan 569–11 Jan 568
180 Amasis 3 12 Jan 568–11 Jan 567
181 Amasis 4 12 Jan 567–11 Jan 566
182 Amasis 5 12 Jan 566–11 Jan 565
183 Amasis 6 12 Jan 565–10 Jan 564
184 Amasis 7 11 Jan 564–10 Jan 563
185 Amasis 8 11 Jan 563–10 Jan 562
186 Amasis 9 11 Jan 562–10 Jan 561
187 Amasis 10 11 Jan 561–9 Jan 560
188 Amasis 11 10 Jan 560–9 Jan 559
189 Amasis 12 10 Jan 559–9 Jan 558
190 Amasis 13 10 Jan 558–9 Jan 557
191 Amasis 14 10 Jan 557–8 Jan 556
192 Amasis 15 9 Jan 556–8 Jan 555
193 Amasis 16 9 Jan 555–8 Jan 554
194 Amasis 17 9 Jan 554–8 Jan 553
195 Amasis 18 9 Jan 553–7 Jan 552
196 Amasis 19 8 Jan 552–7 Jan 551
197 Amasis 20 8 Jan 551–7 Jan 550
198 Amasis 21 8 Jan 550–7 Jan 549
199 Amasis 22 8 Jan 549–6 Jan 548
200 Amasis 23 7 Jan 548–6 Jan 547
201 Amasis 24 7 Jan 547–6 Jan 546
202 Amasis 25 7 Jan 546–6 Jan 545
203 Amasis 26 7 Jan 545–5 Jan 544
(continued on next page)
276 leo depuydt

Table (cont.)
Era of Ruler+ Length of 365-day
Nabonassar Regnal Year Egyptian Year

204 Amasis 27 6 Jan544–5 Jan 543


205 Amasis 28 6 Jan543–5 Jan 542
206 Amasis 29 6 Jan542–5 Jan 541
207 Amasis 30 6 Jan541–4 Jan 540
208 Amasis 31 5 Jan540–4 Jan 539
209 Amasis 32 5 Jan539–4 Jan 538
210 Amasis 33 5 Jan538–4 Jan 537
211 Amasis 34 5 Jan537–3 Jan 536
212 Amasis 35 4 Jan536–3 Jan 535
213 Amasis 36 4 Jan535–3 Jan 534
214 Amasis 37 4 Jan534–3 Jan 533
215 Amasis 38 4 Jan533–2 Jan 532
216 Amasis 39 3 Jan532–2 Jan 531
217 Amasis 40 3 Jan531–2 Jan 530
218 Amasis 41 3 Jan530–2 Jan 529
219 Amasis 42 3 Jan529–1 Jan 528
220 Amasis 43 2 Jan528–1 Jan 527
221 Amasis 44 2 Jan527–
+ uncertainvii –1 Jan 526
222 uncertainvii 2 Jan 526–1 Jan 525
223viii uncertainvii 2 Jan 525–
+ Cambyses 5 –31 Dec 525
224 Cambyses 6 1 Jan 524–31 Dec 524
225 Cambyses 7 1 Jan 523–31 Dec 523
226ix Cambyses 8 1 Jan 522–death in 522x
+ Darius I’s accession in 522xi–
accession year –31 Dec 522
227ix Darius I’s 1 Jan 521–
accession year or Year 1? –ca. 13 Apr 521xii
Darius I 1 ca. 13 Apr 521–
–30 Dec 521
228 Darius I 2 31 Dec 521–30 Dec 520
229 Darius I 3 31 Dec 520–30 Dec 519
230 Darius I 4 31 Dec 519–30 Dec 518
231 Darius I 5 31 Dec 518–29 Dec 517
232 Darius I 6 30 Dec 517–29 Dec 516
233 Darius I 7 30 Dec 516–29 Dec 515
234 Darius I 8 30 Dec 515–29 Dec 514
235 Darius I 9 30 Dec 514–28 Dec 513
236 Darius I 10 29 Dec 513–28 Dec 512
237 Darius I 11 29 Dec 512–28 Dec 511
238 Darius I 12 29 Dec 511–28 Dec 510
(continued on next page)
saite and persian egypt, 664 bc‒332 bc 277

Table (cont.)
Era of Ruler+ Length of 365-day
Nabonassar Regnal Year Egyptian Year

239 Darius I 13 29 Dec 510–27 Dec 509


240 Darius I 14 28 Dec 509–27 Dec 508
241 Darius I 15 28 Dec 508–27 Dec 507
242 Darius I 16 28 Dec 507–27 Dec 506
243 Darius I 17 28 Dec 506–26 Dec 505
244 Darius I 18 27 Dec 505–26 Dec 504
245 Darius I 19 27 Dec 504–26 Dec 503
246 Darius I 20 27 Dec 503–26 Dec 502
247 Darius I 21 27 Dec 502–25 Dec 501
248 Darius I 22 26 Dec 501–25 Dec 500
249 Darius I 23 26 Dec 500–25 Dec 499
250 Darius I 24 26 Dec 499–25 Dec 498
251 Darius I 25 26 Dec 498–24 Dec 497
252 Darius I 26 25 Dec 497–24 Dec 496
253 Darius I 27 25 Dec 496–24 Dec 495
254 Darius I 28 25 Dec 495–24 Dec 494
255 Darius I 29 25 Dec 494–23 Dec 493
256 Darius I 30 24 Dec 493–23 Dec 492
257 Darius I 31 24 Dec 492–23 Dec 491
258 Darius I 32 24 Dec 491–23 Dec 490
259 Darius I 33 24 Dec 490–22 Dec 489
260 Darius I 34 23 Dec 489–22 Dec 488
261 Darius I 35 23 Dec 488–22 Dec 487
262xiii Darius I 36 23 Dec 487–
–at least until 24 Nov 486xiv
+ Xerxes I’s at the latest by 1 Dec 486–
accession year –22 Dec 486
263xiii Xerxes I’s 23 Dec 486–
accession year or Year 1? –ca. 5 Apr 485xv
Xerxes I 1 ca. 5 Apr 485xv–21 Dec 485
264 Xerxes I 2 22 Dec 485–21 Dec 484
265 Xerxes I 3 22 Dec 484–21 Dec 483
266 Xerxes I 4 22 Dec 483–21 Dec 482
267 Xerxes I 5 22 Dec 482–20 Dec 481
268 Xerxes I 6 21 Dec 481–20 Dec 480
269 Xerxes I 7 21 Dec 480–20 Dec 479
270 Xerxes I 8 21 Dec 479–20 Dec 478
271 Xerxes I 9 21 Dec 478–19 Dec 477
272 Xerxes I 10 20 Dec 477–19 Dec 476
273 Xerxes I 11 20 Dec 476–19 Dec 475
274 Xerxes I 12 20 Dec 475–19 Dec 474
275 Xerxes I 13 20 Dec 474–18 Dec 473
(continued on next page)
278 leo depuydt

Table (cont.)
Era of Ruler+ Length of 365-day
Nabonassar Regnal Year Egyptian Year

276 Xerxes I 14 19 Dec 473–18 Dec 472


277 Xerxes I 15 19 Dec 472–18 Dec 471
278 Xerxes I 16 19 Dec 471–18 Dec 470
279 Xerxes I 17 19 Dec 470–17 Dec 469
280 Xerxes I 18 18 Dec 469–17 Dec 468
281 Xerxes I 19 18 Dec 468–17 Dec 467
282 Xerxes I 20 18 Dec 467–17 Dec 466
283xvi Xerxes I 21 18 Dec 466–
–Day x in 4–8 Aug 465xvii
+ Artaxerxes I’s by 5 Aug 465 at the earliestxviii–
accession year? –16 Dec 465
284xvi Artaxerxes I’s 17 Dec 465–
accession year or Year 1? –ca. 13 Apr 464xix
Artaxerxes I 1 ca. 13 Apr 464–16 Dec 464
285 Artaxerxes I 2 17 Dec 464–16 Dec 463
286 Artaxerxes I 3 17 Dec 463–16 Dec 462
287 Artaxerxes I 4 17 Dec 462–15 Dec 461
288 Artaxerxes I 5 16 Dec 461–15 Dec 460
289 Artaxerxes I 6 16 Dec 460–15 Dec 459
290 Artaxerxes I 7 16 Dec 459–15 Dec 458
291 Artaxerxes I 8 16 Dec 458–14 Dec 457
292 Artaxerxes I 9 15 Dec 457–14 Dec 456
293 Artaxerxes I 10 15 Dec 456–14 Dec 455
294 Artaxerxes I 11 15 Dec 455–14 Dec 454
295 Artaxerxes I 12 15 Dec 454–13 Dec 453
296 Artaxerxes I 13 14 Dec 453–13 Dec 452
297 Artaxerxes I 14 14 Dec 452–13 Dec 451
298 Artaxerxes I 15 14 Dec 451–13 Dec 450
299 Artaxerxes I 16 14 Dec 450–12 Dec 449
300 Artaxerxes I 17 13 Dec 449–12 Dec 448
301 Artaxerxes I 18 13 Dec 448–12 Dec 447
302 Artaxerxes I 19 13 Dec 447–12 Dec 446
303 Artaxerxes I 20 13 Dec 446–11 Dec 445
304 Artaxerxes I 21 12 Dec 445–11 Dec 444
305 Artaxerxes I 22 12 Dec 444–11 Dec 443
306 Artaxerxes I 23 12 Dec 443–11 Dec 442
307 Artaxerxes I 24 12 Dec 442–10 Dec 441
308 Artaxerxes I 25 11 Dec 441–10 Dec 440
309 Artaxerxes I 26 11 Dec 440–10 Dec 439
310 Artaxerxes I 27 11 Dec 439–10 Dec 438
311 Artaxerxes I 28 11 Dec 438–9 Dec 437
312 Artaxerxes I 29 10 Dec 437–9 Dec 436
(continued on next page)
saite and persian egypt, 664 bc‒332 bc 279

Table (cont.)
Era of Ruler+ Length of 365-day
Nabonassar Regnal Year Egyptian Year

313 Artaxerxes I 30 10 Dec 436–9 Dec 435


314 Artaxerxes I 31 10 Dec 435–9 Dec 434
315 Artaxerxes I 32 10 Dec 434–8 Dec 433
316 Artaxerxes I 33 9 Dec 433–8 Dec 432
317 Artaxerxes I 34 9 Dec 432–8 Dec 431
318 Artaxerxes I 35 9 Dec 431–8 Dec 430
319 Artaxerxes I 36 9 Dec 430–7 Dec 429
320 Artaxerxes I 37 8 Dec 429–7 Dec 428
321 Artaxerxes I 38 8 Dec 428–7 Dec 427
322 Artaxerxes I 39 8 Dec 427–7 Dec 426
323 Artaxerxes I 40 8 Dec 426–6 Dec 425
324xx Artaxerxes I 41xxi 7 Dec 425–
–death (Feb 424?)
ephemeral rulersxxi death Artaxerxes I–
–6 Dec 424
325xx ephemeral rulersxxi 7 Dec 424–Darius II’s
accession on Day x in
25 Dec 424–13 Feb 423
Darius II’s Darius II’s accession
accession yearxxi –ca. 11 Apr 423
Darius II 1xxi ca. 11 Apr 423–6 Dec 423
326 Darius II 2 7 Dec 423–6 Dec 422
327 Darius II 3 7 Dec 422–5 Dec 421
328 Darius II 4 6 Dec 421–5 Dec 420
329 Darius II 5 6 Dec 420–5 Dec 419
330 Darius II 6 6 Dec 419–5 Dec 418
331 Darius II 7 6 Dec 418–4 Dec 417
332 Darius II 8 5 Dec 417–4 Dec 416
333 Darius II 9 5 Dec 416–4 Dec 415
334 Darius II 10 5 Dec 415–4 Dec 414
335 Darius II 11 5 Dec 414–3 Dec 413
336 Darius II 12 4 Dec 413–3 Dec 412
337 Darius II 13 4 Dec 412–3 Dec 411
338 Darius II 14 4 Dec 411–3 Dec 410
339 Darius II 15 4 Dec 410–2 Dec 409
340 Darius II 16 3 Dec 409–2 Dec 408
341 Darius II 17 3 Dec 408–2 Dec 407
342 Darius II 18 3 Dec 407–2 Dec 406
343xxii Darius II 19 3 Dec 406–
–at least until ca. 17 Sep 405xxiii
Darius II 19 and ca. 18 Sep 405–
Artaxerxes II’s accession year? –1 Dec 405xxiii
(continued on next page)
280 leo depuydt

Table (cont.)
Era of Ruler+ Length of 365-day
Nabonassar Regnal Year Egyptian Year

344xxii Darius II 19 or 2 Dec 405–


Artaxerxes II’s accession year? –ca. 10 Apr 404xxiii
Artaxerxes II 1 ca. 10 Apr 404–
–1 Dec 404xxiii
345 Artaxerxes II 2 2 Dec 404–1 Dec 403
346 Artaxerxes II 3 2 Dec 403–1 Dec 402
347 Artaxerxes II 4 2 Dec 402–30 Nov 401
348 Artaxerxes II 5 1 Dec 401–30 Nov 400

About this time, Persia lost control of Egypt.xxiv The native kings of Dyns.
28–30 ruled for about seven decades. Their reigns are not day-exact, except
Nectanebo II’s of Dyn. 30. The following provisional estimates (see Lloyd,
CAH 2, vol. 6, 358) are probably correct to within one or two years.xxv

Dynasty 28
Amyrtaios ca. 404/3–398/7
Dynasty 29
Nepherites I ca. 398/7–392/1
Achoris ca. 392/1–379/8
Psammuthis brief reign
Nepherites II brief reign
Dynasty 30
Nectanebo I (Nectanebes) ca. 379/8–361/0
Teos/Tachos ca. 361/0–359/8

The reign of the third king of Dyn. 30 and the last native ruler of Egypt,
Nectanebo II (Nectanebos), is again day-exact, with high probability, as follows.

390 Nectanebo II 1 21 Nov 359–20 Nov 358xxvi


391 Nectanebo II 2 21 Nov 358–19 Nov 357
392 Nectanebo II 3 20 Nov 357–19 Nov 356
393 Nectanebo II 4 20 Nov 356–19 Nov 355
394 Nectanebo II 5 20 Nov 355–19 Nov 354
395 Nectanebo II 6 20 Nov 354–18 Nov 353
396 Nectanebo II 7 19 Nov 353–18 Nov 352
397 Nectanebo II 8 19 Nov 352–18 Nov 351
398 Nectanebo II 9 19 Nov 351–18 Nov 350
399 Nectanebo II 10 19 Nov 350–17 Nov 349
400 Nectanebo II 11 18 Nov 349–17 Nov 348
401 Nectanebo II 12 18 Nov 348–17 Nov 347
402 Nectanebo II 13 18 Nov 347–17 Nov 346
403 Nectanebo II 14 18 Nov 346–16 Nov 345
(continued on next page)
saite and persian egypt, 664 bc‒332 bc 281

Table (cont.)
Era of Ruler+ Length of 365-day
Nabonassar Regnal Year Egyptian Year

404 Nectanebo II 15 17 Nov 345–16 Nov 344


405 Nectanebo II 16 17 Nov 344–16 Nov 343
406 Nectanebo II 17 17 Nov 343–16 Nov 342
407 Nectanebo II 18 17 Nov 342–15 Nov 341
408 Nectanebo II 19 16 Nov 341–15 Nov 340xxvii
409 Nectanebo II 20 16 Nov 340–15 Nov 339xxvii

The following years are Egyptian years derived from the Canon (see section 3
above).

409 Artaxerxes III 20 16 Nov 340–15 Nov 339


410 Artaxerxes III 21 16 Nov 339–15 Nov 338
411 Arses 1 16 Nov 338–14 Nov 337
412 Arses 2 15 Nov 337–14 Nov 336
413 Darius III 1 15 Nov 336–14 Nov 335
414 Darius III 2 15 Nov 335–14 Nov 334
415 Darius III 3 15 Nov 334–13 Nov 333
416 Darius III 4 14 Nov 333–13 Nov 332
417 Alexander 1 14 Nov 332–13 Nov 331
and so on and so on
i
Taharqa presumably came to the throne some day in this Egyptian year. There
is a possibility that Year 1 is 12 Feb 691–11 Feb 690 (see section 2 above). No doc-
uments dating to Taharqa’s Years 1 and 2 are known (cf. Chapter III. 11, section 3).
ii
Psammetichus I’s reign presumably began in this Egyptian year. No documents
dating to his Years 1–8 are known.
iii
The earliest date for Necho II is 19 Nov 610 BC, or Year 1 Month 11 Day 1,
in stela Leyden V 18–9 (Kienitz, Geschichte, 155, with note 6, and 157–158).
iv
The latest date for Necho II is 4 May 595 BC, or Year 16 Month 4 Day 16, in
stela Louvre 193 (Kienitz, Geschichte, 155, with note 2, and 158). The earliest date for
Psammetichus II is 23 Nov 594 BC, or Year 1 Month 11 Day 9, in stela Louvre 240
(Kienitz, Geschichte, 155, with note 4, and 158).
v
Psammetichus II died on 9 Feb 589 BC, or Year 7 Month 1 Day 23. The next
day is here taken as Day 1 of Apries’s reign.
vi
The latest date for Apries is 19 Oct 570 BC, or Year 20 Month 10 Day 10, in
abnormal hieratic pBM 10113 (M. Malinine, Choix de textes juridiques, I (Paris, 1953),
17). The same document even anticipates a regnal Year 21 for Apries. The earliest
date for Amasis is 11 Jul–9 Aug 570 BC, or Year 1 Month 7, in a stela from the
northwest Delta (Kienitz, Geschichte, 158, with note 4; G. Maspero, “Sur deux stèles
récemment découvertes”, RecTrav 15 (1893), 84–86, at 86; E. Edel, “Amasis und
Nebukadnezar II”, GM 29 (1978), 13–20, at 13).
It has hitherto remained without explicit notice, as far as I know, that the latest
date for Apries is later than the earliest date for Amasis, although S. P. Vleeming (“The
Sale of a Slave in the Time of Pharaoh Py”, OMRO 61 (1980), 1–17, at 6, note 20)
rightly calls the Apries date “rather high.” It now becomes possible to buttress the
282 leo depuydt

veracity of Herodotus’s account (II, 169) about the overlap of the reigns of Apries and
Amasis. After a short period with two rival Pharaohs, Amasis first vanquished Apries
but then let him rule as coregent before eventually killing him. It is not clear whether
Apries was a rival ruler or a nominal coregent on 19 Oct 570 BC, the date of pBM
10113. It is significant that pBM 10113 is, like all abnormal hieratic texts, from Thebes,
which was Apries’s power base as rival ruler, whereas the Amasis stela mentioned
above is from the Delta, Amasis’s original power base.
vii
For the chronology of the transition from Amasis to Cambyses, see section 2
above. Amasis could still have been in power on 1 Jan 526 BC as last day of his Year
44. Greek sources mention a king Psammetichus, commonly styled as “III,” who ruled
six months between Amasis and Cambyses. The sole native documents ever assigned
to his reign have recently been re-dated to a Psammetichus “IV,” who may have ruled
part of Egypt in the 480s BC (see E. Cruz-Uribe, “On the Existence of Psammetichus
IV,” Serapis 5 (1980), 35–39; cf. also P. W. Pestman, “The Diospolis Parva Documents:
Chronological Problems concerning Psammetichus III and IV”, in: Grammata Demotika,
H.-J. Thissen & K.-Th. Zauzich, eds., (Würzburg, 1984), 145–155; S. P. Vleeming,
The Gooseherds of Hou (Pap. Hou) (Leuven: Studia demotica 3, 1991), 3–4).
viii
With this Egyptian year, Year 223 from Nabonassar, the present table joins
Ptolemy’s Canon, which earlier lists rulers of Babylon (see Chapter III. 11, section
2.1). In the Canon, whose years are all full 365-day Egyptian years, Year 223 from
Nabonassar is entirely Year 5 of Cambyses.
ix
In Ptolemy’s Canon, Year 226 from Nabonassar is entirely Year 8 of Cambyses
and Year 227 entirely Year 1 of Darius I.
x
Cambyses was still recognized in Babylon in April 522 BC and, according to the
Behistun inscription, did not die till after 1 Jul 522 BC (Parker & Dubberstein, Chronology,
14).
xi
The accession to the throne probably occurred between about 29 Sep 522 BC,
the date of Bardiya’s defeat according to the Behistun inscription, and about 22 Dec
522 BC, the date of the earliest Babylonian tablet of Darius I’s reign (Parker &
Dubberstein, Chronology, 15). Again, lunar dates are mostly not known to the exact
Julian day but cannot be more than one to two days off.
xii
It is not known whether the time from the Egyptian new year of 1 Jan 521 BC
to the Babylonian new year of about 13 Apr 521 BC was called Year 1 in Egypt or
considered part of a numberless accession year (for an argument in support of the lat-
ter, see L. Depuydt, “Regnal Years and Civil Calendar in Achaemenid Egypt;” JEA
81 (1995), 151–173, at 164). In Babylon, for sure, regnal Year 1 did not begin before
the first Babylonian new year of the reign.
xiii
In Ptolemy’ Canon, Year 262 from Nabonassar is entirely Year 36 of Darius I
and Year 263 entirely year 1 of Xerxes I.
xiv
The latest date for Darius I is about 24 Nov 486 BC (Parker & Dubberstein,
Chronology, 17). The earliest known date for Xerxes I is about 1 Dec 486 BC (M.W.
Stolper, “Babylonian Evidence for the End of the Reign of Darius I: A Correction”,
JNES 51 (1992), 61–62; cf. Depuydt, JEA 81, 157, note 22).
xv
It is not clear whether, in Egypt, the time from the Egyptian new year of 23 Dec
486 BC, to the Babylonian new year of about 5 Apr 485 BC was called Year 1 or
considered part of as numberless accession year. Cf. note xii above.
xvi
In Ptolemy’s Canon, Year 283 from Nabonassar is entirely Year 21 of Xerxes
and Year 284 entirely Year 1 of Artaxerxes I.
xvii
Xerxes I may have been murdered on one of these days. The source is an eclipse
text (Parker & Dubberstein, Chronology, 17).
xviii
This is the first day after the earliest possible date for the murder of Xerxes I
(see note xvii). The earliest date for Artaxerxes I is 2 Jan 464 BC in an Aramaic
papyrus (“B2.2” in B. Porten and A. Yardeni, Textbook of Aramaic Documents from Ancient
Egypt ( Jerusalem, 1986–99). Since the date is from Aswan, Artaxerxes I was probably
saite and persian egypt, 664 bc‒332 bc 283

in power already in 465 BC and perhaps even before the Egyptian new year of 16
Dec 465 BC. His Babylonian regnal Year 1 certainly began about 13 Apr 464 BC.
His Egyptian year may have begun earlier, but hardly before the Egyptian new year
of 17 Dec 465 BC (cf. notes xii and xix).
xix
It is not known whether the period from the Egyptian new year of 17 Dec 465
BC to the Babylonian new year of about 13 Apr 464 BC was called Year 1 in Egypt
or considered part of a numberless accession year (cf. notes xii and xviii).
xx
In Ptolemy’s Canon, Year 324 from Nabonassar is entirely Year 41 of Artaxerxes
I and Year 325 entirely Year 1 of Darius II.
xxi
For problems pertaining to the chronology of the transition from Artaxerxes I to
Darius II, see L. Depuydt, “The Date of Death of Artaxerxes I”, WdO 26 (1995),
86–96; id., JEA 81, 159, note 28. Arguments can be produced for the following pos-
sible scenario: (1) Artaxerxes I dies in Feb 424 BC; (2) ephemeral kings including
Xerxes II and Sogdianus rule for several months, while documents keep being dated
according to Artaxerxes I, whose fictional Babylonian Year 41 began about 24 Apr
424 BC; (3) Darius II assumes power some day in 25 Dec 424 BC–13 Feb 423 BC,
perhaps rather near the end of this period. Darius II’s Year 1 had certainly begun by
about 10 April 404 BC, the Babylonian new year.
xxii
In Ptolemy’s Canon, Year 343 from Nabonassar is entirely Year 19 of Darius II
and Year 344 entirely Year 1 of Artaxerxes II.
xxiii
It is certain that Artaxerxes II’s Babylonian regnal Year 1 began on about 10
April 404 BC, the Babylonian new year. The latest date for Darius II is about 17 Sep
405 BC, the date of Louvre cuneiform tablet AO 17603 (cf. Depuydt, JEA 81, 159,
n. 29). It is not known when in the period from about 17 Sep 405 BC to about 10
Apr 404 BC the transition from Darius II to Artaxerxes II happened.
xxiv
The latest date is 18 Jan 401 BC (see Chapter III. 11, section 3).
xxv
Ptolemy’s Canon continues at this point with regnal Years 6–46 of Artaxerxes
II, measured by the Egyptian calendar. The New Year’s days of these 41 full Egyptian
civil years, Years 349 to 389 from Nabonassar, are as follows: 1 Dec 400–398 BC in
years 349–51 from Nabonassar (= Artaxerxes II’s regnal Years 6–8); 30 Nov 397–4
BC in 352–5 Nab. (= 9–12); 29 Nov 393–0 BC in 356–9 Nab. (= 13–6); 28 Nov
389–6 BC in 360–3 Nab. (= 17–20); 27 Nov 385–2 BC in 364–7 Nab. (21–4);
26 Nov 381–78 BC in 368–71 Nab. (= 25–28); 25 Nov 377–4 BC in 372–5 Nab.
(= 29–32); 24 Nov 373–0 BC in 376–9 Nab. (= 33–6); 23 Nov 369–6 BC in 380–3
Nab. (= 37–40); 22 Nov 365–2 BC in 384–7 Nab. (= 41–4); 21 Nov 361–0 BC in
388–9 Nab. (= 45–6).
xxvi
It is not known when Nectanebo II assumed power in this Egyptian year.
xxvii
There is no hieroglyphic evidence for Years 19 and 20 of Nectanebo II.
II. 12 THE CHRONOLOGY OF NUBIAN KINGDOMS
FROM DYN. 25 TO THE END OF THE
KINGDOM OF MEROE

Karola Zibelius-Chen

The era of the independent kingdom of Kush in Nubia is broadly


divided into four parts:
1. Pre-Dyn. 25, i.e. the epoch of the preliterate rulers of el-Kurru with
approximately 6 reigns up to and including Alara, known from their
tombs in the necropolis of el-Kurru (ca. 885/835 BC1 to ca. 765 BC).2
2. Dyn. 25, simultaneously a segment of Egyptian history—and differing
from Manetho and Egyptological tradition—plus the Kushite Kashta
(Gen. 1) as its first king, since he secured the Thebaid by having
his daughter Amenirdis I adopted by the Egyptian God’s Wife
Shepenupet I, daughter of Osorkon III.3 Thus his daughter suc-
ceeded Shepenupet in her official role at Thebes. The Assyrians
effectively ended the dynasty, and Psammetichus I expelled the last
Kushites at the time of Tanwetamani (ca. 655 BC).
3. The Napatan Period, subdivided into an Early Napatan Period
(Atlanersa (Gen. 7) to Malowi-Amani (Gen. 19),4 ca. 653 until the
mid-5th century BC) and the Late Napatan period, from Talakhamani
and including the so-called Neo-Ramesside rulers with 5 kings up
to Sabrakamani5 (mid 5th century to early 3rd century BC).

1
See T. Kendall, Meroitica 15 (1999), 97. A long chronology (most recently L. Török,
Meroitica 14 (1999), 149ff.) no longer seems tenable after Kendall’s article.
2
Conventional dates reckoned back from the Tang-i Var inscription (see below)
with the accession of Shebitku in 707: Shabaka, 15 years (highest known date: statue
of Jty, BM 24429: Leclant, Enquêtes, 15ff.); Piankhi, 31 years inferred; Kashta (= Manetho’s
Ammeris Aithiops), 12 years assumed; Alara, 20 years presumed; plus roughly 50 to
100 years for the five reigns between Alara and the foundation of the el-Kurru necrop-
olis. (For the reinstatement of the reading Piankhi, see C. Rilly, BIFAO 101 (2001), 351ff.)
3
So R. G. Morkot, Meroitica 15 (1999), 195f., contra Kitchen, TIP 3, § 122, who
assumes that Amenirdis I was installed by her brother Piankhi.
4
I prefer this transcription to Malowiebamani (Macadam, Kawa I, 54 (25), with
reproducing Meroitic -b indicating a plural). I suppose instead that is a
phonetic determinative, derived from Egyptian w #j (Wb I, 246).
5
Contrast Hofmann, Chronologie, 64ff.
the chronology of nubian kingdoms 285

4. The Meroitic Period with the Early Meroitic period from Arkamaniqo/
Ergamenes I (Gen. 33) up to Queen Sanakadakhete (ca. early 3rd
century until late 2nd century BC), the Middle Meroitic period (late
2nd century BC to late 1st or early 2nd century AD), followed by
the Late Meroitic Period (late 1st or early 2nd century AD until the
end of the Meroitic kingdom).6
In the southern part of the kingdom, the post-Meroitic Period follows7
with off-shoots of Meroitic culture surviving the end of the Kingdom
between 330 and 370 AD (when the royal necropolis at Begerawiya
North and the West cemetery at Begerawiya were finally abandoned)8
and the Ballana culture (so-called X-Group) in Lower Nubia with royal
necropoleis at Qustul and Ballana which are beyond the chronological
range concerning us here.
This historical division into four periods ultimately depends on link-
ing the seriation of the tombs in the royal necropoleis of the kingdom
to other monuments and epigraphic data. Individual rulers are num-
bered according to their relative positions in a scheme of generations
(Gen.), especially since existing royal tombs cannot be assigned in every
case to specific named rulers, and since some known rulers cannot yet
be associated with any tomb. The cemeteries are el-Kurru (Ku.), Nuri
(Nu.), Gebel Barkal (Bar.), Begerawiya South (Beg. S., with only two
kings’ tombs but other royal and non-royal burials), Begerawiya North
(Beg. N.) and Begerawiya West (Beg. W., used by lesser members of
the royal family and commoners). The excavator Reisner9 established
a typology, based on the architectural evolution of the tombs, and he
identified tomb groups.10 Reisner’s archaeological seriation of the tombs,

6
Török, Meroe City—An ancient African capital: John Garstang’s excavations in the Sudan
(London, 1997), 13–14.
7
As viewed from a political perspective; see Török, in: Welsby, Research, 142ff.
Whether the cultural aspect of the transitional phase should be described as “post-
pyramidal” (P. Lenoble, SARS Newsletter 3 (1992), 9ff.; idem, MNL 25 (1994), 113f.;
idem, in: Welsby, Research, 157ff.), remains open.
8
The end of the Meroitic state and the abandonment of the royal cemetery in Beg.
N. ca. 360/370 AD was approximately contemporaneous with the military campaign
of the Axumite king Ezana against the Noba, unless Ezana’s conversion to Christianity
was earlier, around 330 AD. If so, the demise of Meroe will have been prolonged, from
330 to 360/70 AD. See Török’s summary of the issues, in: Welsby, Research, 142ff.
9
G. A. Reisner, “Preliminary Report on the Harvard-Boston excavations at Nûri:
the kings of Ethiopia after Tirhaqa”, HAS II (1918), 1–64; idem, SNR 2 (1919), 35–67,
237–253; idem, JEA 9 (1923), 34–77, 157–160.
10
Török summarizes Reisner’s criteria in: ANRW II 10, 169ff.
286 karola zibelius-chen

which envisoned a continuous succession of 5 plus 68 rulers, 850 BC


to 355 AD,11 was revised by Dunham when he published the excava-
tions.12 Further revisions13 were undertaken by Macadam14 and, for the
Meroitic material in particular, by Hintze15 and Wenig.16 The last also
drew on the iconography of the tomb chapel reliefs which provides
information about the tomb-owner (king, queen, prince), resulting in
the elimination of some pyramid owners from the list of reigning mon-
archs, but also on additional criteria for sequencing and thus dating
tombs. Wenig’s research eliminated Reisner’s First Collateral Meroitic
Dynasty of Napata, and integrated its rulers into the main line. The
problem of the Second Collateral Dynasty of Napata, however, cannot
be considered resolved. To include it, too, in the main line as Wenig
proposes, seems justified, by demonstrable kinship relationships between
pyramid owners in Barkal and Begerawiya North, aside from the fact
that most of the rulers of the 1st century BC (to which the tombs
belong) are attested both in the north and in Meroe. But if so, tem-
porary changes in the royal burial grounds must be assumed, which
could perhaps be explained in the context of the contemporaneous
political situation.17
While Napatan internal chronology can generally be considered reli-
able,18 despite some specific problems and questions,19 the same cannot

11
The seriation, attempted by K. Grzymski, Meroitica 7 (1984), 222ff., using com-
puter-aided quantitative analysis is generally considered unsuccessful.
12
RCK I, II, IV, V. RCK IV, 6f.: 5 plus 72 rulers.
13
On the typology for the last 20 generations, see F. Hinkel, ZÄS 108 (1981), 105ff.;
idem, ZÄS 109 (1982), 27ff.; 127ff.; idem, Bulletin du Centenaire, Supplement BIFAO 81
(1981), 379ff.
14
Kawa I and II; idem, Allen Memorial Art Museum Bulletin 23 (1966), 42–71.
15
Studien, 33 and Meroitica 1 (1973), 127–144: 5 plus 67 generations.
16
MIO 13 (1967), 1–44 and Meroitica 1 (1973), 147–160: 5 plus 73 rulers; idem,
Africa in Antiquity II (Brooklyn, 1978), 16f. and Festschrift für Steffen Wenig zum 65. Geburtstag
(Nürnberger Blätter zur Archäologie. Sonderheft Sudan, 1999), 181–182: 6 plus 76 rulers.
17
Summary: Hintze, Meroitica 1 (1973), 135ff., with discussion of the conceivable
historical scenario; see also Wenig, Meroitica 1 (1973), 152ff.
18
Even in view of new data from the field, such as the finds at Gebel Barkal, with
Bar. P-26 from the mid-6th century BC. This tomb is decorated with an astronomi-
cal ceiling: F. Berenguer, Kush 17 (1997), 108ff.; eadem, in: Welsby, Research, 55–63.
Berenguer refers to P-26 as a royal tomb. It has only two chambers, yet contains a
serekh with a name; there were no foundation deposits.
19
The difficulties concern, e.g., the assignment of Nu. 20 and the position of
Amanibakhi (Gen. 26) in the sequence; the stela from his chapel and an offering table
were discovered in Nuri (RCK II, 269, fig. 213; R. J. Leprohon, CAA Boston Museum
of Fine Arts 3, 127), but no tomb can yet be assigned to him. The sequence and relative
chronology of the Neo-Ramesside rulers are also problematic, but according to their
the chronology of nubian kingdoms 287

be said of the Meroitic material, where the situation is significantly


more complex. In a fundamental review, Hofmann underscored the
necessity of systematically utilizing Hellenistic and Roman imports in
the tomb inventories for dating purposes,20 but her work leads to a
number of controversial problems, the most important being Hofmann’s
attribution of royal pyramids to persons whom she considers ruling
kings. She bases her attibutions on offering tables with royal benedic-
tion formulas K, L, and C’, but the offering tables were not found in
the royal necropolis itself. Furthermore, royal benediction formulas could
also presumably have been used for members of the royal family.21
Török discusses the chronological issues in the context of the publica-
tion of the Fontes Historiae Nubiorum.22 A further difficulty in ascribing
pyramids to known rulers ignores the fact that some pyramids were
torn down (e.g., Beg. N. 53) or built over in antiquity.
Wenig’s proposal that stylistic elements be given more attention, and
Hofmann’s call for a more systematic examination of imports among
the funeral deposits were virtually ignored, because the poorly preserved
tomb chapels at Begerawiya have been inadequately explored archae-
ologically. The challenges presented by interdisciplinary research essen-
tial for dating the imports and evaluating their chronological and
geographical distribution in the Mediterranean are also considerable.
Nevertheless the results of the clearance, reconstruction, and docu-
mentation of the pyramids of Meroe undertaken by Friedrich Hinkel
and Janice Yellin and others, should eventually produce new evidence
relevant to chronological issues.23
The relative chronology of the royal tombs, epigraphic sources, and
other monuments furnish the foundations for establishing the sequence
of the kings of Kush. Since the mechanism of succession remains unclear,

titularies and the language of the texts (for which see now C. Peust, Das Napatanische:
ein ägyptischer Dialekt aus dem Nubien des späten ersten vorchristlichen Jahrtausends (Göttingen,
1999), 70–71) they belong at the end of the Napatan period. According to Morkot,
in: Centuries of Darkness, P. James et al., eds., (London, 1991), 216f., the Neo-Ramesside
rulers should be assigned to the period before Dyn. 25; more cautious idem, The Black
Pharaohs (London, 2000), 146–150; see further below.
20
Hofmann, Chronologie, passim (p. 192: 6 plus 69 rulers).
21
Hintze, Studien, 62–63; Rilly, MNL 28 (2001), 81.
22
See the paragraphs for the reigns under discussion in FHN I–III.
23
Hinkel began studying the pyramids in 1976; there are 169 relief scenes at his
disposal compared to 52 previously available; see Hinkel in: K. Bard, ed., Encyclopedia
of the Archaeology of ancient Egypt (London, 1999), s.v. Meroe; idem, Meroitica 7 (1984),
310ff.
288 karola zibelius-chen

and since succession from one generation to the next is paralleled by


collateral succession among brothers and cousins, it is virtually impos-
sible to project life spans and lengths of reigns from a change of sov-
ereign. Genealogies established for individual sovereigns are largely
based on circumstantial evidence or on conjecture about possible rules
of succession; they are rarely certain and apply only to specific cases.24
For example, Assyrian sources designate Tanwetamani as the son both
of Shabaka25 and of Taharqa’s sister.26 His presumed mother is Qalhata,
who should accordingly have been Taharqa’s sister and Shabaka’s con-
sort. Her titles define her only as mwt nswt, snt nswt, ˙nwt n t #-stj and
possibly z #t nswt (?),27 but not ˙mt nswt which is, however, not among
the titles of any Kushite royal mother documented to date. Either these
women were not royal wives, or the title ˙mt nswt was deemed irrele-
vant for them, possibly because it was considered a lower ranking title,
for even in those cases where the husband may have been king, he
was dead at the time his son was proclaimed king. The reconstruction
of kinship ties is rendered even more difficult since it is not known
whether the Kushites employed the terminology in a literal sense, or
whether terms such as snt “sister” and sn “brother” had broader con-
notations.28 For the period after Aspalta, who traced his maternal line
back seven generations,29 there is either very little genealogical or his-
torical information, or none at all. The evidence and conjectural data
suggest only that fully adult men ascended the throne.30

24
For a current, reliable study, see Morkot, Meroitica 15 (1999), 218–219, compar-
ing idem, in the Preprint of the 7th International Conference for Meroitic Studies, Part 3,
Appendix, 1992, 1–39 and especially Morkot’s remarks on the royal succession in Dyn.
25, Meroitica 15 (1999), 202–209.
25
Egyptologists, except M. A. Leahy, GM 83 (1984), 43ff., consistently consider
Shabaka in error for Shebitku. However, it is improbable that the Assyrians erred in
the filiation of their opponent, citing a king who had been dead for 43 years.
26
See R. Borger, Beiträge zum Inschriftenwerk Assurbanipals (BIWA), 24 (A II 22 B II
10) and 214f.
27
A. Lohwasser, Meroitica 19 (2001), 180f.
28
Cf. D. Apelt, Meroitica 12 (1990), 23ff.
29
In his coronation stela: N. Grimal, Quatre stèles napatéennes au Musée du Caire: JE
48863 et 48866 (Cairo: MIFAO 106, 1981), 30, 12–31, 3. The cartouches were all
defaced. For one possible reconstruction and the problems it entails, see Morkot, Meroitica
15 (1999), 196–200 and my review of Meroitica 15, OLZ 98 (2003), 441.
30
Arike-Amanote, e.g., states that he was 41 years of age (Kawa IX, l. 4). Tanwetamani,
Shabaka’s son, may have been in his mid-40’s. If Taharqa left Nubia in 701 as a 20
year old (Kawa V, l. 20) to participate in the battle of Eltheke (K. A. Kitchen, in:
Fontes atque pontes: eine Festgabe für Helmut Brunner (Wiesbaden: ÄUAT 5, 1983, 249ff.),
then he may have been about 31 when he ascended the throne; Sabrakamani (Kawa
XIII. l. 2) gives his age as 39 (?).
the chronology of nubian kingdoms 289

Similarly, it is not known whether the ruling sovereign influenced


the nomination of a crown prince, if indeed such an institution existed
(no source preserves a term for it).31 Kawa Stelae IV, 7ff. and V, 13ff.
state that Shebitku appreciated Taharqa more than any of the other
royal “brothers”, but this need not necessarily imply that Shebitku con-
sidered Taharqa crown prince.32 On the other hand, the Assyrian sources
report that in 671 BC, Esarhaddon captured Taharqa’s crown prince
Ushanahuru.33 But other Assyrian sources refer only to his son.34 Kushite
sources, and particularly the election stela of Aspalta, appear to testify
against the practice of designating a successor.
The only reign in more than twelve centuries of Kushite history
whose length and absolute dates are known is Taharqa’s (690–664 BC),
fifth king of Dyn. 25 and eleventh (?) sovereign of Kush. The dates
and lengths of his predecessors’ reigns, especially those of his immedi-
ate predecessor Shebitku (who had ascended the throne by 707 as the
Tang-i Var inscription shows—see further below) and the others back
to Alara (the first Kushite ruler whose name is known),35 must be reck-
oned from this point, as well as those of his successor Tanwetamani
and the later Early Napatan rulers up to Aspalta. All other dates for
Kushite rulers, particularly pre-Dyn. 25, but also those of the Napatan
period after Aspalta, are estimates. There is a single certain date for
the entire Meroitic period: 10 April 253 AD, in regnal year 3 of
Teqorideamani (Beg. N. 28).36 All other year dates of this period are
based on logical deductions from indirect testimony and historical con-
siderations, frequently deriving from Classical sources. Some of these

31
I doubt that ˙wn “youth” was such a designation (as Macadam suspected, Kawa
I, 53 [6]). Taharqa, e.g., travelled to Egypt in the middle of the ˙wnw nfrw (plural),
according to Kawa IV, l. 8. The Adoption stela, l. 4, describes Psammetichus II as jΔj
jw' Gb dm≈ ps“wy m ˙wn, and even Amasis is ˙wn nfr. Cf. also Zibelius-Chen, in:
T. Kendall, ed., Nubian Studies 1998 (Boston, 2004), 468.
32
As Macadam assumed, Kawa I, 17 (19).
33
See R. Borger, Die Inschriften Asarhaddons, AfO Beih. 9 (1967), 99.
34
In Chronicle 1 (A. K. Grayson, Assyrian and Babylonian Chronicles [Locust Valley,
1975], 85) there is only a reference to his (i.e., Taharqa’s) son and brother. See also
J. Börker-Klähn, Altvorderasiatische Bildstelen (Mainz, 1982) Nos. 217. 218. Whether the
Zincirli stela also depicts Ushanahuru is uncertain. According to Börker-Klähn, ibid.,
No. 219, the person wearing an uraeus could be Taharqa; however, Taharqa was not
himself captured by Esarhaddon.
35
For the nameless predecessors of Alara and their chronological classification, see
Kendall, Meroitica 15 (1999), 3–117.
36
Demotic graffito of Pasan (Ph. 416).
290 karola zibelius-chen

estimates are well founded while others are conjectures. As a rule,


neither the accession dates of Kushite kings nor their reign lengths are
known, and only rarely can they be inferred.
The reign of Taharqa (Gen. 5) depends upon the absolute dates of
Dyn. 26 and those of its first king, Psammetichus I, whose first regnal
year falls in 664 BC.37 The stela Louvre 192 from the burial of an
Apis bull in the Serapeum mentions that it was born in Taharqa’s year
26 and enthroned in the same year on 9/IV/peret; the bull died on
20/IV/shemu in year 20 of Psammetichus I, having lived 21 years. In
our calendar, the bull’s death occurred on February 25, 644, and its
enthronement on September 11, 665. Accordingly, Taharqa’s year 26
is the Egyptian year which lasted from February 6, 665 to February
4, 664, when he died in the course of his 27th regnal year, having
reigned 26 full years between 690 and 664 BC.38
The virtually certain sequence of Taharqa’s predecessors in Dyn. 25
is supported by the relative chronology of their pyramids, their inscrip-
tions, and their representations. Assyrian synchronisms, together with
the Tang-i Var inscription, show that Shebitku (Gen. 4) was already
king in 707, or at the latest in 706. After Yamani of Ashdod fled by
sea from the Assyrians in 711 and had dwelt “like a thief ” in “the
area of Egypt at the border by/to Kush”, he must have been handed
over to the Assyrian king Sargon II by the king of Melu¢¢a/Kush in
the year 706 at the latest. According to the Tang-i Var inscription and
the version from Malatya,39 this king is Shebitku. There is no evidence
in favour of a supposed vice-royalty in Kush with Shebitku regent for
Shabaka ruling in Egypt, nor for suggesting a coregency between the
two kings, theorized to salvage the old chronology with its “anchor
date” of 712 for the postulated campaign of Shabaka against or into

37
Beckerath, Chronologie, 41, 84–88.
38
Beckerath, Chronologie, 91. According to Beckerath, SAK 29 (2001), 1 his first year
began antedating on 12 February 690 BC.
39
G. Frame, Orientalia 68 (1999), 40 [20]; N. Na"aman, NABU 3 (1999), 63 [65].
For the chronological implications and interpretation of the historical/political events,
see B. U. Schipper, Israel und Ägypten in der Königszeit (Fribourg, 1999, OBO 170), 200ff.;
K. Dallibor, Der antike Sudan 11 (2001), 41ff.; D. Kahn, Orientalia 70 (2001), 1ff. and
above all, A. Fuchs, Die Annalen des Jahres 711 v. Chr. nach Prismenfragmenten aus Ninive
und Assur, The Neo-Assyrian Text Corpus Project (Helsinki 1998), 124ff. who reached
the same conclusion about the date of Yamani’s extradition before the Tang-i Var
inscripton was made public. Note that D. B. Redford, Orientalia 68 (1999), 58ff. and
K. A. Kitchen, in: Bietak, ed., SCIEM Haindorf 1996/98, 50f., as well as Beckerath,
SAK 29 (2001), 1ff., are sceptical of the value of these new sources.
the chronology of nubian kingdoms 291

Egypt.40 Consequently, D. Kahn rejects such ideas.41 A division of power


is documented neither for Piankhi nor Taharqa. That the Assyrians
would have referred to the junior, less powerful ruler, is improbable.
And why indeed should they not have mentioned both rulers of Egypt
and Kush in their propaganda? It is more logical to suppose that
Shebitku at his accession handed over Yamani, after he had spent some
time in Egyptian-Kushite territory. Such an act might be viewed as a
friendly gesture, since Yamani was an outlaw in Assyrian eyes, while
in Egypt he could have stirred up trouble, perhaps seeking political
support against the Assyrians. Both domestic and foreign policy could
thus have provided motives for his extradition. Simultaneously, it might
have served as a warning for the Delta princes to keep them in line.
This scenario is, however, pure speculation, since neither the place of
Yamani’s sojourn nor what he did is known. Regardless, Shebitku did
not want Yamani on his territory, and he failed to find asylum for him
elsewhere, although committing himself to the anti-Assyrian alliance at
Eltekeh shortly thereafter (701). Here, the petty rulers stood in battle
array beside the Kushite army and their allies, as the Assyrian king
Sennacherib specifically states;42 the Delta rulers were thus still to be
reckoned with, despite more than 30 years of Kushite rule in Egypt.
Shebitku’s reign can be calculated at a minimum of 17 years, 707–690.
Africanus gives him 14 and Eusebius, 12. By contrast, his highest known
date from Egypt is year 3.43 His involvement in the Near East can only
be grasped by studying Assyrian and Old Testament sources, as is also
true for Taharqa.44
The accession of Taharqa’s successor in 664 is certain. According to
the Dream Stela ll. 6f.,45 Tanwetamani (Gen. 6) proceeded to Egypt
in his first year, passing through Napata and Elephantine on the way
to reclaim the Delta which had been lost to the Assyrians in 667.
Indirect confirmation comes from Assyrian sources which refer to the
change of rulers, noting that Tanwetamani made Thebes and Heliopolis
his major bastions. Assyrian sources report that Tanwetamani was

40
So, Beckerath, SAK 29 (2001), 4–5.
41
Orientalia 70 (2001), 6–7.
42
E. Frahm, Einleitung in die Sanherib-Inschriften, AfO Beiheft 26 (1997), 59.
43
Nile Level mark No. 33 in Karnak; Kitchen, TIP 3, § 126.
44
On Assurbanipal’s battles against Taharqa and Tanwetamani, see the edition of
textual sources by Borger (n. 26), 210–215.
45
Grimal (n. 29), 7, 11f.
292 karola zibelius-chen

expelled from Thebes shortly thereafter (probably within only a mat-


ter of months).46 With the installation of Nitocris as God’s Wife in
Thebes by Psammetichus I ca. 655, Tanwetamani’s rule in Upper Egypt
ended. There is no hint of how long he continued to rule Kush. Using
Greek sources, Burstein47 plausibly argues that Tanwetamani did in fact
return once to Memphis, where he fought Psammetichus I who had
engaged the aid of Carian mercenaries. Tanwetamani was buried in
the royal necropolis at Kurru, in Ku. 16.
Psammetichus I fortified Elephantine and campaigned in LN,48 but
the identity of his Kushite opponent is not certain; chronologically, it
could be either Atlanersa or Senkamanisken.49 With the withdrawal of
the Kushites from Egypt, Nubia sank into obscurity. Until the mid-4th
century AD, which marks the end of the Meroitic kingdom, there were
at least another 61 kings, but dated monuments are associated with
few of them. Synchronisms with Egypt provide chronological fixpoints
soley for the early Meroitic Period. Otherwise, there exists only Reisner’s
relative chronology, based on the typological criteria of the royal tombs
and emended since only in some specific cases. Estimates of average
reign lengths have been made, working with the number of kings
between accepted fixpoints of the absolute chronology and by com-
parison with Egyptian data of different periods; but these remain hypo-
thetical. The proposed dates are merely suggested reference points
indicating approximate positions, by contrast to Egyptian chronology
for the LP, which is certain after 664 BC.
The sequence of kings for the generations following Tanwetamani
until Malonaqen (Gen. 12), based on archaeological seriation and inscrip-
tions, is resonably reliable. The succession Atlanersa—Senkamanisken
follows from the addition of the latter’s name to the barque stand of
Atlanersa (MFA 23.728).50 Senkamanisken also completed Temple B
700 at Barkal, begun by Atlanersa. Since the succeeding kings Anlamani51
and Aspalta52 were both children of Nasalsa, it is clear that they were

46
Borger (n. 26), 214 (B § 13, II 10–17).
47
JSSEA 14 (1984), 31ff.
48
L. Habachi, Oriens Antiquus 13 (1974), 323ff.
49
A fragment of an offering table belonging to Senkamanisken was found in Memphis:
Zibelius-Chen, Meroitica 15 (1999), 712.
50
Dunham, Barkal Temples, 32 (15); cf. also Reisner, ZÄS 66 (1930), 91–92.
51
Kawa VIII, representation in the lunette and ll. 22f.
52
Stela of Madiqen from year 3, representation in the lunette, and on the stela of
Khaliut. For the latter, see M.B. Reisner, ZÄS 70 (1934), 40 (13).
the chronology of nubian kingdoms 293

brothers. Indirect evidence correlates Aspalta with Psammetichus II who


campaigned against Nubia in 593, his 3rd regnal year.53 Traces of
destruction at Napata which affected the monuments of Kushite rulers
up to and including Aspalta (with later monuments untouched) are asso-
ciated with Psammetichus’s incursion. Ash layers were found in the cor-
responding levels of Palace 1200 and in Temple B 500. Fires were
likewise confirmed in the temple and “treasury” at Sanam, where the
latest royal name found is Aspalta’s. The chronological position of
Aspalta thus inferred54 was confirmed by the discovery of statues of
Taharqa, Tanwetamani, Senkamanisken, Anlamani and Aspalta in a
cachette in the temple precinct at Doukki Gel (Kerma),55 all showing
deliberate damage (including the removal of royal insignia).
Generations 11 to 27, following Aspalta, span about three centuries,
from the second quarter of the 6th century to the end of the 4th cen-
tury BC (ca. 570–315 BC). Burials have been confirmed for 15 of the
17 kings postulated. One tomb (Ku. 1), with a very large pyramid, is
still unattributed, as is Ku. 2, presumably the burial of the consort or
mother of Ku. 1’s owner. The temporary return from Nuri to the old
royal necropolis at el-Kurru may reflect a brief change of dynasty. King
Amanibakhi’s mortuary stela and his offering table were discovered
reused in Nuri 100,56 but the king’s pyramid has not yet been identified.
He is tentatively placed before Nastasen (Gen. 27).57 Only a few royal
monuments are known in this sequence before Nastasen, and those
citing regnal years are quite rare. The highest known year for Aspalta
is 3. For Arike-Amanote year 25+x is attested58 and from Kawa

53
For stelae of the king at Shellal, Karnak, and Tanis, see P. Der Manuelian, Living
in the Past (London 1994), 337ff. Psammetichus II ascended the throne on 19/I/595
BC (W. Barta, ZÄS 119, 1992, 89); his campaign in year 3 thus took place in 593 BC.
54
Most recently, Kendall, Kush 17 (1997), 232ff.; but cf. Török, FHN I, 230–231,
who considers the evidence insufficient for concluding that the Egyptian army reached
and destroyed Napata. Török thinks that the destruction could have resulted from a
natural disaster or local political conflicts, in which case the survival of the Khaliut
stela undamaged and in situ is remarkable. Unfortunately it bears no date. Also undam-
aged are the stela of Madiqen (dated to year 3; perhaps originally erected in Sanam),
the inscriptions on the sphinx of Defeia, and the tomb and sarcophagus of Aspalta.
Thus Aspelta could hardly have been the victim of persecution, as frequently claimed.
55
For the find, see C. Bonnet, Genava 51 (2003), 267ff.; D. Valbelle, RdE 54 (2003),
199.
56
RCK II, 269 fig. 213; for the stela, see also Leprohon (n. 19), 127ff.
57
So now also Török, FHN II, 465.
58
Kawa XII.
294 karola zibelius-chen

inscription IX, 4 it emerges that his predecessor was Talakhamani


(Nu. 16). Thus the typological sequence postulated by Reisner has
been confirmed by epigraphic evidence.59 The mention of the ruler
Malowi-Amani in Kawa IX, 12 demonstrates that he should be con-
sidered the immediate predecessor of Talakhamani.60 The annals of
Harsijotef (Gen. 23) fall in his year 35,61 and those of Nastasen (Gen.
27) in his year 8.62
Archaeological remains confirm that after the campaigns of Psam-
metichus I, LN was subject to Egyptian hegemony. Even under the
Achaemenids, in Dyn. 27, the fort at Dorginarti remained for some
time under northern control. Kushites appear as tribute bearers in the
Apadana reliefs at Persepolis under Darius I, but their costume does
not correspond to the Kushites’ own imagery. On the statue of Darius
I from Susa, t #-n˙sj is listed among the occupied countries, along with
Egypt and Libya. Cambyses is supposed to have attempted an inva-
sion of (Upper) Nubia; and Xerxes deployed an Ethiopian contingent
in his invasion of Greece.63 The inscriptions of the Kushite kings
Harsijotef and Nastasen reveal that they fought in LN, confirming that
the region was then no longer subject to northern rule. Apparently,
Kushite control extended northwards during the era of Egyptian inde-
pendence from Persian rule, Dyns. 28–30 (404–342 BC). Harsijotef
(Nu. 13) is thus placed around 404. It seems that Kushite kings also
remained active in LN during the last phase of the Second Persian
Period in Egypt, as Nastasen names ›mbswdn as the enemy in LN.64
Hintze, citing linguistic and historical arguments, equated this foe with
the Egyptian rival king Khababash,65 a proposal that strongly influenced
the chronological position assigned to Nastasen and resulted in his being
dated ca. 335–315 BC.66 Although this identification of Nastasen’s

59
Macadam’s proposal of a coregency between Talakhamani and Arike-Amanote
was justly refuted by Vinogradov, Atti I. VI Congresso Internazionale di Egittologia (Turin,
1992/93), 635ff.
60
So, too, Reisner HAS II (1919), 58.
61
L. 1: Grimal (n. 29), 42, 2.
62
L. 1: Peust (n. 19), 34.
63
L. A. Heidorn, in: W. V. Davies, ed., Egypt and Africa: Nubia from prehistory to Islam
(London, 1991), 206ff.; Morkot, in: Achaemenid History Workshop 6 (Leiden, 1991), 324ff.
64
L. 39; on the reading of the name, see most recently, Peust (n. 19), 210–211.
65
Studien, 17–20.
66
On Khababash, who should probably be dated to 343–332 BC, see recently
F. Hoffmann, Ägypten (Berlin, 2000), 87 note 145. See also Peust (n. 19), 210–211. For
the chronology of the Second Persian Period in Egypt, see D. Devauchelle, Transeuphratène
10 (1995), 35–43 (non vidi ).
the chronology of nubian kingdoms 295

opponent (presumably a LN prince), is most assuredly inaccurate,67 the


position of Nastasen in the sequence is unaffected. In terms of the gen-
eral time scale, placing him somewhere in this era seems to be entirely
reasonable, given the dates proposed for the early Meroitic king Arkamani-
qo and the insertion of the so-called Neo-Ramesside rulers after Nastasen.
The five “Neo-Ramesside” kings, Gen. 28–32, have been placed
between Nastasen, the last king to be buried in Nuri (Nu. 15), and the
first Meroitic king to be buried in Begerawiya South (Beg. S. 6). They
are Aktisanes,68 Ary(amani),69 Kash . . . meri-Amun,70 Arike-Pi(ankhi)-qo,
and Sabrakamani.71 But using archaeological criteria, it is possible to
associate only three of the five with burials, viz. the anepigraphic Barkal
pyramids Bar 11, 14, and 15, which belong between Nu. 15 and Beg.
S. 6, the first royal burial at Meroe.72 Ascribing the pyramids to specific
Neo-Ramesside kings is, however, completely hypothetical. Moreover,
these kings are viewed as a group and placed in the same period solely
on account of their names.73 Their proximity to Amanislo (Gen. 34),
buried in Beg. S. 5, and their position before Arnekhamani mrj-Jmn
(Gen. 36), the builder of the Lion temple at Musawwarat es-Sufra may,
however, be viewed as certain. Arnekhamani, buried in Beg. N. 53,
was the last king to bear the epithet mrj Jmn (see below).74 Amanislo
and Amani-tekha75 (Gen. 35) also used the epithet. However, it is not
clear whether all five Neo-Ramesside kings form a single group in
the period between Nastasen (Gen. 27) and Arkamani-qo (Gen. 33).

67
Hintze identified the last part of the name as Meroitic wte (for wte, see now Rilly,
MNL 27 [2000], 105 n. 5). It is, however, highly improbable that such a positive epi-
thet would be attached to the name of an opponent; see A. S. Spalinger, ZÄS 105
(1978), 147. Furthermore, Khababash was chiefly active in LE.
68
For Gtsn/Ktsn see K.-H. Priese, in: Fs Hintze, 343ff. Beckerath, Handbuch 2, 275,
“Atiaa-qo” is to be corrected; likewise, the mistaken data for H, G, and N names of
Arike-Amanote (Beckerath, Handbuch 2, 273). Diodorus I, 60 records an Ethiopian king
Aktisanes, who might be identical with Gtsn/Ktsn. Diodorus took the name from Hekataios
of Abdera, a contemporary of Ptolemy I, whose work must have been composed before
300 BC; thus Aktisanes should be dated to before 300, or at the very latest, before
282 BC.
69
Kendall’s identification as Alara (Meroitica 15 [1999], 64) cannot be maintained.
70
Cf. Beckerath, Handbuch 2, 275 Arkamanikasch . . . (?).
71
All but Aktisanes are documented only in Kawa. See Macadam, Kawa I, 72ff.
(XIII, XIV, XV), 90 (XLV).
72
See Wenig, MIO 13 (1967), 1ff.
73
I.e., the epithet mrj-Jmn written in the cartouche with the nomen of Kash . . .,
Ary(amani) and Sabrakamani.
74
Priese, in: Fs Hintze, 351ff. See also above, note 14.
75
Probably restored correctly as Amani-tekha mrj-Jmn by Beckerath, Handbuch 2, 277.
296 karola zibelius-chen

Peculiarities of writing and diction in the inscriptions of Sabrakamani


and Arike-Pi(ankhi)-qo led Priese to date them earlier in the Napatan
era, before Nastasen,76 whereas Hofmann places Sabrakamani and
Kash . . . amani in the Meroitic period (as Gen. 39 and 40), assigning
them the later pyramids Bar. 7 and Bar. 9 in the north group at Gebel
Barkal.77 But this would imply that these two kings temporarily revived
the tradition of using the epithet mrj-Jmn in their cartouches after a
lapse of several generations. In the absence of additional arguments,
this proposal does not seem very appealing. Positioning the Barkal pyra-
mids 7 and 9 between the necropoleis of Nuri and Begerawiya led to
the rejection of the idea of a First Collateral dynasty at Napata.
Typologically, the tombs Beg. S. 6 und Beg. S. 5 at Meroe imme-
diately follow the pyramids of the southern group at Barkal. They
belong to kings Arkamani-qo (Gen. 33) and Amanislo (Gen. 34). According
to Diodorus III, 6, the Ethiopian king Ergamenes was a contemporary
of Ptolemy II (285–246 BC) and the first to oppose the priesthood and
their custom of ritual regicide. This information resulted in the iden-
tification of Ergamenes with Arkamani-qo, and thus to establish him
approximately in the 2nd quarter of the 3rd century BC. The trans-
fer of the royal necropolis from Barkal to Meroe in Arkamani-qo’s
reign also contributed to his identification with Ergamenes, as the move
could have been related to the latter’s opposition to the priesthood.78
After Amanislo, the sequence of royal tombs at Meroe continues
in the northern area of Begerawiya. Of 41 tombs, at least two must
be assigned to princes, and one to a non-reigning queen. The uncer-
tainty of ascribing tombs in this cemetery results from the use of two-
chambered tombs for kings after the turn of the 2nd to the 1st century
BC. Previously kings owned three-chamber tombs while two-chamber
tombs were used for non-ruling members of the royal family. Since the
non-ruling members continued to be buried in two-chamber tombs, it
is difficult to identify the burial of a king when the reliefs in the tomb
chapels have been destroyed, as is frequently the case; or when tomb
stelae and offering tables are not preserved in situ. Furthermore, kings
are usually not distinguished by the use of their title qore in the

76
In: Fs Hintze, 352f.
77
Chronologie, 65–66.
78
F. Hintze, Die Inschriften des Löwentempels von Musawwarat Es Sufra (Berlin, 1962),
16f.
the chronology of nubian kingdoms 297

nomination of the funerary texts. The series of tombs in the northern


cemetery was interrupted for three generations by a move to Barkal,
but then the sequence was resumed and maintained in the north ceme-
tery until the end of the Meroitic kingdom.
Associating individual pyramids with known kings to establish a
chronological sequence is thus a particularly difficult task in this ceme-
tery. The first tomb, Beg. N. 4, belongs to Amani-tekha who, distin-
guished by the epithet mrj-Jmn in his cartouche,79 follows Amanislo
closely. The chronological position of Arnekhamani (Gen. 36), builder
of the Lion Temple at Musawwarat es-Sufra, who has been assigned
the destroyed pyramid Beg. N. 53, has been deduced from the repeated
changes of epithet in his cartouche. Mrj-Jmn he first altered to 'n¢ ≈t
mrj-Jmn, and then to 'n¢ ≈t mrj-Jst. This last epithet naming Isis is used
in Egypt only by Ptolemy IV (221–205 BC). Presumably Arnekhamani
ascended the throne during the reign of Ptolemy III (246–221 BC),
and subsequently changed his name, when Ptolemy IV became pharaoh.80
Arqamani/Ergamenes II (Gen. 37; Beg. N. 7) and Adikhalamani/
Tabirqo?81 (Gen. 38; Beg. N. 9) used the same epithet. They should
therefore be successors of Arnekhamani, especially since Arqamani might
be identical with the son of Arnekhamani who appears in the reliefs
of the Lion Temple at Musawwarat es-Sufra.82 A chronological niche
for Arqamani and Adikhalamani is easily defined, since both built in
the Dodekaschoinos in LN,83 and can only have done so (for political
and chronological reasons) during the Theban secession between 207/6
and 186 BC.

79
See above, note 75.
80
Hintze (n. 78), 14f. The Egyptian rival king Ór-wn-nfr likewise takes the epithet
'n¢-≈t mrj-Jst along with others at his coronation in Thebes, 205 BC, see W. Huß,
Ägypten in hellenistischer Zeit (München, 2001), 446 and P. W. Pestman, Chronologie égyp-
tienne d’après les textes démotiques (Leiden, 1967), 44.
81
Possibly a mortuary name of Adikhalamani.
82
Hintze, (n. 78), 25 (10). However, Hofmann, Chronologie, 57, rightly argues that
Prince Arka cannot be Arqamani (Dakke; RCK IV,
fig. D no. 24 G) because the orthography of the name is quite different, but she does
concede that his chronological position after Arnekhamani could be correct.
83
Chapel of Ergamenes in the temple of Dakke (G. Roeder, Der Tempel von Dakke,
Cairo, 1911) and chapel of Adikhalamani in Debod (G. Roeder, Debod bis Kalabsche, Cairo,
1911); see also a stela of the latter in Philae (A. Farid, MDAIK 34 [1978], 53–56). For
the alleged joint constructions of Meroites and Ptolemies in the Dodekaschoinos, see
Török, Handbook, 210–211, 428–431.
298 karola zibelius-chen

The tombs continuing the archaeological sequence in the northern


cemetery can be assigned to specific individuals only with difficulty.
Relevant epigraphic material is generally lacking, and attributions on
the basis of fragmentary and displaced offering tables can be no more
than provisional. A ruling queen must have been interred in the pyra-
mid Beg. N. 11 (Gen. 41), where the reliefs depict a woman with royal
insignia. Shards with Demotic and Meroitic cursive characters have
been found in the debris, supporting attribution of it to Sanakadakhete.
The earliest known inscription in Meroitic comes from Temple F which
she built in Naqa. An iconographic detail leads Hofmann to correlate
the pyramid with the later part (after 145 BC) of the reign of Ptolemy
VIII (170–163, 145–131, and again 127–116 BC).84 A block from a
pylon with traces of a Horus name with the component k #-n¢t comes
from Beg. N. 20 (Gen. 44). Ptolemy IX (116–107 and 88–80 BC) and
Ptolemy XII (80–57 and 55–51 BC) use k #-n¢t in their Horus names,
suggesting to Wenig that the pyramid Beg. N. 20 belongs to the Meroitic
king Taneyidamani, who is known from many monuments and must
belong to this period.85 However, Hofmann and Török consider him
the immediate successor of Sanakadakhete, and therefore assign him
pyramid Beg. N. 12 (Gen. 42).86
After Beg. N. 20, the royal family moved the burial ground to Barkal
for three generations, before returning to Begerawiya North for the
remainder of Meroitic history. Only one of the three tombs at the lat-
ter site can be assigned, viz. Bar. 6 to the ruling queen Nawidemaka.
Other identifications depend upon differing interpretations of the his-
torical events that may have led to the transfer of the necropolis.87
Since the assignment of the tomb group Beg. N. 11, 12, 13, and 20
is already hypothetical, and the sequence of kings accordingly contested,
any interpretations amount to mere circular reasoning until new mate-
rial comes to light.
Epigraphic documentation for three personalities—king Teriteqase,
Kandake Amanirenase and Paqara Akinidada—at the temple of Dakke88

84
Chronologie, 77–78.
85
MIO 13 (1967), 43.
86
Hofmann, Chronologie, 78. Török notes palaeographic details favouring Taneyidamani
for Sanakadakhete’s immediate successor, Handbook, 205; idem, FHN II, 664.
87
But not, in my opinion, an active northern policy (cf. Hofmann, Chronologie, 66ff.),
since the kings of Dyn. 25 had a decided northern policy with the seat of government
in Egypt, yet were buried in Kush.
88
REM 0092; REM 0093.
the chronology of nubian kingdoms 299

is chronologically very significant. Queen Amanirenase, who apparently


reigned after the death of Teriteqase, is attested together with Akinidada
in Kawa and in Hamadab.89 Akinidada may be the son of the one-
eyed Kandake mentioned by Strabo.90 Akinidada is also named with
the reigning queen Amanisakheto (Beg. N. 6) on a stela in Qasr Ibrim.91
These four individuals must have been not only contemporaries, but
also apparently associated with historical events linked to a more aggres-
sive northern policy, which led to the Meroitic conquest and pillage of
Philae and Syene and provoked the conflict with the Romans, resulting
in Petronius’s campaign of 23 BC. With the peace of Samos 21/20 BC,
both parties were satisfied with the Roman garrison’s withdrawal from
Qasr Ibrim while the Meroites recognized the border at Maharraqa.92
Only in 297 AD did Diocletian move the frontier back to Syene, but
exactly how the Romans ruled LN during their years of sovereignty,
and the degree to which the Meroites may have been involved, can-
not be established with certainty. Demotic graffiti in the Dodekaschoinos,
dating for the most part to the 1st and 3rd centuries AD, testify to the
activities of Meroitic officials in the region. The cult of Isis at Philae
was particularly important to them, and defensive actions against the
aggressive and marauding Blemmyes will have been a common con-
cern of Romans and Meroites. The Kharamadoye text in the Mandulis
temple at Kalabsha is the last Meroitic inscription in the region;93 it
dates to the beginning of the 5th century AD.94 Whether King Aqraka-
mani (#qrg-#mn #) should also be assigned to the period of conflict with
Petronius remains unclear.95 A Demotic inscription of a Meroite in
Dakke belongs to his reign and that of his royal mother, the pr-'#t Naytal
(Njtl ).96

89
REM 0628; REM 1039.
90
Strabo 17, 1. 53f.; see FHN III, 828ff. (190).
91
REM 1141; cf. also REM 0705. 0706 in the temple T at Kawa.
92
Summary of the historical events according to Roman sources: Welsby, The Kingdom
of Kush: the Napatan and Meroitic Empires (London, 1996) 68ff.; Török, Handbook, 448ff.
93
For the dates, see Török, FHN III, 1105f.
94
That the Kharamadoye inscription might have been written in a language other
than Meroitic (as Peust (n. 19), 75, maintains) is highly improbable. It is precisely post-
positions such as -dik or -lw (which Peust cites as indicators of a different language)
which are known Meroitic language morphemes, like the lexeme qore “king”. A glance
at the beginning of the inscription, which clearly reveals Meroitic diction as far as line
9, should have discouraged him from pursuing this proposal. For analyses of the inscrip-
tion, see Bibliographie REM 0094 and Millet, MNL 30 (2003), 57–72.
95
Török in: FHN II, 686 (161), without, however, going into A. Burckhardt’s argu-
ments in Meroitica 8 (1985), 76 and n. 12, where this graffito is assigned to the late
1st or early 2nd century AD.
96
Dak. 17; cf. Burckhardt, Meroitica 8 (1985), 99.
300 karola zibelius-chen

After Amanisakheto come the great Meroitic builders Natakamani


(Beg. N. 22) and Amanitore (Beg. N. 1) under whose reign a “Re-
Egyptianization” in art and script occurred. The new excavations of
the Berlin Egyptian Museum discovered two stelae of Amanisakheto in
the temple of Amun at Naqa,97 built by Natakamani and Amanitore,
which may necessitate a reassessment of the sequence of these rulers,
unless the two stelae actually come from an earlier building in Naqa,98
or were perhaps shifted to the Amun temple and left there. Both ste-
lae show the queen with Amesemi, once before Amesemi’s consort
Apedemak (REM 1293), and once simply together (REM 1294). Studying
details of the representations and the finds in the tombs, Hofmann99
dates Natakamani to the 2nd half of the 1st century AD.
It is uncertain, although entirely plausible, that Sorakarora, who occa-
sionally appears together with Natakamani and Amanitore and may
have initially occupied the position of a pqr(tr), should be placed in the
sequence of rulers after Natakamani. A relief at Gebel Qeili shows
Sorakarora with full royal insignia.100
Links with the Roman Empire are lacking for the following epoch,
and there are only a few kings attested in inscriptions and by offering
tables. The small number of names which belong to known rulers, and
those which have been proposed as names of rulers, are mere compi-
lations in list form. Only the pyramids Beg. N. 17, 18, and 19101 and
the late pyramid Beg. N. 28 (of Teqorideamani) can be associated by
inscriptions with owners and their sequence established beyond doubt,
following Beg. N. 16. Beg. N. 17, 18, and 19 date to the period from
somewhat after the mid-1st century AD to the mid-2nd century. How-
ever, two offering tables were found in Beg. N. 16, one belonging to

97
REM 1293 and 1294.
98
Naqa may already have been documented in the stela of Nastasen as trrqt; see
Zibelius, Afrikanische Orts- und Völkernamen in hieroglyphischen und hieratischenTexten (Wiesbaden,
1972), s.v. twrkt(t) and trrqt.
99
Chronologie, 128ff.; see also Török, Handbook, 461ff.; idem, FHN III, 898f.
100
M. Zach (GM 136 [1993], 89ff.) considers Sorakarora a rival of King Natakamani;
Hofmann (Chronologie, 128) removes him from the list of kings and assigns Beg. N. 15
to him; Török (FHN III, 910f.) considers him a ruling king since he wears a royal cos-
tume at Gebel Qeili; Zach (GM 145, 1995, 105ff.) concurs. See also Wenig (n. 16),
17; idem, Fs Wenig (n. 16), 182.
101
Amanitenamomide (Beg. N. 17), Queen Amanikhatashan (Egyptian hieroglyphs,
Beg. N. 18), Tarekeniwala (Beg. N. 19, in the debris of which was found an offering
table of Ariteneyesebokhe). Beg. N. 29 probably Takideamani, and certainly Teqoridea-
mani (Beg. N. 18).
the chronology of nubian kingdoms 301

Amanitaraqide102 and the other to Arayesebokhe. Palaeographic and


typological features allow earlier pyramid Beg. N. 16 to be ascribed to
Amanitaraqide. Arayesebokhe is dated to a later period, since his offering
table is typologically younger, and Beg. N. 36 is assigned to him, but
there are significant discrepancies in his chronological position.103 Further-
more, it is by no means clear that both these individuals were kings
at all: no other monuments have yet surfaced for either.104 The assign-
ment of pyramids in the following period also remains entirely hypo-
thetical. For the specific identification of the burials in Hofmann’s 6th
group of pyramids in Begerawiya North, which she dates from the mid-
2nd century to the mid-3rd century, she relies on offering tables with
royal benediction formulae found in the West necropolis at Begera-
wiya, on the assumption that they had been moved.105 This clearly
illustrates the difficulties and uncertainties surrounding the identifications
of rulers and their relative chronological positions.106
King Amanikhareqerema, who left us the “omphalos” (i.e., a naos) of
Napata,107 two ram statues108 and a rounded stone with his name found
at Naqa,109 has as yet no tomb at all,110 unless Beg. N. 37 (Gen. 67)111

102
His parents are Mn¢doke (B) and Pisekr (C). The facsimile REM 0816 has Pisekdo
instead (C), but it is not possible to decide which reading is correct on the basis of
the photograph. Pisekara appears in the king lists because kings’ fathers are presumed
to have been kings too, but this is by no means certain.
103
For example, Török, FHN III, 912–914 places Arayesebokhe as Gen. 56 with
Beg. N. 36 immediately after Amanitaraqide with Beg. N. 16. The same sequence
appears in Török, Handbook, 205, but with Beg. N. 36 assigned to Amanitaraqide and
Beg. N. 16 to Arayesebokhe. Hintze, Studien, 33 placed Amanitaraqide in Gen. 48 and
Arayesebokhe much later in Gen. 60, while Wenig (n. 16), 17; idem, Fs Wenig (n. 16),
182) assumes Gens. 56 and 69, and Hofmann, Chronologie, 192, assigns them, respec-
tively, to Gens. 52 and 63.
104
See Rilly, MNL 28 (2001), 81.
105
Chronologie, 155.
106
For example, inserting Maloqorebara (Meroitic chamber in Philae, REM 0101 l.
1) in the king list is fraught with difficulties; cf. Török’s discussion, FHN III, 1028f.
Even if a royal name should be expected in REM 0101 (so Török) the context of the
inscription (td¢e:Mloqorebr:qoretl§idemni . . . .) does not suggest that Maloqorebara should
be considered a king. At our present understanding of the language qoretl§idemni
/qoretala§idemani/ is to be read as qorese-l l§idemni.
107
REM 1004.
108
REM 0001 and REM 1151.
109
REM 1282, with the writing Mn¢ereqerem (contra Rilly, MNL 28 [2001], 71 n. 1),
see the figure, K. Kroeper & D. Wildung, ANM 9 (2002), 146 pl. VI a (inverted);
G. & J. Hallof, BzS 7 (2000), 169ff.; C. Carrier, MNL 27 (2000), 2 and figs. 4–5.
110
Török, Handbook, 206 (Gen. 62).
111
Wenig (n. 16), 17; idem, Fs Wenig (n. 16), 182.
302 karola zibelius-chen

or Beg. N. 41 (Gen. 56) is his.112 Proposed dates range from the


2nd half to the end of the 2nd century. When doubts arose about the
late dates for Amanikhareqerema, Rilly attempted to check the chrono-
logical position proposed for him by applying palaeographic criteria to
the inscription on the stone REM 1282.113 To avoid circular reasoning,
he relied on statistical analyses of the cursive royal inscriptions after
Natakamani. His study includes 18 documents, analysed according to
strict methodological criteria, while excluding archaeological and icono-
graphic factors. Rilly concludes that the later dates for Amanikhareqerema
should be reconsidered, since he may belong instead to the end of the
1st century.
We have almost no names for the kings reigning between Amani-
khareqerema and those of around a century later. Many princes and
queens were included in the king-list for the last centuries of the king-
dom of Meroe, none of whom reigned autonomously. In the corpus
studied by Rilly they are: Amanitaraqide, Arayesebokhe, Amanikhedolo,
Mashaqadakhela, Temelordeamani, who may have been a half-brother
of King Teqorideamani, Pat..rapeamani and Amanipilade.
A maximum list of kings after Natakamani (and Sorakarora?) would
include 23 rulers with a total of 24 pyramids. In a minimal list, only
the following individuals can be considered kings:
Amanikhareqerema, end 1st century AD
Amanitenamomide (Beg. N. 17)
Amanikhatashan (Beg. N. 18)
Takideamani (?)
Tarekeniwala (Beg. N. 19)
Ariteneyesebokhe
Teqorideamani (Beg. N. 28), accession 248/249 AD
Yesebokheamani
Teqorideamani can be dated precisely; he is named in the Demotic
graffito Ph. 416 of Pasan mentioning the third year of the ruler Tqrrmn
and year 253 (in our reckoning) under Trebonius Gallus. The king’s
name, written in Meroitic Teqorideamani, was found in his pyramid,
Beg. N. 28. We also possess an offering table REM 0829, dedicatory
texts REM 0408 through REM 0410, and the graffito REM 1261 in

112
Hofmann, Chronologie, 160.
113
MNL 28 (2001), 71ff.
the chronology of nubian kingdoms 303

Meroe inscribed with his name. He may have reigned 20 years.114 The
inscription of his official Pasan and those of other Meroites attest that
the personnel of the Isis temple was partially under the authority of
Meroitic officials. It is probably in this connection that the represen-
tations in the Meroitic chamber at Philae showing a Meroitic delega-
tion should be understood.115 One of the successors of Teqorideamani
was Yesebokheamani, who is attested in a dedicatory inscription on
Hadrian’s gate at Philae, where he is explicitely titled qore.116 He thus
belongs most probably to the period of Roman withdrawal from the
Dodekaschoinos.117 His lion statue was found in Qasr Ibrim in LN,118
and he is also known from a stela in the Apedemak temple at Meroe,
with an inscription identifying him as king.119
The sequence of pyramids in Begerawiya North ends with Beg. N.
25. This tomb belonged to a queen who is depicted wearing royal
insignia and sitting on a lion throne to receive an offering of incense.120
The pyramid Beg. N. 25 represents the last burial in the royal necrop-
olis at Begerawiya North, marking the end of the Meroitic kingdom.

114
Deduced by Hofmann, Chronologie, 168, based on Ph. 68, where year 20 of a
Meroitic king is mentioned.
115
REM 0097–0111.
116
REM 0119 and REM 0120. On qoro < qore-lo see Rilly, MNL 26 (1999), 79ff.
117
See Török, FHN III, 1050; contrast Hofmann, Chronologie, 189.
118
Not in REM. See J. M. Plumley, JEA 52 (1966), 12, pl. 4, 3. The writing can-
not be confirmed on the basis of the published photograph, but see Hallof, JEA 89
(2003), pl. 23.
119
REM 0407, l. 2 without the adjunct amni only as yesebo§e:qore.
120
RCK III, pl. 23.
II. 13 CHRONOLOGICAL LINKS BETWEEN
THE CUNEIFORM WORLD OF THE ANCIENT
NEAR EAST AND ANCIENT EGYPT

Jörg Klinger

The multiple sources available for the reconstruction of the chrono-


logical background of ancient Near Eastern history are a major factor
distinguishing these cultures from Egypt, where ultimately there is but
one single historical thread. By contrast, Mesopotamia and the neigh-
bouring regions offer a series of at least partially independent sources.
There are admittedly important differences in the distribution of these
sources over time and space such that phases with abundant material
contrast with others which are less satisfying. For certain periods, there
is simply insufficient material to draw any conclusions, regardless of the
number of different threads available.
The middle of the second millennium is such a period, and this has
led to the conundrum of three parallel chronological systems, each
based on the Venus tablet recording year 6 of Ammisaduqa of Babylon.1
These Venus observations seem to have offered the basis for the astro-
nomical calculations, since this observational event repeats itself every
56/64 years, and thus the date means that the year 6 of Ammisaduqa
can be set on a spectrum ranging from 1641 to 1577 BC. This fix
point became a pivot upon which the various chronologies turned,2

1
The texts are part of the omina series Enùma Anu Enlil; cf. esp. E. Reiner &
D. Pingree, Babylonian Planetary Omens. Part I: The Venus Tablet of Ammisaduqa (Malibu,
1975). For an evaluation of the data, and the link with the lunar dates in particular,
cf. also, V. G. Gurzadyan, “On the Astronomical Records and Babylonian Chronology”,
Akkadica 119–120 (2000), 180ff. and the literature listed there.
2
Fundamental for establishing the chronological frameworks were the work of
F. Cornelius, “Die Chronologie des Vorderen Orients im 2. Jahrtausend v. Chr.”, AfO
17 (1954–56), 294ff., with a short survey of the state of the chronological discussion
up to that time, and a plea for the “Low Chronology”, along with B. Landsberger,
“Assyrische Königsliste und ‘Dunkles Zeitalter’”, JCS 8 (1954), 31–45, 47–73, 106–133
and A. Goetze, “On the Chronology of the Second Millennium BC”, JCS 11 (1957),
53–61, 63–73, favouring the “Middle Chronology”, along with the response from
F. Cornelius, “Chronology. Eine Erwiderung”, JCS 12 (1958), 101–104. A standard
table with dates for Mesopotamian rulers following the Middle Chronology was pre-
pared by J. A. Brinkman for A. L. Oppenheim & E. Reiner, Ancient Mesopotamia. Portrait
of a Dead Civilization (Chicago, 1977), 335–348.
chronological links between the cuneiform world 305

with the later periods firmly anchored, and the earlier periods floating
freely, the date for the fall of Babylon being 1595 (according to the
Middle chronology) or 1531 BC (according to the Low chronology).3
This point has now been lost since the observations are no longer
viewed as reliable,4 with the result that the long debate about the rel-
ative merits of the Middle or Low chronologies must now be aban-
doned, having become pointless as the various alternatives for an absolute
date are no longer separated by intervals of 64 years. Instead, we now
have a moving link separating a block of 500 years5 from the later
periods where the chronology is relatively reliable.6

3
The conquest of Babylon by Murshili I is an important event in the Hittite ver-
sions, but cannot be used chronologically. The Babylonian records indicate that it took
place at the “time of Samsuditana”; Agum (II?) recovered of the plundered statues
only 24 years later. The relevant references have been collected by G. Wilhelm,
“Murshili”, RlA VIII: 434–435; on the chronological evaluation, cf. A. Goetze, JCS
11, 65–73, and for the difficulties of the Kassite kings named Agum, cf. esp. J. A.
Brinkman, Materials and Studies for Kassite History (Chicago, 1976), I, 95, 97.
4
Cf. most recently, L. Sassmannshausen, “Babylonian Chronology of the 2nd Half
of the 2nd Millennium BC”, in: H. Hunger & R. Pruzsinszky, eds., Mesopotamian Dark
Age Revisited. Proceedings of an International Conference of SCIEM 2000, Vienna 8th–9th November
2002 (Vienna, 2004), 65.
On the evaluation of the dates, cf. P. J. Huber, “Astronomy and Ancient Chronology”,
Akkadica 119–120 (2000), 160–174, who still maintains that a statistical analysis of the
possibly faulty dates is possible, and that the ensuing results which he considers to be
compatible with a date for Ammisaduqa 1 are 1582 or 1516.
5
The preservation of the very different strands from the beginning of the Ur III
period (2111–2003 BC, Middle Chronology) to the end of the Old Babylonian period
are so tightly woven that the relative chronology for the history of the first half of the
second millennium can be viewed as certain; the prospective of either larger hiati or
a significant reduction can be excluded. For the Ur III period, with its rich harvest of
chronological data, cf. W. Sallaberger “Ur III-Zeit”, in: W. Sallaberger & A. Westenholz,
Mesopotamien: Akkade-Zeit und Ur III-Zeit (Fribourg, 1999), esp. 123ff.
6
Explicitly: the use of a date 1531 or 1595 BC for the Hittite conquest of Babylon,
which is certainly relevant for the beginning of Hittite history, is today a mere con-
vention; limiting it to a period of 10–20 years is possible. Cf. the basic observation by
C. Kühne, “Imperial Mittani: An Attempt at Historical Reconstruction”, SCCNH 10
(1999), 203 n. 1, who in the meantime assumes a decade between “1550 and 1540
as a transitional decade for the end of the Old Babylonian Dynasty”. Referring to the
running dendrochronological analyses by P. I. Kuniholm (cf. also S. W. Manning,
B. Kromer, P. I. Kuniholm & M. W. Newton, “Anatolian Tree Rings and a New
Chronology for the East Mediterranean Bronze-Iron Ages”, Science 294 (2001), 2532–35;
M. W. Newton & P. I. Kuniholm, “A Dendrochronological Framework for the Assyrian
Colony Period in Asia Minor”, Türkiye Bilimler Akademisi Arkeoloji Dergisi 7 (2004), 165ff.
and the current reports of the Aegean Dendrochronology Project at http://www.arts.cor-
nell.edu/dendro/), he proposes 1541 BC +/– 37 years as a possible date for the Babylon
campaign.
306 jörg klinger

It was just a question of time until a new attempt would be made


to revise the entire chronology of the period before the first millen-
nium. The model proposed by Gasche et al. in 19987 thus gave momen-
tum to a new and intense debate which endures up to the time of
writing.8 Whether a new chronological framework similar to that pre-
vailing with the Middle Chronology will emerge remains to be seen.
New sources are still not available, with the exception of the gradual
clarification of the royal succession in the Middle Hittite period, which
can at least offer some kind of orientation around the middle of the
second millennium.9 At present, the earlier chronological proposals are
maintained as conventions.
The greatest lack in the cuneiform sources thus continues to be the
lack of adequate data10 combined with the absence of an overarching
system of year names.11
In Assyria, the traditional principle of naming years and counting
according to the holder of a certain office (lìmu) was maintained, but
the sequence of officials in the books of eponyms eliminated the need
for naming the years. This system was already established in the first
quarter of the second millennium, and became a standard dating method,
as can be seen in the eponym dates in the thousands of Old Assyrian
texts from the kàrum-period in Anatolia, and as is confirmed by the
recent discovery of two tablets with lìmu lists.12 With the overlap between

7
Gasche et al., Dating.
8
Thus a recent attempt at an ultra-long chronology, cf. C. Eder, “Assyrische
Distanzangaben und die absolute Chronologie Vorderasiens”, AoF 31 (2004), 191–236.
9
There are a whole series of astronomical observations completely independent of
these which are highly significant and could contribute to gaining a higher resolution
for the DeltaT (Deceleration of the rotation of the earth) factor which is so important
and difficult in these calculations. In addition, there are specific cases where chrono-
logically absolute dates can be checked or controlled using calculable dates. Here, we
will merely refer to the detailed investigations by Stephenson, Eclipses. Specifically rel-
evant for the Ancient Near East are chapters 4–7; cf. also below Locher, chapter III. 9.
10
Cf. A. Ungnad, “Datenlisten”, RlA II: 131–194 and Oppenheim (n. 2), 145–146.
11
An illustration of selection of such year names can be found in translation in
W.H.Ph. Römer, in: TUAT I, Fasc. 4, (Gütersloh, 1984), 337ff. For the year names,
cf., above all, the relevant article by A. Ungnad, “Eponymen”, RlA II: 412–457, and
the compilation of the literature with possible additions in Gasche et al., Dating, 47
n. 200 and the short survey in Sallaberger (n. 5), 231–237.
12
On the discovery of the lists and an initial evaluation of their chronological rel-
evance, cf. K. R. Veenhof, “Old Assyrian Chronology”, Akkadica 119–120 (2000), 137–150;
these are now published: K. R. Veenhof, The Old Assyrian List of Year Eponyms from Karum
Kanish and its Chronological Implications, Atatürk Supreme Council for Culture, Language
and History—Publications of the Turkish Historical Society Serial VI—No. 64 (Ankara,
2003. The discovery of additional texts of this genre has since been announced.
chronological links between the cuneiform world 307

these lists and the eponym chronicle from Mari, a period of almost
exactly 200 years is not absolutely precisely dated in relative terms.
This allows the end of kàrum Kanesh II to be placed into a relative
chronological context, and this permits further deductions about the
overall chronological framework of the whole period. If we allow that
the eponym chronicle ends with the death of Shamshi-Adad I, and the
final 20 or more names in the kàrum-texts cannot be identified, the end
of kàrum II will have been roughly in year 11 or 12 of Shamshi-Adad
I.13 This means that the restoration of a few more fragmentary entries
in the Assyrian king-lists contributes significantly to the second major
source for the establishment of an Ancient Near Eastern chronology.14
Yet here too, it must be conceded that the later periods are better doc-
umented and more reliable, than both the earlier era, and the neigh-
bouring regions. These lists are the spine of all dates stretching back
to the 12th century, complemented by eponym-lists (generally complete
fort he first half of the first millennium),15 and themselves anchored
absolutely with a solar eclipse, dated to 15 June 763 BC, recorded in
the eponym-lists.16 The general compatibility of the versions of the
Assyrian king-lists thus leaves only small margins in dating the reigns
of the Assyrian kings back to the 12th century.17

13
For details, cf. Veenhof, Akkadica 119–120, esp. 139 and Veenhof (n. 12), 61–62.
Veenhof dates the accession of Shamshi-Adad to the year 1833 BC, on the basis of
the Middle Chronology. Important is also his observation that the Distanzangaben of the
later Assyrian tradition relate to Irishum I and Shamshi-Adad I, meaning that these
are therefore probably realistic, effectively negating the hypothesis that the Distanzangaben
are not chronologically relevant. Following the traditional Middle Chronology, Veenhof
specifies the following Old Assyrian dates: the accession of Irishum I in 1974 BC and
the death of Shamshi-Adad I in 1776 BC.
14
On the king-lists, cf. above all, the article by A. K. Grayson, “Königslisten und
Chroniken. B. Akkadisch”, RlA VI: 86–135; by comparison, the Sumerian lists are less
useful, cf. D. O. Edzard, “Königslisten und Chroniken. A. Sumerisch. 1.5. Historischer
Wert”, RlA VI: 81–82.
15
Fundamental for the Neo-Assyrian eponyms is A. Millard, The Eponyms of the
Assyrian Empire 910–612 BC (Helsinki, 1994); for the Middle Assyrian period, H. Frey-
dank, Beiträge zur mittelassyrischen Chronologie und Geschichte (Berlin, 1991) is indispensable.
For the more recent eponym lists, cf. the literature cited by Gasche et al., Dating, 47
n. 204.
16
The course of this total solar eclipse and other data is easily accessible on Espenak’s
NASA eclipse website: http://sunearth.gsfc.nasa.gov/eclipse/. For the specifications of
this solar eclipse, cf. also Stephenson, Eclipses, 126–127.
17
For this, cf. J. Boese & G. Wilhelm, “A““ur-dàn I., Ninurta-Apil-Ekur und die mittel-
assyrische Chronologie”, WZKM 71 (1979), 19ff.; the error margin for Ashshur-resha-ishi
I (1132–1115 BC) towards the end of the 12th century is +/– 2 years.
308 jörg klinger

In comparison with the Assyrian sources, the Babylonian material is


far poorer. The Babylonians did indeed move relatively rapidly to a
more elegant and relatively more accessible means of dating, namely
simply counting the years sequentially according to the king’s reign,
which became the norm as early as the first half of the 14th century.18
However, there is a comparative dearth of written sources,19 so that
the Babylonian chronology must be established by synchronisms with
Assyrian history, where works such as the synchronic history are impor-
tant—if partial20—witnesses.
The general situation is more or less comparable in all of the other
areas, as a consistent sequential system of dates was not used anywhere,
or is at least not preserved. This means that any attempt to establish
a precise chronology for the history of the ancient Near East must rely
initially on the Assyrian data, and only then on that of their neigh-
bours, and their predecessors. Once these relative chronologies have
been established, the second step is to identify the best possible syn-
chronisms with the Assyrian chronology. Due to the multitude of sources,
and especially the diplomatic correspondence of certain periods, it is
possible to establish not merely one-to-one links, but actually to weave
a fabric of relations which can actually be established without any inter-
nal contradictions.
Arriving at absolute dates depends upon a very different and diverse
set of conditions. For the first millennium, people and events can be
dated with near precision, as the uncertainty is highly constrained. For
the earlier periods, some closed blocks can be isolated in which the
relative chronology is certain, but anchoring these to the absolute
dating of the first millennium leaves a considerable margin of error.
There is a significant contrast with respect to the precise chronologi-
cal synchronisations between Mesopotamia and Egypt for the first and
second millennia. For the first millennium BC, where the Assyrian data
is reliable and precise, the correlation difficulties lie with Egypt, whereas
before this, the reverse is true.

18
Cf. Brinkman (n. 3), 402f.; probably during the reign of Kadashman-Enlil II.
19
A. K. Grayson, “Königslisten und Chroniken. 3. Kings Lists”, RlA VI: 89ff. offers
a survey of the Babylonian sources; along with the fragments of a synchronic list, 121ff.
20
This covers the period from the middle of the 2nd millennium through Adad-
nerari III in the 8th century; edited by A. K. Grayson, Assyrian and Babylonian Chronicles
(Locust Valley, 1975).
chronological links between the cuneiform world 309

Despite the millennia during which these civilisations existed, docu-


ments confirming direct contacts and exchange between Mesopotamia
and it neighbours on the one hand, and Egypt on the other, are rather
rare and restricted to certain historical phases. Although one can assume
that at all times there will have been people who were travelling back
and forth or trading between these lands, this has left virtually no
trace in the sources. We only have access to major state activities—
diplomatic contacts or military engagements—which were generally
exceptional.21 The lucky find of the Amarna archive appears to be a
period of particularly intensive contacts, but may actually have been
repeated at very different times. The multi-facetted diplomatic archives
from Hattusha can serve as a hint that cum grano salis the same conti-
nuity of epistolary exchange may have taken place with other powers,
and even in times which appear to us quite obscure.22
Despite such obstacles, there is a general consensus among scholars
on the basic framework of the chronological framework which can be
established using the basis provided by the Assyrian records, including
also the reduction of ca. one decade in the Assyrian chronology pro-
posed some time ago.23 Another constraint is the revised model pro-
posed by H. Gasche et al., which would result in a significant reduction
even with respect to the Low chronology. Affecting the ancient Near
East up to the middle of the second millennium BC, it touches a period
for which the Assyrian sources do not provide complete cover.24

21
The seemingly comprehensive documentation in the Mari archives does not reveal
one single direct link to Egypt, cf. A. Malamat, Mari and The Early Israelite Experience
(Oxford, 1989), 61f. c. n. 125. An informative survey on the geographical horizon of
international relations in the Amorite period will be found in B. Lafont, “Relations
intenationales, alliances et diplomatie au temps des royaumes amorrites”, Amurru 2
(2001), 213ff., where Egypt likewise fails to appear.
22
The first fragment of a cuneiform letter, presumably from the diplomatic corre-
spondence of Ramesses II, was found in 2003 in the excavations at Pi-Ramesses in
the Delta, cf. E. B. Pusch & S. Jakob, “Der Zipfel des diplomatischen Archivs Ramses’
II.”, Ä&L 13 (2003), 145–153.
23
Fundamental is the contribution by Boese & Wilhelm (n. 17) as well as the fur-
ther and generally positive reception with the literature to be found by Freydank
(n. 15), 11 n. 3. Freydank (n. 15; 34) confirms that although no definitive conclusions
can be drawn from what is hitherto known from the Middle Assyrian eponyms, he
tends to assign the two kings relatively short reigns.
24
The arguments in Gasche et al., Dating, are supported not only by philological
historical sources, but fundamentally based on archaeological criteria. For a critical
evaluation, cf. G. Colbow, “Syrian Chronology in the Old and Early Middle Babylonian
Periods”, Akkadica 119–120 (2000), 103–116.
310 jörg klinger

For the period from the end of the 14th century, the dates of this
new proposal do not differ significantly from those used hitherto, as a
comparison between the Boese/Wilhelm shortened Assyrian chronol-
ogy and the “ultra-low chronology” produced by Gasche et al. shows.25
The only point up for discussion between Tiglathpileser I (1114–1076
BC) and Ashshur-nerari II (1414–1408 BC) is the reign of Ashshur-
dan I, and his predecessor Ninurta-apil-Ekur, which has led to the pro-
posed reduction of Assyrian chronology by 10 years, as mentioned.26
The interpretation of the expression tuppishu is less significant, since it
concerns only the insertion of an additional year.27
The further back in Assyrian history one goes, the greater the diver-
gences and the greater the room for differing chronological assump-
tions so that Gasche et al. favour bringing the end of the First Dynasty
of Babylon down to 1499, rather than 1531 or 1595 BC. At present,
from the philological standpoint, only the statements of the periodic
intervals (Distanzangaben) can serve as control mechanisms. Unfortunately,
there is no consensus on their value and chronological relevance. Gasche
et al. logically argue that the temporal intervals preserved as Distanzangaben
are either incorrect or they try to provide a specific interpretation for
the relevant interval.28 An inscription of Shalmaneser I concerning his

25
Compare the table in Boese & Wilhelm (n. 17), 38 with that of Gasche et al.,
Dating, 62. The work by Freydank (n. 15; 188–189), offers another survey of the reigns
of the Middle Assyrian kings following the standard dates and the shortened Assyrian
chronology. In the following, we follow the shortened chronology offered by Boese &
Wilhelm, as this provides better data when compared with Egyptian chronology.
26
For the details, cf. Boese & Wilhelm (n. 17), 23ff.; the shortening was not adopted
by Gasche et al., Dating, 51 c. n. 223, probably because the collation of the relevant
passage (cf. J. A. Brinkman, Or 42 [1973], 300 n. 15 and particularly on this, cf. Boese
& Wilhelm (n. 17; 24) in the “Nassouhi-Kinglist” (= Copy A of King List No. 9,
J. A. Brinkman, RlA VI: 101ff.; the passage is in III 41f., RlA VI: 111) was interpreted
as a confirmation although the reading—“46”—there, which Gasche et al., Dating, 51,
used their table is described as “expressively optimistic”, cited from Boese/Wilhelm
and not the actual “26+[×]” originally published by Brinkman, and cited by him again
in the RlA VI: 111.
27
Cf. the references cited by J. A. Brinkman, RlA VI: 112, who translates “his
tablet”, for Ninurta-tukulti-Ashshur and Mutakkil-Nusku; W.v. Soden, AHw, 1304b,
assumes that the corresponding entries are “short reigns of less than a year”, while
Freydank (n. 15), 33–34 assumes either a form of coregency or a distinction of the
actual exercise of power during the formally still existing reign of Ashshur-dan I,
whereby the corresponding entries in the king lists do not record an independent reigns.
Gasche et al., Dating, 53–54, likewise reckon with 0 years.
28
Cf. Gasche et al., Dating, 57: “Our analysis of the Distanzangaben makes it appar-
ent that no firm chronological conclusions can be drawn from them.” This statement
only applies to the model they propose.
chronological links between the cuneiform world 311

reconstruction of the Ekursagkurkurra specifies a period of 580 years


separated his reign from that of Shamshi-Adad I. Understood in this
fashion, this would be an obstacle to a major chronological change.29
The authors therefore propose that the passage should be interpreted
as meaning that the number “580” does not refer to the interval between
Shamshi-Adad I and the author Shalmaneser I, but also that the pre-
ceding period of 159 years between the construction of the temple by
Irishum I and its first renovation by Shamshi-Adad should be sub-
tracted from the figure,30 and therefore the 580 years thus represent
the entire period between the construction and reconstruction. This
necessarily leads to a substantial reduction in the chronology, a cen-
tury and a half in fact.31
It must be stated that this significant reduction not only entails sug-
gesting a reduction of the length of the Assyrian period of almost 200
years,32 whereby roughly a quarter of a century will have passed since
the end of the Third Dynasty of Ur,33 but also that a close link between
the dynasty ruling in Yamkhad probably allows a link between the Old
Hittite Period and the earlier periods,34 and that this simply cannot be
reduced at will.35 The links in the chronological system of the first half

29
Simply rejecting the Distanzangaben completely is apparently not appropriate as
can be surmised from the analysis of the dates by Boese & Wilhelm (n. 17), 29ff.;
indeed Veenhof ’s demonstration, based on the recently discovered Old Assyrian lìmu-
Lists, is highly significant (cf. Veenhof n. 12; 139, c. n. 7).
30
It is precisely this source, the reliability of which can be assumed; cf. the last
note.
31
For an analysis of the data in a diametrically opposed sense, cf. now Eder (n. 8),
194ff.
32
Veenhof (n. 12), 139f. indicates that the period from Irishum I, year 1, to the
death of Shamshi-Adad I was exactly 199 years, resulting from the recently discovered
lìmu-Lists allowing an overlap with the eponym-lists from Mari.
33
Cf. Veenhof (n. 12), 141.
34
Based on the genealogy beginning with Samsuiluna of Babylon, who was at least
partially contemporary with Abba"el and over the immediate descendents Yarim-Lim
II—Niqmepa—Yarim-Lim III—Hammurapi, who will have a contemporary of Hattushili
I or Murshili I.
35
In addition there is a solar eclipse listed in the Mari chronicle for the year after
the birth of Shamshi-Adad I, for which C. Michel & P. Rocher, “La chronologie du
IIe millénaire revue à l’ombre d’une éclipse de soleil,” JEOL 35–36 (1997–2000), esp.
124 proposed that the most likely match would be the eclipse of 1795 BC. Taking
the Distanzangabe separating Shalmaneser I and Shamshi-Adad I, it follows that 1234
BC + 580 years = 1814 BC, placing the accession at the age of 18, resulting in a
birth around 1832–33 which matches quite well with the solar eclipse of 24 June 1832
BC. When revising her position to take account of the recent dedrochronological infor-
mation, C. Michel (“Nouvelles données pour la chronologie”, N.A.B.U. 2002/1, 17–18)
312 jörg klinger

of the second millennium between Assyria and Babylon, with the death
of the Assyrian king Shamshi-Adad I during the second decade of the
reign of the Babylonian king Hammurabi36 ultimately allows the syn-
chronisation of Hittite history with Mesopotamian history through the
campaign of Murshili I leading to the end of the First Dynasty of
Babylon in year 31 of Samsuditana. Dating this event to the year 1499
BC creates major problems for Hittite history.37
The question of the form of the Assyrian calendar is less serious than
the differing evaluation of the Distanzangaben, but nevertheless impor-
tant. The issue is whether it depended upon lunar months and whether
or not intercalary months were used to match the solar year,38 and
thus whether or not calculations lead to a difference in the number of
years in the chronology with respect to solar years. Fortunately, there
is evidence favouring the use of solar years in Assyria during the sec-
ond millennium and not just later.39
For the present purpose, these points do not have any significant
impact since there are no direct synchronisms before the Amarna

shifted her support to the identification of this eclipse. It is difficult to believe that the
dates from such different sources could merely coincidentally converge into such a
coherent chronology. It follows that a shortening of the chronology as proposed by
Gasche et al., is excluded. Instead, this seems to suggest a chronological model slightly
higher than the “Low Chronology” and slightly lower than the “Middle Chronology”.
Michel (above) proposes shortening the “Middle Chronology” by 15 years, and some
such variant would be entirely compatible with the Hittite royal succession, and par-
ticularly that of the Middle Hittite period.
36
This applies although the precise year of the death of Shamshi-Adad I is still dis-
cussed, as it may have been in year 12, 13 or 17 of the reign of Hammurabi; cf.
Gasche et al., Dating, 64 and the literature cited in n. 265, as well as D. Charpin, in:
D. Charpin, D. O. Edzard & M. Stol, Mesopotamien: Die Altbabylonische Zeit (Fribourg:
OBO 160/4, 2004), 155–156 (n. 713), 193.
37
As is the new suggestion by Gasche et al., Dating, 64f., 77ff.; they did not attempt
to integrate the difficulties of Hittite history between Murshili I and Tutkhaliya I (cf.
Gasche et al., Dating, note 92). G. Beckman, “Hittite Chronology”, Akkadica 119–120
(2000), 19ff. discussed precisely these issues at length and came to the conclusion “that
the Anatolian evidence does not favour the suggested downward revision of the date
of the fall of Babylon”.
38
The practice is known since the Ur III period, and letters from the Old Babylonian
period record an intercalary month based on royal decrees, but the continuous usage
of some rules cannot be documented, at least partially because of the lack of sources.
In Assyria, the regular use of calibrating by using the lunar calendar is only known
from the 1st millennium, cf. H. Hunger & E. Reiner, “A Scheme for Intercalary
Months from Babylonia”, WZKM 67 (1975), 21ff. and H. Hunger, “Kalender”, RlA
V: 297–303.
39
Cf. the negative position taken by Veenhof (n. 12) 147, joining the arguments in
Gasche et al., Dating.
chronological links between the cuneiform world 313

period.40 We can thus use the dates for the Assyrian kings as proposed
by Boese/Wilhelm as the basis for the presentation of the synchronisms
with Egyptian history.
The earliest contacts are recorded in the Amarna letters41 which refer
to earlier events, and specifically to the relations between Egypt and
the Hurrian Mittani state, the most important major power in Western
Asia in the 15th century. After intense diplomatic contacts under
Amenhotep II a dynastic marriage was arranged during the reigns of
Thutmose IV and Artatama (I), which should be dated to the first
decades of the 14th century according to the Egyptian chronology. The
tradition was maintained by Shuttarna II and Amenhotep III who mar-
ried a Mittani princess in his year 10, and later a daughter of Tushratta,
who had since become ruler. In Shuppiluliuma I of Hatti, Tushratta
met a dangerous foe in the first decade of the second half of the 14th
century. In light of the complete absence of any original Hurrian or
Mittani state sources, not much more chronological information can
be gleaned from the information of the Amarna letters.42
It is only with the 14th century43 letters of the Assyrian king Ashshur-
Uballit I to Amenhotep IV (EA 15 and 16) that a real synchronism
can be won from the Amarna letters.44 There would appear to be con-
siderable room for debate since the Assyrian ruled for more than 30

40
The situation is different when considering the indirect relations which can be
adduced from the interpretation of historical developments, as in, e.g., the question of
whether the Hittite advance into northern Syria took place during a period of Mittani
dominance or was in fact favoured by Egyptian advances; cf. J. Klinger, “Synchronismen
in der Epoche vor Suppiluliuma I.—einige Bemerkungen zur Chronologie der mittel-
hethitischen Geschichte”, in: O. Carruba, Cl. Mora & M. Giorgeri, eds., Atti del II.
Congresso Internazionale di Hittitologia (Pavia: Studia Mediterranea 9, 1995), 235ff. Ultimately,
such reconstructions depend upon the chronology selected, and thus easily feed into
circular logic.
41
Cf. J. A. Knudtzon, Die El-Amarna Tafeln (Leipzig, 1915) or W. L. Moran, The
Amarna Letters (Baltimore, 1992).
42
For a summary, cf. Kühne (n. 6), and St. de Martino, “Il regno hurrita di Mittani:
profilo storico-politico”, in: La Parola del Passato LV (2000), 25ff. and with particular
reference to the Amarna correspondence, B. M. Bryan, “The Egyptian Perspective on
Mittani”, in: R. Cohen & R. Westbrook, eds., Amarna Diplomacy. The Beginnings of
International Relations (Baltimore, 2000), 71–84 and P. Artzi, “The Diplomatic Service
in Action: The Mittani File”, idem, 205–211.
43
Assyria will have freed itself from Mittani hegemony shortly before the end of
the 15th century, which allowed Ashshur-bel-nisheshu (1409–1400 BC) to renew the
alliance with Babylon, under Kara-indash.
44
Mentioning an exchange of letters between Ashshur-nadin-a¢¢e II and Amenhotep
III.
314 jörg klinger

years,45 but the contents of the first letter place it relatively near the
start of his reign (1353 BC).46 Thereafter follows a gap of several cen-
turies in the historical record of direct contacts between the two powers.
For further links between the cuneiform-using states and Egypt, we
can turn to Babylonian and Hittite sources. As preserved, the Amarna
correspondence documents contacts over several generations—as in let-
ters between the Babylonian king Kadashman-Enlil I and Amenhotep
III, and their successors Burnaburiash II and Amenhotep IV where the
Egyptian must have come to the throne during the reign of the sec-
ond Babylonian king, as the latter is in touch with both of these
pharaohs. The preserved letters contain indications that of the earlier
Kassite kings, Kara-indash had contact with Thutmose IV, and Kurigalzu
I with Amenhotep III.47 Since the reigns of the Kassite Kings before
Kadashman-Enlil are unknown these vague hints appear unpromising,
yet one can actually create a chronological framework since we know
that Kara-indash reigned at the same time as the Assyrian king Ashshur-
bel-nishu (1407–1399 BC), which puts some limits for the not partic-
ularly long reign of Thutmose IV.48
The sources are more precise for Kadashman-Enlil I49 and Burna-
buriash II,50 as the reign of the latter is known, and the synchronic

45
In the absence of detailed information about the temporal relationship between
the predecessors of Eriba-Adad I and Amenhotep III and his successor Amenhotep
IV, we can only deduce that the exchange must have taken place around 1381 BC.
46
Beckerath, Chronologie, 61 places this “shortly after his accession to the throne”,
which he dates to 1355 BC, and thus according to his chronology, dating the acces-
sion of Amenhotep IV to the end of 1351 BC, allowing an indirect synchronism, as
EA 15 does not name the pharaoh. Quite apart from this, there is sufficient time in
the reign of Amenhotep IV for the second letter to have been addressed to him, despite
the uncertainties with respect to the reading of the name.
47
Above all, Burnaburiash II to Amenhotep IV in EA 10, 8ff. The details of the
Babylonian succession add to the uncertainty, but do not have any significant impact
in this context, cf. Brinkman (n. 3), 169 and J. A. Brinkman, “Istanbul A. 1998, Middle
Babylonian Chronology, and the Statistics of the Nippur Archives”, ZA 73 (1983),
67–74. For the beginnings of the correspondence with Babylon, cf. C. Kühne, Die
Chronologie der internationalen Korrespondenz von El-Amarna (Neukirchen-Vlyn: AOAT 17,
1973), 52f. n. 244.
48
Recently, J. A. Brinkman, “Nazi-Marrutta“”, RlA 9: 190b, placed Kadashman-
Harbe I ca. 1405, and Kurigalzu I ca. 1390 BC; cf. also the chronological table by
Sassmannshausen (n. 4), esp. 67–69.
49
The assumption of a 15-year reign is not, however, compelling, as Brinkman
(n. 3), 142f. c. n. 27 specifically stresses.
50
Following Brinkman, this would be 1359–33 BC, as recently in RlA 9: 190b, i.e.,
the same dates as those in his table published in Oppenheim & Reiner (n. 2), 338. It
should be noted that Brinkman (n. 3), 32 n. 89 already observed the proposed dates
chronological links between the cuneiform world 315

history links the end of his reign with the Assyrian chronology.51 As he
actually corresponded with Amenhotep III,52 but the exchange of let-
ters only began after his 30th year,53 the entire reign of Amenhotep
IV would thus parallel that of Burnaburiash II, who must have reigned
until the accession to the throne of the successor of this pharaoh, and
thus to the year 1323 BC according to the shortened chronology.54 His
accession to office must then have taken place in 1349 BC, when
Amenhotep III would still have been reigning according to the syn-
chronism of the Amarna letters. However, some of the numbers remain
uncertain due to factors which have not yet been eliminated in the
reigns of the Kassite rulers, and thus the figures cannot be taken as
absolute limits. Discrepancies with Egyptian dates55 can easily be explained
as lying in the Babylonian sources. We must therefore confirm that the
known synchronisms between Egypt and Babylon cannot provide and
exact chronological fix points due to the imprecision inherent in the
frame of reference, or in the uncertainties in the chronological frame-
work itself, although they do place some further limits on the range of
variation.56
According to sources in internal Hittite history, contacts between the
Egyptian and Hittite kings were so intense that a treaty already clarified
relations between the two in the 15th century. Fragments of this accord—
known as the Kurushtama-treaty—are preserved,57 and its significance

ultimately depend upon the reigns of the Assyrian kings, and that aside from a few
uncertain factors mentioned there that any reduction must be accompanied by a cor-
responding change there; however, there still remains a margin of +/– 5 years.
51
For the details cf. Brinkman (n. 3), 418ff.
52
This is the most probable interpretation of EA 6 according to the collation by
Kühne (n. 47), 129 c. n. 642.
53
Cf. most recently, Beckerath, Chronologie, 66.
54
Cf. also Brinkman (n. 3), 6 n. 1.
55
The reduction in the Babylonian chronology in this case would stand in contra-
diction to the assumed accession of Amenhotep IV ca. 1351/50 BC, as Burnaburiash
would only have ascended to the throne in the following year, and thus a synchro-
nism with the older of the two pharaohs would be excluded, although precisely this
is fact reliably attested. The margin of uncertainty in this phase remains the +/– 5
years mentioned.
56
A fragmentary economic text found during the excavations at Assur in 2001 (Ass.
2001.D–2217) documents a direct synchronism between Melishipak of Babylon (ca.
1186–1172 BC following J. A. Brinkman, “Meli-shipak”, RlA VIII: 52) and Ninurta-
apil-Ekur of Assyria (for whom, cf. E. Cancik-Kirschbaum, AoF 26 [1999], 215ff.); cf.
the preliminary report by E. Frahm, MDOG 134 (2002), 75.
57
The few sources available are assembled under CTH 134. KUB 40.28 (= 134.C)
is a fragmentary text preserved in a Middle Hittite copy. A. Ünal’s (RlA VI: 373)
316 jörg klinger

recorded later in the context of the Deeds of Shuppiluliuma I,58 yet


the hints do not suffice to provide a precise context.59 According to the
sources, the only possible Hittite ruler would be Tutkhaliya I,60 as oth-
erwise the Hittite history of this period was dominated by internal
conflict with a correspondingly weak foreign policy, and thus actions
in Syria cannot be expected.61 The pharaoh who was party to this
treaty can only be traced through the Hittite sources,62 and for the
moment these are inadequate to define the period when Tutkhaliya I
reigned,63 beyond specifying that it might have been around the third

speculation that these Kurushtama-fragments could merely be part of the Egyptian cor-
respondence of Shuppiluliuma I is thus impossible.
58
D” Frag. 28 (cf. H. G. Güterbock, “The Deeds of Shuppiluliuma as Told by his
Son Murshili II”, JCS 10 (1956), 41–68, 75–98, 107ff.). A great many details are still
unresolved, and thus it is impossible to state with certainty that it merely a question
of a single treaty and not possibly a series of international accords; for this issue, cf.
D. Sürenhagen, Paritätische Staatsverträge aus hethitischer Sicht (Pavia: Studia Mediterranea,
1985), 22ff. Equally uncertain is the exact placement of those fragments 29 and 30
discussed by H. G. Güterbock of D”, which both mention Egypt.
59
Likewise belonging to the Middle Hittite period in the draught letter in Hittite,
where neither the name of the Hittite king nor that of the Pharaoh to whom it was
addressed is preserved. Given its date, the text can be assigned to the period before
Shuppiluliuma I. As preserved, it can be deduced that this letter followed another
exchange and the Hittite kings complains of a deterioration in relations which has now
been resolved. Tutkhaliya I is one possible author, but this remains pure speculation
in the absence of further information. For the text itself, cf. E. Edel, “Bo 92/129, ein
neues Brieffragment in hethitischer Sprache der Korrespondenz zwischen Ägypten und
Hatti”, ZA 86 (1996), 114–7; since published as KBo 31.40.
60
There is still some doubt about the number of kings named Tutkhaliya who
reigned in the period immediately before Shuppiluliuma I. In this case I assume the
sequence: Tutkhalija I—Arnuwanda I—Tutkhalija II—Shuppiluliuma I; cf. also simi-
larly, H. Klengel, Geschichte des Hethitischen Reiches (Leiden: HdO I. 34, 1999), 103.
61
Cf. the literature cited by Klengel (n. 60), 106f. (on source [A8]) c.n. 93f., who
likewise stresses this identification is purely hypothetical.
62
W. Helck, Die Beziehungen Ägyptens zu Vorderasien im 3. und 2. Jahrtausend v. Chr.
(Wiesbaden: ÄA 5, 19712), 166 suggested either Thutmose III or Amenhotep II, but
did not exclude Thutmose I. In any case, it should be evident that the events detailed
here are incompatible with the proposed date of 1499 BC for the campaign of Murshili
I to Babylon, as they took place more than a century later, and thus well into the
14th century. The synchronism between Idrimi and Murshili I recently proposed by
W. van Soldt, “Syrian Chronology in the Old and Early Middle Babylonian Period”,
Akkadica 119–120 (2000), 111 or between Idrimi and Zidanta I, by Kühne (n. 6), 214
n. 67 are incompatible with the chronological scheme proposed here, as is the attempt
to make a historical link between Murshili I and Thutmose I; for this, cf. F. Zeeb,
“The History of Alala¢ as a Testcase for an Ultrashort Chronology of the Mid-2nd
Millennium BCE”, in: Hunger & Pruzsinszky (n. 4), 91f.
63
The proposed reigns of some 20–25 years each for the three kings before
Shuppiluliuma I in the chronological table of the exhibition catalogue, Die Hethiter und
ihr Reich (Stuttgart, 2002), 312ff. are largely hypothetical or rely upon an historical
chronological links between the cuneiform world 317

quarter of the 15th century.64 We can, however, define the number of


generations separating Tutkhaliya I and Shuppiluliuma I, since the lat-
ter was the son of the former’s grandson (Tutkhaliya II).65 Given the
difficulties of counting generations, the internal between them should
be more than 50 and less than 100 years.66
Similarly difficult is precisely delimiting the reign of Shuppiluliuma
I,67 whose long reign will have stretched from that of Amenhotep III
until perhaps the reign of Aya.68 It is only certain that he was in direct
contact with Amenhotep IV. Whether he was also in touch with his
father depends upon identifying the pharaoh concealed behind the
apparently unreliable cuneiform transcription Hurija in EA 41 written
by Shuppiluliuma I.69 Any further attempts to establish a more exact

image dependent upon the short chronology, which simply cannot be assumed given
the very limited historical information at our disposal; this also necessarily applies to
the proposed accession date ca. 1420 BC for Tutkhalija I.
64
Dating Thutmose I to the first decade of the 15th century means that he should
be most probably understood as a contemporary of Tutkhaliya I, as the now reliable
reconstruction of the royal succession in the Middle Hittite period tends to bring
Tutkhaliya I closer to Shuppiluliuma I than was the case earlier. On the other hand,
the land donation documents definitely attest that allowance must be made for more
generations between the reign of Tutkhaliya I and the end of the Old Hittite period
(or the beginning of the Middle Hittite period, with the reign of Telpinu as the thresh-
old) than had been appreciated earlier.
65
A detailed discussion of the Middle Hittite king lists is impossible here for obvious
reasons.
66
For a detailed account of the specific problem as concerns Hittite history, cf.
G. Wilhelm, “Generation Count in Hittite Chronology”, in: Hunger & Pruzsinszky
(n. 4), 71–79; Eder (n. 8), 224ff. likewise reveals the great differences possible in the
interpretation of the material.
67
Cf. Klengel (n. 60), 147: “The precise moment of the accession cannot be fixed;”
the various possibilities were discussed by G. Wilhelm & J. Boese, “Absolute Chronologie
und die hethitische Geschichte des 15. und 14. Jahrhunderts v. Chr.”, in: Aström, Acts,
73–117, esp. 76ff. themselves pleading for a much later accession, possibly even after
the accession of Amenhotep IV, whose reign they set at 1343–1322 or 1318 BC.
68
Most recently discussed in detail by Wilhelm & Boese (n. 67) suggesting a significant
reduction in the length of the reign rather than the 40 years traditionally accepted.
This has not been accepted as the sources simply do not allow for a definitive con-
clusion, but the issue must remain open. Cf. also G. Wilhelm, “Probleme der hethiti-
schen Chronologie”, OLZ 86 (1991), cols. 471ff. The issue was most recently discussed
by J. Freu, “La chronologie du règne de Suppiluliuma: Essai de mise au point,” in:
P. Taracha, ed., Silva Anatolica (Warsaw, 2002 = Fs Popko), 87ff. He opposes a dras-
tic reduction in the reign of Shuppiluliuma I, and assumes a 30–40 year reign—
dependent upon a coregency between Amenhotep III and Amenhotep IV (esp. 107).
69
A decision depends largely upon the author’s historical interpretation, but there
is a tendency to exclude Amenhotep III, probably following Wilhelm & Boese (n. 67),
96ff.: W. L. Moran (n. 41), 114f. and Klengel (n. 60), 139 (on source [A 10]).
318 jörg klinger

chronological ordering of the historic events thus depends upon the


interpretation of the Dakhamunza-affair, that is of who made the aston-
ishing offer of a dynastic marriage to the unbelieving Hittite king—
and a generally accepted solution has still not emerged.70 The widow’s
name is not preserved in the Hittite version, and the throne name,
Nib¢ururıƒa, of the dead pharaoh is recorded in the cuneiform version
of the Deeds of Shuppiluliuma I, but this was only written down on
the orders of his son Murshili II. The question is therefore identifying
the name of the pharaoh whose widow sent the proposal to Shuppiluliuma
I.71 Can the cuneiform Nib at the beginning of the name be only an
Egyptian nb, or is a nf(r) also possible?72 According to my judgement,
in the absence of a sufficient quantity of cuneiform attestations which
could offer a corresponding orthographic principle to which one could
refer, the linguistic and orthographic arguments simply cannot be
resolved.73 Quite apart from this, there is no means of knowing whether
a Hittite copyist faced with an unfamiliar name written with what was
at that time the quite alien sign /nap/ may not have chosen to divide
the signs syllabically as NI-IB. Given that the quite adequate tran-
scription of Nb-¢prw-R' as mNi-ib-¢u-ru-ri-ja (KUB 34.24+ rev. IV
18)74 is in principle correct, we can assume that the throne name of

70
The literature here is almost endless, we cannot go into detail here, as for some
time now the Hittite sources have simply not shed any new light on the matter. The
most recent discussion of the question is that of M. Gabolde, D’Akhenaton à Toutânkhamon
(Lyon, 1998), to which we refer here. M. Gabolde offers a summary of his thesis: “Das
Ende der Amarnazeit”, in: A. Grimm & S. Schoske, eds., Das Geheimnis des goldenen
Sarges (Munich, 2001), 9–42. Basing himself on new epigraphic finds and new inter-
pretations of previously neglected (or differently interpreted) materials, he identifies the
widow as Meritaten who replaced Nefertiti near the end of the reign of Amenhotep
IV, and when he died, turned to Shuppiluliuma I. This compels him to identify
Nib¢ururiƒ as Nfr-¢prw-R' Amenhotep IV, and to assign the letter EA 9 to him. For
this, cf. M. Eaton-Krauss & R. Krauss, [review of Gabolde], BiOr 58 (2001), col. 96
and furthermore W. J. Murnane, OLZ 96 (2001), 11ff.
71
All of the preserved texts relevant to the remarkable episode were assembled by
Th.P.J. van den Hout, “Der Falke und das Kücken”, ZA 84 (1994), 60–88, esp. 61f.
72
For a long time, a reading of nb- favoured Tut'ankhamun, but this changed when
Krauss, Amarnazeit, 9ff. argued in favour of nfr-, and thus for Amenhotep IV.
73
Definitely favouring the first possibility is most recently T. Bryce, “The Death of
Niphururiya and its Aftermath”, JEA 76 (1990), 97–105, whereby he shares what can
be viewed as the Hittitological consensus, as argued by Wilhelm & Boese (n. 67), 100f.;
cf. also van den Hout (n. 71), 84f.; E. Edel, Die ägyptisch-hethitische Korrespondenz aus
Boghazköi in babylonischer und hethitischer Sprache II (Opladen, 1994), 23 n. 3; and G. Meyer,
GM 126 (1992), 87–92.
74
In contrast KBo 5.6 obv. III 7 has mPí-ip-¢u-ru-ri-ja-a; an error of PÍ for NE?
chronological links between the cuneiform world 319

Tut'ankhamun was intended, as alternative readings are difficult to jus-


tify, and actually rely upon a further series of assumptions. Assuming
this identification of the pharaoh, we can project a series of dates for
the end of the reign of Shuppiluliuma I. After the victorious campaign
with which the Hittites responded to the murder of the Hittite prince
Zannanza,75 Egyptian prisoners brought the plague to Hattusha, and
Shuppiluliuma I himself died shortly thereafter, possibly due to the epi-
demic.76 The death of Shuppiluliuma I—and the accession to the throne
of his two sons Arnuwanda II and Murshili II—can thus be tied to
the Egyptian chronology, and Shuppiluliuma I can only have lived for
2–4 years after the death of Tut'ankhamun. On the other hand, this
also enables us to verify the veracity of the interpretations due to the
solar eclipse dating to the reign of Murshili II, whose reign can thus
be given absolute dates. This means that the two systems must be
formed in a fashion which allows an accession to the throne for Murshili
II in 1321,77 since the death of Shuppiluliuma I must be limited to
either 1323–1322 or 1326–1325 BC.78 In contrast to the first men-
tioned contacts between Egypt and the Hittites, the Dakhamunzu episode
might—with all due caution—provide a date relevant to the entire
chronological discussion, allowing a link for the two independent chrono-
logical systems.
For the decades after the Amarna period, the sources on Hittite-
Egyptian relations vary widely. Although the age of Murshili II is among
the most productive in terms of overall Hittite historiography, there
would appear to be few real opening for the study of relations with
Egypt. In terms of their political calculations, the Hittites did not feel

75
The thesis argued by M. Gabolde, that Zannanza actually briefly reigned under
the throne name of Smenkhkare is not probable; cf. the sceptical judgement of Eaton-
Krauss & Krauss, BiOr 58 (2001), col. 96.
76
As noted, what can be won from the Hittite sources was assembled by van den
Hout (n. 71), 85ff.
77
Cf. the detailed discussion by Wilhelm & Boese (n. 67), 105ff. Nevertheless, the
assumption that the solar eclipse took place in the spring (ibid, pp. 106f.) is not oblig-
atory as it actually relies upon what is actually a rather dubious restoration of KUB
14.4 IV 25f. by E. Forrer, but cf. A. Götze, KlF 1, 1930, 405. Therefore, nothing
stands in the way of the total eclipse of 24 June 1312 BC.
78
Favouring the latter date is van den Hout (n. 71), 88, who arrives at 17 years
from the death of Nibkhururija = Tut'ankhamun to the solar omen (with 6 years of
the reign of Shuppiluliumas I + 1 year of Arnuwanda II + 10 years of the reign of
Murshili II), and thus assumes a death in 1325 BC; his calculations are, however,
based on the solar eclipse of 13 April 1308 BC.
320 jörg klinger

either challenged or constrained by Egypt. This only changed significantly


when Ramesses II came to the Egyptian throne, changing the politi-
cal constellation during the reign of Muwatalli II, the son and succes-
sor of Murshili II.79 The temporary removal of the Hittite capital from
Hattusha to Tarkhuntashsha had significant repercussions for the tex-
tual record such that the reign of Muwatalli is badly documented. The
lack of relevant documents can be interpreted as meaning that some
of the documentation was not returned when Hattusha became the
capital again under Muwatalli’s successor. This may be the reason why
that period of increasing contact which culminated in the battle of
Kadesh, so heavily stressed in the Egyptian tradition, appears to be
marginal in the Hittite tradition. A further consequence of our limited
understanding of this period in Hittite history is that at the pinnacle
of the Hittite-Egyptian conflict, the battle of Kadesh in year 5 of
Ramesses II, we can only specify that his opponent was Muwatalli II.
The treaty signed in the wake of the conflict in year 21 of Ramesses
II was negotiated and sealed by Hattushili III.80 Nor are there any
sources from the Hittite capital Hattusha which can allow this decisive
event to be dated, despite the abundance of letters from both king-
doms dating to before and after the signing of the treaty, and despite
the wealth of details we have for the internal history of the Hittites
during this important age.
Nevertheless, the Egyptian sources do allow some key dates, throw-
ing light on the transfer of power from Urkhi-Teshshup (= Murshili
III), whose reign was ended by the usurpation of Hattushili III during
these 16 years. It is unclear how long after year 5 of Ramesses II, that

79
The end of the reign of Murshili II can only be approximately delimited since
for the later years of the reign the preservation of the annals is significantly worse than
for the phase up to ca. regnal year 20. Nevertheless, we can be relatively certain that
Murshili remained on the throne for some three decades. A. Ünal, “Muwatalli II”,
RlA VIII: 524–527, esp. 524 assigns Muwatalli II a reign of “ca. 1290–1265 BC”, but
postulates that he either fell (or was mortally wounded) at the battle of Kadesh, and
therefore dying then or shortly thereafter, yet he likewise dates this to ca. 1274 BC
(ibid., 527)—which would be nearly a decade earlier. The source used by A. Ünal in
this connection, linking the death of the Hittite king with this campaign (citing ABoT
57 obv. 8 is probably an error, as the passage, and as KBo 4.10+ obv. 40’ff. clearly
confirms, relates to the removal of the capital and not to the death of the king), refer
only to the death without placing this in any identifiable context.
80
For the text of the treaty, cf. E. Edel, Der Vertrag zwischen Ramses II. von Ägypten
und Hattushili III. von Hatti (Berlin: WVDOG 95, 1997) prefaced (p. 1) with a short
account of the prehistory of the treaty.
chronological links between the cuneiform world 321

Urkhi-Teshshup, the direct and legitimate heir of Muwatalli II, came


to the throne as Murshili III, and how long he reigned.81 It was dur-
ing his reign that the removal of the capital was cancelled, for reasons
beyond our knowledge. The sources are silent about Murshili’s other
activities,82 and direct synchronisms with other kings cannot be proved.83
Nevertheless, a draught letter from the reign of Hattushili III to the
Assyrian king—which could be either Adad-nerari I or Shalmaneser
I84—mentions contacts between his predecessor and the Assyrian recip-
ient of the royal letter, and the correspondence between Hattushili III
and Ramesses II suggests that Murshili III was also in touch with this
Egyptian king, although these cannot be documented either. Just how
dependent the interpretations of the fragmentary letters are upon the
assumed historical background is illustrated in exemplary fashion by
the fragmentary draught of a letter (KUB 23.102). It is generally assumed
that the intended recipient was Adad-nerari I, given the clear hints at
a very tense relationship between Assur and Hatti, and the Hittite king
dismisses any equality between the two. This suggests that the letter
should be dated to a period of increasing Assyrian power, but a concrete

81
The immediate circumstances of the usurpation and its possibly not entirely unan-
imous acceptance have been frequently discussed, but the partial character of the
sources available mean that the details still remain obscure. In view of the partial
descriptions by one of the actors, namely the usurper Hattushili III, it is significant
that there is a seal impression naming Muwatalli and the later king Murshili with his
given name, which could potentially imply a relatively early attempt at securing the
succession; for a detailed survey of the entire situation, cf. most recently, Ph. Houwink
ten Cate, “Ur¢i-Tessub revisited”, BiOr 51 (1994), 239ff.
82
Summarized by Klengel (n. 60), 226ff.
83
The reason for this may be that the lack of information forthcoming from the
fragmentary sources available, e.g., the diplomatic correspondence cannot generally be
assigned to or related to Murshili III. On the general tenor of the later sources on
Murshili III, cf. also the literature cited Klengel (n. 60), 257 n. 501.
84
This is the letter KBo 1.14; cf. A. Hagenbuchner, Die Korrespondenz der Hethiter. 2.
Teil: Die Briefe mit Transkription, Übersetzung und Kommentar (Heidelberg: THeth 16, 1989),
No. 195 (obv. 15f.) and A. Harrak, Assyria and Hanigalbat (Hildesheim: TSO 4, 1987),
69ff., who pleads for Adad-nerari I while A. Ünal, Hattushili III., Teil I, Hattushili bis
zu seiner Thronbesteigung, Bd 1: Historischer Abriß (Heidelberg: THeth 3, 1974), 6, favours
Shalmaneser I. By contrast, following M. B. Rowton, “The Background of the Treaty
between Ramesses II and Hattushili III”, JCS 13 (1959), 1–11, or M. B. Rowton, “The
Material from Western Asia and the Chronology of the Nineteenth Dynasty”, JNES
25 (1966), 249–258, the only possible candidate would be Adad-nerari I. This proposal
is, however, based exclusively on an interpretation of the material based upon the con-
tent, and a corresponding reconstruction of the historical context—which is by means
the only possibility. The reference to Murshili III while avoiding the throne name
Ur¢i-Te““up only permits us to conclude that one of his successors wrote the letter.
322 jörg klinger

synchronism does not follow. It is not irrelevant that we can hardly


expect to find draught letters from the reign of Muwatalli II in the
archives of Hattusha,85 as this favours attributing the letter to Murshili
III. An uncertain but possible synchronism of the latter with Adad-
nerari I would follow, leading to a throne change before 1264 BC (the
death of Adad-nerari I). Were one to date the letter to Hattushili III,
the length of the reign of Murshili would shrink further.86 It is certain
that Murshili III came to the throne after 1275, and remained on it
until perhaps 1263 BC, but a shorter reign is more probable. The (at
least) “7 years” during which Hattushili III allegedly remained loyal87
according to his “Apology” need not be taken as a real measure of
time; although possible, it cannot be independently verified. This means
that the accession to the throne of Hattushili III cannot be determined
with precision. Neither the Egyptian chronology nor Egyptian sources,
including the correspondence of Ramesses II, offer additional aid.88 As
the synchronism with Shalmaneser I can only offer a date after 1263,
then the letter KBo 1.14 becomes interesting. If this letter is from
Hattushili III and addressed to Adad-nerari I, then—and only then—
would we be able to place the beginning of the reign of Hattushili III
in the period after 1268 and before 1264/3 BC.89
Although generally viewed as a letter from Tutkhaliya IV to Tukulti-
Ninurta I, which would provide another hook for the end of the reign

85
Following Hagenbuchner (n. 84), 263; for a discussion of the correct placing of
the text, cf. Klengel (n. 60), 204 c. n. 304.
86
A. Hagenbuchner, THeth 16, reads in obv. 16 [A-BI A-BI-J]A-ma! A-BI-JA-ja
“my father and [m]y [grandfather], however”, yet the position of the -ma would be
unusual, aside from the fact that a collation with the photo suggests that the sign
should rather be read as BA, so that the point of departure should be A-BA A-BI-JA-
ja, and thus exactly that expression used in KBo 6.28 obv. 16 by Hattushili III with
reference to his grandfather Shuppiluliuma I, during whose reign, Assyria first attempted
to free itself from Mittani hegemony. By contrast, in his 9th regnal years his succes-
sor Murshili II was forced to ward off an assault by Adad-nerari I which reached as
far as the eastern bank of the Euphrates at Carchemish; the Assyrians managed to
take Taide, Wasashatta’s capital, as mentioned in KUB 23.102 obv. I 1. Regardless,
Ur¢i-Te““up was the grandson of Murshili II.
87
Cf. H. Otten, Die Apologie Hattusilis III. Das Bild der Überlieferung (Mainz: StBoT
24, 1981), 20f.: “And given the estimation of my brother, I refrained from any actions,
and was obedient for 7 years” (rev. III 62; on the verbal form in question, cf. Most
recently J. Tischler, HEG III, 421).
88
The peace treaty with Hattushili III was sealed in regnal year 21 (i.e., 21/XI/1259
BC) of Ramesses II, and was followed by a widely publicised dynastic marriage in reg-
nal year 34.
89
Cf. Boese & Wilhelm (n. 67), 36 n. 65.
chronological links between the cuneiform world 323

of Hattushili III, the fragmentary letter KUB 26.70 cannot really shed
light on the matter. We must bear in mind that in terms of content,
this would suggest an exchange of letters between Urkhi-Teshshup/
Murshili III and Shalmaneser I, but this can nevertheless have taken
place after the usurpation by Hattushili III.90 Even using the fragments
of the diplomatic correspondence with Assur, we simply lack the nec-
essary cornerstones required to date the reigns of the Hittite kings of
the 13th century.91 Hattushili III and Tutkhaliya IV reigned during the
times of Shalmaneser I, but we cannot tell whether his 30 year reign
began before the fall of Murshili III, or only thereafter, although the
latter possibility is more probable. Tutkhaliya IV definitely reigned in
Hatti when Tukulti-Ninurta I came to the throne in Assur. Whether
he was still in office when Shuppiluliuma II became king cannot be
confirmed as nothing can delimit the length of the reign of the last
king known from the archives of Hattusha, according to which the
Hittite empire ended sometime shortly after 1200 BC.92 Nor are there
any hints at contacts to any of the rapidly disappearing kings during
the uneasy period after the murder of Tukulti-Ninurta, i.e., Ashshur-
nadin-apli to Ninurta-apil-ekur and Ashshur-dan II.
The correspondence with the Babylonian kings is ever worse pre-
served than that with the Assyrians, who only allow a few additional
points in a more precise temporal organization of the Hittite kings of

90
For this, cf. primarily, S. Heinhold-Krahmer, “Zu Salmanassars I. Eroberungen
im Hurritergebiet”, AfO 35 (1988), 94 n. 181, and now Beckerath, Chronologie NR, 26
c. n. 144 (following up on Hornung’s earlier work).
91
Little of value can be gained from the fragments (KBo 28.59–63) of a Middle
Assyrian royal letter sent to Hattusha, but one can still appreciate that the subject is
the behaviour of ”agarakti-”uria“‚ of Babylon which suggests that one can assume that
it was composed between the later part of the reign of Shalmaneser I and the begin-
ning of that of Tukulti-Ninurta I; for such a reading, cf. W. von Soden, “Weitere mit-
telassyrische Briefbruchstücke aus Hattusas”, in: E. Neu & Ch. Rüster, eds., Documentum
Asiae minoris antiquae (Mainz, Fs Otten, 1988), 346. The reference to Tutkhaliya does
suggest dating the correspondence to his reign, yet this does not allow its chronolog-
ical limits to be defined any further. Finally, it is unclear just what the reference to a
period of “100 years” signifies (KBo 28.61 rev. 9’).
92
The date from Emar for year 2 of the Babylonian king Melishipak—which cor-
responds to the year 1187 BC, cannot be used for the abandonment of the Hittite
capital Hattusha as it is by no means certain that the “Sea Peoples” were responsible
for the end of Emar, cf. now Klengel (n. 60), 318 c. n. 35. The destruction of Ugarit
thus remains the terminus ante quem, but this only allows an extension into the reign
of Siptah, i.e., until 1197 BC (cf. the literature cited by Klengel (n. 60), 318, no. 33,
as well as Ras Shamra-Ougarit XI, [1995] with the text published by S. Lackenbacher,
RSO XI, 77ff.).
324 jörg klinger

the 13th century, allowing for some synchronisms supported by the


sources. There is a letter from Kadashman-Turgu I to Hattushili III
(KUB 3.71),93 and these two would appear to have signed a treaty with
each other, as indicated by a draught letter prepared at the Hittite
court (KBo 1.10+9), but apparently during the reign of Kadashman-
Enlil II and to be assigned to Hattushili III who mentions the acces-
sion of the young Babylonian ruler in the text.94 The change of power
in the Kassite royal house is set at about 1264/63 BC, but from the
Kadesh treaty it is clear that Hattushili III came to the throne before
the year 21 of the reign of Ramesses II. Reducing this date by a decade,
following the Assyrian regnal years, would appear to be difficult, but
there nevertheless remains the margin of uncertainty, amounting to
almost a decade for the Babylonian kings, to which repeated reference
has been made, and which cannot be further reduced for the moment.

93
For this, the best discussion remains E. Edel, Ägyptische Ärzte und ägyptische Medizin
am hethitischen Königshof. Neue Funde von Keilschriftbriefen Ramses’ II aus Bo<azköy (Opladen,
1976), 123ff.
94
Above we mentioned the indirect references relating to ”agarakti-”uria“, from the
later correspondence with Tukulti-ninurta I. Whether this Babylonian king’s reign was
contemporary with that of Tutkhaliya IV (the presumed recipient of the letter) is uncer-
tain, as the relationship between the time the letter was written and the references to
Tukulti-ninurta I who conquered Babylon during the reign of Kashtiliash IV (1232–1225
BC) is not clear. It depends upon whether the events mentioned in the text took place
at that time, or whether the letter mentioned events which had taken place at some
point in the past.
PART III

ABSOLUTE CHRONOLOGY
III. 1 RADIOCARBON DATING AND
EGYPTIAN CHRONOLOGY*

Sturt W. Manning

1. Introduction: History of Field

In the beginning, the historical chronology of Egypt was held to offer


a test for the utility of the radiocarbon dating method; measurements
were thus run on several ancient Egyptian samples and the ability to
achieve ages relatively close to the historical age demonstrated that
radiocarbon dating worked in approximate terms (or was not ‘beyond
reasonable credence’).1 Egyptian samples thus comprised part of the
original ‘curve of knowns’ published in Arnold and Libby2 to show that
the radiocarbon method worked, approximately, over the last several
thousand years.
Over the next few decades a number of radiocarbon ages were
obtained on Egyptian samples. Egyptian chronology continued to be
considered as the known age, and radiocarbon was being compared—
tested. Radiocarbon technology through the 1960s was not capable of
delivering ages of sufficient accuracy or precision to be of actual util-
ity to Egyptologists.3 In 1970 Säve-Söderbergh and Olsson published a
well thought out critical analysis of radiocarbon dates from Egypt.4
They highlighted problems of poor association between samples and
presumed historical context (or age), of contamination, and of the need

* The final text of this paper was submitted 19 August 2003. The paper employs
the then standard IntCal98 radiocarbon calibration dataset. A new IntCal04 dataset has
since been published in early 2005 (Radiocarbon vol. 46(3), 2004). Use of the new dataset
would make only small and fairly insignificant changes to the figures and discussions
in this paper. For a comparison of the two calibration curves for the period 500–3500
BC, see Figure III. 1.6 at the end of this paper.
1
W. F. Libby, “Archaeology and radiocarbon dating”, Radiocarbon 22 (1980), 1017–1020.
2
J. R. Arnold & W. F. Libby, “Age determinations by radiocarbon content: checks
with samples of known age”, Science 110 (1949), 678–680.
3
H. S. Smith, “Egypt and C14 dating”, Antiquity 38 (1964), 32–37.
4
T. Säve-Söderbergh & I. U. Olsson, “C14 dating and Egyptian chronology”, in:
I. U. Olsson, ed., Radiocarbon variations and absolute chronology (Stockholm, 1970), 501–511.
328 sturt w. manning

to achieve replication and inter-laboratory checks. But Säve-Söderbergh


and Olsson also noted the uncertainties attending the historical dates,
especially those prior to the second millennium BC.
Overall, Egyptian chronology contributed positively to the develop-
ment of radiocarbon dating in the earlier decades: the apparent discrepan-
cies observed between the radiocarbon age of some third millennium
BC samples (mainly from Egypt) versus their ‘known’ age led to focussed
interest in the investigation of such anomalies.5 Such work, using espe-
cially known age tree-rings, led to the realisation that the relationship
between radiocarbon and solar (calendar) years was neither equivalent
nor fixed.6 The development of increasingly accurate and precise records
of such secular variation in natural radiocarbon levels became the dom-
inant theme in radiocarbon dating for the next generation; already by
the late 1960s to early 1970s calibration curves existed to convert radio-
carbon years to calendar years back to beyond 5000 BC.7
The advent of calibrated radiocarbon dating, which had the effect
of making many prehistoric contexts in Europe older than previously
believed, had a radical impact in prehistoric archaeology—in particu-
lar leading to the replacement of the previous ‘diffusionist’ models.8
Calibration also meant that the radiocarbon ages for Egyptian samples
needed reconsideration, and a series of papers quickly addressed the
radiocarbon dates from Egypt in the light of the initial proposals for
an approximate calibration curve.9 However, although the calibrated
ages made general sense, the radiocarbon dates continued to be of

5
W. F. Libby, “Accuracy of radiocarbon dates”, Antiquity 37 (1963), 7–12.
6
H. deVries, “Variation in the concentration of radiocarbon with time and location
on Earth”, Koninklijke Nederlandse Akademie van Wetenschappen (Series B) 61 (1958), 94–102.—
H. E. Suess, “Secular variations of the cosmic-ray-produced carbon 14 in the atmos-
phere and their interpretations”, Journal of Geophysical Research 70 (1965), 5937–5952.
7
E.g. H. E. Suess, “Bristlecone-pine calibration of radiocarbon time 5200 BC to
present”, in: Olsson (n. 4), 303–312.
8
C. Renfrew, “The tree-ring calibration of radiocarbon: an archaeological evalua-
tion”, PPS 36 (1970), 280–311.—Idem, Before civilisation. The radiocarbon revolution and pre-
historic Europe (London, 1973).
9
I. E. S. Edwards, “Absolute dating from Egyptian records and comparison with
carbon-14 dating”, Philosophical Transactions of the Royal Society of London A 269 (1970),
11–18.—R. M. Derricourt, “Radiocarbon chronology for Egypt and North Africa”,
JNES 30 (1971), 271–292.—R. M. Clark & C. Renfrew, “Tree-ring calibration of radio-
carbon dates and the chronology of ancient Egypt”, Nature 243 (1973), 265–270.—
R. D. Long, “Ancient Egyptian chronology, radiocarbon dating and calibration”, ZÄS
103 (1976), 30–48.—R. M. Clark, “Bristlecone pine and ancient Egypt: a re-appraisal”,
Archaeometry 20 (1978), 5–17.
radiocarbon dating and egyptian chronology 329

neither the accuracy nor precision to be of any real use to Egyptology;


furthermore, the routine radiocarbon technology of the time required
large sample sizes that were often problematic or impossible in terms
of acquisition from archaeological excavations or from monuments or
museums.
Over the subsequent quarter of a century radiocarbon dating has
dramatically improved in terms of accuracy, precision, and sample size
requirements.10 In tandem, the natural and anthropogenic cycles and
variations in atmospheric radiocarbon levels have become quantified in
considerable detail.11 The necessity of careful archaeological and other
analysis to ascertain the security of association between the sample to
be dated and the context for which a date is required is now fully
appreciated (seminal paper by Waterbolk 1971).12 Programmes of inter-
laboratory checking have greatly improved general standards in the
field.13 New technologies like accelerator mass spectrometry (AMS) per-
mit dating of tiny samples,14 and several routine radiocarbon labora-
tories refined accuracy and precision to what is termed ‘high-precision’
level. Today the leading high-precision laboratories can demonstrate
both good correspondence between measured ages and known real tree-
ring ages, and good agreement between the laboratories, within the
presently possible precision margins of c.2‰—that is within c.10–20
radiocarbon years for the periods discussed in this paper.15 The key

10
R. E. Taylor, Radiocarbon dating, in: R. E. Taylor & M. J. Aitken, eds., Chronometric
dating in archaeology (New York, 1997), 65–96.
11
T. F. Braziunas, I. E. Fung & M. Stuiver, “The pre-industrial atmospheric 14CO2
latitudinal gradient as related to exchanges among atmospheric, oceanic, and terres-
trial reservoirs”, Global Biochemical Cycles 9 (1995), 565–584.—M. Stuiver & T. F. Braziunas,
“Anthropogenic and solar components of hemispheric 14C”, GRL 25 (1998), 329–332.—
I. Levin & V. Hesshaimer, “Radiocarbon—a unique tracer of global carbon cycle
dynamics”, Radiocarbon 42 (2000), 69–80.
12
H. T. Waterbolk, “Working with radiocarbon dates”, PPS 37 (1971), 15–33.
13
E.g. International Study Group, “An inter-laboratory comparison of radiocarbon
measurements in tree-rings”, Nature 298 (1982), 619–623.—E. M. Scott, A. Long &
R. S. Kra, eds., “Proceedings of the international workshop on intercomparison of
radiocarbon laboratories”, Radiocarbon 32(3) (1990), 253–397. And further international
inter-comparison exercises since.
14
H. E. Gove, “The history of AMS, its advantages over decay counting: applica-
tions and prospects”, in: R. E. Taylor, A. Long & R. S. Kra, eds., Radiocarbon after
four decades: an interdisciplinary perspective (New York, 1992), 214–229.
15
E.g. M. Stuiver, P. J. Reimer, E. Bard, J. W. Beck, G. S. Burr, K. A. Hughen,
B. Kromer, G. McCormac, J. van der Plicht & M. Spurk, “INTCAL98 radiocarbon
age calibration, 24,000–0 cal BP”, Radiocarbon 40 (1998), 1041–1083.—B. Kromer,
S. W. Manning, P. I. Kuniholm, M. W. Newton, M. Spurk & I. Levin, “Regional 14CO2
offsets in the troposphere: magnitude, mechanisms, and consequences”, Science 294
330 sturt w. manning

outcome of the latter development was the creation in the mid-1980s


of a high-precision calibration of the radiocarbon timescale for the BC
period.16
Shaw17 quickly tried out high-precision calibration for existing Egyptian
samples using the Irish Oak data of Pearson et al.18 He found the cal-
ibrated ages to be in general agreement with the historical chronology,
but did not see them as able to offer a useful alternative. Shaw was
uncomfortable with the ‘wiggles’ in the calibration curve, and the sit-
uation where a given radiocarbon age could yield two or more calen-
dar age ranges. It was Hassan and Robinson19 who finally brought
methodological sophistication and chronometric hygiene to bear for
Egyptian radiocarbon dates. They reanalysed the corpus of radiocar-
bon data from Egypt against the 1986 high-precision calibration curve.
They found that with suitable samples radiocarbon often could yield
results compatible with the historical chronology;20 and they highlighted
the ability of radiocarbon to date directly a whole range of Egyptian
contexts not closely tied into written records and the chronology of the
pharaohs—a hint of the future real relevance of radiocarbon to (espe-
cially non-élite) Egyptian archaeology and its chronology. But they also
concluded that the existing corpus of radiocarbon data as of 1987 was
not, with a few exceptions, fully satisfactory—they instead looked for-
ward to better measurements in the future and then the fulfilment of
the promise of radiocarbon dating.21

(2001), 2529–2532.—F. G. McCormac, A. G. Hogg, T. G. F. Higham, J. Lynch-


Stieglitz, W. S. Broecker, M. G. L. Baillie, J. Palmer, L. Xiong, J. R. Pilcher, D. Brown
& S. T. Hoper, “Temporal variation in the interhemispheric 14C offset”, GRL 25 (1998),
1321–1324.
16
G. W. Pearson, J. R. Pilcher & M. G. L. Baillie, “High precision 14C measure-
ments of Irish oaks to show the natural 14C variations from 200BC to 4000BC”,
Radiocarbon 25 (1983), 179–186.—G. W. Pearson, & M. Stuiver, “High-precision cali-
bration of the radiocarbon time scale”, 500–2500 BC, Radiocarbon 28 (1986), 839–862.—
M. Stuiver & G. W. Pearson, “High-precision calibration of the radiocarbon time scale,
AD 1950–500 BC”, Radiocarbon 28 (1986), 805–838.
17
I. M. Shaw, “Egyptian chronology and the Irish Oak calibration”, JNES 44 (1985),
295–317.
18
Pearson et al. (n. 16).
19
F. A. Hassan & S. W. Robinson, “High Precision radiocarbon chronometry of
ancient Egypt, and comparisons with Nubia, Palestine and Mesopotamia”, Antiquity 61
(1987), 119–135.
20
See also B. Weninger, “Theoretical radiocarbon discrepancies”, in: D. A. Hardy
& A. C. Renfrew, eds., Thera and the Aegean world III. 3: Chronology (London, 1990),
216–231.—B. Weninger, Studien zur dendrochronologischen Kalibration von archäologischen
14
C-Daten (Bonn, 1997).
21
Cf. (n. 19) at p. 129.
radiocarbon dating and egyptian chronology 331

But sadly there have been at best limited attempts to provide such
better quality radiocarbon data for the periods after the Archaic—where
increasingly good data exist.22 Instead, publications by leading Egyptian
chronological specialists concerned with the 3rd through earlier 1st mil-
lennia BC in the late 1990s through 2003 largely dismissed or ignored
radiocarbon evidence;23 with Kitchen stating that ‘“science” cannot
solve the intricate problems of detailed Egyptian successions, and the
cross-links with the neighbouring Near East; texts alone can do that’.24
Such scholars cannot see any use for radiocarbon dating versus the
believed-in dating accuracy and precision available from textual evidence.
Ironically, the potential modern relevance of radiocarbon to Egyptology
has been brought to the fore by a set of publications in the 1990s,
which sought to question and reject the standard chronological syn-
thesis and instead to propose a radically different (lower) Egyptian
chronology for the second and earlier first millennia BC.25 These writers
appreciated that radiocarbon dating offered an independent check on
their claims—they thus sought to dismiss or downplay radiocarbon dat-
ing evidence.26 In reality, however, radiocarbon evidence from the east
Mediterranean indicated the reverse: that the range of the standard
chronology was correct.27 Radiocarbon perhaps had a use after all for
Egyptian chronology.

22
J. Görsdorf, G. Dreyer & U. Hartung, “14C dating results of the Archaic Royal
Necropolis Umm el-Qaab at Abydos”, MDAIK 54 (1998), 169–175.—S. H. Savage,
“AMS 14Carbon Dates from the Predynastic Egyptian Cemetery, N7000, at Naga-ed-
Dêr”, JAS 25 (1998), 235–249. Cf. A. R. Millard & T. A. H. Wilkinson, “Comment
on ‘AMS radiocarbon dates from the Predynastic Egyptian Cemetery, N7000, at Naga-
ed-Dêr’ by S. H. Savage”, JAS 26 (1999), 339–341.—S. H. Savage, “Towards an AMS
radiocarbon chronology of Predynastic Egyptian ceramics”, Radiocarbon 43 (2001),
1255–1277.
23
E.g. Beckerath, Chronologie, 55–56.
24
K. A. Kitchen, “Ancient Egyptian chronology for Aegeanists”, Mediterranean Archaeology
and Archaeometry 2.2 (2002), 5–12 at p. 11.
25
P. James, I. J. Thorpe, N. Kokkinos, R. Morkot & J. Frankish, Centuries of darkness
(London, 1991).—D. M. Rohl, A test of time. 1: The Bible—from myth to History (London,
1995).
26
James et al. (n. 25), 321–325.—P. James, N. Kokkinos & I. J. Thorpe, “Mediterranean
chronology in crisis”, in: M. S. Balmuth & R. H. Tykot, Sardinian and Aegean chronol-
ogy: towards the resolution of relative and absolute dating in the Mediterranean (Oxford: Studies
in Sardinian Archaeology V, 1998), 29–43.—Rohl (n. 25), 384–387.
27
E.g. S. W. Manning & B. Weninger, “A light in the dark: archaeological wiggle
matching and the absolute chronology of the close of the Aegean Late Bronze Age”,
Antiquity 66 (1992), 636–663.—S. W. Manning, B. Weninger, A. K. South, B. Kling,
P. I. Kuniholm, J. D. Muhly, S. Hadjisavvas, D. A. Sewell & G. Cadogan, “Absolute
age range of the Late Cypriot IIC period on Cyprus”, Antiquity 75 (2001), 328–340.
332 sturt w. manning

2. Radiocarbon Dating and the Historical Timescale

The problem, historically, is that earlier radiocarbon dating offered at


best large possible date ranges for any given measurement and these
dates thus seemed an order of magnitude less accurate or precise than
those available from the historical chronology of Egypt (the point made
by von Beckerath).28 And, as the review of Weinstein showed,29 up until
the early 1980s it is true that radiocarbon simply lacked the ability to
supply the precision required in calendar years to be relevant to the
existing, quite refined, historical chronology for the ancient Near East.
But the advent of high-precision calibration curves from the mid-1980s,
and increased accuracy and precision for standard radiocarbon dates,
dramatically changed the situation. It was now possible to approach
the precision of the historical chronology, and radiocarbon dating could
thus offer an independent chronology free from the assumptions and
step-wise logic transfers inherent in the existing chronological synthesis
for Egypt and, there from, for the whole east Mediterranean.30
Recent developments emphasise this position. Following the first inter-
nationally recommended high-precision calibration curves of 1986, a
second internationally recommended extension, refinement, and revi-
sion was made available in 199831—and another (IntCal04) has been
published while this text was in press.32 Radiocarbon calibration datasets
from tree-ring records from the east Mediterranean33 have confirmed
the local relevance of the standard northern hemisphere calibration for
most periods (such work has inaddition identified some intervals of pos-
sible regional/temporal variations for further study linked to key periods
of short-term solar irradiance minima and climate change issues).34 The

28
Beckerath, Chronologie, 56.
29
J. M. Weinstein, “Radiocarbon dating in the southern Levant”, Radiocarbon 26
(1984), 297–366.
30
H. J. Bruins & W. G. Mook, “The need for a calibrated radiocarbon chronol-
ogy of Near Eastern archaeology”, Radiocarbon 31 (1989), 1019–1029.
31
Stuiver et al. (n. 15).
32
P. J. Reimer et al., “Radiocarbon calibration from 0–26 cal kyr BP”, Radiocarbon
46, 3 (2004), 1029–1058.
33
Kromer et al. (n. 15).—S. W. Manning, B. Kromer, P. I. Kuniholm & M. W.
Newton, “Anatolian tree-rings and a new chronology for the east Mediterranean Bronze-
Iron Ages”, Science 294 (2001), 2532–2535.—S. W. Manning, M. Barbetti, B. Kromer,
P. I. Kuniholm, I. Levin, M. W. Newton & P. J. Reimer, “No systematic early bias
to Mediterranean 14C ages: radiocarbon measurements from tree-ring and air samples
provide tight limits to age offsets”, Radiocarbon 44(3) (2002), 739–754.
34
See B. Kromer, M. Korfmann & P. Jablonka, “Heidelberg radiocarbon dates for
radiocarbon dating and egyptian chronology 333

development and application of stratified archaeological ‘wiggle-match-


ing’ techniques have in turn allowed the exploitation of both (i) refined
archaeological knowledge (stratigraphy) and (ii) the now refined and
specific history of past natural atmospheric levels of radiocarbon entailed
in these calibration curves in order to yield accurate and highly pre-
cise calendar age ranges for sets of seriated samples.35 Radiocarbon dat-
ing has thus moved from a resolution of century-scale at best, to now
being capable of decadal scale resolution. Radiocarbon can thus now
have relevance at the ‘historical’ timescale.
This does not mean that everything is now simple and clear; there
remains plenty of scope for ambiguities and inconsistencies—as illus-
trated in several of the papers in the recent Bruins et al. volume in
Radiocarbon.36 In particular, without selection of directly relevant sam-
ples from primary contexts (e.g. short- to shorter-lived samples from
secure and specific archaeological contexts relevant to the archaeological
event/phase for which a date is sought), and then proper pre-treatment,
and accurate and precise measurement in the laboratory, nothing use-
ful will be gained. Old wood is clearly a major problem with some
Egyptian samples37 (see Section 4 below). The need for quality control
at radiocarbon laboratories in terms of known-age blind checks is widely
appreciated these days. Attention is increasingly moving now to the
consideration of the integrity of the sample itself as offering only the

Troia I to VIII and Kumtepe”, in: G. A. Wagner, E. Pernicka & H.-P. Uerpmann,
eds., Troia and the Troad: scientific approaches (Berlin, 2003), 52–53.
35
E.g. Manning and Weninger (n. 27).—C. E. Buck, J. A. Christen & G. N. James,
“BCal: an on-line Bayesian radiocarbon calibration tool”, Internet Archaeology 7 (1999),
[http://intarch.ac.uk/journal/issue7/buck/]—C. E. Buck, C. D. Litton & S. J. Shennan,
“A case study in combining radiocarbon and archaeological information: the early
Bronze-Age settlement of St. Veit-Klinglberg, Land Salzburg, Austria”, Germania 72
(1994), 427–447.—B. Weninger, “Stratified 14C dates and ceramic chronologies: case
studies for the Early Bronze Age at Troy (Turkey) and Ezero (Bulgaria)”, Radiocarbon
37 (1995), 443–456.—C. Bronk Ramsey, “Radiocarbon calibration and analysis of
stratigraphy: the OxCal program”, Radiocarbon 37 (1995), 425–430.—J. A. Christen &
C. D. Litton, “A Bayesian approach to wiggle-matching”, JAS 22 (1995), 719–725.—
J. A. Zeidler, C. E. Buck & C. D. Litton, “The integration of archaeological phase
information and radiocarbon results from the Jama River Valley, Ecuador: a Bayesian
approach”, Latin American Antiquity 9 (1998), 160–179.—C. Bronk Ramsey, J. van der
Plicht & B. Weninger, “‘Wiggle matching’ radiocarbon dates”, Radiocarbon 43 (2001),
381–389.
36
H. J. Bruins et al., “Near East Chronology: archaeology and environment”, Radio-
carbon 43 (2001), 1147–1390.
37
M. Lehner, S. Nakhla, Z. Hawass, G. Bonani, W. Wölfli, H. Haas, R. Wenke,
J. Nolan, W. Wetterstrom, “Dating the pyramids”, Archaeology 52(5) (1999), 26–33.
334 sturt w. manning

age of interest. Thus does the sample remain intact with only the radio-
carbon age from the time the sample was exchanging with the atmos-
phere, or have contaminating materials become included? And have
there been processes of post-depositional diagenesis at work which are
relevant? The need to investigate bone samples to ensure good colla-
gen preservation is already appreciated and various strategies have been
adopted.38 Although typically not likely to be a significant issue in
general, the need to confirm removal of potentially contaminating humic
material from archaeological wood/charcoal/seed samples should be a
focus of further work.39

3. Radiocarbon and Egypt: An Example of Historically Relevant Data

Integrated archaeological and radiocarbon analyses in other parts of


the world carried out over the last decade have shown that, with high
quality sampling and analysis, it is possible and practical to resolve
chronology accurately and precisely down to the near-historical timescale.
Although there has not yet been a significant body of work for Egypt
after the Archaic period (e.g. refs. Section 1 above), it is important to
appreciate that radiocarbon is now capable of offering relevant and
independent dating for the OK through TIP. What is needed are mod-
ern research programmes. To demonstrate that radiocarbon can poten-
tially provide useful data which can either confirm and test historical
chronology (where available), or can provide near-historical level dat-
ing for those many other archaeological contexts in Egypt for which
secure historical dates are not available, I review one example. The
lack of my ability to note several good examples reflects the history of
the field (cf. previous sections), and the failure so far to exploit radio-
carbon where it could be most useful.

38
R. E. M. Hedges, “Bone diagenesis: an overview of processes”, Archaeometry 44
(2002), 319–328.—G. J. Van Klinken, “Bone collagen quality indicators for palaeodi-
etary and radiocarbon measurements”, JAS 26 (1999), 687–695.
39
See already D. Alon, G. Mintz, I. Cohen, S. Weiner & E. Boaretto, “The use
of Raman spectroscopy to monitor the removal of humic substances from charcoal:
quality control for 14C dating of charcoal”, Radiocarbon 44 (2002), 1–11.
radiocarbon dating and egyptian chronology 335

Tell el-Amarna
Among existing radiocarbon dates from Egypt, one suite demands atten-
tion.40 These are five dates on a range of materials (bone, horn, skin,
wood and charcoal) collected specifically and carefully for a high-quality
programme of radiocarbon dating41 from modern excavations at Tell
el-Amarna, the short-lived capital of Egypt for most of Akhenaten’s
reign, founded in his 5th regnal year or ca. 1350/1346 BC.42 The city’s
relative chronology is based on seventeen successive vintages documented
in its epigraphical record, fourteen of Akhenaten himself (years 4 to
17), and three belonging to his successors.43 The city was deserted before
the delivery of an eighteenth vintage. The specific context of the sam-
ples taken for radiocarbon analysis was a midden probably deposited
early within the site’s history and thus it would date during the 13
years of Akhenaten’s reign at the site.44 Hence the historical date range
might be narrowed to between c.1350/1346 BC to 1338/1334 BC.
The Amarna radiocarbon ages on both known shorter-lived samples
(skins, bone and horn) and on the potentially longer-lived wood and
charcoal samples, offer a tight and coherent set of results entirely con-
sistent with the historical dates and disprove any radically different
chronology: Figure III. 1.1. We can see that the final interpretation of
the radiocarbon data is very much determined by the shape of the
radiocarbon calibration curve in the 14th–13th centuries BC: see Figure
III. 1.2. There is a sharp ‘wiggle’ upwards centred 1325 BC (confirmed
for the east Mediterranean from Anatolian trees).45 The Amarna data
(bone sample Q-2505 perhaps apart) clearly do not match the peak of
the wiggle, and thus could lie on either side. At 2s (95.4%) confidence,
we see almost equal probability for either 1389–1329 BC or 1323–1260
BC. The former range (and especially the most likely sub-range at 1s
confidence of 1373–1338 BC) matches the historical age estimate very
closely. In support, we might note what seems to be an anomaly in
the five-date set. The wood sample Q-2401 yields the second youngest
(i.e. second most recent) radiocarbon age, and the animal bone sample

40
Also Hassan & Robinson (n. 19), 123.
41
V. R. Switsur in: Kemp, Amarna Reports I (1984), 178–188.
42
Murnane & Van Siclen, Stelae, 73–86; Kitchen, “Chronology I”, 39–52.—Kitchen
(n. 24).
43
Cf. above; Chapter II. 8.
44
Cf. Switsur (n. 41), 181–182.
45
Cf. n. 33.
336 sturt w. manning

Sequence
Boundary _Bound
Phase Tell El-Amarna
Q-2401 3035±35BP
Q-2402 3055±35BP
Q-2403 3050±35BP
Q-2404 3025±35BP
Q-2405 3088±35BP
Boundary _Bound
C_Date Amarna Context Earliest -1350
C_Date Amarna Context Latest -1334

2000BC 1800BC 1600BC 1400BC 1200BC 1000BC 800BC


Calibrated Date

Sequence {A=118.2%(A'c= 60.0%)}

Boundary _Bound

Phase Tell El-Amarna

Q-2401 111.7%

Q-2402 110.9%

Q-2403 111.9%

Q-2404 108.6%

Q-2405 96.4%

Boundary _Bound

2000BC 1800BC 1600BC 1400BC 1200BC 1000BC 800BC


Calendar Date

Figure III. 1.1.A. Calibrated calendar ages for the radiocarbon data reported from
Tell el-Amarna, Egypt (Switsur, n. 41, 178–188) compared to the historical date for
the context (see text). The upper and lower lines under each histogram indicate respec-
tively the 1s (68.2%) and 2s (95.4%) calibrated age ranges. B. Sequence analysis (solid
histograms) of the Amarna data (with the individual probabilities from A. indicated by
the hollow histograms) as a phase within calculated boundaries. The Amarna data are
entirely consistent with the historical age estimate for the context. Calibration and
analysis employing OxCal 3.9 (Bronk Ramsey, n. 35 and later versions, with curve
resolution set at 4) and INTCAL98 (n. 15). Q-2401, wood; Q-2402, charcoal; Q-2403,
skin; Q-2404, horn; Q-2505, bone. Weighted average of all five data: 3050±16 BP (1),
weighted average of just the three definitely shorter-lived samples 3054±20 BP (2). 2s
(95.4%) confidence calibrated ranges respectively (1) 1388–1331 BC (46.6%), 1322–1260
BC (48.8%), and (2) 1393–1260 BC (94%), 1228–1222 BC (1.4%).
radiocarbon dating and egyptian chronology 337

3200BP Tell el-Amarna weighted average : 3050±16BP


Radiocarbon determination

68.2%probability
3150BP 1372BC (35.2%) 1338BC
1318BC (22.6%) 1295BC
3100BP 1274BC (10.5%) 1264BC
95.4%probability
3050BP 1388BC (46.9%) 1331BC
1322BC (48.5%) 1260BC
3000BP
2950BP
2900BP

1500CalBC 1400CalBC 1300CalBC 1200CalBC 1100CalBC


Calibrated Date
Figure III. 1.2. Calibrated probability distribution for the weighted average radiocarbon
age from the five measurements on samples at Tell el-Amarna reported by Switsur,
n. 41). For discussion, see text. Calibration and analysis employing OxCal 3.9 (Bronk
Ramsey, n. 35 and later versions, with curve resolution set at 4) and INTCAL98
(n. 15).

Q-2405 yields the oldest age, and, in general, the average age of the
likely longer-lived samples (wood and charcoal) at 3045±25BP, is ( just)
younger than the average age of the shorter-lived samples (animal skin,
horn and bone) at 3054±20BP. Yet one would expect the wood sam-
ple to be older in real calendar terms than the animal bone sample
(by a few years or even a few decades or more). Out of the dating
possibilities for each sample, the only way for this likely correct sam-
ple relationship to occur is for the wood sample to date around the
earliest of its three potential intercept ranges with the calibration curve
at c.1368–1360 BC (and not c.1315–1289 BC or c.1280–1262 BC), and
for the animal bone sample to date around the later of its two possi-
ble intercept ranges at c.1336–1320 BC (and not c.1394–1375 BC).
And, plausibly, for the other three samples to date around or in between
these preferred ranges. In turn, the mid to later 14th century BC date
range is most likely for the Amarna samples. This is exactly compatible
with, and in support of, the standard Egyptian chronology, and, via
338 sturt w. manning

the cuneiform text linkages attested at Amarna,46 this finding in turn


supports and requires the standard Assyrian-Babylonian chronological
range for this period.47 Hence again radiocarbon provides useful inde-
pendent support to Egyptian and ancient Near Eastern chronology, and
disproves attempts to install radical chronological alternatives.

4. Past Radiocarbon Fluctuations (the Shape of the Calibration Curve), the Old
Wood Problem, and Egyptian OK Radiocarbon Dates

A study by Haas et al.,48 which indicated radiocarbon ages for vari-


ous OK monuments several centuries earlier than expected, was widely
seen as both a problem,49 and by some as a good reason to avoid
radiocarbon dating in Egyptology. The Haas et al. finding was largely
repeated in the followup study by Bonani et al.50 But it is not at all
clear that there is any unknown ‘problem’. A key issue is the history
of past natural radiocarbon levels; there was in effect a plateau in radio-
carbon levels in the period 2900–2500 BC. This means that radiocar-
bon ages for the period 2900–2500 BC typically could intercept at
several places with the radiocarbon calibration curve (i.e. several calendar
periods have similar radiocarbon ages). For example, if we consider the
OK monuments thought to be constructed c.2600–2500 BC, then the
wood employed will, at the latest, have its outermost ring dating then,
and the rest of the relevant tree will be progressively older. Depending
on species and source of the wood, one might expect an average offset
of several decades to a century, give or take a range, for an average
wood sample (e.g. compare the +50 ±50 old wood adjustment esti-
mated by Vogel et al.).51 Thus the ‘average’ wood used in a monument

46
Summary in Beckerath, Chronologie NR, 23–24; cf. above, Chapter II. 13.
47
J. A. Brinkman, Materials and studies for Kassite History (Chicago, 1976).
48
H. Haas, J. Devine, R. Wenke, M. Lehner, W. Wölfli & G. Bonani, “Radiocarbon
chronology and the historical calendar in Egypt”, in: O. Aurenche, J. Evin & F. Hours,
eds., Chronologies du Proche Orient/Chronologies in the Near East. Relative chronologies and absolute
chronology 16,000–4,000 B.P. (Oxford: BAR Int. Ser. 379, 1987), 585–606.
49
Hassan & Robinson (n. 19), 129.
50
G. Bonani, H. Haas, Z. Hawass, M. Lehner, S. Nakhla, J. Nolan, R. Wenke,
W. Wölfli, “Radiocarbon dates of Old and Middle Kingdom monuments in Egypt”,
Radiocarbon 43 (2001), 1297–1320.
51
J. S. Vogel, W. Cornell, D. E. Nelson & J. R. Southon, “Vesuvius/Avellino, one
possible source of seventeenth century BC climatic disturbances”, Nature 344 (1990),
534–537.
radiocarbon dating and egyptian chronology 339

built in the reign of Cheops in the mid or third quarter of the 26th
century BC (conventional date ranges) likely dates during the first quar-
ter of the 26th century BC give or take about 50 years—let us say
2587±50 BC in broad terms. If we simulate the radiocarbon age, and
its calibration, for 2587±50 BC, we get a result like that shown in
Figure III. 1.3. And what we find is that the shape of the calibration
curve (which represents the history of past natural variations in atmos-
pheric radiocarbon levels) yields a calibrated age that seems 100–300
years too old in the main and only just includes the real date at the
very end of the calibrated range at 95.4% probability. But we cali-
brated the ‘correct’ radiocarbon age! The point is that radiocarbon
dating of single context events in this period is problematic because of
the history of natural radiocarbon variations. Only use of another
approach (like wiggle-matching)52 can overcome this limitation.
We can in fact generalise the potential and problems of OK radio-
carbon dating by simulating radiocarbon ages for calendar years across
this period. Figure III. 1.4 shows two runs of simulated dates at ±50
years dating precision for the period 2750–2300 BC at 25 year inter-
vals and including also the weighted average radiocarbon ages deter-
mined and used for calibration by Bonani et al.53 for the Pyramid of
Snofru at Maidum and the Pyramids of Cheops, Khephren and Mycerinus
at Giza. Each run of a simulation produces different data from within
the possible range. Thus note how the calibrated age for 2525 BC at
±50 precision can vary quite a bit from a ‘low’ date range in Figure
III. 1.4.A to a ‘high’ date range in Figure III. 1.4.B. Samples near a
slope in the calibration curve have more such potential for movement;
other samples are much more stable. What we see is that the four sets
of Dyn. 4 pyramid data lie entirely within the expected calibrated range
for real dates from c.2750 BC to 2600 BC; they could be consistent
with data from as late as c.2475 BC, but clearly prefer a date range
starting around 2600 BC and older (compare also Figure III. 1.3 where
the data want to lie on the plateau 2850–2600 BC and not so much
on the slope following c.2600 BC). Such an outcome seems entirely
plausible for the non-specific wood/charcoal samples (including ‘flecks

52
For examples at this time period, see e.g. B. Weninger, “Die Radiocarbondaten”,
in: M. Korfmann, ed., Demircihüyük: Die Ergebnisse der Ausgrabungen 1975–78. II. Natur-
wissenschaftliche Untersuchungen (Mainz, 1987), 4–13.—Weninger (n. 35).
53
Cf. n. 50.
340 sturt w. manning

of carbon in mortar’ such as from the Cheops pyramid shown in Lehner


et al.)54 from the Cheops, Khephren and Mycerinus monuments built
in the 26th to early 25th century BC where average sample age is
probably of the order of c.50±50 years at the time of inclusion into
the monument. (We therefore see that the radiocarbon ‘dates’ thus can
be valid/correct—but they date the ‘old’ wood (etc). and not the cultural/
historical target date wanted: the building of the pyramid monument).
The Pyramid of Snofru at Maidum55 provided data where six of the
seven dates are closely comparable—SMU-1412 on a ‘log’ is either
aberrant or very old wood notwithstanding the stated dating of its ‘outer
rings’—and five of the determinations are stated to date outer rings
from wood from the burial chamber (see Lehner et al.)56 or shaft thereto.
Thus these samples might be expected to derive from closer to the
construction period of the monument (with this period usually assumed
to start at year 2 of the reign, onwards). Following the ‘historical’
chronology, work on this monument began c.2600 BC (Stadelmann)57
or 2638/2588 BC (Beckerath),58 2616 BC (Kitchen)59 or 2574 BC
(Baines & Málek).60 The calibrated age range of the average of these
six similar 14C ages given by Bonani et al.61 (2855–2583 BC at 1s, and
2860–2579 BC at 2s) is entirely compatible at the end of its range (for
why it will be just the end, see Figure III. 1.3 above) with the ‘his-
torical’ age estimates (and especially not the lowest of these). The cal-
ibrated probability distribution is entirely similar with a real age of
c.2600 BC (see Figure III. 1.4). One may therefore conclude that the
radiocarbon ages are approximately valid.
The plateau in radiocarbon levels clearly creates difficulties for nar-
row dating for OK samples. However, we may make some progress
with current debates. For example, Spence proposed a rather lower
OK chronology based on a hypothetical stellar alignment used by the
pyramid builders.62 She proposed dates of 2526±7 BC for the start of

54
Lehner et al. (n. 37), 31 bottom left illustration
55
Bonani (n. 50), 1304.
56
Lehner et al. (n. 37), 31 top right illustration
57
R. Stadelmann, “Beiträge zur Geschichte des Alten Reiches. Die Länge der
Regierung des Snofru”, MDAIK 43 (1986), 229–240.—R. Stadelmann, Die Großen
Pyramiden von Giza (Graz, 1990).
58
Beckerath, Chronologie.
59
Kitchen (n. 43).
60
J. Baines & J. Malek, Atlas of ancient Egypt (Oxford, 1980).
61
Bonani (n. 50).
62
K. Spence, “Ancient Egyptian chronology and the astronomical orientation of
radiocarbon dating and egyptian chronology 341

work at the Snofru pyramid at Maidum, 2480±5 BC for the Cheops


pyramid, 2448±5 BC for the Khephren pyramid and 2415±10 BC for
the Mycerinus pyramid. However, if one examines Figure III. 1.4, it
is evident that the radiocarbon data from these monuments are less
consistent with such a very low chronology unless very, very old wood
is always assumed. The range of simulated calibrated ages for samples
dating 2450–2400 BC are not at all similar with the radiocarbon ages
obtained from the Khephren and Mycerinus samples. The latter clearly
date much older wood, wood preferably 100–150 years older. The
more traditional range of ‘historical’ chronology estimates provides more
suitable dates (allowing for a plausible average old wood factor where
relevant; cf. also below, Chapter III. 4).
Apart from the general calibration issue discussed above, some other
issues may also be noted with regard to the Bonani et al. data sets.63
This team has published an enormous number of radiocarbon dates
from OK and MK monuments. There are wide spreads of ages in sev-
eral of the sets—suggested by the team involved themselves to be partly
if not largely accounted for by an ‘old wood’ issue, as all available trees
in the region, of widely varying ages, were consumed by the pyramid
builders and as older settlement debris was recycled in fires.64 While
this is plausible in many cases, nonetheless, some samples are clearly
aberrant for unspecified reasons. It is undoubtedly the case that the
association of measured age for the sample (biological age unless other
contaminating processes were involved) versus the date for monument
construction is not demonstrated or clear in a number of instances (e.g.
‘charcoal’ from mudbricks or from mortar—see Bonani et al.65—may
easily represent ‘old’ or re-used tree-rings). The limestone and mortar
associated with a number of samples may also provide a source of
old carbon—for example with reference to samples from ‘flecks of
carbon in mortar’ such as from the Cheops pyramid.66 It is certainly
interesting that the two secure datasets from early second millennium
BC MK monuments (Pyramid of Senwosret II at Illahun and Pyramid
of Amenemhet III at Dahshur), a new phase of pyramid building after
a significant interval, yielded calibrated ages compatible with historical

pyramids”, Nature 408 (2000), 320–324. Cf. D. Rawlins, K. Pickering and K. Spence,
“Astronomical orientation of the pyramids”, Nature 412 (2001), 699–700.
63
Cf. n. 50.
64
Lehner et al. (n. 37), 33.
65
Cf. n. 50, 1297–1298.
66
Lehner et al. (n. 37), 31 bottom left illustration.
342 sturt w. manning

4500BP R_Simulate Khufu average wood age : 4154±50BP (2587BC)


Radiocarbon determination
4400BP 68.2% probability
2880BC (13.5%) 2830BC
4300BP 2820BC ( 5.2%) 2800BC
2790BC (41.9%) 2660BC
4200BP 2650BC ( 7.5%) 2620BC
95.4% probability
4100BP 2880BC (95.4%) 2580BC

4000BP
3900BP
3800BP

3200CalBC 3000CalBC 2800CalBC 2600CalBC 2400CalBC 2200CalBC


Calibrated Date

Figure III. 1.3. Simulated radiocarbon age for a calendar date of 2587±50 BC using
OxCal 3.9 (Bronk Ramsey, n. 35 and later versions with curve resolution set at 4) and
INTCAL98 (n. 15). Note: every simulation run produces a slightly different outcome—
this is a ‘typical’ output based on a number of runs. This simulated calendar age is an
example of possible typical age of ‘average’ wood from a monument built in the reign
of Cheops (reign starts 2604/2593/2554/2551 BC and ends 2581/2570/2531/2528
BC: Beckerath, Chronologie; Kitchen n. 43; Baines & Málek, n. 60), given a typical +50
±50 year ‘old wood’ adjustment. The calibrated radiocarbon age intersects with the
c.2900–2500 BC plateau in the radiocarbon calibration curve, and offers several 1s
ranges within a large 2s range c.2880–2580 BC. We thus see as a function of the nat-
ural history of radiocarbon fluctuations that real dates in the early 26th century BC
yield calibrated radiocarbon ranges mainly apparently too early, with the real date just
creeping into the last few years of the calibrated age range. It is noteworthy that the
average radiocarbon age from the 45 samples used for the Pyramid of Cheops at Giza
is in fact 4147±10BP (Bonani et al., n. 50, 1315)—almost exactly the 4154BP radiocarbon
age derived by the simulation of a calendar date of 2587±50 BC, as shown above. Thus
it appears that the Cheops data do, on average, yield a plausible age for wood used
in his reign (with this average wood typically +50 or so years in age, give or take a
range, versus the actual use date—and the large range within the Cheops data, see
below, indicates such a range, or more, in the real wood ages, apart from any conta-
minating factors from associated mortar/limestone). Because of the history of radiocar-
bon variations (the 2900–2500 BC plateau), only the very last part of the calibrated
range indicates the real age. These data, and the other similar OK data in Bonani et al.
(n. 50) where the ‘real’ age at best creeps into the end of the calibrated age range, or
lies shortly afterwards, therefore do not provide evidence of any additional offset beyond
old wood and the calibration outcomes given the history of natural radiocarbon levels
c.2900–2500 BC. They provide no evidence at all for claims of hypothetical 100–300
years too early offsets in Mediterranean radiocarbon ages based on claims of putative
upwelling of old carbon (Keenan)67—something for which there is no positive evidence
within an order of magnitude (Manning et al. 2002, n. 33).

67
D. J. Keenan, “Why early-historical radiocarbon dates downwind from the Medi-
terranean are too early”, Radiocarbon 44(1), (2002), 225–237.
radiocarbon dating and egyptian chronology 343

estimates.68 The data from Archaic contexts also yielded radiocarbon


ages largely in keeping with approximate ‘historical’ estimates69—as have
other recent studies.70 These periods both have helpful, non-plateau,
radiocarbon calibration curve shapes, and may also plausibly have had
less of an exhausted natural supply of wood—contrast the peak OK
period of pyramid construction which probably forced much recycling
of old material.71
For the third millennium BC, Bonani et al. report 17 date sets of
the OK as older than their stated historical estimate, 6 as compatible,
and 4 as more recent than the historical estimate. This clearly ‘seems’
to be a problem. But, apart from noting that the historical age esti-
mate is commonly regarded as ±100 years for this period, the inter-
pretation of Bonani et al. is based on two inappropriate starting points.
First, there is no allowance for likely average sample age at time of
use (i.e. ‘old-wood’ age for random wood/charcoal samples not known
to be outer tree-rings), and second Bonani et al. use average values for
the radiocarbon age of sample sets which contain significant internal
variation, and this is thus potentially misleading (probably less so as set
size increases). To illustrate: examination of Bonani et al. (n. 50, Fig. 1)
shows the Cheops Pyramid (object number 13) to yield one of the
apparently tighter calibrated age ranges and to be some two centuries
older than the estimated historical age.72 And this despite 46 radio-
carbon dates being reported for the monument.73 But examination of
the 46 radiocarbon data reveals ages varying by 1210 radiocarbon
years!—and even excluding the two gross outliers in the set,74 the age
range left in the set is still 513 14C years! As shown in Figure III. 1.5,
just over one-third of the individual samples—the younger ages—do in
fact offer calibrated ages more or less compatible with the estimated
historical age of 2589–2566 BC,75 and most of the remainder offer

68
Bonani (n. 50), Fig. 1 and p. 1320.
69
Bonani (n. 50), Fig. 1 object numbers 1–5.
70
E.g. Görsdorf et al. (n. 22).
71
Lehner et al. (n. 37).
72
Bonani et al. (n. 50), 1315 use the historical range of 2589–2566 BC—other stan-
dard sources suggest c.2593–2570 BC: Kitchen (n. 43), 48; around 2604/2554 BC to
2581/2531 BC: Beckerath, Chronologie, 188, or 2551–2528 BC: M. Lehner, The com-
plete pyramids (London, 1997), 8.
73
Bonani (n. 50), 1305.
74
Marked by the * and + signs by Bonani (n. 50), 1305.
75
Bonani (n. 50), 1316.
344 sturt w. manning

R_Simulate 2750BC -2750±50


R_Simulate 2725BC -2725±50
R_Simulate 2700BC -2700±50
R_Simulate 2675BC -2675±50
R_Simulate 2650BC -2650±50
R_Simulate 2625BC -2625±50
Snefru at Meydum 4110±23BP
R_Simulate 2600BC -2600±50
R_Simulate 2575BC -2575±50
Khufu at Giza 4147±21BP
R_Simulate 2550BC -2550±50
Khafre at Giza 4173±27BP
R_Simulate 2525BC -2525±50
R_Simulate 2500BC -2500±50
Menkaure at Giza 4127±25BP
R_Simulate 2475BC -2475±50
R_Simulate 2450BC -2450±50
R_Simulate 2425BC -2425±50
R_Simulate 2400BC -2400±50
R_Simulate 2375BC -2375±50
R_Simulate 2350BC -2350±50
R_Simulate 2325BC -2325±50
R_Simulate 2300BC -2300±50

3500CalBC 3000CalBC 2500CalBC 2000CalBC


Calibrated Date
Figure III. 1.4 (A and B). Two outputs of simulations of calibrated radiocarbon ages
for calendar years 2750–2300 BC at ±50 dating precision and at 25 calendar year
intervals (data from OxCal 3.9 with curve resolution set at 4, Bronk Ramsey, n. 35
and INTCAL98, n. 15). Included also are the weighted average radiocarbon ages used
for calibration by Bonani et al., n. 50, 1314–1316) for the Pyramid of Snofru at
Maidum, the Pyramid of Cheops at Giza, the Pyramid of Khephren at Giza and the
Pyramid of Mycerinus at Giza. For discussion, see text.
radiocarbon dating and egyptian chronology 345

B
R_Simulate 2750BC -2750±50
R_Simulate 2725BC -2725±50
R_Simulate 2700BC -2700±50
R_Simulate 2675BC -2675±50
R_Simulate 2650BC -2650±50
R_Simulate 2625BC -2625±50
Snefru at Meydum 4110±23BP
R_Simulate 2600BC -2600±50
R_Simulate 2575BC -2575±50
Khufu at Giza 4147±21BP
R_Simulate 2550BC -2550±50
Khephren at Giza 4173±27BP
R_Simulate 2525BC -2525±50
R_Simulate 2500BC -2500±50
Menkaure at Giza 4127±25BP
R_Simulate 2475BC -2475±50
R_Simulate 2450BC -2450±50
R_Simulate 2425BC -2425±50
R_Simulate 2400BC -2400±50
R_Simulate 2375BC -2375±50
R_Simulate 2350BC -2350±50
R_Simulate 2325BC -2325±50
R_Simulate 2300BC -2300±50

3500CalBC 3000CalBC 2500CalBC 2000CalBC


Calibrated Date
346 sturt w. manning

variously a little to quite a bit older ages—‘old wood’76—would appear


the obvious first hypothesis.77 Such a pattern: younger ages corresponding
to, or close to, context date and older ages reflecting old wood is quite
common and expected when dealing with wood/charcoal samples.78
Similar observations may be made about the data sets for: Step Pyramid
of Djoser at Saqqara, Temple Complex associated with the Step Pyramid,
Pyramid of Sekhem-khet at Saqqara,79 Pyramid of Khephren at Giza,
Pyramid of Ra'djedef at Abu Rawash, Sphinx Temple of Khephren
at Giza (n. 50, 1306), Pyramid of Mycerinus at Giza, Mortuary Temple
of Shepseskaf at South Saqqara (n. 50, 1307), Mortuary Temple and
Pyramid of Sahure at Abusir (n. 50, 1309) and Pyramid of Teti at
Saqqara (n. 50, 1310). In contrast, it is notable that the radiocarbon
ages from a modern excavation at the Royal Production Centre at
Giza offer both a reasonably consistent set, and calibrated ages more
recent than the surrounding OK datasets from the monuments.80
We have already noted the case of the Pyramid of Snofru at Maidum,
where six of the determinations date outer rings from wood from the
burial chamber or shaft thereto. And thus the usual old-wood effect is
likely minimised. The calibrated age range of the average of these six
similar 14C ages given by Bonani et al. (2855–2583 BC at 1s, and
2860–2579 BC at 2s) is entirely compatible at the end of its range (for
why it will be just the end, see Figure III. 1.3 above) with the ‘his-
torical’ age estimate employed by Bonani et al. (n. 50, 1314) or those
estimated by Kitchen (n. 43) or Beckerath, Chronologie, (higher range),
and also overlaps with the date for the accession of Snofru c.2600 BC
and his earlier reign, and thus the construction of this Snofru’s first (of
three) pyramids, given by Stadelmann (n. 57). One may observe that
the stated calibrated range ends +8/+4 years from the start of the
lower ‘historical’ age estimate for Snofru by Lehner from Baines &
Málek (n. 60);81 this is hardly a significant difference, and the wood in

76
M. B. Schiffer, “Radiocarbon dating and the ‘old wood’ problem: the case of the
Hohokam chronology”, JAS 13 (1986), 13–30.
77
See Lehner (n. 37), esp. 31–33.
78
For an example from Troy II, see Kromer et al. (n. 34), 48 and Fig. 4.
79
Bonani (n. 50), 1303.
80
Bonani (n. 50), Fig. 1 object 12, contrasted with other objects 10–19. One might
speculate that the samples from this context, which are not from major architecture/
monuments and their creation industries, do not therefore suffer so much from an
average old wood problem.
81
Lehner (n. 72).
radiocarbon dating and egyptian chronology 347

question could easily have been cut a few years earlier than the start
of Snofru’s reign if the lower dates are to be preferred, just as it could
have been cut during his reign if the slightly higher dates are preferred.
We have already seen how even the correct radiocarbon age for a date
around the early 26th century BC only includes the real calendar age
within the very end of the calibrated range, as much of the dating
probability ends up on the plateau in the radiocarbon curve over the
preceding couple of centuries (compare Figure III. 1.3 above). Thus
these dates for the Pyramid of Snofru at Maidum are entirely consis-
tent with the estimated ‘historical’ age.
In sum, these OK radiocarbon dates do not in fact indicate any
problem with radiocarbon dating and Egyptian chronology; instead they
nicely illustrate the importance and impact of the shape of the cali-
bration curve in dating, and they highlight the need to obtain organic
samples directly associated with, and relevant to, the human context
for which a date is sought. Wood and charcoal samples especially can
easily be older, or even much older, than their final deposition con-
text depending on tree species and the uses and perhaps re-uses of the
wood. Add in calibration taphonomy and correct radiocarbon ages for
organic materials can appear to yield dates that are centuries too old
for their historical/archaeological context (see Figure III. 1.3). Aquatic
samples, which may include a water/marine reservoir radiocarbon age
(versus solely the normal atmospheric reservoir radiocarbon age repre-
sented in normal terrestrial plants, and animals eating these), must also
be treated with care and caution—again this may explain some of the
apparently aberrant radiocarbon ages obtained on ‘reed’ samples in
Egypt.
As evident from the Amarna example in Section 3, and other stu-
dies cited in Sections 1 and 2, or other studies in the literature,82 in
appropriate circumstances high-quality radiocarbon data from Egypt
and the east Mediterranean region can provide accurate and precise
dates).83 When issues occur, such as the completeness of removal of age
contamination by humic material,84 or the old wood offsets evident in
the extensive OK radiocarbon measurements published by Bonani et al.

82
M. Lange, “Wadi Shaw 82/52: 14C dates from a peridynastic site in northwest
Sudan, supporting the Egyptian historical chronology”, Radiocarbon 40 (1998), 687–692.
83
See also Manning et al. (2002, n. 33).
84
Alon et al. (n. 39).
348 sturt w. manning

Sequence
Boundary _Bound
Phase Khufu
charcoal
ETH-0302 4260±80BP
ETH-0303 4300±90BP
ETH-0304 4245±85BP
ETH-0305 4355±90BP
ETH-4226 4195±105BP
ETH-0306 4320±85BP
ETH-0228 4350±125BP
SMU-1417 4359±241BP
ETH-0227 4360±125BP
ETH-13762 3927±58BP
ETH-13754 3980±57BP
ETH-13756 4143±61BP
ETH-13757 4225±79BP
ETH-13761 3928±54BP
ETH-13763 3937±61BP
ETH-0307 4440±90BP
ETH-4227 4215±105BP
ETH-13770 4087±53BP
ETH-13771 4187±60BP
ETH-13775 4190±52BP
ETH-13777 4313±57BP

6000BC 5000BC 4000BC 3000BC 2000BC 1000BC


Calibrated Date
Figure III. 1.5 (A and B). Calibrated calendar ages for the 46 radiocarbon data reported
from the Pyramid of Cheops at Giza (data from Bonani et al. n. 50, 1305). The esti-
mated historical age employed by Bonani et al. (n. 50, 1315) is 2589–2566 BC. The
upper and lower lines under each histogram indicate respectively the 1s (68.2%) and
2s (95.4%) calibrated age ranges. Calibration employing OxCal 3.9 (Bronk Ramsey,
n. 35 and later versions, with curve resolution set at 4) and INTCAL98 (n. 15). For
discussion, see text.
radiocarbon dating and egyptian chronology 349

Sequence
Phase Khufu
Phase ETH-13778 4156±58BP
ETH-13779 4062±61BP
ETH-0308 4300±85BP
ETH-4228 4390±110BP
ETH-13783 4237±62BP
ETH-13784 4068±54BP
ETH-13785 4083±53BP
ETH-13782 3984±55BP
ETH-13787 4197±49BP
ETH-13791 3810±60BP
ETH-0309 4420±100BP
ETH-13800 4195±55BP
ETH-13799 4128±58BP
ETH-13801 4189±60BP
ETH-13802 4174±61BP
ETH-13803 4062±60BP
ETH-13804 4254±59BP
ETH-13805 4267±54BP
ETH-0311 4395±85BP
ETH-0312 5020±130BP
ETH-0334 4440±320BP
ETH-13900 4068±60BP
reed
ETH-0313 4330±125BP
Boundary _Bound

7000BC 6000BC 5000BC 4000BC 3000BC 2000BC 1000BC


Calibrated Date
350 sturt w. manning

(n. 50), they in fact lead us to consider important topics concerning


taphonomy, sample diagenesis, social history, economic processes, and
the environment. As Lehner et al. conclude with regard to the OK
issue:85
If the fair agreement of our 1995 results with historical dates and previ-
ous radiocarbon dates for the Archaic period and with the historical dates
for the MK hold, the problematic OK dates are boxed in. And therein
may lie a hint of multifaceted old wood effects for a period, especially
from Djoser to Mycerinus, when any and all wood resources may have
been consumed at a whole other order of magnitude than before or after
the giant pyramid-building projects . . . our project . . . now has us think-
ing about forest ecologies, site formation processes, and ancient industry
and its environmental impact . . .

5. Caution: The Need to Make Only Secure Historical Associations is


Paramount

Radiocarbon dating determines the age of an organic sample. The asso-


ciation of such a sample and its radiocarbon age with history/archaeology
is the task of the archaeologist. And one has to be careful and rigor-
ous. Associations must be demonstrated, not casually assumed. A recent
example illustrates the potential problems and the need to be even
more careful as better precision becomes available in modern radio-
carbon dating.

Shoshenq I and Radiocarbon?


A key synchronism for the standard chronology of Egypt (and wider
Near Eastern history) concerns the identification of the important Egyp-
tian pharaoh Shoshenq I with the Shishak attested in the Bible (I Kings
14:25–26; II Chronicles 12:3–4) as invading Judah and Israel in the
5th year of Rehabeam.86 Rehabeam year 5 is in turn dated c.926/925
BC by linking the attested names and reign lengths of the 10th–9th
century BC kings of Israel and Judah with recorded synchronisms in
the 9th century BC between the Israelite kings Ahab and Jehu and the

85
Lehner (n. 37), 33.
86
Kitchen, TIP, esp. xliv, 72–76, 287–302; Kitchen (n. 24), 7–8; Beckerath, Chronologie
NR, 30–34; idem, Chronologie, 68–70.
radiocarbon dating and egyptian chronology 351

Assyrian king Shalmaneser III. Since the chronology of the Assyrian


kings is effectively absolute back to the 10th century BC, this enables
precise calendar dates to be applied (of respectively c.853 BC for the
last year of Ahab’s reign and c.841 BC for the first year of Jehu’s
reign).
In an important and controversial paper, Bruins et al. (2003) recently
reported sets of high-precision radiocarbon dates, allied with an inter-
pretative stratified archaeological wiggle-matching analysis, from the site
of Tel Rehov in Israel.87 These samples, on high-quality short-lived
samples, provide the basis for a high-resolution chronology for the site
in the 12th through 9th centuries BC. But Bruins et al. also suggested
that the date for the destruction of Stratum V at Tel Rehov could be
associated with the campaign of Shoshenq I and thus their date for
this stratum—c.940–900 BC—was argued to support this proposed
identification, and in turn the standard Egyptian chronology or one
very close to it (with the Shoshenq I invasion dated c.926/925 BC—
see above, or various slight alternative calculations, such as the 918 BC
of Miller and Hayes,88 cited by Bruins et al. (n. 35) 317, or 927 BC in
Barnes89 or 922–921 BC in Hayes and Hooker.90
However, the critical logical step was missing. There is no evidence
at all that the Stratum V destruction links with Shoshenq I—this is
merely an unproven and (unnecessary) assumption incorporated into
the Bruins et al. paper and its dating model (and so leads to a circu-
lar argument). The dating of the site and the dating of Shoshenq I
are separate until and unless clear evidence can be produced to show
that Shoshenq I caused the specific Stratum V destruction horizon
dated by the radiocarbon measurements. Archaeologists must always
be aware that non-rigorous and specific (i.e. documented) assumptions
that try to bring archaeological and historical evidence together (the
event-historical model) are often inherently problematic because the
respective evidence types represent fundamentally different facets of

87
Bruins et al. n. 35.
88
M. Miller & Hayes, A history of ancient Israel and Judah (Philadelphia, 1986).
89
W. H. Barnes, Studies in the chronology of the divided monarchy of Israel (Cambridge,
Mass., 1991).
90
J. H. Hayes & P. K. Hooker, A new chronology for the kings of Israel and Judah and
its implications for Biblical history and literature (Atlanta, 1988).
352 sturt w. manning

historical reality.91 In the Shoshenq I case, it is fair to note that much


is less than certain and very different narratives are possible based on
the same limited and likely non-contemporary ‘historical’ evidence/
tradition.92
This example highlights the need to delineate clearly what is the tar-
get date and how and why the relevant organic samples do (or do not)
provide associated dating evidence when radiocarbon dated. Without
such chronometric care results are not credible, and conclusions may
turn out to rest on foundations of sand.

6. Conclusions

High quality radiocarbon dating offers an important but as yet not fully
exploited resource for Egyptology. It provides an increasingly accurate
and precise test for the historical chronology and can actively inform
and resolve disputes in less certain or ambiguous periods. Available dat-
ing accuracy and precision from radiocarbon should in principle—i.e.
on suitable short-lived samples from primary contexts dated at good
precision—offer a chronological precision for the third millennium BC
of the order of, or better than, the historical age estimates—which, for
this period, are often regarded as having a significant error margin of
up to a century. It could test and resolve claims for significantly differing

91
A. M. Snodgrass, “Archaeology”, in: M. Crawford, ed., Sources for ancient history
(Cambridge, 1983), 137–184.—Idem, An archaeology of Greece: the present state and future
scope of a discipline (Berkeley, 1987), 37–66.
92
I. Finkelstein, “The campaign of Shoshenq I to Palestine: a guide to the 10th
century BCE polity”, ZDPV 118 (2002), 109–135.
Postscript. Since the present text was submitted in August 2003, there have been sev-
eral further publications (and much discussion) taking this topic now well beyond the
initial publication of Bruins et al. (2003) cited above. However, the logic/methodology
point noted in the text remains relevant as outlined (and has since been accepted by
the Bruins et al. authorship—I wish to thank Hendrik Bruins and Amihai Mazar for
friendly, constructive, and productive discussion). For the latest (AD 2005) situation on
the analysis of the important Tel Rehov datasets, see now (i) Mazar, A. et al. “Ladder
of Time at Tel Rehov: Stratigraphy, Archaeological Context, Pottery and Radiocarbon
Dates”, and (ii) Bruins, H. J. et al. “The Groningen Radiocarbon Series from Tel
Rehov: OxCal Bayesian Computations for the Iron IB–IIA Boundary and Iron IIA
Destruction Events”, both papers in Radiocarbon Dating and the Iron Age of the Southern
Levant—the Bible and Archaeology Today, edited by Thomas Levy and Thomas Higham,
Equinox Publishing, Ltd., London (2005), 193–255, and 271–293 respectively.
radiocarbon dating and egyptian chronology 353

dates, such as those suggested from speculative astronomical conjecture


by Spence (n. 62).
For the second millennium BC radiocarbon may also be able to
assist. For many non-elite contexts it may offer the best means of dat-
ing. For the chronology of the pharaohs it will be of less need, as the
dates in the second millennium BC are relatively accurately and pre-
cisely determined from a combination of so-called ‘dead-reckoning’ (the
compilation of documented names of Egyptian kings and various other
persons and attested years of reign/office backwards from an agreed
starting point fixed against the Greco-Roman timescale)93 and analysis
of some records of astronomical observations (e.g. Krauss;94 Beckerath).95
Recent scepticism, and claims to reject for example all lunar data
(Wells96—approvingly cited by e.g. Kitchen),97 have been shown to
be based on incorrect or partial understanding of the data and their
analysis.98
The leading scholars immersed in the details argue that this combi-
nation of historical data and astronomical evidence forms a closely
dated chronological system for the second millennium BC, with only
at most a few years to a decade or so error range, and with several
likely absolute placements therein, such as the accession of Tuthmosis
III in 1479 BC.99 Nonetheless, one role for radiocarbon will be to offer
an independent check and verification of these chronologies. Without
this, complete certainty will never be possible given that there are gaps
and uncertainties/ambiguities in the evidence (textual or astronomical),
and key assumptions/interpretations have been made by modern schol-
ars. Radiocarbon dating is direct and independent, and can cut through
circular debates and assumptions.

93
Kitchen, TIP; idem, “Chronology I”; idem n. 24.
94
Krauss, Sothis.
95
Beckerath, Chronologie, 41–51.
96
R. A. Wells, “The role of astronomical techniques in ancient Egyptian chronology:
the use of lunar month lengths in absolute dating”, in: J. M. Steele & A. Imhausen,
eds., Under one sky: astronomy and mathematics in the ancient Near East (Münster, 2002),
459–472.
97
Kitchen, n. 24 at p. 11.
98
R. Krauss, “Ronald A. Wells on astronomical techniques in Ancient Egyptian
Chronology”, DE 57 (2003), 51–56.
99
Kitchen, TIP; n. 43; n. 24; Beckerath, Chronologie NR; 1997; cf. below, Chapter
III. 8.
354 sturt w. manning

High-quality radiocarbon dating also offers the independent means


to test and reject the several publications of the last decade which have
argued that conventional Egyptian (and wider ancient Near Eastern)
historical chronology is incorrect.100
Radiocarbon dating should become the friend of Egyptologists. Whereas
in its origins Egyptology helped to test radiocarbon dating and to expose
the need for calibration, modern radiocarbon dating now offers the
means to test, support, extend, and even to refine Egyptian chronol-
ogy. Certain periods like the OK will be problematic if samples or con-
texts are dated in isolation thanks to the unhelpful plateau in the
calibration curve (see Figure III. 1.3 above); but by exploiting tech-
niques like seriated archaeological wiggle-matching (ideally of short-lived
samples tied securely to the context for which a date is sought), even
this time period can be made to yield an accurate and precise cali-
brated radiocarbon chronology by taking advantage of the shape of the
calibration curve.101 Radiocarbon dating also offers the route to engage
with all those many Egyptian archaeological contexts not specifically
linked with the (largely elite centred) textual record (compare the sim-
ilar but still largely unfulfilled hope expressed twenty years ago by
O’Connor.102 The entirety of Egyptian archaeology can then be inte-
grated into an accurate and precise near-historical level timeframe.

100
E.g. J. Goldberg, “Centuries of darkness and Egyptian chronology: another look”,
DE 33 (1995), 11–32.—G. Hagens, “A critical review of dead-reckoning from the 21st
Dynasty”, JARCE 33 (1996), 153–163.—James et al. (n. 25), (n. 26).—Rohl n. 25.—
P. Van der Veen & W. Zerbst, Biblische Archäologie am Scheideweg? (Holzgerlingen, 2002).
101
See e.g. Weninger (n. 52); Weninger (n. 35).
102
D. O’Connor, “New Kingdom and Third Intermediate Period”, in: B. G. Trigger,
B. J. Kemp, D. O’Connor & A. B. Lloyd, Ancient Egypt: a social history (Cambridge,
1983), 183–278, esp. 185.
radiocarbon dating and egyptian chronology 355

Appendix

Figure III. 1.6. The new (AD 2005) IntCal04 radiocarbon calibration curve (black) at
1s for period 500 BC to 3500 BC, compared to the previous IntCal98 curve (grey)
as used in this paper. There is little significant difference—the main change is that the
IntCal04 curve is a little more smoothed. Data from Reimer et al. (n. 32) and Stuiver
et al. (n. 15).
III. 2 ABSOLUTE CHRONOLOGIES:
LUMINESCENCE DATING OF EGYPTIAN ARTEFACTS

Christian Goedicke

The phenomenon of luminescence in solids has been used as a dating


method in archaeology since the late 1960s. Since then the method
developed as an important and universal tool alongside radiocarbon
dating and dendrochronology. It is based on the property of a num-
ber of minerals to store and release radioactive decay energy. Two
minerals frequently occurring in archaeological contexts, quartz and
feldspar, show this effect very distinctly which makes the application of
luminescence analysis in archaeology particularly useful. In the field of
the geological sciences luminescence dating has caused a decisive increase
of knowledge and is regarded meanwhile as indispensable. In the fol-
lowing it will be discussed in which parts of archaeology the use of the
method may be inadequate.

Luminescence Dating: Basic Principles

Radioactive traces (U-238, U-235, Th-232, K-40, Rb-87) are constituent


in clay, in soil, and in rocks. During spontaneous decay, these elements
release energy into the environment. When quartz or feldspar occurs
in the environment, the emitted energy may be stored in the crystal
lattice of these minerals. After an archaeological storage time the stored
energy can be released in form of light. The longer the archaeological
storage time, the larger the accumulated energy (phys. the dose) and
the stronger the light signal. Luminescence dating is a dosimetric dating
method requiring no external calibration. The necessary zero-setting
event is achieved by heating or exposure to light when the energy
stored in the lattice over geological times is zeroed and the accumulation
can begin again.
Due to different ranges and to different effectiveness in producing
luminescence, the three types of radioactive decay radiation (alpha-,
beta- and gamma radiation) need to be taken into account separately.
Consequently, more than one physical quantity must be measured in
absolute chronologies 357

the laboratory and, hence, the number of measurements results in a


complex error of the age which is one of the major differences from
C-14 dating. The total error of a thermoluminescence single date of
an object amounts to between 7 to 12% of the total age expressed in
years. In absolute numbers this amounts to roughly ±450 years for an
OK date. The range of uncertainty is much smaller, if several samples
from the same archaeological context are analyzed; the so-called con-
text error can be reduced to approx. ±5% which still corresponds to
±230 years for the OK.
Thermoluminescence ages are calculated according to the following
formula:

archaeological dose
age = (Eq. 1)
a·dose-ratea + dose-rateb + environmental dose-rate

internal dose-rate external dose-rate

A feature peculiar to Egyptian artefacts is the low radioactive trace-


element content which explains the low internal dose-rates frequently
encountered in Egyptian artefacts. Consequently, thermoluminescence
ages become dependent on the ratio of the internal to external dose-
rate (see Eq. 1). Should the two dose-rates come close to or equal each
other, the external dose-rate may determine the thermoluminescence
age. Fig. III. 2.1 illustrates the effect of the dose-rate ratio on the ther-
moluminescence age.
Thermoluminescence produces the best results when applied to arte-
facts as they are excavated, i.e. when the measurement of the envi-
ronmental dose-rate can be made on-site. No access to the site limits
the usefulness of the method, e.g. analysis of museum objects of unknown
provenance cannot produce definitive results. The humidity of a sam-
ple also requires access to the site, as humidity is a correcting factor
for the age of a sample. Variation of the humidity during storage times
has to be taken into account.
358 christian goedicke

Fig. III. 2.1. Effect of the environmental dose-rate on thermoluminescence age; the
example shows a typical ceramic sample from the OK. The fraction of the environmental
dose-rate most probably prevailing in Egypt lies between ca. 15 and 30.

Luminescence techniques can be used to date:


A. fired materials: ceramics (minimum firing temperature 500°C), stones
(firesides), casting cores of bronzes
B. unfired materials: sediments (aeolian, fluvial, colluvial), mortar

How Useful is Luminescence Dating for Dynastic Egyptian Objects? Answers


Based on Published Dates

Considering the usual error margins of 7–12%, thermoluminescence


dating of Egyptian artefacts cannot contribute much to chronological
evidence. The method may prove useful for a piece which cannot be
attributed to any period on stylistic criteria. Even the range of context
errors exceeds the possible imprecision of the Egyptian chronology
which in its present state is better than 5% in all periods. Hence it is
not surprising that published examples of luminescence dating for
Egyptian artefacts are few. Recent culling of various bibliographical
sources including CAS (chemical abstracts services) turned up the fol-
lowing papers.
absolute chronologies 359

Thermoluminescence Dates Quoted in the Literature

Two nearly identical papers deal with pottery of the fourth dynasty.1
The instrumentation used and the reported error of 2.3% cast some
doubt on the seriousness of both; the standard calculation will not result
in an error that small.
A French team notes that the pigment Egyptian Blue results from
a thermal reaction thus making dating of this material possible.2
The attempt to date a large chunk of Egyptian Blue from Karnak is
described resulting in a reasonable date. However, dating pigments
from paint-layers is not feasible due to dosimetry problems arising from
paint-layer thickness which usually is smaller than the range of beta-
particles.
Some unpublished results of the author (Rathgen-Forschungslabor in
Berlin) on potsherds from excavations at Abu Minshat Omar can be
mentioned here. A comparison of the results with radiocarbon data
from the same site revealed a systematic deviation towards younger
ages.
Samples of vitrified silt from the Early Dynastic tombs at Abydos
turned out to be undatable. Secondary firing resulted in a high degree
of vitrification making luminescence dating impossible.
For earlier periods with a comparatively fluid chronological frame-
work luminescence dates, even including the above quoted errors, may
furnish a valuable chronological contribution. However, even for this
period luminescence has rarely been used. Whittle’s early study dealt
with potsherds from Hemamieh and Qurna-Tarif.3 Environmental
dose-rate data were obtained in the laboratory by alpha-counting of
soil samples. An archaeological evaluation of the data obtained was
not attempted. Two additional studies considered Middle Palaeolithic

1
S. A. Elfiki, M. S. Abdelwahab, N. Elfarawamy & M. A. Elfiki, “Dating of ancient
Egyptian pottery using thermoluminescence techniques”, Nuclear Instruments and Methods
94 (1994), 91–94.—M. S. Abdel-Wahab, S. A. Elfiki, M. A. Elfiki, M. Gomaa, S. Abdel-
Kariem & N. El-Faramawy, “Annual dose measurements and TL-dating of ancient
egyptian pottery”, Radiation Physics and Chemistry 47 (1996), 697–700.
2
M. Schvoerer, M. C. Delavergne & R. Chapoulie, “The thermoluminescence of
Egyptian Blue”, Nuclear Tracks and Radiation Measurements 14 (1988), 321–327.
3
E. H. Whittle, “Thermoluminescent dating of Egyptian pre-dynastic pottery from
Hemamieh and Qurna-Tarif ”, Archaeometry 17 (1975), 119–122.
360 christian goedicke

settlements.4 Dating Neolithic cultures by thermoluminescence and radio-


carbon in the Sudan Nile valley was the aim of a French working
group.5

Outlook

Luminescence techniques other than thermoluminescence, e.g. optically


stimulated luminescence (OSL), are not very likely to improve the
situation; conversely, the sensitive dependence on the dose-rate ratio
will become even more important. Corresponding to the zero-setting
event (bleaching by sun-/daylight) these techniques are more appro-
priate for use in geological sciences. About half a dozen studies devoted
to sediment dating and shoreline development have been done.6

4
N. Mercier, H. Valladas, L. Froget, J. L. Joron, P. M. Vermeersch, P. Van Peer
& J. Moeyersons, “Thermoluminescence dating of a Middle Palaeolithic occupation at
Sodmein Cave, Red Sea Mountains (Egypt)”, Journal of Archaeological Science 26 (1999),
1339–1345.—P. M. Vermeersch, E. Paulissen, S. Stokes, C. Charlier, P. Van Peer,
C. Stringer & W. Lindsay, “A Middle Palaeolithic burial of a modern human at
Taramsa Hill, Egypt”, Antiquity 72 (1998), 475–484.
5
P. Guibert, C. Ney & M. Schvoerer, “Datation croisée TL/C14 de cultures néo-
lithiques de la Vallée du Nil, Soudan, sites d’El Kadada et d’El Ghaba”, Archéologie du
Nil Moyen 5 (1991), 129–141.—P. Guibert, C. Ney, F. Bechtel, M. Schvoerer & F. Geus,
“TL and radiocarbon dating of neolithic sepultures from Sudan: intercomparison of
results”, Radiation Measurements 23 (1994), 393–398.
6
O. Bubenzer & A. Hilgers, “Luminescence dating of Playa Sediments of the
Egyptian Plateau, Western Desert, Egypt”, Quarternary Science Reviews 22 (2003), 1077–
1084.—H. M. El-Asmar & P. Wood, “Quaternary shoreline development: the north-
western coast of Egypt”, Quaternary Science Reviews 19 (2000), 1137–1149.
III. 3 DENDROCHRONOLOGY

Otto Cichocki

Definition

Dendrochronology is the science of dating tree rings. It includes inves-


tigations of the information provided by/contained in the structures of
dated ring sequences and its applications to environmental and histor-
ical problems/questions.1

History

The astronomer Andrew S. Douglass expected to find growth reactions


to the sun-spot cycle, when he measured the width of tree rings of old
pine trees. What he found was a similar pattern of ring width com-
mon to trees growing in the same area at the same time. He applied
this discovery to develop a method to date the remains of pueblo settle-
ments in the American Southwest.2
Subsequently the method was continually improved, especially by
computer based statistical analysis in laboratories worldwide.

Biological Background

In regions with seasonal variation the growth of trees and other woody
plants is not an invariable process but influenced by climatic factors.
Particularly outside the tropical and subtropical zones, growth stops

1
M. Kaennel & F. Schweingruber, Multilingual Glossary of Dendrochronology. Terms and
Definitions in English, German, French, Spanish, Italian, Portugese and Russian (Bern, 1995), 91
(modified).
2
A. E. Douglass, “The secret of the southwest solved by talkative tree rings”, National
Geographic Magazine 54 (1929), 737–770. Idem, “Dating Pueblo Bonito and other ruins
of the southwest”, National Geographic Society. Contributed technical papers. Pueblo Bonito series
(Washington, 1935), 1–74.
362 otto cichocki

almost completely in winter, causing a distinct border between each


annual increment. Depending on genus and species, these “tree-rings”
are composed of different cells in different arrangements. Early wood
is springwood formed at the beginning of the growing season for con-
ducting liquids, while late wood grows in late summer and autumn and
is composed mainly of cells with thicker walls. Both together form a
tree-ring. The thickness of the ring is more or less correlated with
precipitation (if this is the determining factor), with density corresponding
to the average temperature of the growth season. Of course, other cli-
matic factors, but also insect damage, wounding or the climatic situa-
tion of previous years, affect growth as well. Therefore each tree ring
is the result of a very complex accumulation of influences.
Trees growing in tropical and subtropical zones also form layers of
different cell arrangement, but since they lack distinct borders, these
layers cannot be measured for dating purposes. As annual thickness
growth is the result of the reaction of trees to ecological influences, in
an area with similar influences the growth reactions of different trees
on comparable stands in the same year will be similar.

Sampling, Data Acquisition, and Synchronisation

Old wood may be preserved in different conditions:


Dry preservation: wood is almost unaltered, but sometimes fragile.
It can be found in buildings (used to reinforce walls, construct ceilings,
roofs, lintels, or doors), in burials (coffins, other wooden objects, burial
chambers), as the ground of a painting (icons, Fayum portraits), sculp-
tures and other objects.
Wet-wood preservation: wood is almost unaltered, but very sensitive
to drying (necessitating storage in water, in a cool and dark environ-
ment). It can be found in rivers, lakes, caves, at the bottom of wells
(often together with other organic remains of importance).
Charcoal preservation: burning changes wood—it is chemically resis-
tant, but very fragile. If thicker beams are visible, they should be exca-
vated separately, bandaged and parts plastered together and packed
in plastic but not sealed to allow a very slow drying. Cool and dark
storage is best.
To measure the thickness of rings in a particular piece of wood, a
cross-section is required. A disc can be cut from wooden architectural
elements or a core removed with a drill. After smoothing the surface,
dendrochronology 363

the ring borders become visible. In most cases mixing samples of different
wood species is not recommended when constructing a standard, so
wood species analysis (identification of anatomical structures of a very
small sample with the help of a light microscope) of all objects avail-
able for investigation should be the initial step.
If the surface (square or longitudinal) of a wooden object is not cov-
ered with paint and has been well smoothed when made in antiquity,
the rings can often be measured directly after preliminary cleaning.
Measuring devices have been developed for this purpose since most
items in museum collections are only available for non-invasive “on the
object” measurements.3 VideoTimeTable equipment uses a digital video
camera with macro lens, which is moved along the surface by a step
motor. A live video image, displayed on a laptop, can be measured
immediately. Plain surfaces can be investigated with the help of a
modified high-resolution flatbed scanner. A device capable of drilling
a 5mm diameter hole for endoscopic measurement of ring width was
designed to measure objects with paint or other surface treatment.
At least two radii of a sample must be measured. The arithmetical
mean value of the ring widths is calculated to compensate for biolog-
ical diversity (e.g. elliptic ring shape, single growth deviations). The
result is a list of mean growth-ring thickness for each year contained
in the sample. A sample must contain at least 70 rings; otherwise in
many cases the statistical methods used for dating cannot work. Exceptions
to this rule are separate pieces of a single stem (e.g. charcoal), collected
as a unit. In this case much shorter overlaps allow reconstructing the
whole ring sequence of the stem because of almost identical ring pat-
terns. It is necessary to collect and measure as many promising sam-
ples as possible from one complex (a “time-unit”) in order to calculate
a reliable mean value list for further analysis.

Constructing Standards

Two samples grown at the same time in the same climatic conditions
will show a statistically significant correlation of their distribution of
wide and thin rings (respectively their mean values). Such tree-ring

3
B. Knibbe & O. Cichocki, “Developing new tools for the SCIEM 2000 project”,
Dendrochronologia, in press.
364 otto cichocki

patterns are significant for a certain time period, and a certain climatic
region, but in many cases only for a single genus (or species) of trees.
Cross dating (the statistical comparison of two mean value lists) aims
at the synchronization of different samples and determining relative
time span between them.
If two samples do not completely overlap in time, their mean pat-
tern will be longer than one of the individual samples. Many success-
fully cross-dated overlapping samples from different periods—the youngest
one being of known age (e.g. a recently cut tree)—allow constructing
a tree ring standard (a dated chronology) valid for one species and a
limited growth area. If a gap is open (a sample is missing), the isolated
part is called a floating chronology.

Dating of Samples

Dating a sample means comparing a sample with a standard on a year-


to-year basis to find the best matching position. In many cases statis-
tical tests result in more than one possible date. To find the correct
position, “pointer years” and additional statistical tests are used. When
setting up a standard by cross-dating all contained samples, certain
years form pronounced peaks in the graph. If more than 75 percent
of the rings from one year have the same trend (increasing or decreas-
ing growth compared with the preceding ring), this year with especially
strong influence on growth is called a “pointer year” and is specially
marked. In dubious cases, in a second match, only these pointer years
are compared with the corresponding rings of the sample. The correct
position will then show a significantly higher correlation of these spe-
cial years than the other theoretically positions.
The skeleton plot is a special method for dry areas using minimum
pointer years. Here, only very narrow rings are observed, as they are
formed by drought, that makes moisture the factor limiting growth. In
years with poor environmental conditions some species cannot create
distinct rings in the entire circumference of a stem or branch. If these
missing rings go undetected, the sample will not match well with the
standard. If this occurs when the standard is being set up, results will
be negatively effected. Detection is possible by comparing many sam-
ples, because the same ring is not usually missing in all of them. The
best way to detect incomplete rings is by comparing different radii on
cut discs.
dendrochronology 365

Another misleading growth pattern may occur in years with a tem-


porary period of low temperature or drought during the growing sea-
son which produces a “false ring”. This “late wood formation” within
the ring has a much smoother outer border than the real late wood
formed at the end of the growth season. Hence, it can be detected
under the microscope. As climatic factors are the reason certain rings
are formed, it is possible to extract a basic part of this information
contained in the ring. Dendroclimatology is concerned with recon-
structing precipitation and temperature as major climatic factors.4
Wood grows annual layer upon annual layer. To avoid confusion,
the age of a sample is usually defined as the absolute age of the
outermost ring preserved in the sample. If this ring is the last grown
before the death of the tree (the so-called waney edge, in some cases
preserved with bark), it is possible to date the cutting year and to iden-
tify the cutting season. Problems can arise from samples of wood in
secondary use or from undetected repairs, for they may give a false
higher or lower age, respectively, for their context.

Dendrochronology in Egypt and the Near East

The wood species to be investigated depend on their occurrence in


the objects available for investigation and their suitability for this
special method. Wood species analyses on Egyptian objects have been
carried out on coffins and other objects in the British Museum,5 on
statues, wooden toilette objects, musical instruments and objects from
the Coptic period in the Louvre,6 and on various objects in the Munich
collection.7

4
F. H. Schweingruber, Tree Rings. Basics and Application of Dendrochronology (Dordrecht,
1988).
5
W. V. Davies, “Ancient Egyptian Timber Imports. An Analysis of Wooden Coffins
in the British Museum”, in: W. V. Davies & L. Schofield, eds., Egypt, the Aegean and
the Levant. Interconnections in the Second Millennium BC (London, 1995).
6
A. Nibbi, Ancient Egypt and Some Eastern Neighbours (New Jersey, 1981).
7
D. Grosser et al., “Holz—ein wichtiger Werkstoff im Alten Ägypten”, in: S. Schoske
et al., eds., Anch—Blumen für das Leben. Schriften aus der ägyptischen Sammlung, Heft 6,
Staatliche Sammlung ägyptischer Kunst (München, 1992).
366 otto cichocki

Table III. 3.1

Type of Wood Nibbi 1981 Grosser 1992 Davies 1995


Louvre Munich and others British Museum

Ficus sycomorus 32 138 230


Tamarix sp. > 100 106 158
Acacia sp. 35 42 53
Cedrus sp. 5 (7) 44 88
Juniperus sp. 2 10 16
Pinus sp. 9 3 37
Cupressus sp. – 2 7

Table III. 3.1 provides a rough idea of the range of wood species used
in ancient Egypt for various purposes. Unfortunately the majority of
these species grow in Egypt (Ficus sycomorus, Tamarix sp., Acacia sp.)
and since they do not form distinct ring borders, they are unsuitable
for dendrochronological analysis. Only the gymnosperms can be used:
Cedrus libani grows in Mediterranean mountain climate and has distinct
ring borders. Its growth age is said to be as much as 500 years. Due
to over-deforestation there are only a few small areas where it grows
today in Lebanon, larger ones in Turkey (Taurus, Antitaurus), and in
Syria. The sub-species Cedrus libani atlantica grows in the Atlas Mountains
of Algeria and Morocco, Cedrus libani brevifolia in a very limited area in
the mountains of the island of Cyprus. There are no anatomical features
that distinguish these three sub-species of cedrus.
Cedar was imported to Egypt from the time of the Old Kingdom.
Different interpretations of Egyptian texts mentioning imports of wood
from certain countries, continue to fuel a heated debate about the origin
of cedar found in Egypt.8 Another contentious issue is the correct trans-
lation of different hieroglyphic terms obviously characterizing different
species of wood.
Nili Lipschitz carried out wood species analysis and dendrochrono-
logical investigations on several historical and archaeological sites in
Israel.9 She worked with wood of Cedrus libani, Cupressus semper-

8
Davies (n. 5); R. Meiggs, Trees and Timber in the Ancient Mediterranean World (Oxford
1982); Grosser et al. (n. 7); Nibbi (n. 6); idem, “Some Remarks on the Lexicon Entry:
Zeder, Cedar”, DE 7 (1987), 13–27.
9
N. Lipschitz, “Overview of the Dendrochronological and Dendroarchaeological
Research in Israel”, Dendrochronologia 4 (1986), 37–58.
dendrochronology 367

virens, Pinus nigra, Quercus cerris and Pistacia khinjuk. Dating the
roof beams of the El-Aqsa mosque in Jerusalem produced a 231 year
floating chronology with 14-C dates of the Byzantine period.10
The Aegean Dendrochronology Project (directed by Peter I. Kuni-
holm) is working on many archaeological sites in Turkey, Greece, and
Italy. In the International Treering Data Base, standards for the 2nd
millennium AD from Turkish forests are published for Cedrus libani,
Juniperus sp., Quercus conferta, Pinus nigra, Pinus sylvestris and Abies
nordmanniana.11 A floating chronology for Bronze Age and Iron Age
by 14-C wiggle matching is based on data for different tree species.12,13
Fritz Schweingruber published several modern standards for Cedrus
libani brevifolia from Cyprus in the International Treering Data Base.
A dendrochronological attempt at dating for Egypt continues/is car-
ried out within the long-term scientific project “The Synchronization
of Civilizations in the Eastern Mediterranean in the Second Millennium
BC” at the dendrolab of VIAS, University of Vienna. The dendro-
chronological investigations try to develop a dating approach indepen-
dent of other archaeological and scientific methods.14 First results are
a 507 year floating Cedrus libani chronology for the 2nd millennium
BC, and several shorter floating chronologies. To date this chronology
includes/subsumes the coffin of Sebekhetepi, the inner coffin of Gua,
and a canopic box (British Museum, London); the garden model of
Meketre (Metropolitan Museum of Art, New York); the coffins of Ashait,
Djehutinakht, Khnumhotep, Neferi, the outer coffins of Amenemhat
and Mesehti, and the shrine of Auibre Hor (Egyptian Museum, Cairo).
Wood samples from modern trees (Besharre, Barouk, Horsh Ehden)
and from several buildings from Deir el Kamar and Qadisha valley
(Lebanon) cover the centuries back to 1369 AD for further construc-
tion of a standard for absolute dating. Analysis of Bronze age charcoal

10
S. Lev-Yadun, “The Origin of the Cedar Beams from Al-Aqsa Mosque: Botanical,
Historical and Archaeological Evidence”, Levant 24 (1992), 201–208.
11
http://www.ncdc.noaa.gov/paleo/ftp-treering.html
12
P. I. Kuniholm, “A Date-List for Bronze Age and Iron Age Monuments based
on combined Dendrochronological and Radiocarbon Evidence”, in: M. J. Mellink, ed.,
Aspects of Art and Iconography: Anatolia and Its Neighbors. Studies in Honor of Nimet Özgüç
(Ankara, 1993), 371–373. Idem et al., “Anatolian tree rings and the absolute chronol-
ogy of the eastern Mediterranean, 2220–718 BC”, Nature 381 (1996), 780–783.
13
S. Manning et al., “Anatolian Tree Rings and a New Chronology for the East
Mediterranean Bronze-Iron Ages”, Science 294 (2001), 2532–2535.
14
O. Cichocki, “Cèdres libanais comme instrument de datation en Egypte”, Proceedings
of the 31CAANE Conference 2002; in press.
368 otto cichocki

samples from Arqa (Lebanon), Qatna, and Ebla (Syria) will enable test-
ing of the standard for the 2nd millenium BC with 14C and should
aid in determining whether Cedar wood was imported from Mount
Lebanon or from other forests.
Dendroclimatology studies the impact of climate on trees through
patterns of growth. As different factors (moisture, temperature, length
of vegetation period) interact in a very complex system, the investi-
gated parameters (ring width, early wood/late wood ratio, density) con-
tain a varying mean of signals. Investigations in the Near East carried
out by a team lead by Ramzi Touchan reconstructed precipitation for
southern Jordan. Currently this project has expanded to take in Turkey,
Lebanon, and Syria.15 Knowledge about climate impact on ancient
economies may aid analyses of political crises or warfare, which in turn
influenced wealth and long-distance trade.

15
R. Touchan, D. Meko & M. K. Hughes, “A 396-Year Reconstruction of Precipi-
tation in Southern Jordan”, Journal of the American Water Resources Association 35(1) (1998),
49–59.
III. 4 DATES RELATING TO SEASONAL PHENOMENA
AND MISCELLANEOUS ASTRONOMICAL DATES

Rolf Krauss

The rising and falling of the Nile and the harvesting of grain and fruit
occur regularly at certain times within the solar year. When their dates
were recorded in terms of the Egyptian calendar, they can be con-
verted into absolute dates. Occasions of quarrying expeditions have also
been traditionally considered examples of seasonal dates, since Egyptologists
conjectured that such work was not undertaken during the summer.1
Even if so, the conversions yield intervals that are too broad to be of
much use for chronology. For if cool weather lasts for 120 days, from
November to February, then a “cool season date” spans an interval of
4 × 120 years = 480 years. Although there might have been a ten-
dency to send expeditions to quarries during the cooler months, there
are nevertheless attestations for expeditions at the hottest time of the
year.2 Thus conversion of expedition dates can result in chronological
contradictions.

Dates of the Nile Flood

The Nile flood results principally from monsoon rains that fall over the
Ethiopian plateau between mid-May and September.3 In modern times
the dams constructed at Aswan beginning around 1900 have prevented
the annual flooding of the Nile Valley. Data recorded in the 19th cen-
tury and the Middle Ages provide information about the onset and
duration of the flood which are crucial for evaluating pharaonic dates.
The Nile sunk to its lowest level in April/May; towards the end of

1
Meyer, Chronologie, 177–178; idem, Nachträge, 20; C. J. Eyre, “Work and organiza-
tion of work in the Old Kingdom”, in: M. A. Powell, ed., Labor in the Ancient Near East
(New Haven, Connecticut, 1987), 16.
2
Beckerath, Chronologie, 53.
3
W. Willcocks, Egyptian Irrigation (London, 1889), 10.
370 rolf krauss

May and the beginning of June it began to rise. The biography of


Weni, a Dyn. 6 official, seems to mention a low water date.4 Weni
reports that he brought an altar from Hatnub to the pyramid of Merenre'
within 17 days in III Shemu, “although there was no water on the
Δzw”. Lieblein, who dated Merenre' far too early at ca. 2525 BC,5 cal-
culated that III Shemu 17 corresponded to March 5 Greg.,6 a time
when the river falls fast. Gardiner paraphrased Weni’s description with
“when the river was at its lowest”.7 If so, Merenre' would have reigned
between ca. 2817 and 2694 BC when III Shemu 30 coincided with
May 31 (Greg.) and III Shemu 1 with April 1 (Greg.). Following
Gardiner and using the standard chronology for the OK, Eyre cor-
rectly calculated the Gregorian months December and January as cor-
responding to III Shemu,8 but he wrongly designated these months as
“a time of low, if not the lowest, water”.9 There are various possibili-
ties to resolve the contradictions in Weni’s report: the translation of
Δzw as sandbanks may be wrong;10 the flood might have been low and
run off very early that year; Weni may even have exaggerated.
Another low water date refers to a difficult passage through the chan-
nels of the Semna rapids at Uronarti in 19 Senwosret III.11 Correlated
with the Dal inscription of year 10 of Senwosret III, the Uronarti
inscription “provides evidence for an extraordinary variability in the
Nile levels of late winter during the reign of Senwosret III”.12 Seasonal
dates that diverge from the statistical mean cannot be used to estab-
lish absolute chronology. In the following paragraphs the basic data
are presumed to be samples of the statistical mean.
64 Gregorian maximum flood dates on record from the Middle Ages
and the 19th and early 20th centuries refer to the Nile gauge at Roda
(Old Cairo), whereas the pharaonic high flood dates refer to Karnak
temple. At Roda the maximum height was reached between September

4
Urk. I, 108.—Meyer, Chronologie, 178; Nachträge, 20 n. 1.
5
J. Lieblein, “Eine chronologische Bestimmung”, ZÄS 44 (1907), 101–102.—Cf.
Borchardt, Mittel, 89 n. 3, for a similar approach.
6
The correct correspondance of III Shemu 17 in 2525 BC is March 7 Greg.
7
Gardiner, Egypt, 97.
8
If Merenre' reigned in ca. 2222 ± 6 BC, then III Shemu corresponded in his
reign to ca. December 8 to January 7 (Greg.).
9
Eyre (n. 1), 16.
10
Cf. J. Vandier, La Famine dans l’Egypte ancienne (Cairo: RAPH 7, 1936), 74–77.
11
Borchardt, Mittel, 91.
12
B. Bell, AJA 79 (1975), 238.
dates relating to a seasonal phenomena 371

25 and October 5.13 The earliest maximum date is August 14 and the
latest October 27. Using Borchardt’s incomplete Nile flood data, Beckerath
presumed that the maximum occurred at Roda as early as August 25
and as late as October 27, but 4 days earlier at Luxor.14 For the lat-
ter, Borchardt relied on the fact that in 1925 the maximum travelled
from Luxor to Roda within 4 days.15 However, the speed differed con-
siderably in other years: “The Nile at Assuân reaches its maximum
about the 5th September, and would under ordinary circumstances be
at its highest in Lower Egypt about the 11th September, but as the
basins of Upper Egypt are being filled in August and September, and
emptied in October, the maximum in Lower Egypt is ordinarily about
the 10th October.”16 The delay apparently resulted from interaction
between the filling of the irrigation basins and the size of the flood.
For example, in 1887 the Nile reached its maximum at Aswan on
September 1, at Armant on September 6, and at Roda on September
25.17 In that case 19 days elapsed between maxima at Luxor and Roda,
instead of 4 days as Borchardt had presumed.
Application of the 19th century flood data for Roda to the pharaonic
period presupposes a comparable irrigation system. Irrigation works of
some kind are first attested in the FIP,18 but very little information sur-
vives for later pharaonic history. There is no indication that a fully
developed system of basin irrigation existed in the NK or even earlier.
Under these circumstances it is preferable to use the maximum flood
dates of Aswan, instead of Roda, when calculating dates for the Luxor
region. The maximum required was at most 4 days for the distance
Aswan: Luxor.19 There are 34 maximum dates for Aswan on record;20
the earliest is August 18, the latest October 1, yielding maximum dates
for Luxor between August 21/22 and October 4/5. Based on a com-
parison of the dates at Aswan and Roda, it follows that the maximum

13
W. Popper, The Cairo Nilometer (Berkeley, 1951), 87–88.
14
Borchardt, Annalen, 7; Borchardt, Mittel, 90; Beckerath, Chronologie, 52.
15
Borchardt, Mittel, 90 n. 5.
16
Willcocks (n. 3), 10.
17
Willcocks (n. 3), 31, 184–185.
18
E. Endesfelder, “Zur Frage der Bewässerung im pharaonischen Ägypten”, ZÄS
106 (1979), 37–51; W. Schenkel, Die Bewässerungsrevolution im Alten Ägypten (Mainz, 1978),
35–36.
19
For varying velocities of the Nile, see W. Willcocks, Egyptian Irrigation2 (London,
1913), 143–144.
20
H. G. Lyons, The Physiography of the River Nile and its Basin (Cairo, 1906), 289–290.
372 rolf krauss

gauge occurred between 4 days (1882) and 63 days (1894) at Roda


later than at Aswan.21 Borchardt’s Roda-based maximum dates for
Luxor tend to be much too late, yielding an interval of ca. 4 × 64
years = 256 years for a pharaonic maximum date. By contrast, the
Aswan-based interval for maximum dates at Luxor results in a period
of ca. 4 × 45 years = 180 years. Records for only three events at high
flood are preserved:
(1) Sebekhotep VIII, Epagomenae of year 4: flooding of Karnak temple.22
(2) Osorkon III, year 3, III Peret 22:23 flooding of Karnak temple.
(3) Shebitku, year 3, I Shemu 5: royal visit to Karnak temple after a
high flood.24
(1) Provided Sebekhotep VIII ruled ca. 1600±150 years, then August
22 (Greg.) corresponds to September 5 ( Jul.) and October 5 (Greg.) to
October 19 ( Jul.). These Julian dates correspond to the E(pagomenae)
as follows:
E 1 = Oct 19 Jul in E 1 = Sept 5 Jul. in
1709/06 BC 1532/29 BC
E 5 = Oct 19 Jul in E 5 = Sept 5 Jul. in
1693/90 BC 1516/13 BC
The flood in 4 Sobekhotep VIII occurred, then, between 1709 and
1513 BC; the mean year is 1611 BC. By contrast, Beckerath obtained
the limits 1869 and 1534 BC and the mean year 1701 BC.
(2) Provided Osorkon III ruled around 700 BC, then August 22
(Greg.) corresponds to August 29 ( Jul.) and October 5 (Greg.) to October
13 ( Jul.). The Julian dates for III peret 22 are:
III peret 22 = October 13 ( Jul.) in 861/858 BC
III Peret 22 = Aug 29 in 681/678 BC

21
For the maxmimum dates at Roda see Lyons (n. 20) 321 and M. Clerget, Le
Caire I (Cairo, 1934), 44 (correct “1880 août” to “1880 septembre”).
22
L. Habachi, SAK 1 (1974), 207–214; J. Baines, Acta Orientalia 36 (1974), 39–54;
idem, Acta Orientalia 37 (1976), 11–20.
23
Reading after Schott; see Borchardt, Mittel, 91.
24
J.v. Beckerath, JARCE 5 (1966), 53; idem, GM 136 (1993), 7; idem, Chronologie
92, 52.
dates relating to a seasonal phenomena 373

Therefore the flood in 3 Osorkon III occurred between 861 and 678
BC; the mean year being 770 BC.25 By contrast, Beckerath obtained
the limits 932 and 677 BC and 805 BC as the mean year.
(3) The text that relates to Shebitku’s appearance in Karnak men-
tions the Nile level “of his time (m h'w.f )”. Borchardt interpreted the
visit as coincidental with a maximum flood whose occurrence was slightly
delayed.26 He conceded that the maximum flood could have occurred
before the king’s visit. I Shemu 5 corresponded to October 9 (Greg.)
in 705 BC = 3 Shebitku.27 It cannot be excluded that the visit coin-
cided with a much delayed maximum flood, although more probably
the maximum occurred earlier, in September.

Opening of the Basin Canals in the Theban Area?

Ramesside graffiti record dates when “this day, the water returned/came
to inundation”28 which were thought to refer to the falling of the flood
or the feast “night of the drop” or the onset of the flood.29 More
recently Janssen believes the dates to refer to the opening of the basin
canals that took place in the 19th century around August 12. He sup-
poses that “the situation in Pharaonic times was not essentially different
than in the nineteenth century AD,” and that the Ramesside flood
dates also fell around August 12 (Greg.).30 Beckerath presumes that the
dams were opened depending on the height of the flood, with the mean
August 12 (Greg.).31 Both authors overlook the recorded indiviual dates
for the opening of basin canals in the 19th century. In Kena province,
and thus at Luxor, the dates deviated from August 12 by +3 days and
–6 days at most.32 August 12 was a deadline set by the Khedive that

25
Borchardt, Mittel 91 n. 6, noted that the day coincided with a procession of Amun
(line 5 of the text); he expected a full moon. He may have been correct coinciden-
tally, if the Tepi Shemu date I Shemu 6, year 18 (Fitzwilliam 68d) is correctly ascribed
to Osorkon III as Kruchten, Annales, 144, 240 suggests.
26
Borchardt, Mittel, 91.
27
Cf. above Jansen-Winkeln, Chapter II. 10, and Zibelius-Chen, Chapter II. 12.
28
J. J. Janssen, JNES 46 (1987), 125–136.
29
Meyer, Nachträge, 39–42; with additional literature.
30
Janssen (n. 28).
31
Beckerath, Chronologie, 52.
32
Willcocks (n. 19), 335.
374 rolf krauss

could be missed if a harvest of sorghum was imminent in a basin.33


The interpretations of Janssen and Beckerath presuppose that the sys-
tem of basin irrigation existed already in Ramesside times and was
managed as in the 19th century. No NK sources support this idea.
The Ramesside flood dates may refer instead to the flooding of the
valley after the Nile had breached the levees.

Grain Harvesting Dates

Ploughing and sowing were done as soon as the flood had receded.
According to one MK source, the fields were measured for tax assess-
ment when the seed had sprouted,34 whereas NK sources seem to indi-
cate the measuring of ripening fields.35 Thus the measuring of fields
cannot be used to establish absolute chronology.
Traditionally, Egyptologists assumed that barley was harvested between
the end of February and the beginning of March (Greg.), and wheat
in April (Greg.).36 By contrast, in pharaonic times the harvest could
have began as early as February and ended as late as in May (Greg.).37
The delivery of grain occurred later, after time-consuming threshing,38
winnowing, and cleaning.39 Most references to collection and delivery
of grain are Ramesside and date to summer and fall months (Greg.);40
they are only suited for a rough determination of Ramesside chronology.41
By contrast, the correct harvest date is crucial for the hypothetical core-
gency of Thutmose III with Amenhotep II.42

33
Willcocks (n. 19), 38 (304).
34
P. Smither, JEA 27 (1941), 74–76.—R. Krauss, “Detailfragen der altägyptischen
Getreidewirtschaft”, in: Form und Mass. Fs für Gerhard Fecht, J. Osing & G. Dreyer (eds.),
(Wiesbaden, 1987), 268–269.
35
Cf. J. J. Janssen, BiOr 49 (1986), 356, and W. Helck, LÄ II, 152–153.
36
Cf. for example, Beckerath, Chronologie, 53.
37
R. Krauss, “März, April und Mai als durchschnittliche Monate der Getreideernte
im antiken und neuzeitlichen Ägypten” DE 27 (1993), 27–34
38
For threshing dates as late as August/September (Greg.), see Krauss (n. 37).
39
H. Felber, “Die Daten in den demotischen Ackerpachtverträgen der Ptolemäerzeit
und das landwirtschaftliche Jahr”, APF, Beiheft 3 (1997), 281–289.
40
For example, pTurin 1895+2006 lists collection dates between May 29 and
September 4 (Greg.); cf. A. H. Gardiner, JEA 27 (1941), 22–37.
41
Beckerath, Chronologie, 53.
42
Cf. above Hornung, Chapter II. 8, and below Krauss, Chapter III. 8.
dates relating to a seasonal phenomena 375

Wine Delivery Dates

Since the Middle Ages and down to the present, grapes ripen in Egypt
towards the end of June;43 vintaging and wine-making occurred in
August (Greg.).44 If the same conditions prevailed in the NK, the sealing
of the wine jars would have taken place after fermentation towards the
end of August/beginning of September (Greg.).45 Jar labels mention the
regnal year of the sealing; the month is very seldom cited.46 The labels
name the chief vintner and the wine estate, but not the ruling king.
Great quantities of dated wine jars were excavated at Malqata47 and
Amarna, at the Ramesseum,48 and in Deir el-Medina.49 At Amarna wine
jar labels of 1+16 successive vintages were found. The labels document
the later part of Akhenaten’s reign, and the reigns of two successors.50
A wine jar from the funerary temple of Amenhotep II bears the date
“regnal year 26”.51 If the year is ascribed to Amenhotep II, then the
sealing of the jar took place in ca. II Akhet,52 10 months after the
beginning of year regnal 26.53 It may be surmised that the reign of 25
y + 10 m that Manetho (Flavius Josephus) ascribes to Mephramuthoses,
the precursor of Thmosis < Thutmose IV, belongs to Amenhotep II.

Harvesting of Flax

In the tomb of the nomarch Djehutinakht (Bersheh 1), a scene of har-


vesting flax is dated to IV Akhet 23.54 The nomarch is datable to
around year 31 of Senwosret I.55 Meyer56 and then Borchardt57 relied

43
Krauss, SAK 23 (1996), 237 n. 67.
44
C. Pellat, Cinq calendriers égyptiens (Cairo, 1986), 247, s.v. vin.
45
R. Krauss, MDOG 129 (1997), 227; idem, SAK 23 (1996), 238–239; with addi-
tional literature.
46
Cf. Hornung, Untersuchungen, 78 n. 51.
47
W. C. Hayes, JNES 10 (1951), 41–56.
48
W. Spiegelberg, ZÄS 58 (1923), 25–36; Beckerath, Chronologie NR, 40.
49
Beckerath, Chronologie NR, 40.
50
See above Hornung, Chapter II. 8.
51
Urk. IV, 1365; Beckerath, Chronologie NR, 94; idem, Chronologie, 109.
52
II Akhet 1 = September 9 Jul. = August 26 Greg., in 1400 BC.
53
See above Hornung, Chapter II. 8.
54
PM IV, 177.
55
H. G. Fischer, LÄ II (1977), 414.
56
Meyer, Nachträge, 189–20.
57
Borchardt, Mittel, 89–90.
376 rolf krauss

on information in the Description de l’Égypte that flax harvesting took


place in the region of Asyut at the beginning of April and lasted 8 to
10 days. But rather than this isolated date, the general date for flax
harvesting in Egypt should be used, viz. March to April.58 Furthermore,
the date is not necessarily contemporaneous with Djehutinakht, for the
scene with its date could have been copied from another earlier tomb.
IV akhet 23 corresponds to March 15 Greg. = March 31 Jul. in
1893/1890 BC and to April 15 Greg. = May 2 Jul. in 2021/2018 BC.
Harvesting of flax as early as March 15 Greg. or later is compatible
with the low chronology (Senwosret I reigned 1920–1875 BC), but also
with a slightly higher one.

Seasonal Finds in the Tomb of Tut'ankhamun

Provided the times of flowering and ripening are known and if the
material did not come from storage, the season when plant mate-
rial was deposited in a tomb can be determined in terms of the Egyp-
tian calendar. By far the most important example is the tomb of
Tut'ankhamun with its great diversity of plant remains. The flowers
that were used in the wreaths (Picris asplenoides L., Centaurea depressa,
Whitania somnifera (L.) and Nymphaea caerulea Sav.) blossom in April;
the Christthorn fruit (Zizyphus spina Christi), of which great quantities
were found, also ripens in April. The absence of oil seed and fruit
common in other tombs, which was harvested in the summer months
(Balanites Aegyptiaca, Ricinus communis; Cyperus esculentus, Punica
granatum and Ficus carica), indicates that the burial took place when
these fruits were already used up. Thus it is to be concluded that the
burial took place in April (Greg.).59 Provided the burial followed a reg-
ular mummification process of 70 days, the king would have died
between January 21 (Greg.) and February 20 (Greg.). If he died in or
ca. 1322 BC the interval corresponds to III peret 18/IV peret 18 which
would also subsume the accession date of Aya, otherwise not known.60

58
R. Krauss, BSEG 15 (1991), 79; with additional literature.
59
P. E. Newberry, in: H. Carter, The Tomb of Tutankhamen II (London, 1927), 189–196;
R. Germer, Die Pflanzenmaterialien aus dem Grab des Tutanchamun (Hildesheim: HÄB 28,
1989), 4–26; R. Krauss, “Nochmals die Bestattungszeit Tutanchamuns”, SAK 23 (1996),
227–254.
60
Cf. above Hornung, Chapter II. 8.
dates relating to a seasonal phenomena 377

Tut'ankhamun’s death in January/February raises an obstacle to his


identification with Nipkhururia, who apparently died at the end of sum-
mer or early in the fall.61

Solar and Lunar Eclipses

There are no reports of solar and lunar eclipses from pharaonic Egypt.
Thus ancient Egyptian data are absent in modern astronomical analy-
sis of recorded eclipses.62 A text from year 15 of Takelot II was thought
to refer to a lunar eclipse. If the verb in question is understood as
s≈m.f, then the translation is “the sky did not devour the moon”,63 if
the form is interpreted as n s≈mt.f, then the translation is “before the
sky swallowed the moon”.64 Regardless, it was already known in the
19th century that there was no lunar eclipse which would fit the stan-
dard chronology of the period;65 it seems possible that the text refers
to a delayed appearance of the new crescent.66
A solar or lunar eclipse supposedly occurred when Psammetichus I
died.67 The source is the Demotic papyrus Berlin 13588, written in late
Ptolemaic or early Roman times.68 The papyrus relates how a priest
heard that the sky swallowed the disk ( jtn) when Psammetichus I died;
later the priest copied the “Book of Breathing” onto the mummy wrap-
pings of Psammetichus I. The reported time of the supposed eclipse
does not suit the solar eclipse of September 30, 610 BC; instead, it
would fit the lunar eclipse of March 22, 610 BC. The context of the
eclipse is fictitious, insofar as it would date the “Book of Breathing”—
a creation of the Ptolemaic period—to the time immediately after the

61
Cf. above Klinger, Chapter II. 13.
62
Stephenson, Eclipses, 58–59, and J. M. Steele, Observations and Predictions of Eclipse
Times by Early Astronomers, (Dordrecht, 2000), 6.
63
Caminos, Chronicle, 88–89.
64
K. Jansen-Winkeln, SAK 21 (1994), 127.
65
Cf. Kitchen, TIP, 181–182; Beckerath, Chronologie, 41–42.
66
R. Krauss, “Die Bubastiden-Finsternis im Licht von 150 Jahren Forschungs-
geschichte”, MDAIK 63 (2007), in press.
67
O. Neugebauer, The Exact Sciences in Antiquity2 (Providence, 1957) 95 n. 2; E. Hornung,
ZÄS 92 (1966), 38–39; M. Smith, OLP 22 (1991), 101–109.
68
W. Erichsen, “Eine neue demotische Erzählung”, Akademie der Wissenschaften und
der Literatur Mainz. Abhandlungen der geistes- und sozialwissenschaftlichen Klasse 1956,
Nr. 2, 49–81.
378 rolf krauss

death of Psammetichus I. An eclipse that is reported in a fictitious tale


cannot be deemed historical.69
Eclipses that were recorded in the Near East such as the solar eclipse
of 15 June 763 BC (see above Klinger, p. 308) are of indirect impor-
tance to Egyptian chronology. The solar omen of year 10 of Murshili
II would be important for the chronology of the Amarna period and
its aftermath, if it were indeed an eclipse. The text says that the sun
god “gave an omen”; the kind of heavenly phenomenon (eclipse?) is
open.70

Miscellaneous Astronomical Phenomena

The observations of the planets Mercury (Seth) and Venus (Horus, Eye
of Horus) that are reflected in the Cairo “Calendar of lucky and unlucky
Days” are datable to 1298/97 BC, whereas the manuscripts are at least
80 years younger.71
There are, for example, no Egyptian reports about periodic comets,
singular Novae or occultations. Whether the “star miracle of Thutmose
III” reflects an actual event or phantasy is moot.72
J. Herschel dated the Giza pyramids on the assumption that the ori-
entation of the ascending corridors was determined by the position of
the “polar star” at the time when the pyramids were laid out.73 Herschel
subsequently retracted his proposal and accepted the explanation that
the angle of the corridors was chosen to facilitate their filling with large
blocks after the burial of the pyramid’s owner.74 But this was over-
looked by later astronomers and Egyptologists who criticized his orig-
inal idea.75

69
Krauss (n. 66).
70
Ph. van den Hout, Purity of Kingship (Leiden, 1998) 42ff.; D. Schwemer, in: Die
Hethiter und ihr Reich. Das Volk der 1000 Götter. Exhibition catalogue (Stuttgart, 2002),
140–145, at 144.
71
R. Krauss, “The Eye of Horus and the Planet Venus: Astronomical and Mythological
References”, in: Under One Sky, J. M. Steele & A. Imhausen, eds., (Münster: AOAT
297, 2002), 193–208.
72
Urk. IV 1238; cf. D. Meeks, LÄ IV, 117–118.
73
J. Herschel, Outlines of Astrononomy 6 (London, 1859), 205–207.
74
M. Brück, Journal of the British Astronomical Association 105.4 (1995), 161–164.
75
E.-M. Antoniadi, L’astronomie égyptienne depuis les temps les plus reculés jusqu’à la fin de
l’époque Alexandrine (Paris, 1934), 146; J.-Ph. Lauer, Le problème des pyramides d’Egypte (Paris,
1952), 187–188.
dates relating to a seasonal phenomena 379

According to a hypotheses proposed by K. Spence, the pyramids of


Dyns. 4 and 5 were oriented towards north utilizing the “simultaneous
transit method”, i.e. projecting the chord joining two circumpolar stars
on opposite sides of the celestial north pole at the moment of their
respective upper and lower culmination.76 In the course of decades and
due to precession the method would yield first values east (west) of
north, then true north, later values west (east) of north. If the method
was applied in the 26th and 25th century BC to the stars Mizar (z UMa)
and Kochab (b UMi), then the resulting pattern seems to fit known
pyramid alignments.77 But the orientation of Djedefre'’s pyramid which
became known in 2001,78 does not fit into the sequence, and thus the
hypothesis cannot be correct.79
Moreover, Spence presumes that it was the west sides of the pyra-
mids which were astronomically aligned; in fact, the corridors on or
parallel to the central axis are decidedly closer to true north than the
west sides and it is likely that they were the features to be astronom-
ically aligned. However, the orientation of the corridors does not adhere
to a scheme that could have been related to precession and thus the
pyramids cannot be dated by this means.80
Finally, it should be emphasized that the angles of the air shafts in
the “Great Pyramid” (pyramid of Cheops) are useless for dating the
structure astronomically.81 Another senseless exercise involves using
the presence or omission of certain planets in the representation of the
astronomical ceiling in the tomb of Senenmut for dating purposes.82

76
K. Spence, Orientation in Ancient Egyptian Royal Architecture (unpubl. Diss., Cambridge,
1997); idem, Nature 408 (2000), 320–324.
77
Cf. J. A. Belmonte, “On the orientation of the Old Kingdom pyramids”, Archaeo-
astronomy 26 ( JHA XXXII) (2001), S1–S20.
78
E. Aubourg & C. Higy, “Détermination de l’orientation de la pyramide de
Rêdjédef ”, BIFAO 101 (2001), 457–459.
79
Cf. also A.-A. Maravelia, “L’horizon astral de Khéops. Archéoastronomie, égyp-
tologie . . . et quelques scénarios de science-fiction”, Tôzai 5 (2000), 11–37.
80
R. Krauss, “¿Las ilusiones perdidas? Recientes intentos en Arqueoastronomía en
Egipto”, BAEDE 16 (2006), in press.
81
See R. Krauss, “Los canales en la pirámide de Kheops”, BAEDE 13 (2003),
55–66, rejecting Brück (n. 74).
82
As C. Leitz, Studien, 35ff. does; accepted by Beckerath, Chronologie NR, 94
n. 581; refuted by R. Krauss, GM 146 (1995), 61–70.—O.v. Spaeth, Centaurus 42 (2000),
159–179; countered by C. Leitz, Centaurus 44 (2002), 140–142.
III. 5 ASTRONOMY ON THE HORIZON AND DATING —
A TOOL FOR ANCIENT EGYPTIAN CHRONOLOGY?

Juan Antonio Belmonte

Astronomy can offer a third dating option, over and above the analy-
sis of lunar and Sothic dates, viz. the possibility of dating monuments
using solar, lunar or stellar alignments depending upon the variation
of stellar coordinates due to precession or the variation of ecliptic obliq-
uity. For such analysis, astronomy on the horizon is the most relevant
tool. The pioneering work in this area is Norman Lockyer’s The Dawn
of Astronomy,1 considered today by some archaeoastronomers as the first
“serious” book in their discipline. The author made ample use of pre-
cession to date Egyptian temples to support the long chronology which
was accepted in his day (Dyn. 1, ca. 5000 BC). When Egyptologists
discarded such chronologies any possibility of their using archaeoas-
tronomy as a chronological tool disappeared with it. In the 1970s Gerald
Hawkins2 reopened the discussion; although the topic was promoted by
Edwin Krupp,3 there was no noticeable response from Egyptologists.
Astronomical alignments are either directed towards the horizon or
towards lower zenithal distances, including alignments towards zenith
pass. The latter has seldom figured in archaeoastronomical studies,
although it has the advantage of offering fewer problems. But its exis-
tence is difficult to demonstrate, except in the case of zenith pass for
which Mesoamerica furnishes several good examples, but Egypt none.4

1
J. N. Lockyer, The Dawn of Astronomy (London, 1894).
2
See e.g. G. S. Hawkins, Beyond Stonehenge (New York, 1973); idem, “Astroarchaeology:
The Unwritten Evidence”, in: A. Aveni, ed., Archaeo-astronomy in Precolombian America
(Austin, 1975), 131–162.
3
E. C. Krupp, In search of the Ancient Astronomers (New York, 1977), 208–219; idem,
“Egyptian Astronomy: Temples, Traditions, Tombs”, in: Archaeoastronomy and the Roots
of Science. AAAS Sym. 71 (Boulder, 1984); idem, Beyond the Blue Horizon (Oxford, 1991).
4
For Mesoamerica, see A. F. Aveni, Skywatchers of Ancient Mexico2 (Austin, 1990); for
a hypothetical use of zenith passage in Egypt, see J. A. Belmonte, “Some open ques-
tions on the Egyptian calendar: an astronomer’s view”, Trabajos de Egiptología 2 (2003),
7–56.
astronomy on the horizon and dating 381

The controversial theories relying on simultaneous star transit to explain


pyramid alignments for dating purposes exemplifies the problem.5 Part
of the “Cosmology of Nut” system, present in the tomb of Ramesses
IV and the Osireon of Abydos, and the Ramesside clock devices in
the tombs of Ramesses VI, VII and IX probably also related to stars
in positions far from the horizon.6
Astronomy dealing with phenomena near the horizon includes heli-
acal risings (and settings) which are basic for the understanding of
decanal star clocks.7 However, in what follows I shall concentrate on
possible stellar or (luni)solar alignments on or near the horizon as a
means of dating pharaonic monuments.
In Fig. III. 5.1 there are three problems, two of atmospheric origin
(refraction and extinction), and one of a topographic nature (rough
horizon). In the case of (luni)solar observations, the size of the solar
disc presents a fourth problem. The azimuth a of the rising or setting
of a celestial object can be calculated in principle by using simple spher-
ical trigonometry. Consequently, if a building were oriented with ref-
erence to a certain celestial body,8 it might be possible to calculate the
date of the building’s foundation.
Simple spherical trigonometric calculations would apply only to a
planet without atmosphere and with a flat surface. In reality, a celes-
tial body is never seen rising or setting at a; Figure 1 illustrates the
following actual possiblities:
a’: The setting of the star, if only refraction is taken into account,
e.g., at a flat desert horizon. This also varies because refraction is espe-
cially dependent on atmospheric conditions such as humidity, temperature,

5
K. Spence, “Ancient Egyptian Chronology and the Astronomical Orientation of
Pyramids”, Nature 408 (2000), 320–324.—Challenged by A. A. Maravelia, “L’horizon
astral de Khéops”, Tozai 5 (2000), 11–37, and J. A. Belmonte, “On the orientation of
the Old Kingdom pyramids”, Archaeoastronomy Supplement, JHA 26 (2001), S1–S20.
6
See C. Leitz, Altägyptische Sternuhren (Louvain: OLA 62, 1995); J. A. Belmonte, “The
Ramesside star clocks and the ancient Egyptian constellations”, in: Calendars, symbols
and orientations: legacies of astronomy in culture, Proceedings of the SEAC 9th Annual Meeting,
Uppsala Astronomical Observatory Report 59 (2003), 57–65.
7
On the decans see J. A. Belmonte, “The decans and the ancient Egyptian sky-
lore: an astronomer’s approach”, Proceedings of the INSAP III Meeting; Memorie della Societa
Astronomica Italiana 63 (2001), special vol. I, 43–57.
8
See e.g., Leitz’ controversial Studien, and J. A. Belmonte, JHA 26 (2001), S1–S20.
See also the only slightly outdated study of Z. Zaba, L’orientation astronomique dans l’an-
cienne Égypte, et la précession de l’axe du monde (Prague: Archiv Orientální Suppl. 2, 1953).
382 juan antonio belmonte

Sun

Refraction

Extinction

Rough Horizon a
a a
1/2°

a
a
h=0°
Figure III. 5.1: Real setting track (black) of a celestial object (here: Venus) under
extreme atmospheric conditions. (On the basis of a photograph taken on southern
Tenerife, Canary Islands, Lat. 29º.) Compare the theoretical astronomical setting azimuth
a to the actual azimuth at last visibility of the star at a or the azimuth under ideal
atmospheric conditions only affected by low refraction and rough horizon a’’’. The
theoretical star track is in white; topographical features and size of the sun disk are
at the same scale; the slightly elliptical shape of the disk is due to refraction.

and the presence of dust or haze near the horizon, an extremely


frequent occurrence in Egypt.
a’’: The azimuth of the actual last visibility of the star is due not
only to refraction but particularly to atmospheric extinction. According
to a rule of thumb, a star becomes visible above the horizon, if its
angular height is at least equal to its magnitude. Accordingly, under
the best circumstances only Sirius, Vega, Rigil Kentaurus or Arcturus,
astronomy on the horizon and dating 383

or the brightest planets, would be visible at 0º altitude. However, this


too, is very dependent on atmospheric conditions; dust clouds (“calimas”)
or haze can severely affect visibility. In particular, Figure 1 represents
a setting of Venus in southern Tenerife when the planet had the mag-
nitude minus 2. However, sighting of Venus was lost when it was still
more than 2º above the horizon. It was a dusty day, as often occurs
at Saharan latitudes like the Canary Islands or Egypt.
a’’’: Setting azimuth of the star taking into account refraction and
rough horizon. This would have been the actual setting azimuth of
Venus on this occasion, provided the atmosphere would have been
much clearer and more stable on that night.
a’’’’: Theoretical setting azimuth considering only rough horizon and
no atmosphere. This value is obtained from standard azimuth (ames) and
height (hmes), with measurements taken either with a theodolite or a
tandem when aligning a specific structure. For the sun or the moon,
one should consider, apart from parallax, the size and shape of the
disc (the latter dependent also on refraction and extinction) which at
a rough horizon, can substantially change the position of the last con-
tact event (or first contact for rising), and, consequently, the alignment
of a building.
On this basis I can affirm without reservations that a precision of
½º in determination of azimuth is perhaps the best one can expect for
solar or very bright star observations near the horizon in Egyptian lat-
itudes. For fainter stars, like those of the Foreleg (ms¢tyw) or Orion (s #˙),
or important asterisms, like the Pleiades (¢#w), the error in azimuth can
range between one and several degrees of arc.9 Because of this variation,
Haack’s theory of pyramid orientation was not taken seriously and his
discovery of the error versus time trend ignored.10 For the same reason
Isler11 and Edwards12 were forced to abandon horizontal astronomy in
favor of a cast shadow system or an artificial horizon, respectively.

9
See the figures in J. A. Belmonte & M. Hoskin, Reflejo del Cosmos: Atlas de
Arqueoastronomía del Mediterráneo Antiguo (Madrid, 2002), 25. They are based on R. M.
Sinclair & A. Sofaer, “A method for determining limits on the accuracy of naked-eye
locations of astronomical events”, in C. L. N. Ruggles, ed., Archaeoastronomy in the 1990s
(Loughborough, 1993). See also B. E. Schaefer, “Atmospheric extinction effects on stel-
lar alignments”, Archaeoastronomy Supplement JHA 10 (1986), S32–S42.
10
S. C. Haack, “The Astronomical orientation of the Egyptian Pyramids”, Archaeo-
astronomy Supplement, JHA 7 (1984), S119.
11
M. Isler, “An ancient method of finding and extending direction”, JARCE 26
(1989), 191–206.
12
I. E. S. Edwards, The Pyramids of Egypt 3 (Harmondsworth, 1993).
384 juan antonio belmonte

Does this mean that astronomical alignments are completely useless


for chronological issues? The answer is yes and no. A recent study by
Gabolde13 illustrates the problem for a solar alignment. There is gen-
eral agreement that Amun’s temple at Karnak was aligned towards the
rising sun on a winter solstice.14 However, Gabolde’s attempt to fix
the date as the winter solstice following July 27 in 1946 BC, during
the reign of Senwosret I, for laying the temple axis should be viewed
with extreme caution. The change of solar declination due to variation
of ecliptic obliquity is of the order of only 0”.46 per year. Therefore
the change of ½º occurs only after 3900 years, which is approximately
the time elapsed between Senwosret I and the present. Thus the
minimum error equals the variation, making dating tenuous.15 Either
those who set the axis were extremely good topographers and precise
technicians working under perfect atmospheric conditions, or I should
conclude that we are faced with a very lucky situation, one in which
fine calendrical determination and timing, splendid weather and the
technical skill of ancient Egyptians conspired to yield an almost per-
fect alignment. If standard chronology did not support this epoch, I
would have never been confident about fixing the date of Senwosret
I’s tenth regnal year just using solar alignment.
For stars, the situation is different. Precession can substantially change
the coordinates of a certain star and thus its rising and setting azimuths
or, in general, its position in relation to the local horizon (e.g. angu-
lar height and moment of culmination). Sirius is exemplary for the case
of horizontal astronomy. Between 3000 BC and 500 AD, the declina-
tion of Sirius changed from –22º.7 to –15º.8 and its rising azimuth
changed from 116½º to 108⅓º for the latitude of Cairo (30º). The ris-
ing azimuth would change from 115º to 107⅓º for the latitude of
Philae or Aswan (nearly 24º), where important temples of Isis or Satet
were located, the divinities who manifest themselves in Sirius.16 Presuming

13
L. Gabolde, Le “Grand Château d’Amon” de Sésostris I er à Karnak (Paris: MAIBL, N.S.
17, 1998), 123–137.
14
For this controversial topic, see M. Shaltout & J. A. Belmonte, “On the orienta-
tion of ancient Egyptian temples: (1) Upper Egypt and Lower Nubia”, JHA 36 (2005),
273–298, with references; for a different view see R. Krauss, “¿Las ilusiones perdidas?
Recientes intentos en Arqueoastronomía en Egipto”, BAEDE 16 (2006), in press.
15
According to my personal impression the foundation blocks of the original MK
temple do not permit a much better precision even with the use of a theodolite.
16
See R. A. Wells SAK 12 (1985), 255–302, for a study of the orientation of the
Satet temple at Elephantine with reference to Sirius.
astronomy on the horizon and dating 385

an error of ½º, we can obtain a precision on average of nearly a quar-


ter of a millennium or, under excellent atmospheric conditions, per-
haps a little better. Obviously, the azimuth dating device is not very
precise, but not useless. If there is textual evidence, as, for instance, in
Dendera and perhaps in Philae, we might be able to date alignments
to Sirius within some centuries.
In the case of fainter stars their coordinate variation and their mag-
nitudes are decisive, but, in general, the results are less reliable than
for Sirius.17 Consequently, for other stars or planets, such us the bright
stars of the Foreleg, and for other epochs, we can make estimates per-
haps to the nearest century in the best cases, even though, on some
occasions, we might be very lucky, as in the example of Karnak.
Astronomy near the horizon cannot be used as an appropriate, and
certainly not as a definitive tool by itself for establishing the precise
parameters of pharaonic chronology. However, the approach is useful
nevertheless, for it provides insights into the role of astronomy within
the culture of ancient Egypt, particularly in the religious sphere.18

17
Except for stars where precession can produce a substantial change in coordinates.
Alnilan (e Ori), the central star of the s #˙ constellation, is a good example; its decli-
nation changed from –17º.3 in 3000 BC to –3º.4 in 500 AD. A still better example
is Arcturus (probably to be identified as a bright star of the constellation mnjt), which
changed from 48º.6 to 27º.6 in the same period. In these cases, ½º error in azimuth
would permit a precision of 110 and 60 years, respectively, for the latitude of Cairo.
18
See, for example, G. de Young, “Astronomy in Ancient Egypt”, in: H. Selin, ed.,
Astronomy across Cultures (Amsterdam, 2000), 475–508; J. A. Belmonte, “Astronomía y
arquitectura: el papel de los astros en la cultura y el arte del antiguo Egipto”, in M. A.
Molinero & D. Sola, eds., Arte y Sociedad del Antiguo Egipto (Madrid, 2000), 109–136.
III. 6 LUNAR DAYS, LUNAR MONTHS,
AND THE QUESTION OF THE ‘CIVIL-BASED’
LUNAR CALENDAR

Rolf Krauss

The Days of the Lunar Month

The ancient Egyptians observed the phases of the moon; they counted
and named the days of the lunar month as well. Early on Brugsch
compiled a list of the names of the lunar days; they are readily acces-
sible in Parkers’s Calendars.1
The earliest attestations for lunar days occur in private and royal
inscriptions of Dyns. 4 and 5. The Palermo stone preserves the earliest
royal example: LD 6 = snwt, a day of offering at Heliopolis in regnal
year 6 of Weserkaf.2 The lunar days ps≈ntyw, #bd, snwt, dnjt and smdt,
i.e. LD 1, 2, 6, 7 and 15 are attested as days of rituals in the Pyramid
Texts.3 Spalinger has collected and analysed the private feast lists of
all periods.4 The early lists mention the lunar days #bd and smdt as well
as s #≈,5 but not ps≈ntyw. As Spalinger notes, “when one descends in
time from the Old Kingdom to the very last phases of Pharaonic civ-
ilization, the number of lunar-based feasts diminishes”.6 But lunar days
are attested throughout Egyptian history and can be utilized for chrono-
logical analysis, if they are combined with dates of the civil calendar.7

1
Parker, Calendars, 11–12; cf. also Belmonte, “Questions”, 35.
2
Wilkinson, Annals, 153–155.—The reading of A. Roccati, La littérature historique sous
l’Ancien Empire égyptien (Paris, 1982), 43, “dnjt: quarter day” in the entry for year 6 of
Weserkaf (Cairo fragment) is baseless, cf. Wilkinson, Annals, 219.
3
For example PT § 657, 716, 794, 1260, 1711, 2056.
4
Spalinger, Lists, 23–24; 28–29; 33.
5
Spalinger, Lists, 101–103.
6
A. J. Spalinger, Studies, VIII, cf. idem, BSEG 19 (1995), 40.
7
For the non-chronological background to the Egyptian lunar days, see Spalinger,
“Dating”, 383–387.
lunar days, lunar months 387

Beginning of the Lunar Month

Historians presumed that the ancient Egyptian lunar month began on


new crescent day down until 1864 when Brugsch suggested that the
lunar month started with conjunction. He cited a Ptolemaic text in
Karnak: “He (Khonsu, the moon-god) is conceived on ps≈ntyw; he is
born on #bd; he grows old after smdt”.8 Ninety years later Parker para-
phrased the text as follows: The moon-god is conceived in the dark-
ness of invisibility on the first day of the lunar month, he is born as
the new crescent on the second day, and he wanes after the day of
full moon, the 15th day.9 Parker also cited an earlier, MK text with a
similar assertion: “I know, O souls of Hermopolis, what is small on
[#bd ] and what is great on [smdt]; it is Thoth.” Parker commented:
“Thoth is, of course, the moon, small on the day of new crescent and
great on the day of full moon”.10
Brugsch’s contemporaries were less enthusiastic than Parker.11 Only
Mahler, and later Sethe, accepted conjunction as the beginning of the
lunar month. Subsequently others disagreed, arguing that conjunction
is not observable and thus cannot have marked the beginning of the
lunar month.12 Around 1920 Borchardt realized that the Egyptian lunar
month must have begun with the first day of invisibility after old (or
last) crescent day,13 i.e. with an observable event.14 Shortly thereafter
Schoch came to the same conclusion independently.15 Parker argued
in detail that the Egyptians reckoned the lunar month from the first
calendar day of the moon’s invisibility, coinciding with the day of con-
junction in ca. 88% of the cases, the day before conjunction in ca.
10.5% and the day after conjunction in ca. 1.5%.16 Parker based his

8
H. Brugsch, Matériaux pour servir à la reconstruction du calendrier des anciens ègyptiens
(Leipzig, 1864), 58–60; Parker, Calendars, 9, 12.
9
Parker, Calendars, 9–10.
10
Cf. also Book of the Dead, title of Spell 135.
11
Cf. Parker, Calendars, 9.
12
For example, Meyer, Chronologie, 49–50; D. R. Fotheringham, PSBA 18 (1896),
101; and E. F. Edgerton, AJSL 53 (1937), 195.
13
L. Borchardt, OLZ 28 (1925), 620 n. 2; idem, Mittel, 19, 30 n. 10.
14
Actually two events: the first day of invisibility can only be recognized with cer-
tainty if the crescent had been observed the day before.
15
C. Schoch, Die Neumondfeste (Berlin-Steglitz, 1928); Reprint: Astronomische Abhandlungen.
Ergänzungshefte zu den Astronomischen Nachrichten Bd. 8, Nr. 2 (Kiel, 1931), B11–B13;
cf. W. Hirschberg, Zeitschrift für Ethnologie 66 (1933–34), 245.
16
Parker, Calendars, 9–13.
388 rolf krauss

argument on the ancient texts cited above and on the correspondances


of the days of Egyptian terminology for the lunar month with lunar
phases.17 There is more circumstantial evidence. For example, in the
Pyramid Texts ps≈ntyw and #bd occur together as first and second day
of a statue ritual that is also known from the Neferırkare' archive.18
On ps≈ntyw the royal statues were dressed, and they “appeared” on #bd.
Apparently, the statues’ “appearance” equated metaphorically with the
appearance of the new crescent on #bd . Thus the earliest known instances
of ps≈ntyw seem to identify this day as one of invisibility.
It is astronomically possible that a last crescent is visible in south-
ern Egypt on a certain day, but not in northern Egypt.19 In such a
case, the counting of the lunar days would have been out of step for
one month. In Egyptian latitudes the new crescent appears in ca. 70%
of the cases after a single day of invisibility, in ca. 30% of the cases
after invisibility lasting two days.20 This circumstance could be under-
stood to suggest that the lunar month was reckoned from the second
day of invisibility. But, if the second day of invisibility were counted
as lunar day 1, how was the first day of invisibility counted? The
observers did not, and could not, know beforehand whether the moon
would be invisible for one or two days. This would have been clear
only on the evening of the day that followed the first calendar day of
morning invisibility: if the new crescent was observable, the moon had
been invisible for a single day; if the new crescent did not become
visible, the moon would be invisible for two days. In other words, the
theory that the lunar month could have begun on the 2nd day of inivis-
ibility would mean that for about 36 hours, or 1½ calendar days,
neither the observers nor the Egyptians who relied on them (for mak-
ing offerings on the appropriate lunar days)21 would have known whether
the current calendar day was to be counted as lunar day **31 or **0
or 1 or 2. If this impracticable if not to say nonsensical procedure had

17
Parker, Calendars, 12.
18
Posener, Archives I, 52–57.—For a chronological analysis of these lunar dates see
below Chapter III. 8 (end).
19
For example, the old crescent of December 23 in 1828 BC was visible below
Coptos, but not in ME and LE.
20
Parker, Calendars, § 44.
21
Cf. for example, the lunar days 1, 2, 4, 5, and 6 enumerated by King Ahmose
(Urk. IV 24.4–7) as appropriate for offerings to the dead.
lunar days, lunar months 389

obtained, then a second day of invisibility that was recorded as apparent


“lunar day 1” would be equivalent to a positively incorrect lunar date
(see below, Chapter III. 8). Down to the present no discrepancy of this
kind has been detected.

Lunar Calendar: ‘Civil-based’ Lunar Year

Observation of lunar phases, counting of lunar days and lunar-civil


double dates do not constitute a lunar year that comprises and counts
successively 12 or occasionally 13 lunar months. The existence of an
Egyptian lunar year was first suggested by Brugsch;22 his idea met with
disapproval.23 Borchardt developed Brugsch’s ideas, but it was Parker
who argued consistently in favor of two lunar calendars, one referring
to the heliacal rising of Sirius and one to the first day of the civil cal-
endar. The ruling principle, as formulated by Parker, was that the
beginning of the lunar year must not lie before the beginning of the
civil year. To conform to this rule, a 13th lunar month had to be inter-
calated from time to time, usually each third year. This lunar calendar
ran parallel to the civil year, and therefore it may be called a civil-
based lunar calendar.
For his interpretation Parker relied on the lunar cycle in pCarlsberg
9, which dates from 144 AD or later. The cycle consists of 309 lunar
months or 25 civil calendar years. It begins on I Akhet 1 to which it
supposedly returns after 25 years.24 The cycle dates are correct in only
70% of the cases, as Parker was well aware.25 The cycle comprises nine
“great” years of 13 lunar months each and 16 “small” years of 12
months. Nowadays there are doubts whether it reflects a system that
was actually used.26 It is equally doubtful whether the same cycle was

22
Cf. in detail, Depuydt, Calendar, 153–157.
23
Cf. for example, Ginzel, Handbuch I, 168.
24
W. Barta, ZÄS 106 (1979), 1–10, computed a series of 25 year cycles beginning
in 2388 BC and ending in 155 AD. Because he misunderstood Parker, Barta used
astronomically incorrect first cycle years; for a correction see idem, GM 94 (1986),
7–12. Furthermore he did not compute first lunar days, but the respective conjunc-
tion days, which renders his cycle tables useless.
25
Parker, Calendars, 25–26.
26
L. Depuydt, “The Demotic Mathematical Astronomical Papyrus Carlsberg 9
Reinterpreted”, OLA 85 (1998), 1277–1297.
390 rolf krauss

used for the Macedonian year, which was in effect during in Ptolemaic
times alongside the Egyptian calendar.27
As proof for the existence of the civil-based lunar calendar Parker
cited double dates that not only counted civil and lunar days, but also
months differently.28 The earliest example is the date of an oath taken
before the moon god Khonsu on a calendric full moon day in year 12
of Amasis (559 BC).29 According to Parker, the date is expressed in
civil and lunar terms as “regnal year 12 of Amasis, (civil month) II
Shemu 13, being the 15th lunar day of (lunar month) I Shemu”.30 This
interpretation implies that the months of a lunar year were counted
from the first LD 1 after (civil) I Akhet 1.
In 1997 L. Depuydt published a study of the civil-based lunar cal-
endar,31 “which subsumes and supercedes previous studies”.32 He con-
cluded that “there is no doubt about the existence of the civil-based
lunar calendar, first discovered by Brugsch. The proof is of the best
kind: astronomical. Civil and lunar double dates are not numerous,
but they provide unquestionable evidence for the existence of this
calendar”.33 But as early as 1955 Gardiner had challenged Parker’s
views, rejecting the existence of any lunar calendar.34 In the late 1980s
Spalinger began analyzing various aspects of Parker’s calendric studies;
he concluded that no civil-based lunar calendar existed.35 And recently,
when J. A. Belmonte scrutinized Depuydt’s arguments in favor of the
civil-based lunar calendar, he also concluded that the data are liable
to different interpretation.36 Thus there is no consensus among the
specialists.
As Spalinger has pointed out, it is possible to analyse presumed civil-
lunar double dates without considering the validity of the civil-based
lunar year itself:37 “. . . chronographers do not need it, . . . Parker and

27
A. Jones, “On the Reconstructed Macedonian and Egyptian Lunar Calendars”,
ZPE 119 (1997), 157–166.
28
Parker, Calendars, 26.
29
R. A. Parker, MDAIK 15 (1957), 208–212.
30
Depuydt, Calendar,
31
Depuydt, Calendar.
32
J. P. Allen, The Heqanakht Papyri (New York: PMMAEE 27, 2002), 135 n. 42.
33
Depuydt, Calendar, 217.
34
A. H. Gardiner, “The Problem of the Month-names”, RdE 10 (1955), 9–31.
35
Cf. for example, A. J. Spalinger, in: Hommages à J. Leclant (BdE 106/4, 1994), 364
n. 4.
36
Belmonte, “Questions”, 14–15.
37
A. J. Spalinger, “Ancient Egyptian Calendars: How many were there?”, review
article of Depuydt, Calendars, JARCE 39 (2002), 241–250, at 250.
lunar days, lunar months 391

later Egyptologists have never used it . . .” The assertion is correct, in


general, but the Heqanakht papyri may represent an exception. In his
letters Hekanakht refers to the months “f-bdt, rk˙-'#, and ¢nt-§ty-prtj. The
internal chronology of the letters depends on whether these months are
civil or lunar. For example, Spalinger sees “no indications of a lunar
calendar operating in this correspondance”;38 Belmonte likewise asserts
that the months are not lunar, but “are clearly mentioned in a civil
calendar context”.39 By contrast, the very use of these month names in
the Heqanakht papyri suggests to Allen “that farmers also followed the
lunar calendar—understandably so, since the phases of the moon were
much easier for them to keep track of than the artificial sequence of
numbered days in the civil calendar.”40 Allen identifies this presumed
lunar calendar as the civil-based lunar calendar, without discussing the
possibility that it might be the Sothis-based lunar calendar.41 Hekanakht
would have meant September/October (Greg.) when he referred to the
month ¢nt-§ty-prtj as the 10th month of the civil calendar or the civil-
based lunar calendar; or March/April (Greg.) if ¢nt-§ty-prtj was the 10th
month of the Sothis-based lunar calendar.42 In the former case Hekanakht
“postponed the beginning of the household’s new salary schedule to
late September”, although it would have been possible to start “two or
even three lunar months earlier”.43 Perhaps this and other difficulties
could be resolved by assuming that Hekanakht meant months of the
Sothis-based lunar calendar, rather than the civil-based lunar calendar.
This discussion illustrates how the vagueness of the ancient sources, as
exemplified in the calendric material contained in the Hekanakht papyri,
are capable of very different interpretations, depending on the bias of
the scholar.

38
A. Spalinger, “Calendrical Evidence and Hekanakhte”, ZÄS 123 (1996), 85–96;
esp. 89.
39
Belmonte, “Questions”, 21.
40
Allen (n. 32), 135–136.
41
W. Barta, ZÄS 110 (1983), 19.
42
Based on Allen’s assignment of year 8 in the papyri to Senwosret I.
43
Allen (n. 32), 137 n. 47.
III. 7 LONG-TERM VARIATION IN THE MOTIONS
OF THE EARTH AND THE MOON

Kurt Locher

Of the three sidereally defined periods—the year, the lunar month, and
the day—, the year is the most stable in terms of modern methods of
time-keeping such as atomic time based on molecular or electron-
tilting oscillations. Its slight long-term variation is irrelevant for Egyptian
chronology, because the artificial 365.0-day civil year, which was kept
through all epochs, would always yield an unambiguous number of
days elapsed between any two historical dates of relative chronology.
This is far from true for the lunar month and the day: The number
of days elapsed between any pair of the same moon phases many cen-
turies apart is affected by two long-term variations. The slowing of the
earth’s rotation over time must be studied with the greatest possible
care, both empirically, using historical astronomical data gathered from
non-Egyptian records, and theoretically, by calculating its physical cause
(the amount of kinetic energy released from rotational momentum and
transformed into heat by tidal friction). Thanks to Stephenson’s recent
extensive work1 we now know that the actual number of days elapsed
from any OK date until today differs by just under one day from what
it would have been if the velocity of the earth’s rotation remained
constant.2
Long-term variation in the orbiting period of the moon has essen-
tially the same physical cause: the slowing of the earth’s rotation means
a decrease in angular momentum, which the motion of the body exert-
ing the causing force, i.e. the moon, must compensate by acceleration.
The total number of lunar months elapsed between any OK date and
today differs from the number they would have amounted to without
such an acceleration by roughly one hundredth of a month.3

1
Stephenson, Eclipses.
2
Ibidem, Figs. 1.6, 2.1, 14.1, 14.2.
3
Ibidem, § 2.6.2.
long-term variation in the motions of the earth 393

Since both differences result from an integration over time of a lin-


ear effect, each increases quadratically with respect to the time elapsed.
Moving back in time from today towards the OK, the result reduces,
e.g., to one fourth if we move back halfway.

Relevant Peculiarities of the Motion of the Moon

There are also periodic short-term variations in lunar motion. One of


them arises from the elliptic shape of its orbit and the dynamical con-
sequences implied which results in variations in speed.
Since the sun appears, simply spoken in the ancient way, to move
through exactly 12 zodiacal signs in the course of a year, and since
there are 12 (exceptionally 13) new moons within this same period, the
sign in which a new moon is seen from earth is the one adjacent (to
the left as seen from northern latitudes) to the sign where the preced-
ing new moon was seen.
Thus every month the breadth of roughly one sign is covered twice
by the moon moving through the zodiac; if the moon happens to be
in that part of its orbit where its speed is smallest (both absolute and
apparent-angular), the month will be considerably longer than average.
Such minimum speed occurs near apogee (apo ghn, “away from the
earth”), the point on the elliptic orbit most distant from the earth. Since
Ptolemy noted the effect of this variable velocity on the position of the
moon (or a planet), the phenomenon has been called “the anomaly”
(an-omalow, “uneven”).4 Calculating the anomaly in the case of the
moon’s orbit is a complex procedure, because it is considerably per-
turbed dynamically by the gravity of a third body (the sun); however,
this perturbation is negligible in the case of a planetary orbital ellipse.
The perturbation in the moon’s case is the reason why the apogee is
spatially (more exactly: sidereally) not always on the same side of the
earth, but moves slowly around it, completing a full cycle in roughly
9 years, a fact which has been termed apsidal motion (aciw, “apse” (of
the ellipse) since Hipparchus.
The fact that the beginning of an Egyptian lunar month depended
on the observation of last visibility of the waning moon at dawn must
also be taken into account. Unlike all other effects considered above,

4
Ptolemy, Almagest, § IV 2, Edition Heiberg (Leipzig, 1898), 269.
394 kurt locher

the success of such an observation depended not only on the lunar


coordinate measured along the ecliptic, but also at right angles to it, i.e.
on the ecliptical latitude. Egypt’s geographic position, which is to the
north of the equator, favours observation of the moon if it is situated
north of the ecliptic, by contrast to a symmetrical position south of it,
especially for marginal situations which always occur near the horizon.5
The orbit of the moon is inclined against the ecliptic by roughly 5°,
so that the moon crosses it twice every month at points called the
nodes. Like the apogee, these nodes are not fixed, but they revolve
around the earth in a cycle of roughly 18 years.
The position of the moon near a node is crucial for the occurrence
of an eclipse; both solar and lunar eclipses are relevant for the chronol-
ogy of most ancient civilizations, but, exceptionally, not for Egypt.
Oddly enough, there are very few, if indeed any reports of eclipses
from Egypt. For that reason, eclipse theory need not be included here.

5
Ibidem, Heiberg, 270.
III. 8 LUNAR DATES

Rolf Krauss

Computation of Old and New Crescent

An initial attempt to compute old and new crescent was made by the
astronomer K. C. Bruhns around 1880.1 He based his calculations, which
were unsuccessful, on observations recorded at Athens by the astronomer
J. Schmidt. 15 years later F. Wislicenus also admitted defeat, asserting:
“if the sky is clear, but under otherwise differing astronomical condi-
tions, the first appearance of the crescent can occur 1 to 3 days after
conjunction”.2 It was K. Fotheringham who first succesfully calculated
old and new crescent, utilizing Schmidt’s observations, in addition to
others.3 The basic parameters of Fotheringham’s calculation are lunar
altitude and lunar and solar azimuth.4 For the old or new crescent to
be considered visible, it must have a minimal altitude h which is depen-
dent on the distance D in azimuth of the sun and moon at the moment
when the centre of the sun is in the mathematical horizon; the posi-
tion of the crescent is computed geocentrically, i.e. without parallax.5
M. Maunder,6 P. V. Neugebauer and C. Schoch7 improved upon the
minimal altitudes h of Fotheringham. In an earlier version of his astro-
nomical tables P. V. Neugebauer incorporated crescent visibility criteria
of Fotheringham and Maunder;8 in his later Astronomische Chronologie,

1
A. Mommsen, Chronologie: Untersuchungen über das Kalenderwesen der Griechen insonderheit
der Athener (Leipzig, 1883), 69–80.
2
W. F. Wislicenus, Astronomische Chronologie: ein Hülfsbuch für Historiker, Archäologen und
Astronomen (Leipzig, 1895), 29.
3
Mommsen (n. 1), 69–73.
4
K. Fotheringham, “On the Smallest Visible Phase of the Moon”, Monthly Notices
of the Royal Astronomical Society 70 (1910), 527–531.
5
Contemporary astronomers somtimes overlook the fact that Fotheringham’s model
is geocentric, cf. R. Krauss, DE 57 (2003), 53–54.—For a topocentric model see J. A. R.
Caldwell & C. D. Laney: <http://www.saao.ac.za/~wgssa/as5/caldwell.html>.
6
M. Maunder, Journal of the British Astronomical Association 21 (1911), 355–362.
7
C. Schoch, Planetentafeln für Jedermann (Berlin-Pankow, 1927); idem, in: Neugebauer
(n. 9), I 79, Tafel E 21.—Use of Schoch’s data is now known as “Indian method” fol-
lowing adoption by the Indian Astronomical Ephemeris in 1966.
8
P. V. Neugebauer, Hilfstafeln zur Berechnung von Himmelserscheinungen. Tafeln zur astronomi-
schen Chronologie III (Leipzig, 1925), VII, Tafel 14.
396 rolf krauss

devised for the use of historians and astronomers, he used Schoch’s


criteria.9 Between the publication of Neugebauer’s Tables in 1929 and
ca. 1990 Egyptologists used them for astronomical computations involv-
ing solar, lunar, planetary, and stellar data. Neugebauer anticipated
that his Tables would become obsolete after half of a century.10 And
indeed the situation changed in the 1980’s when astronomical software
to be used with computers became available. (All astronomical calcu-
lations for Chapters III. 8 and 10 were made with the program UraniaStar
Release 1.1 [M. Pietschnig & W. Vollman, Vienna, 1995]. The program
was developed under the supervision of the astronomer Hermann Mucke.
For the program’s reliability, see M. G. Firneis & M. Rode-Paunzen,
“Progress-Report on Egyptian Astrochronology”, in: Bietak, ed., SCIEM
Haindorf 2001, 48).
Especially for lunar positions, modern astronomical computation yields
results different from Neugebauer’s Tables. One reason is that research
has significantly changed the values for Dt (delta t), the difference between
Universal Time and Terrestrial Time that results from the slowing of
the earth’s rotation.11 Regardless, in many cases the ancient date of
observability of old or new crescent remains the same, whether calcu-
lated with the outdated parameters of Neugebauer or according to the
most recent ones.12
In the late 1980’s B. Schaefer, following an earlier attempt by
F. Bruin,13 developed a model for reckoning old and new lunar cres-
cents taking into consideration: (1) the physiology of the human eye,
(2) the brightness of the twilight sky, (3) the surface brightness of the
moon, (4) the extinction in the atmosphere, and (5) the local conditions.
According to Schaefer, there is a general shift from clear skies in
winter to hazy skies in summer, and thus in the northern hemisphere
the minimal altitude is lower in winter and higher in summer. This
rule does not seem to apply to all regions, but it is valid for Egypt.14
Thus Schaefer’s model represents an improvement over the visibility

9
P. V. Neugebauer, Astronomische Chronologie I, II (Leipzig & Berlin, 1929).
10
Neugebauer (n. 9, I), V.
11
See above Locher, Chapter III. 7.
12
Krauss (n. 5), 52–53.
13
F. Bruin, Vistas in Astronomy 21 (1977), 331–358.
14
M. Shaltout, “Study of the Solar Radiation over Menia”, Renewable Energy 23
(2201), 621–639, at 631–634; and personal communication, January 17, 2004.
lunar dates 397

criteria of Schoch and others, as far as they do not take into account
the seasonal variation of extinction and its daily random variability.
Table III. 8.1 lists a selection of Schoch’s and Schaefer’s minimal cres-
cent altitudes h, relating to D. In the Table months apply only to
Schaefer’s figures;15 Schoch’s do not change during the year.
For example, if the distance D in azimuth is 0°, then according to
Schoch the crescent ought to be visible throughout the year if h *
10.4°. By contrast, Schaefer defines the minimal altitude as a mean
value h together with its mean square root error s. If D = 0°, then for
example, in December h = 10.2° ± 0.6°, i.e. for ca. 68% of crescents
h = 10.2° is the minimal altitude; for 16% it would be 10.2° to 10.8°,
and for the remaining 16% it would be 10.2° to 9.6°. If the crescent
is within h ± s, then visibility or invisibility of the crescent depends on
extinction at the time of observation. Thus h ± s is a zone of uncer-
tainty, because extinction cannot be predicted exactly. If the crescent
is above h ± s, then visibility is to be expected; if the crescent is below
h ± s, it ought to be invisible, although an outlier is a rare possibility.

Table III. 8.1


Schaefer Schoch Schaefer Schoch
h (D = 0°) h (D = 0°) h (D = 10°) h (D = 10°)

December 10.2° ± 0.6° 8.8° ± 0.8°


March/Sept 11.0° ± 0.8° 10.4° 9.5° ± 0.9° 9.3°
June 11.6° ± 0.7° 10.1° ± 1.1°
h (D = 15°) h (D = 15°) h (D = 20°) h (D = 20°)
December 7.5° ± 0.7° 6.6° ± 0.7°
March/Sept 8.5° ± 1.2° 8.0° 7.6° ± 0.9° 6.2°
June 9.0° ± 0.8° 7.9° ± 0.8°

15
I owe specific numerical values for h ± s and the permission to use them in pub-
lications to a generous personal communication from Schaefer in November 1999; see
also R. Krauss, “An Egyptian Chronology from Dynasties XIII to XXV”, in: Bietak,
SCIEM Vienna 2003, n. 25; in press.
398 rolf krauss

Negative and Positive Errors in Lunar Observation

Provided the weather was fine, old crescent ought to have been always
visible at dawn of a LD 29. The critical day was LD 30. If the cres-
cent was still visible at dawn of a LD 30, then the current lunar month
had 30 days and the following lunar month began on the next calen-
dar day. If the crescent was not visible at dawn of LD 30, then a new
lunar month had begun and the expired lunar month had 29 days.
The new crescent appears on the second or third calendar day after
morning invisibility; an observer could determine the first day of invis-
ibility with certainty only when he had been able to see the crescent
on the previous morning. According to Parker, mistakes could occur
on either day: “An Egyptian lunar date as given in the civil calendar
and as calculated by modern tables may lack agreement by a day . . . due
to faulty observation or any other reason”.16 Parker left open whether
there is a 50% chance that a lunar date is correct, or whether it is
more probable that a lunar date is correct than wrong. Nevertheless,
it is clear that Egyptian lunar dates exhibit random qualities, and there-
fore analysis of them mandates taking into consideration the mathe-
matical probability of correct and mistaken lunar dates.
By definition, an ancient lunar date is “correct” if it is confirmed by
modern astronomical computation. A Egyptian lunar date is negatively
(or positively) incorrect if the respective lunar phase is one day too
early (or too late). A negative error results when the observer does not
see an old crescent that is visible under favorable meteorological con-
ditions somewhere else. A negatively incorrect lunar date implies the
beginning of an Egyptian lunar month one day too early. A positive
error implies a delay of the beginning of an Egyptian lunar month by
one day, i.e. the month would begin one day too late. A positive error
should not result from direct observation, because a crescent which is
not present cannot be seen. A positive error may occur on a cloudy
day when the observer guesses that the crescent is present above the
clouds. Late or early lunar dates are evidently not symmetrical errors;
therefore the likelyhood of each occurring should not be the same.

16
Parker, Calendars, 211; idem, JNES 16 (1957), 39–40.
lunar dates 399

Error Quotas in Lunar Dates

The three possible sources of error in Egyptian old crescent observa-


tion are: a) the observer himself, who might falsify data; b) observa-
tional difficulties when the crescent is in the zone of uncertainty; c)
observational difficulties resulting from unfavourable weather conditions,
i.e. haziness or overcast skies due to clouds or a sandstorm. Presumably
the accuracy of Egyptian old crescent observation was the same as
Babylonian new crescent observation. In both cultures observations were
made under similar meteorological conditions by experienced observers.17
C. Schoch18 and P. H. Huber19 investigated the accuracy of Babylonian
lunar observation. They concluded that at least 85% of the recorded
lunar dates were correct new crescent observations. Recent analysis sug-
gests that Schoch used not only observed crescents, but predicted ones
as well, whereas Huber’s data consist of observed crescents only.20
F. R. Stephenson’s team analysed the circumstances of visibility for
209 observed Babylonian new crescents21 published by Hunger.22 When
topocentric lunar positions were plotted against different versions of the
geocentric visibility lines of Schoch, 8 of 209 crescents were below the
visibility line. If geocentric lunar positions are plotted, only 2 crescents
are to be found just below Schaefer’s visibility line, which if slightly
adjusted, includes them also.
According to Schaefer’s criteria, ca. 10% of the 209 new crescents
were sighted one day too late, presumably because of high extinction
or some other reason. Thus the relative frequency of negative errors
in these 209 Babylonian crescent observations is ca. 10%. The 95%
confidence interval for the true probability of negative errors in a sam-
ple of 209 new crescents is 6% to 14%. Correspondingly it can be
expected that 86% to 94% of Egyptian old crescent observations were
correct, the quota of negative errors stemming from extinction being
6% to 14%.

17
For the imy-wnwt, the astronomer of Egypt, see J.-L. Fissolo, Égypte, Afrique et Orient
21 (2001), 15–20.
18
C. Schoch, in: S. Langdon & J. K. Fotheringham, eds., The Venus-Tablets of
Ammizaduga (Oxford, 1928), 96–97.
19
P. Huber, Astronomical Dating of Babylon I and Ur III (Malibu: MJNE. Occasional
Papers ¼, 1982), 25ff.
20
L. Fatoohi, F. R. Stephenson & S. Al-Dargazelli, “The Babylonian First Visibility
of the Lunar Crescent: Data and Criterion”, JHA 30 (1999), 51–72 at 64.
21
Cf. n. 20.
22
Sachs & Hunger, Diaries.
400 rolf krauss

This result is supported by a computation (sic) of about 150 old cres-


cents for the years between 2001 and 2013 AD at the latitude of
Illahun, yielding about 84% crescents (95% confidence interval: 78%
to 90%) that should (have) be(en) visible without difficulty; about 16%
of the crescents (95% confidence intervall: 10% to 22%) would (have)
be(en) situated within the zone of uncertainty.23 Presuming that high
and low extinction are equally possible, it is to be expected that in
84% + 8% = 92% of the cases (95% confidence intervall: 89% to
94%) the sightings would (have) be(en) correct, and in 8% (95%
confidence intervall: 5% to 11%) negatively incorrect. Thus it may be
expected that under favourable weather conditions Babylonian and
Egyptian crescent observations yielded ca. 90% correct lunar dates and
ca. 10% negatively incorrect ones.
In the 1990’s Doggett and Schaefer organised “moon watches” to
establish the Lunar Date Line for specific months. Their evidence for
negative mistakes is indirect:24 “Of 520 negative reports, 5 were made
by observers who missed an easy sighting on the following night. We
suspect the rate of negative errors is greater (and probably much greater)
than 1%”. The evidence for positive mistakes is direct: “Of the 20
observers in the northeast [of North America where the moon should
not have become visible], 3 reported sighting the Moon. In all three
cases, the reported time of sighting, orientation of the horns, and direc-
tion of the Moon were grossly in error. The large errors in reported
details confirm that these three observations were positive errors. From
our small sample [from Moonwatch 5], the positive error rate is 15%”.
But, the three reports might just as well be spurious; in fact, no mod-
ern series of observations provides a reliable basis for extrapolating the
quality of professional ancient Egyptian (and Babylonian) old (and new)
crescent observations.

23
Krauss (n. 15).
24
L. Doggett & B. Schaefer, “Lunar Crescent Visibility”, Icarus 107 (1994), 388–403,
esp. 402; see also B. Schaefer, “Lunar Crescent Visibility”, Quarterly Journal of the Royal
Astronomical Society 37 (1996), 759–768, at 760.
25
Climates of Africa, J. F. Griffiths, ed., (Amsterdam: World Survey of Climatology 10,
1972), 84–85, 126–128.
lunar dates 401

Mistakenly Conjectured Lunar Dates

If the sky is overcast, visibility or invisibility of the moon has to be


conjectured. There are statistics available for the average monthly cloudi-
ness at Egyptian sites.25 The mean cloudiness decreases between the
Delta and UE.26 Presuming the climate in Egypt has not changed since
the end of the OK,27 modern regional conditions are applicable to the
MK. Around Illahun the yearly mean cloudiness amounts to 20%, i.e.
during the year the probability that clouds obscure old crescent amounts
to p = 0.2, complemented by the probability of q = 0.8 that old cres-
cent is visible. Under these circumstances it is to expected that over
the years, at most ca. 20% of the old crescents went unobserved.28
The proportion between lunar months of 29 and 30 days is 47:53;
the total of 20% old crescents on overcast days divide into 9.4% old
crescents on a LD 29 and 10.6% on a LD 30. On a overcast LD 29,
the assumption that the crescent is present is correct, the opposite sup-
position negatively incorrect. Thus 4.7% of all lunar dates are correct
and 4.7% negatively incorrect. If it is a LD 30 and the sky is over-
cast, then the assumption concerning the presence of the crescent is (a)
correct, and the opposite assumption (b) negatively incorrect. If it is
not a LD 30 but rather a LD 1 of the next lunar month, then the
assumption that the crescent is present is (c) positively incorrect and
the opposite assumption (d) correct. These possibilities result in 5.3%
correct and 2.65% negatively or positively incorrect lunar dates each.
Thus the 20% conjectured lunar dates divide into 10% correct dates,
7.35% negatively incorrect and 2.65% positively incorrect dates. Altogether
(20% conjectured and 80% observed crescents) there are 82% correct
dates, 15.35% negatively, and 2.65% positively incorrect dates. The
percentages of correct and incorrect observations following from 20%
at Illahun and 12.5% conjectures at Luxor are expressed in Table III.
8.2 as rounded decimal values.

26
For example the mean yearly cloudiness in Cairo amounts to 2 octas or 25%, in
Luxor to 1 octa or 12.5%, (cf. n. 25).
27
Cf. K. W. Butzer, “Klima”, LÄ III, 456.
28
The result was obtained by first computing the probabilities that during 62 lunar
months or ca. 5 years in exactly 0, 1, 2, 3 . . . 62 instances old crescent was hidden
by clouds. According to standard procedure it follows that the crescent was hidden in
at most 12 to 14 of altogether 62 instances, corresponding to a mean of ca. 20%.
402 rolf krauss

Table III. 8.2


Conjectures Correct Negatively Incorrect Positively Incorrect

20% 0.82 0.153 0.026


12.5% 0.85 0.134 0.016

Figs. III. 8.1–3 (below) present the observability of the Illahun lunar
dates in simplified schemes, according to three different absolute dates
proposed for year of 1 Amenemhet III. The lunar positions are plot-
ted against the visibility line of Schaefer. To avoid a three-dimensional
representation only one visibility line is used and each crescent is plot-
ted against it, as if it were its own visibility line. Different values for
Dt are considered,29 resulting in different positions of one and the same
crescent; the positions are represented as a single line corresponding to
a difference in Dt of roughly ± 1 h.
Fig. III. 8.1 represents the positions of the old crescents when the
Illahun lunar dates (ILD) are computed as if 1 Amenemhet III =
1844/43 BC. Fifteen of 21 crescents would have been observable above
the visibility line. There would have been one negatively incorrect obser-
vation, since crescent ILD 10 should have been visible in position 10*.
The crescents ILD 6 and 12 would have been just below the visibility
line, whereas ILD 13, 14 and 15 would have been far below, alto-
gether corresponding to five positively incorrect observations.
Fig. III. 8.2 shows that for 1 Amenemhet III = 1819/18 BC, the
two crescents ILD 2 and 16 would have been negatively incorrect obser-
vations. 19 of 21 ILD would have been observable above the visibility
line and none below.
Fig. III. 8.3 demonstrates that for 1 Amenemhet III = 1794/93 BC
sixteen crescents would have been visible, two would have been unob-
servable as positively incorrect and three would have been missed as
negatively incorrect.
The situation changes if P. J. Huber is followed with Dt computed
according to a formula of Morrison and Stephenson.30 Then Dt is 12.6 h,

29
The lunar positions in Figs. III. 8.1–3 are computed using Spencer Jones’s for-
mula for Dt, based on an uncertainty of ± 1 h.—Cf. H. Spencer Jones, “The Rotation
of the Earth, and the Secular Accelerations of the Sun, Moon and Planets”, Monthly
Notices of the Royal Astronomical Society 99, (1939), 541–558.
30
P. J. Huber, Akkadica 119–120 (2000), 173–176; with reference to L. V. Morrison
& R. L. Stephenson, in: W. Fricke & G. Teleki, Sun and Planetary System (Dordrecht,
1982), 173–178.
lunar dates 403

10

9
7

11
19 2
20
8 4
21
10* 3
17
18 16 5

1
6 visibility line
visibilityline
12

14
15
13
distance
distanceofofsun
sun and
andmoon
moonininazimuth
azimuth

0° 5° 10 ° 15 ° 20 °

Fig. III. 8.1

8 10

18

7 2
3 11
1 14
9 17
20 16
6
15
4 19 2*
16* 12
21
13 5
visibility line
visibilityline

distance
distanceof sun and
ofsun andmoon
moonininazimuth
azimuth

0° 5° 10 ° 15 ° 20 °

Fig. III. 8.2


404 rolf krauss

visibility line

distance of sun and moon in azimuth

Fig. III. 8.3

instead of Spencer Jones’s 10.56 h for ca. 1800 BC. The increased
value for Dt results in a decrease of lunar altitude at the moment when
the centre of the sun is in the horizon. In other words, if the larger
Dt is presumed, those crescents in Figs. III. 8.1–3 that are just below
the visibility line were definitely below it. Of interest is ILD 13 in Fig.
III. 8.2, because it is close to the lower limit of visibility. This crescent
would have tended to be visible if Dt ) 12.6 h, invisible if Dt > 12.6 h.

Partial Repetition of Lunar Dates after 25 Egyptian Years

The lunar positions represented in Figs. III. 8.1–3 are 25 Egyptian


years apart; contrary to expectation the positions differ markedly with
respect to the visibility line. Egyptologists generally assume that lunar
dates repeat after 25 Egyptian years,31 because 25 Egyptian years

31
E.g., K. A. Kitchen, “The Chronology of Ancient Egypt”, World Archaeology 23
(1991), 201–208, at 204: “these moon-risings occur in the ancient calendar every twenty-
five years.”
lunar dates 405

correspond to 309 mean (synodic) lunar months of 29.53059 days each,


the difference being only about an hour:32
25 Egyptian years = 25 × 365 days = 9125 days = 309 mean synodic
months = 309 × 29.53059 days = 9125 days minus 1 hour and
7 minutes33
Actually an observer counts either 29 or 30 full days in a lunar month
and arrives at the length of the mean synodic month by calculation
based on the observation of a great number of lunar months of either
29 or 30 days.
A lunar date repeats on the same calendar day, if 9125 days com-
prise 309 lunar months of which 164 are lunar months of 30 days and
145 are lunar months of 29 days: (164 × 30 days) + (145 × 29 days) =
4920 days + 4205 days = 9125 days. Because the movement of the
moon is irregular, there can be 165 lunar months of 30 days and 144
lunar months of 29 days in a series of 309 lunar months. If so, a lunar
date does not repeat on the same calendar day after 25 years = 9125
days, but rather after 9126 days.34 Or if there are 163 lunar months
of 30 days and 146 lunar months of 29 days in a series of 309 lunar
months, then a lunar date does not repeat after 25 years on the same
calendar day, but rather after 9124 days.
Apparently the irregularity results because the mean synodic move-
ment (a) comprises (b) the anomalistic and (c) draconitic movement of
the moon, which do not share a common period of 9125 days = 25
Egyptian years. The mean anomalistic velocity is not the same after
25 years, whereas the mean draconitic movement results in a different
latitude of the moon.
9125 days = (a) 309 × 29.53059 d + 0.04 d =
(b) 331 × 27.55455 d + 4.44 d =
(c) 335 × 27.21222 d + 8.91 d
On average only about 70% of the dates in a set repeat on the same
day after a single 25 year shift.35 For multiples of 25 years, percentages

32
So for example, Beckerath, Chronologie, 48–49.
33
Within the siderial and synodic months the moon travels at a mean velocity of
13.176° per day. Within 25 Egyptian years = 24.982 siderial years, the sun travels in
the mean 24 × 360° + 353.683°, whereas the moon travels 333 × 360° + 354.272°.
In 25 Egyptian years the positions of sun and moon have decreased by about 6.317°
and 5.728° respectively, whereas their original distance has decreased only by about
0.52°, i.e. the distance that the moon travels in an hour.
34
For a specific example see Krauss, in: Bietak, SCIEM Haindorf 2001, 190–192.
35
Krauss, Sothis, 27; similarly Parker, Calendars, 25–26.
406 rolf krauss

of correct repetitions decrease. Shifts of 2 × 25 and 3 × 25 years yield


exactly repeated dates in only 50% of the cases.36 Under these premises
a large set of Egyptian lunar dates tends to have one solution with a
maximum of correct dates whereas shifts of ± 25 years have less cor-
rect dates. This is exemplified by 37 alternative solutions for the Illahun
lunar dates between 2286 and 1387 BC. The calculations use 37 alter-
natives for 1 Amenemhet III, each separated by 25 years from the next.
As Fig. III. 8.4 shows, the alternatives differ in the percentages of
correct and incorrect dates, both negative and positive. There is one

Fig. III. 8.4

36
Krauss, Sothis, 27.
lunar dates 407

set with ca. 90% correct dates (1 Amenemhet III = 1819/18 BC),
whereas the remaining 10% are negatively incorrect. There are seven
sets with 60% to 80% correct dates and varying percentages of nega-
tively and positively incorrect dates. The other sets display small per-
centages of correct dates, whereas their percentages of negatively or
positively incorrect dates are high. The trend is clearly to more nega-
tively incorrect dates for shifts backwards and to positively incorrect
dates for shifts forwards. An increase in the number of correct lunar
dates after multiples of 150 Egyptian years is to be noted. This results
from the fact that the synodic, anomalistic, and draconitic months have
an approximate common period of 150 Egyptian years. Nevertheless,
instead of yielding a series of various solutions, which are all astro-
nomically equally possible, there is practically only one astronomically
workable solution for the Illahun lunar dates.
The different astronomical possibilities can be evaluated by com-
puting the respective probabilities. The appropriate tool to deal with
lunar dates and their three properties is the trinomial formula for prob-
ability P:
n!
P = x!×y!×z! × p1 × p2 × p3
x y z

n = x + y + z; x = correct lunar dates; y = negatively incorrect lunar


dates; z = positively incorrect lunar dates; p1, p2, and p3 are the proba-
bilities of x, y and z (see Table III. 8.2). Table III. 8.3 contains the
respective probabilities if the ILD are computed for different first years
of 1 Amenemhet III, including Luft’s suggestion that this might be
1855/54 BC.37 As Table III. 8.3 implies, only 1 Amenemhet III =
1819 BC ± 25 years is by any means probable (cf. also Figs. III. 8.1–3).

Table III. 8.3


1 Amenemhet III correct neg. incorrect pos. incorrect probability

1869/68 BC 8 9 4 6.2 × 10–7


1855/54 (Luft) 6 0 15 2.7 × 10–20
1844/43 16 1 4 0.00032
1819/18 19 2 0 0.114
1794/93 17 2 2 0.020
1769/68 3 4 4 0.00027

37
Luft, Fixierung, 228; cf. below Krauss, Chapter III. 10.
408 rolf krauss

Recorded Lunar Dates from Dyn. 5 to Dyn. 22.

Dates of a Lunar Feast (Tepi Shemu) in the Bubastide Period 38

The reign of Psammetichus I began in February 664 BC and the reign


of Taharqa, 26 years earlier, in 690 BC. According to the Tang-i Var
inscription, Shebitku ruled from at least 707/706 BC. The highest
attested date for Shabaka is year 15. Thus 1 Shabaka corresponds to
722/721 BC at the latest. It is possible, but not assured that Shabaka
defeated Bocchoris in the former’s year 2;39 thus 6 Bocchoris might be
723/722 BC = 2 Shabaka or slightly earlier.
The Bocchoris-Apis was the successor of the Apis that died in 37
Shoshenq V. According to data concerning the three Apis bulls buried
between 28 Shoshenq III and 37 Shoshenq V, and presuming that
Pami’s reign ended in a year 7, 95 years elapsed between 1 Shoshenq
III and 37 Shoshenq V (inclusive). If the Bocchoris-Apis was born very
soon after the death of its predecessor and had a life span of 26 years
at most (the maximum life span attested),40 the upper limit for 1 Shoshenq
III is 723/722 BC + 26 + 95 = 844/843 BC. This limit would need
to be adjusted upwards, if Shebitku’s reign began before 707/706 BC
and/or Shabaka occupied the throne longer than 15 years.
When reckoning the lower limit, it must be borne in mind that
Shepsesre' Tefnakhte may or may not have ruled in Memphis as pre-
decessor of Bocchoris for at least 7 full years. Tefnakhte’s initial take-
over of Memphis occurred at the earliest in the course of 38 Shoshenq
V. In his 20th (?) year Piye drove an apparently non-royal Tefnakhte
out of Memphis; subsequently the kings Shepsesre' Tefnakhte and
Bocchoris may have ruled for at least 12 full years (7+5) or more before
the death of the Bocchoris-Apis. Then, the lower limit for 1 Shoshenq
III would be ca. 722 BC + 12 + 2 (?) + 95 = 831 BC or 722 BC +
5 + 2 (?) + 95 = 824 BC, if Shepsesre' Tefnakhte did not rule as king
in Memphis.
Aston concluded that the rival kings Takelot II and Petubaste I ruled
Thebes when Shoshenq III reigned in LE. The synchronism 5 Petubaste

38
For the TIP cf. above Jansen-Winkeln, Chapter II. 9, and Zibelius-Chen, Chapter
II. 12.
39
Krauss (n. 15).
40
J. Vercoutter, “The Napatan Kings and Apis Worship”, Kush 8 (1960), 62–76 at
64; idem, MDAIK 16 (1958), 339–342.
lunar dates 409

I = 12 [Shoshenq III] is attested while [4 Takelot II = 1 Shoshenq


III] or [5 Takelot II = 1 Shoshenq III] is deducible.41 If 1 Shoshenq
III lies between 844 and 824 BC, then 1 Petubaste I falls in 837 to
817 BC and 1 Takelot II in 848 to 827 BC. The lunar dates of the
Tepi Shemu feast provide a means for determining exact dates.42 The
feast is documented at Karnak from the NK to the Saite Period;43
according to MHC 1451ff., it began on LD 1 and lasted till LD 5.
Parker computed the latest known example which is dated to 14
Psammetichus I.44 Vernus and Krauss followed Parker to interprete
Bubastide examples of the Tepi Shemu feast as lunar dates.45 According
to Vernus and Kruchten, the inductions of priests took place during
the Tepi Shemu feast.46 Bubastide examples of the *Tepi Shemu feast
and/or of inductions are:47
*(A) 11 Takelot II: I Shemu 11 (B) 7 Petubaste I: I Shemu [1]
(C) 8 Petubaste I: I Shemu 19 *(D) 39 Shoshenq III: I Shemu 26
The figures in Table III. 8.4 assume that 11 Takelot II fell between
838 and 817 BC. It lists the lunar days which could correspond to the
explicit Tepi Shemu feast date A; there are six possible years in which
LD 1 to 5 (± 1 day) corresponded to A (bold type). (Lunar days are
counted forward as positive from lunar day 1 to 15; starting with the
last lunar day, whether day 30 or 29, the lunar days are counted back-
ward as negative down to lunar day 16.)
Kruchten identified D as a day of the feast of Tepi Shemu.48 The dis-
tance between A and D amounts either to 32 a + 15 d (if 5 Takelot
II = 1 Shoshenq III) or to 33 a + 15 d (if 4 Takelot II = 1 Shoshenq
III). In the latter case the lunar day of D is ca. 11 days later than
that of A, i.e. D is not the date of the Tepi Shemu feast, and neither is
B or C. In the former case the lunar day of D nearly coincides with
that of A; thus A and D are days of the Tepi Shemu feast as well
as the dates B and C. Table III. 8.5 contains the acceptable and the

41
J. v. Beckerath, “Beiträge zur Geschichte der Libyerzeit”, GM 144 (1995), 7–13.
42
Krauss (n. 15); Krauss, Sothis, 168–177, is outdated.
43
Schott, Festdaten, 104–105.
44
Parker, Oracle Papyrus, 7–8.
45
P. Vernus, “Inscriptions de la Troisième Période Intermédiaire (I)”, BIFAO 75
(1975), 24; Krauss, Sothis, 168–177.
46
Vernus (n. 45); Kruchten, Annales, 244 n. 3.
47
G. Daressy, RecTrav 35 (1913), 130; Kruchten, Annales, 239–240: B–D.
48
Kruchten, Annales, 80.—Cf. also Spalinger, “Dating”, 393.
410 rolf krauss

Table III. 8.4

11 Takelot II LD 11 Takelot II LD 11 Takelot II LD 11 Takelot II LD

840 7 834 11 828 15 822 21


839 17 833 –9 827 –5 821 2
838 –3 832 3 826 7 820 12
837 8 831 14 825 –11 819 22
836 –12 830 –6 824 –1 818 4
835 –1 829 5 823 10 817 14

Table III. 8.5


Takelot II A B C D 1 Takelot II A B C D

845 –1 5 4 1 834 –1 5 4 –1
842 3 7 7?8? 3 831 2 7 6 3?
839 5 11 10 6

Table III. 8.6


Combination Probability Combination Probability

cccc 0.52 c--- 0.008


ccc– 0.33 ---- 0.0003
cc-- 0.004

nearest unacceptable possibilities which result when the lunar day equiv-
alents of the feast dates of Shoshenq III and Petubaste I are also com-
puted and correlated with those of Takelot II within the limits for their
first regnal years.
There are only two acceptable alternatives: either 1 Takelot II cor-
responds to 845 or to 834 BC. In 834 BC two of four lunar dates
would have been negatively incorrect (cc--), but in 845 BC there is
only one error of this kind (ccc–). Table III. 8.6 lists the probabilities
for the combinations of correct (c) and negatively incorrect (–) lunar
dates for Luxor as the place where the observations were made (see
Table III. 8.2).
1 Takelot II = 845 BC is definitely preferable, because it is far more
probable that exactly 1, instead of exactly 2, of 4 lunar dates are too
early. If Takelot III = 845 BC, the Julian calendar equivalences for
A–D are:
lunar dates 411

*(A) 11 Takelot II: = LD 30 = November 25, 835 BC


I Shemu 11
(B) 7 Petubaste I: = LD 5 = November 13, 828 BC
I Shemu [1]
(C) 8 Petubaste I: = LD 4 = November 11, 827 BC
I Shemu 19
*(D) 39 Shoshenq III: = LD 1 = December 2, 803 BC
I Shemu 26
The figures imply that 1 Shoshenq III = 841 BC and 38 Shoshenq
V = 743 BC, so that ca. 21 to 20 years elapsed between 38 Shoshenq
V and 6 Bocchoris. (Kitchen maintains that Shoshenq III ruled after
Takelot II.49 Under his premise the Tepi Shemu dates A–D result in a
single astronomical solution: 1 Takelot II = 856 BC; dead reckoning
yields ca. 951/953 BC instead of 945 BC for 1 Shoshenq I which is
not acceptable.)

The Lunar wr“ Date in Year 5 of Shoshenq [I]50

The larger Dakhla stela can be ascribed to Shoshenq I with confidence.51


The text mentions a wr“ feast on the occasion of a procession of the
god Seth on IV Peret 25 in 5 [Shoshenq] I.52 According to other attes-
tations, wr“ designates the lunar month and/or a lunar day, possibly
day 1: Demotic papyrus CG Cairo 30801;53 magical papyrus London-
Leiden (X 22 and XXI 19);54 Spiegelberg, Mythus vom Sonnenauge,
VIII 20;55 Ny Carlsberg Tebtunis-Papyri;56 Demotic Chronicle II,

49
Kitchen, TIP 3 XXIII–XXV.—Contra Kitchen, cf. G. F. Broekman, GM 205
(2005), 21–33.
50
For Dyn. 21 see above Jansen-Winkeln, Chapter II.9.
51
T. L. Sagrillo, The Reign of Shoshenq I of the Egyptian Twenty-second Dynasty (Leuven:
Dissertation, 2005).—R. Krauss, “Das wr“-Datum aus Jahr 5 von Shoshenq [I]”, DE
62 (2005), 43–48.
52
A. Gardiner, JEA 19 (1933), 19–30.
53
Parker, Calendars, § 89–98; cf. also R. Parker, A Vienna Demotic Papyrus on Eclipse
and Lunar Omina (Providence: BES 2, 1959), 8–9.
54
F. Ll. Griffith & H. Thompson, The Demotic Magical Papyrus of London and Leiden
(London, 1904), 79, 139; J. H. Johnson, in: H. D. Betz, ed., The Greek Magical Papyri
in Translation (Chicago, 1986), 209, 213, 230.
55
J. F. Quack, in: Res severa verum gaudium. Fs für Karl-Theodor Zauzich, F. Hoffmann
& H. J. Thissen, eds., (Leuven: SD 6, 2004), 50–51.
56
J. Osing, Hieratische Papyri aus Tebtunis I. Text (Copenhagen: CNIP 17, 1998), 207–210.
412 rolf krauss

9;57 building inscription of Parthenios (Moscow stela).58 Thus it is fea-


sible to search for a match between the date mentioned and a lunar
day appropriate for a procession. The sum of the highest attested reg-
nal dates for Osorkon II, Takelot I, Osorkon I, and Shoshenq I, added
to 841 BC as year 1 of Shoshenq III, yields 934 BC at the latest for
year 5 of Shoshenq I. Between 950 and 930 BC, only IV Peret 25 =
December 5 in 939 BC is an acceptable match for the wr“ date as
shown in Table III. 8.7. Consequently 1 Shoshenq I began in November
943 BC at the latest, and at the earliest in December 944 BC.
The stela records a judgment on the ownership of a well in Dakhla.
According to Gardiner’s understanding of the text, the mother of the
claimant is mentioned as the owner in a document dated to year 19
of a king Psusennes.59 Because at least 80 years separate 5 Shoshenq
I and 19 Psusennes I, it is unlikely that the document was written in
19 Psusennes I, but rather in 19 Psusennes II.

Three Lunar Dates of the Tepi Shemu Feast in Dyn. 21

Two of the priestly inductions known from Dyn. 21 apparently occurred


during the lunar Tepi Shemu feast:60 (a) 2 'Akheperre' setepenre': I Shemu
20, induction of a man, (b) whose son was inducted in 17 Siamun : I
Shemu [1]. E. Young assumed that the two inductions would have
been separated by 20 to 30 years, i.e. a generation.61 Provided that
both dates correspond to lunar days 1 to 5,62 at least 21 y minus 19
d separate them; other astronomical possibilities are 24, 27 and 30
years. The distance of 21 years is methodologically preferable, because
young men were inducted when they were 20 years old;63 furthermore
a distance of 21 years yields 6 regnal years for 'Akheperre' setepenre',
as in the Manethonian tradition for “Osochor”, successor of Amenemope.

57
H. Felber, “Die Demotische Chronik”, in: Apokalyptik und Ägypten, A. Blasius, ed.,
(Leuven: OLA 107, 2002), 76–77.
58
W. Spiegelberg, ZÄS 66 (1930), 422–443; Borchardt, Mittel, 39. Parker, Calendars, 18.
59
Gardiner (n. 52), 28.
60
Kruchten, Annales, 45–48.—Confirmed by Spalinger, “Dating”, 393.
61
E. Young, JARCE 2 (1963), 99–111.
62
Induction date b cannot be later than LD 3, because the 9th lunar month could
not begin before IV Peret 28. On the other hand, the distance between induction
dates a and b implies that b is 3 to 2 lunar days earlier than a.
63
Kruchten, Annales, 206.
lunar dates 413

Table III. 8.7


Year BC IV Peret 25 Year BC IV Peret 25 Year BC IV Peret 25

950 LD 2 943 LD–14 936 LD 3


949 12 942 –3 935 13
948 –7 941 9 934 –6
947 5 940 –11 933 5
946 15 939 1 932 –15
945 –5 938 12 931 –4
944 6 937 –8 930 8

Provided that 1 Shoshenq I = 943 BC and that Psusennes II ruled at


least 19 years, induction date b occurred at the earliest in 962 BC.
Table III. 8.8 contains the LD equivalents for dates a and b.64
Between 981 and 962 BC there are three viable alternatives (bold type).
Only one of them is left, if the oracular text no. 6 of Djehutimose is
taken into consideration.65 According to Kruchten, the oracular text
implies that the last and fifth day of Tepi Shemu fell on I Shemu 10 (or
9 or 8) in year 3 of either Amenemope, Osorkon, or Siamun. If the
induction dates (a) and (b) are correctly ascribed to Osorkon and Siamun
in relative chronology, then the oracular date would correspond to LD
8–10 in 3 Siamun. It would correspond to LD 15–18 in 3 Amenemope
if he ruled 9 full years and a fraction of a year, as usually assumed;
the date would correspond to LD 4–6 if 11 years are ascribed to
Amenemope. By contrast the oracular date coincides with LD 1 to 5
in year 3 of Osorkon, if induction date (a) corresponds to LD 1 to 5,
i.e. the oracular date must be attributed to Osorkon.
In table III. 8.8 only years 970, 973 and 981 BC suit induction date
(b) in 17 Siamun, and 990, 993 and 1001 BC for induction date (a)
in 2 'Akheperre' (Osorkon). The corresponding distances between 17
Siamun and 1 Shoshenq I = 943 BC amount to 26, 29 or 37 years.
If a reign of 19 years is supposed for Siamun on the basis of Manetho,66
then 24, 27 or 35 years are possible for Psusennes II. But because
Djehutimose oracle date no. 6 does not work with 992 and 1000 BC
as year 3 of 'Akheperre' (Osorkon), we are left with 989 BC = 3
'Akheperre' (Osorkon) as the only possibility.

64
The distance is throughout 21 years minus 19 days.
65
J.-M. Kruchten, Le grand texte oraculaire de Djehoutymose: intendant du domain d’Amon
sous le pontificat de Pinedjem II (Brussels: MRE 5, 1986), 237.
66
Cf. above Jansen-Winkeln, Chapter II. 9.
414 rolf krauss

Table III. 8.8


17 Siamun date b 2 Akheperre date a 3 Akheperre oracle date

962 BC LD 28 982 BC LD 1
963 18 983 20
964 7 984 9
965 26 985 28
966 15 986 17
967 4 987 7
968 23 988 26
969 12 989 16
970 2 990 5 989 BC LD 5 (6?)
971 21 991 24
972 11 992 13
973 –1 993 2 992 LD 4
974 19 994 21
975 8 995 10
976 27 996 29
977 16 997 19
978 5 998 8
979 24 999 27
980 14 1000 16
981 3 1001 5 1000 LD 6

Ramesside Lunar Dates67

About 85 years elapsed between 1 Smendes and 10 Amenemope. 200


to 201 years separated the accession of Ramesses II and the latest attes-
tation of Ramesses XI. If these figures are added to 992/991 as year
1 of 'Akheperre Osorkon, then 1 Ramesses II fell in 1279/1277 BC.
The ship’s log pLeiden I 350 records a LD 1 that coincided with II
peret 27 in 52 Ramesses II; on that day the ship moored in Piramesses,
hence the designation Piramesses date.68 This lunar date can be com-
bined with dates that refer to the feast-of-the-valley.69 According to
MHC 135 the feast began on LD 1 in II Shemu. On that day the cult
statue of Amun crossed the Nile, went to the temple of Djeser-akhet,70
toured Deir el-Bahri in a procession, and spent the night in the funer-

67
For the NK see above Hornung, Chapter II. 8.
68
J. J. Jansen, Two Ancient Egyptian Ship’s Logs (Leiden: OMRO Suppl. 42, 1961),
12, 33. Cf. also Beckerath, Chronologie, 51.
69
For a preliminary analysis under the outdated supposition that the reign of Ramesses
V extended into a year 5, see Krauss, Sothis, 136–144.
70
Urk. IV, 929.
lunar dates 415

ary temple of the ruling king 71 where Amun received offerings on LD


1 and 2, according to MHC 159. Four Dyn. 19/20 graffiti from the
Djeser-akhet temple in Deir el Bahri (DB) attest spending the night
or receiving offerings in II or III Shemu;72 the dates imply lunar days
1 or 2.
DB 3: year 7, II Shemu 28; Amun rests in the funerary temple of
Twosre73
DB 10: year 7, III Shemu 9; Amun rests in the funerary temple of
Ramesses III74
DB 9: year 6, III Shemu 9; Amun rests in the funerary temple of
Wesermare setepen ///75
DB 32: year 3, II Shemu 20; Amun rests in the funerary temple of
Ramesses II
Combining the Piramesses date with DB 3 and DB 10 allows the
chronology between Ramesses II and Twosre to be fixed. 76 According
to DB 10 Amun rested in the funerary temple of Ramesses III in a
year 7, presumably of Ramesses III himself.77 The dead-reckoned min-
imum distance between DB 3 and DB 10 amounts to either 9 years +
11 days or 10 a + 11 d. The astronomically correct distance between
DB 3 and DB 10 is 10 a + 11 d = 124 LM – 0.8 d = 3661 days,
resulting in a correct LD 2 for DB 10, provided that 1 Ramesses II =
1279 BC. According to oCG 25293 the highest attested dates for Queen
Twosre are either IV [prt x] and [I] Shemu [y] or IV [Shemu x] and
[IV] Shemu [y] of regnal year 8;78 thus the queen ruled either 323 ±
15 days or 412 ± 15 days subsequent to the date of DB 3. Sethnakhte
ruled into a year 3, corresponding to a reign of at least 731 days or
an arithmetical mean of 912 days. There are 2234 days between
Ramesses III’s accession and DB 10. These distances add up to 323
± 15 (412 ± 15) days + 731 to 912 days + 22343 days = 3288 ± 15
days to 3558 ± 15 days, approximating the distance of 3661 days
between the lunar dates of DB 3 and DB 10.

71
Or in the funerary temple of an earlier king, if the ruling king had not yet a
temple of his own.
72
Marciniak, Inscriptions.—Peden, Graffiti, 106–107, 120–123.
73
KRI IV 376.
74
KRI V 337.
75
KRI VI 102; but cf. Peden, Graffiti, 122 n. 395.
76
For details, see R. Krauss, SAK 24 (1997), 175–177.
77
Thus also KRI IV, 376.—Cf. table III. 10.10 for the astronomical possibility that
DB 10 belongs to Ramesses VII.
78
Beckerath, Chronologie NR, 74–76.
416 rolf krauss

If Amenmesses was a usurper in control of Nubia and UE in the


mid-reign of Sety II, then the interval between the Piramesses date and
DB 3 amounts to 36 a +121 d = 449 LM + 2 d. This figure corre-
sponds to the proper interval between a first LD and a LD 1–2 as a
feast-of-the-valley date (± 1 day). If, by contrast, the interval between
the two dates is lengthened by a chronologically independent 4 year
reign of Amenmesses, then the interval amounts to 40 a + 121 d =
498 LM + 15 d: then DB 3 would correspond to a LD 15 (full moon),
but not to a LD 1 or 2 as expected. If the distance is shortened, then
DB 3 would coincide with a LD 5 or 23. It follows that the reign of
Sety II subsumes that of Amenmesses.
Table III. 8.9 presents astronomically possible years for DB 3, 9, 10,
and 32, provided that 1 Ramesses II = 1279 BC and that the only
unresolved issue in Ramesside chronology is whether Ramesses’ VII
reign extended into a year 9 and/or Ramesses’ VIII reign into a year
2. Possible years are also given for DB 31, a graffito dated to year 22,
II Shemu 22 (sic); written “during the feast-of-the-valley”. Since neither
the resting of Amun nor offerings to him are mentioned, it is unclear
whether the date can be considered to be a LD 1–2.
Table III. 8.9 demonstrates that DB 9 is attributable to Siptah (sic)
and DB 32 to Ramesses VI, whereas DB 10 could be a date of Ramesses
III or VII. The attribution of DB 31 to Ramesses XI is problematic,
because a rock fall might have already destroyed Djeser-akhet during
his reign.79 If attributed to Ramesses II, then the date relates to a LD
3 which might have been a feast day in early Dyn. 19. If attributed
to Ramesses III, “during the feast-of-the-valley” could relate to all of
(lunar) II Shemu as the month of the feast. The Julian calendar equiv-
alents for the lunar days of the graffiti are:
DB 9: III Shemu 9, Year 6
[Siptah]: LD 2: April 21, 1192 BC
DB 3: II Shemu 28, Year 7, Twosre: LD 2: April 10, 1191
DB 10: III Shemu 9, Year 7,
[Ramesses III]: LD 2: April 18, 1181
DB 32: II Shemu 20, Year 3
[Ramesses VI]: LD 2: March 21, 1143
[DB 31: II Schemu 22, Year 22
[Ramesses XI]: LD 2: March 8, 1085]

79
Cf. J. Lipinska, JEA 53 (1957), 28–30; G. Pinch, Votive Offerings to Hathor (Oxford,
1993), 10–11.
lunar dates 417

Table III. 8.9


Ruler
DB 3 DB 9 DB 10 DB 32 DB 31

Ramesses II 27 – 7 3
Merneptah 20 – 28 –
Sety II 27 – 6 –
Amenmesses – – 27 –
Siptah 2 – 10 –
Twosre 2 – – – –
Sethnakhte – – 25 –
Ramesses III 21 2 28 26
Ramesses IV 6 – 14 –
Ramesses V – – 19 –
Ramesses VI 24 5 2 –
Ramesses VII 20 1 28 –
Ramesses VIII – – – –
Ramesses IX 27 9 6 –
Ramesses X – – 20 –
Ramesses XI 25 6 4 2

The DB-graffiti document the feast-of-the-valley in II and III Shemu,


whereas in MHC 135+159, only II Shemu is mentioned. As Table III.
8.10 shows, the feast dates fall in LM 11 or 10. The respective LM
could have been determined by the beginning of the first lunar month
within the civil year. The feast might have begun on LD 1 in LM 11,
if the first LM in the civil year began on I Akhet 15 at the latest; if
after I Akhet 15, then the feast began on LD 1 in LM 10. The ref-
erence to II Shemu in MHC might reflect the situation in the year
when the calendar was devised.
The lunar dates of DB 3, 9, 10 and 32 (DB 31 being not consid-
ered) are astronomically only compatible with 1 Ramesses II = 1314
or 1279 BC. In Table III. 8.11 all alternatives between 1304 and 1265
BC are tested. If 1 Ramesses II were 1290 or 1265 BC, then the DB
graffiti dates would coincide with LD 3–4 (positive errors), instead of
LD 1–2. If 1 Ramesses II were 1268 BC, the graffiti dates would cor-
respond to LD 1–2, but the Piramesses date would be off by +2 days
(positive error). It follows that only 1304 and 1279 BC are astronom-
ically possible for 1 Ramesses II. But 1304 BC is excluded on the basis
of the relative chronology of Dyns. 19 to 21, whereas 1279 BC accords.80

80
Thus the observation of the Piramesses date was negatively incorrect, and for that
reason A. Dodson (BiOr 57, 2000, 51 n. 5) suggested 1265 BC as year 1 of Ramesses
418 rolf krauss

Table III. 8.10


Graffito LD 2 LM LD 1 Graffito LD 2 LM LD 1
of LM 1 of LM 1

DB 9 III S 9 11 I A 13 DB 32 II S 20 10 I A 23
DB 3 II S 28 11 I A 2
DB 10 III S 9 11 I A 1

Table III. 8.11


1 Ramesses II LD of LD of LD of LD of Graffito LD of Graffito
Pir DB 3 DB 10 Year 34 Year 47

1304 BC –1 2 1 4 4
1290 2 3 2 6 5
1279 –1 2 2 4 4
1268 –2 1 –1 4 3
1265 1 3 2 6 5

Supplementary Lunar Dates

A visitor’s graffito, written at a pyramid in Saqqara on IV Shemu 24,


in 34 Ramesses II attests the “day of the feast of Ptah-south-of-his-
wall”;81 Peden presumed that the feast day was a public holiday.82 The
distance between it and the Piramesses date shows that the feast day
coincided with LD 4 or peret Sem, appropriate for a feast day of Memphite
Ptah whose main priest was the Sem.83 Another visitor’s graffito from
Saqqara attests two Memphite officials enjoying a stroll (swtwt) on II
peret 25 in 47 Ramesses II.84 The absence from work of a treasury-
scribe and a scribe of the vizier predicates a public holiday. The dates
of these two graffiti coincide with the same lunar day and thus sup-
port, but do not prove that 1 Ramesses II = 1279 BC.

II, or 1214 BC respectively as year of the Piramesses date when the observation would
have been correct. But Egyptologists should be prepared to find now and then a neg-
atively incorrect lunar date in their sources.
81
G. Jéquier, Deux pyramides du Moyen Empire. Fouilles à Saqqarah (Cairo, 1933), 13–15;
KRI III, 436.
82
Peden, Graffiti, 98–99.
83
L. Borchardt, ZÄS 70 (1934), 97–98, 100 n. 9; idem, Mittel, 52, speculated that
the date was related to full moon.
84
Firth & Quibell, Step Pyramid I, 82–83; KRI III, 148; cf. Peden, Graffiti, 99.
lunar dates 419

A Lunar Date for the Foundation of Amarna?

R. A. Wells presumed that the axis of the Small Temple at Amarna


(Ówt Jtn) was aligned towards the sun when it rose over the entrance
to the royal wadi on IV Peret 13 in 5 Akhenaten.85 On that day the
king took an oath to found Akhetaten.86 It may be presumed that
Akhenaten’s oath was followed by a foundation ceremony. If the cer-
emony took place on a LD 1 (attested for an earlier foundation cere-
mony of Thutmose III at Karnak (see below) and a later one of Ramesses
II at Luxor),87 then the year would have been 1348 BC, when IV Peret
14 (March 6 Jul) coincided with a LD 1.88

The Lunar Date of Amenhotep II

According to pSt. Petersburg 1116A, grain to brew beer for consumption


on a LD1 was allotted at the earliest on III Shemu 6 and at the lat-
est on III Shemu 9 in year 19 or 20 [of Amenhotep II]. Reckoning
from the lunar dates of Thutmose III, Beckerath concluded that the
LD 1 occurred close to III Shemu 9, provided (1) the papyrus dates
to year 20, and (2) there was no coregency of Thutmose III and
Amenhotep II.89 Parker objected to Beckerath’s conclusions on the
following grounds:90 1) In early 19th century Nubia brewing beer
took 3 to 4 days. If 1 Thutmose III = 1490 BC and if there was no
coregency, then 20 Amenhotep = 1417 BC. In that year LD 1 fell on
III Shemu 8, providing insufficient time for brewing beer. 2) Grain of
year 18 was allotted before II Shemu 17, a date corresponding to the
beginning of May (Greg.) around 1400 BC, when grain from the har-
vest of the current year would have been available, the harvest being
over before May (Greg.). It is not to be expected that grain of year
18 would still be in the granaries after the harvest of year 20 became
available. Therefore the papyrus should date to year 19.

85
R. A. Wells, SAK 14 (1987), 313–333; idem, GM 108 (1989), 87–90.
86
Murnane & van Siclen, Stelae, 38–40, 48.
87
D. B. Redford, JEA 57 (1971), 114; cf. KRI III, 346.
88
R. Krauss, GM 103 (1988), 44.
89
J.v. Beckerath, ZDMG 118 (1967), 18–21.
90
R. Parker, “Once again the Coregency of Thutmose III and Amenhotep II”, in:
Studies in Honor of John A. Wilson (Chicago: SAOC 35, 1969), 75–82; accepted by
Beckerath, Chronologie NR, 44; idem, Chronologie, 43–44.
420 rolf krauss

Parker’s arguments are invalid. The NK sources oCG 25780; 8 and


oCG 25782; 4, 7 report consumption of beer on the day after brew-
ing.91 The grain harvest actually continued well into May (Greg.); at
the beginning of the month deliveries from the new harvest would not
yet have arrived at the granaries.92 Thus it is quite possible that pSt.
Petersburg 1116A dates to year 20 of Amenhotep II. Provided 1
Thutmose III = 1479 BC, the alternatives for a LD 1 close to III
Shemu 10 in 20 Amenhotep II are as shown in Table III. 8.12.
If the papyrus dates to year 19, then there was a coregency which
lasted for 2y + 4m. But if to year 20, then there was a coregency of
only four months. There was no coregency at all, if the death date of
Thutmose III in the tomb of Amenemheb (TT 85) is emended from
III Peret 30 to III *Akhet 30 to conform with IV Akhet 1 as acces-
sion date of Amenhotep II.

Lunar Dates of Thutmose III

In 1957 Parker computed the possible solutions for the Akhmenu foun-
dation date under the traditional assumption that it took place in 24
Thutmose III on II peret 30, the day being a LD 1.93 In 1975 Wente
argued that only the order for preparing the foundation ritual was given
on II peret 30 and that LD 1 occurred on III peret 1;94 Parker criti-
cized Wente’s interpretation.95 In 1986 Beckerath published a new trans-
lation of the text, based on a more complete copy.96 He concluded that

Table III. 8.12


year of Amenhotep II LD 1 in III Shemu duration of coregency

19 = 1409 BC 11 2 years + 4 months


20 = 1406 BC 9 4 months or none

91
R. Krauss, OLZ 90 (1995), 241–242.
92
Cf. also Beckerath, Chronologie, 53.
93
R. A. Parker, JNES 16 (1957), 39–43.
94
E. F. Wente, “Thutmose III’s accession and the beginning of the New Kingdom”,
JNES 34 (1975), 265–272.
95
Parker, in: W. K. Simpson, Studies Dunham, 146–148.
96
J. v. Beckerath, “Ein Wunder des Amun bei der Tempelgründung in Karnak”,
MDAIK 37 (1981), 41–49.
lunar dates 421

the order for the foundation ritual was given on II peret 30 and that
the ritual took place on the same day. In other words, if the proper
day for the ritual was a LD 1, then the text implies that II Peret 30
was a LD 1.97 Parker, Wente and Beckerath relied on astronomical cal-
culations that were based on the slightly outdated parameters of P. V.
Neugebauer.98 According to computation with modern parameters, II
Peret 30 was a LD 1 rather than an old crescent day in 1455 BC as
year 24 of Thutmose III, corresponding to 1 Thutmose III = 1479
BC. We do justice to both uncertainties by reckoning with II Peret 30
as well as III Peret 1, as possible first lunar days.
The date of the Battle of Megiddo is (Urk. IV, 657): rnpt-zp 23, tpy
“mw sw 21, hrw ˙ #b n ps≈ntjw r-mtj:99 “Regnal year 23, I Shemu 21, day
of the feast of lunar day 1, exactly.” Although the text asserts the exact
coincidence of the civil date and a LD 1, Parker accepted Faulkner’s
argument for emending the day to I Shemu 20.100 Faulkner had pointed
out that Thutmose III arrived at Megiddo on I Shemu 19 and thus
the battle “should” have taken place on the next day, I Shemu 20, a
day that is not mentioned in the report. By contrast, Wente maintained
that Helck “convincingly” demonstrated that no such emendation of
the text is required.101 Whereas Beckerath accepted Wente’s judgment,
Parker cited Spalinger who he believed had shown that “Helck’s argument
lacks foundation.”102 Under these circumstances it is advisable to pro-
cede using both possible dates for the Battle of Megiddo.
Table III. 8.13 shows that there are two astronomically workable
possibilities: Either 1 Thutmose III = 1479 BC (Megiddo date = I
Shemu 21, Karnak date = II Peret 30 or III Peret 1), or 1 Thutmose
III = 1504 BC (Megiddo date = I Shemu 20, Karnak date = III Peret
1). It is impossible to reconcile 1 Thutmose III = 1504 BC with 1
Ramesses II = 1279 BC and 9 Amenhotep I = 1506 BC (see chapter
III. 10); thus only 1 Thutmose III = 1479 BC remains. In that case

97
Cf. most recently Spalinger, “Dating”, 387–390.
98
Parker did his own computations; Wente used Parker’s computations. Beckerath
asked for the assistance of astronomers who also used Neugebauer’s Tables; there are
(copying?) mistakes in Beckerath’s figures.
99
For the “exact” feast of (lunar) wr“, see Quack (n. 55), 472.
100
Parker, Calendars, 73; cf. idem, JNES 16 (1956), 40, referring to R. O. Faulkner,
JEA 28 (1942), 4; 11.
101
E. F. Wente, JNES 34 (1975), 265, referring to W. Helck, MDAIK 28 (1972),
101–102; see also J. v. Beckerath, MDAIK 37 (1981), 49 n. 18.
102
Parker (n. 95), 146, citing A. J. Spalinger, MDAIK 30 (1974), 221–229.
422 rolf krauss

the battle was fought on I Shemu 21. The text’s assertion that the date
corresponded to a LD 1 “exactly” can imply that both armies waited
for the moon’s invisibility to go to battle.

The Lunar Date of the Ebers Calendar

According to the interpretation of both Parker and Hornung, the ris-


ing of Sothis and the first day of the lunar month wep renpet coincided
in regnal year 9 of Amenhotep I on III Shemu 9. As argued below in
chapter III. 10, it is highly likely that this is correct. Nevertheless, for
now we shall refrain from using III Shemu 9 in 9 Amenhotep I as a
LD 1.

The Monthu Feast in pBoulaq 18 as a Lunar Date103

Gardiner, Borchardt104 and Quirke read the royal name in pBoulaq


18/L as [Sbk]-˙tp and in pBoulaq 18/S as Jmn-m-˙ #t sbk-˙tp,105 identi-
fying this king with Sobekhotep II, the predecessor of Khendjer accord-
ing to TC. By contrast, Ryholt recognizes only “doubtful traces”, arguing
that the TC’s Sobekhotep II was in fact the first ruler of Dyn. 13.106
pBoulaq 18/L,S/ attests a vizier Ankhew who is a known contempo-
rary of Khendjer. By the reign of Sobekhotep III, Ankhew’s son Ressonb

Table III. 8.13

1 Th III LD of Error LD of Error LD of Error LD of Error


I Shemu 21 I Shemu 20 II Peret 30 III Peret 1

1504 BC 2 – 1 0 –1 + 1 0
1493 –1 + –2 ++ –2 ++ –1 +
1490 2 – 1 0 2 – 3 – –
1479 1 0 –1 + 1? 0? 1?? 0??
1468 1 0 –1 + –2 ++ –1 +
1465 2 – 1 0 2 – 3 – –

103
For the MK and SIP, see above Schneider, Chapter II. 7.
104
Cf. Beckerath, Zwischenzeit, 47–49.
105
S. Quirke, The Administration of Egypt in the Late Middle Kingdom (Whitstable, 1990),
27 n. 12.
106
Ryholt, Situation, 319.—Cf. above Schneider, Chapter II. 7.
lunar dates 423

had succeeded him as vizier, and thus pBoulaq 18 and the other sources
that mention Ankhew can be dated prior to Sobekhotep III.107 The
attribution of pBoulaq 18 to one of the kings between Khendjer and
Sobekhotep III (Seth, Inyotef V or Imiramesha)108 is contradicted by
the poor documentation for these kings.109 Thus it is preferable to date
pBoulaq 18 to Khendjer’s immediate predecessor, be he Sobekhotep
II or not.
According to pBoulaq 18/L, in Medamud a Monthu feast was cel-
ebrated on III Akhet 17 and 18 in year 3 of the predecessor of
Khendjer.110 A Monthu feast is also attested in the Illahun archive. The
instances are dated to II Akhet 14 and 22 respectively111 and coincided
with ¢np-“ 'w, a feast on lunar day 2.112 Provided the Monthu feast of
Boulaq 18 on III Akhet 17 was also celebrated on lunar day 2, the
difference of one civil month can be explained like the shift of the w #g j
feast in the Illahun archive: it shifted during the reigns of Senwosret
III and Amenemhet III from II to III Shemu; presumably it took place
in the 2nd lunar month after peret Sepdet.113 If, for example, the Monthu
feast was fixed on LD 2 of the 7th lunar month after peret Sepdet, then
it would have taken place mostly in II Akhet during Dyn. 12, but never
before III Akhet 1 during Dyn. 13.
For an estimate of the period of pBoulaq 18 we rely provisionally
on the TC and the stèle juridique. Provided 1 'Ahmose = 1539/34 BC,
the figures of the TC and the few preserved regnal dates yield 1591/86
BC at the latest for year 1 of Sewadjenre' Nebiriau I. The genealog-
ical and historical information contained in the text of the stèle juridique
indicates an interval of ca. 55 ± 15 years between the first years of
Sewadjenre' and Merhetepre' Ini.114 Considering ca. 2 more years for
the immediate predecessor of Khendjer,115 5 years for Khendjer,116 8

107
Ryholt, Situation, 193, 243.
108
Ryholt, Situation, 194, 244.
109
Cf. Ryholt, Situation, 342.
110
A. Scharff, ZÄS 57 (1922), 63.
111
Luft, Fixierung, 97–99, 114–118, 170.
112
Cf. Luft, Fixierung, 29, 111, 118 (pBerlin 10001, 10248 and 10282); cf. also Krauss,
GM 138 (1994), 84–85.
113
Krauss, Sothis, 94.
114
Beckerath, Untersuchungen, 182; Ryholt, Situation, 233–234.
115
Provided that year 5 of pBoulaq 18/S belongs to the same reign as pBoulaq
18/L with the Monthu feast in year 3; cf. Ryholt, Situation, 194 with n. 688.
116
Control note in the pyramid of Khendjer, cf. Ryholt, Situation, 193–195.
424 rolf krauss

to 9 years for Sobekhotep IV,117 and the figures in the TC for Sobekhotep
III through Merhetepre' Ini, the distance between regnal year 3 of
pBoulaq 18 and year 1 of Merhetepre' Ini amounts to ca. 65 to 67
years. Thus regnal year 3 of pBoulaq 18 fell at the latest in 1712/07
BC ± 16 years. III Akhet 17, the date of the Monthu feast coincided
with a LD 2 in 1745, 1734 and 1684 BC.118 The regnal years that are
preserved in the TC between Wegaf and the predecessor of Khendjer
add up to 8 to 14 years at least; the regnal years of ca. 8 kings are
lost. Thus nothing prohibits equating year 3 of pBoulaq 18 with 1734
BC, provided the last year of Dyn. 12 corresponds to ca. 1760 BC (see
next paragraph).

The Illahun Lunar Dates (ILD)

In 1992 Luft presented the lunar dates that are contained in the Illahun
archive,119 relying on earlier studies of Borchardt, Dévaud, Möller,
Parker and Kaplony-Heckel. Details of Luft’s original presentation have
been corrected in the following list.120 21 Illahun lunar dates (ILD) are
useful for chronological analysis:
ILD 1: pBerlin 10003; 9 [Senwosret III]
ILD 2: pBerlin 10248; 14 Senwosret III
ILD 3: pBerlin 10016; 18 [Senwosret III]
ILD 4: pBerlin 10090; 3 [Amenemhet III]
ILD 5: pBerlin 10056 A; 8 [Amenemhet III]
ILD 6: pCairo 58065; 9 [Amenemhet III]
ILD 7: pBerlin 10103; 29 [Amenemhet III]
ILD 8–19: pBerlin 10056; 30/31 [Amenemhet III]
ILD 20–21: pBerlin 10006; 32 [Amenemhet III]

117
W. K. Simpson, MDAIK 25 (1969), 154–158.
118
In 1709 BC a LD 1 coincided either with III Akhet 15 or 16, resulting in III
Akhet 17 being a LD 3 or 2.
119
Luft, Fixierung.
120
Krauss, in: Bietak, SCIEM Haindorf 2001, 175–178.
lunar dates 425

Commentary

ILD 1: Report of the 4th phyle that withdraws from monthly temple
service and of the 1st phyle that takes over. Luft dates the protocol to
[III Peret 9], the day before the first preserved dated entry. This is
confirmed by the distances between III Peret 9 in 9 Senwosret III and
ILD 3, 5, and 6.
ILD 2: copy of a letter announcing two festivals: one on II Akhet
18, corresponding to LD 2, and another on II Akhet 20, correspond-
ing to LD 4. According to Luft the letter might have arrived on II
Akhet 16, i.e. one day before the lunar month that is referred to would
have begun. Therefore it possible that the lunar dates do not corre-
spond to observation, but to expectation and are off by one day.
ILD 3: letter concerning the moveable w #g j festival. After II Shemu
17, the date of w #g j, there is added: “2-nw n m≈≈j-nt = the second (day)
of LD 15”. Evidently the addendum identifies the date of w #g j as LD
17. Thus II Shemu 17 coincided with LD 17 as the day of w #g j, and
the corresponding LD 1 was II Shemu 1.
ILD 6: list of offerings for the Wagi festival. Here Wagi coincided
with II Shemu 29; if it were LD 17, then the corresponding LD 1 was
II Shemu 13.
ILD 4: The entries mention offerings on III Shemu 15, followed by
“sw3 ˙r III Shemu 16” in turn followed by “LD 1”.121 According to
Luft “sw3 ˙r III Shemu 16” means that III Shemu 16 was skipped.122
If so, the LD 1 mentioned after III Shemu 16 has to be III Shemu
17.123
ILD 5: list of offerings, including those made in year 8.124 Luft rec-
ognized the date IV Akhet 26 in year 8 as a LD 1.125
ILD 7: In a letter dated I Shemu 15, an official complains that a
LD 9 had occurred, without the delivery of a bull for an offering which
had been due. Luft argued that the non-delivery should have prompted
an immediate reaction, “so that in view of the small distances between
the offices the 9th lunar day in all probability coincided with I Shemu
15”, i.e. LD 1 would fall on I Shemu 7.

121
Luft, Fixierung, 86–88.
122
Luft, Fixierung, 86 ff.
123
R. Krauss, OLZ 89 (1994), 10.
124
Luft, Fixierung, 70–73.
125
R. Krauss, GM 138 (1994), 82–83.
426 rolf krauss

ILD 8–19: account for a “year” (rnpt) overlapping regnal years (rnpt
zp) 30 and 31 of Amenemhet III. Borchardt recognized this “year”
(rnpt) as a lunar year of 354 days with the account covering the months
2, 4, 6, 8, 10 and 12. The lunar intervals of 29 or 30 days are indi-
cated by first and last dates linked by nfrjjt-r (until):
II Shemu 26 nfrjjt-r III Shemu 25
IV Shemu 25 nfrjjt-r year 31, I Akhet 1[9]
year 31, II Akhet 20 nfrjjt-r III Akhet 19
IV Akhet 1[9] nfrjjt-r I Peret 18
II Peret 18 nfrjjt-r III Peret 17
IV Peret 17 nfrjjt-r I Shemu 16
Parker interpreted the dates before nfrjjt-r as first days, and the ones
after as last days of an Egyptian lunar month.126 Reckoning the dis-
tances between ILD 8–19 and the other ILD, Luft realized that “LD
2 nfrjjt-r LD1” is meant.127
ILD 20–21: Rt III of pBerlin 10006 contains an account for the 29
days from II Akhet 9 to III Akhet 7 in 32 [Amenemhet III]; the inter-
val is called a “month” (3bd ).128 According to Luft, this “month” des-
ignates a lunar month of temple service which started on a LD 2 and
ended on a LD 1. Rt II records the delivery of offerings for LD 1 and
2 and III Akhet 6 and 7. Parker presumed that the dates coincided
with the lunar days;129 according to Luft, the deliveries were one day
ahead.
Reckoning the distance between the lunar dates shows that the ILD
derive from two different reigns, one of them included ILD 1, 2 and
3 from regnal years 9, 14 and 18 of a king who ruled 19 years and
who is clearly Senwosret III. The remaining lunar dates come from
regnal years 3, 8, 9, 29, 30/31 and 32 of a different king who is just
as clearly Amenemhet III. In the context of the Illahun archive, these
reigns can only be arranged as a reign of at least 32 years that suc-
ceeded a reign of 19 years. Whether or not Senwosret III took Amenem-
het III as coregent at the end of his regnal year 19 is irrelevant since
the following regnal years were counted as Amenemhet III’s (see above

126
Parker, Calendars, 64, after Wheeler.
127
Luft, Fixierung, 233–234.
128
Luft, Fixierung, 42–44.
129
Parker, Calendars, 63.
lunar dates 427

Schneider, Chapter II. 7). Thus the ILD span a period of 42 calendar
years; the earliest is 9 Senwosret III, the latest is 32 Amenemhet III.
The ILD constitute a set of interconnected lunar dates. If one of them
is ascribed to a certain year BC, the others follow. The set can be
expressed either in relation to 1 Senwosret III or 1 Amenemhet III.
As demonstrated above, the astronomically correct solution is 1 Senwosret
III = 1837/36 BC and 1 Amenemhet III = 1819/1818 BC.
In 6 Senwosret III the temple service was organized by civil months;
the phyles were not numbered, but named after civil months.130 Three
years later, in 9 Senwosret III, ILD 1 attests temple service by lunar
months, beginning on LD 1. Some 40 years later, in years 30–32 of
Amenemhet III, the monthly temple service started on a LD 2 and
ended on a LD 1 as Luft has recognized. It is clear that the lunar
month itself and the counting of its days were not affected by the
changes within the temple service. Nevertheless, when analysing the
ILD, there are two sets of dates: a) ILD 8–21, dependent on the begin-
ning of the temple service in the lunar month and b) ILD 1–7, inde-
pendent of it. If 1 Amenemhet III = 1869/68 BC, then six dates of
ILD 1–7 are correct and one is positively incorrect, yielding p = 0.05.
If 1 Amenemhet III = 1844/43 BC then of ILD 1–7 six dates are cor-
rect and one is negatively incorrect, yielding p = 0.32. Thus 1 Amenemhet
III = 1844/43 BC is preferable over 1 Amenemhet III = 1869/68
BC.131

The Khozam Lunar Date

The inscription Cairo JE 43290 equates IV Akhet 25 in year [1] of


an unnamed [Herakleopolitan] king with lunar day 15.132 Fischer asso-
ciated the inscription with Khozam (Coptos) and dated it to the last
years of the Herakleopolitan’s hold over the Coptite nome, i.e. to the
early years of Dyn. 11.133 Dead reckoning yields 2046 BC at the very
latest for the beginning of Dyn. 11. The regnal years of the kings of
early Dyn. 12 are well attested; nevertheless, there is the problem of

130
pBerlin 1050; cf. L. Borchardt, ZÄS 41 (1904), 35–36; Luft, Fixierung, 65–67.
131
Cf. Figures III. 8.1–2.—The probabilities for ILD 8–21 are correspondingly
p = 0.16 and p = 0.014.
132
Goedicke, MDAIK 50 (1994), 72; see above Baud, Chapter II. 5.
133
Fischer, in: Manuelian, Studies Simpson, 267–270.
428 rolf krauss

the historicity of coregencies during the first four reigns. Uncertainties


amount to 16 years, and thus 1 Amenemhet I is either 1940 BC or
1956 BC, if 1 Senwosret III is 1837/36 BC.
The highest attested dates for Dyn. 11 yield 106 years. There are
no regnal years preserved for three rulers, and for at least three others,
the highest attested dates are presumably different from their last reg-
nal years. Thus dead reckoning implies a date well before 2046 or 2062
BC for the beginning of Dyn. 11.
The total of 143 years in the TC for 6 kings of Dyn. 11 represents
a tradition that goes back to the common source of TC and Manetho:
Manetho’s 16 kings of Dyn. 11 who ruled for 43 years are clearly sec-
ondarily garbled from 6 kings ruling 143 years. The highest attested
dates do not contradict the figures in TC, although it gives 49 years
for Horus Wah-ankh Inyotef II, whereas the date of the so-called dog-
stela is his year 50; presumably the TC rounded off from 49 y + x
m + y d. The total of 143 years seems to be of correct magnitude,
but it might be incorrect in detail, if only a few years are missing
because the figures are rounded off.134 Addition of the TC’s 143 years
to 1 Amenemhet I = 1940/39 or 1956/55 BC yields 2083/82 or
2099/98 BC for the beginning of Dyn. 11. These figures must be rec-
onciled with the Khozam lunar date that corresponds to the following
absolute dates as shown in Table III. 8.14:

Table III. 8.14


Year BC LD of IV Akhet 25 Year BC LD of IV Akhet 25
2053 15 2092 14
2064 16 2103 15
2067 14 2114 16
2078 15 2117 14
2089 16 2128 15

2053 BC, the earliest astronomical possibility for the Khozam lunar
date, is historically too early. The years 2078, 2089, and 2103 BC are
compatible with the suppositions made here about Dyns. 12 and 11.135

134
K. Ryholt, Ä&L 14 (2004), 152.
135
The year 2092 BC should not be considered, because it would imply a positively
incorrect lunar date.
lunar dates 429

Lunar Dates of Late Dyn. 5136

Analysis of the lunar dates preserved in the Neferirkare' archive pre-


supposes an estimate of the minimum distance between Dyns. 6 and
11. We rely on the judicious estimate of Hayes and Fischer that the
FIP lasted between 30 and 40 years.137 Baud estimates a “generation”
or ca. 30 years for Dyn. 8 and at least 83 to 85 years for Dyn. 6 (if
only the attested (m-)¢t zp years are considered) or a minimum of 147
to 153 years (assuming a regular biennial count). He notices the imbal-
ance of zp years and (m-)¢t zp years during the reign of Pepy II, but
tends to accept a regular biennial count throughout Dyn. 6 neverthe-
less (see above, Chapter II. 5).
Baud’s point of view is supported by the following argument. From
all of Dyn. 6, 10 m-¢t zp years and 12 zp years are preserved. At least
72 zp years actually occurred, i.e. 83.3% of the zp years are missing
from the record. There is no reason why more or less m-¢t zp years
than zp years should be missing. Thus it is likely that at most 60% of
the regnal years of Dyn. 6 were zp years and at least ca. 40% m-¢t zp
years; if so, at least 120 to 144 regnal years must be attributed to Dyn.
6, implying a regular biennial count and thus the dynasty’s duration
for 147/153 years.138 If the Khozam lunar date corresponds to 2078
or 2103 BC, our figures for the FIP, and Dyns. 6 and 8, result in ca.
2293 or 2318 BC at the latest for the first year of Dyn. 6.
The highest date for Wenis is IV Shemu in rnpt zp 8,139 corresponding
to regnal year 16 at most. If Wenis reigned for 16 years, his year 1
fell between ca. 2334 and 2309 BC at the latest, whereas a 44 year
reign of Isesi would have begun between 2378 and 2353 BC.
Posener-Kriéger identified two lunar dates in the Neferirkare' archive:140
(I) pLouvre E.25279 recto: LD 1 and 2 = IV Shemu 17 and 18, cor-
responding to the year before rnpt zp 4 that is mentioned on the verso,
i.e. regnal year 7 or 6;141 (II) pBM 10735 recto: II Shemu 18, statue

136
For Dyns. 4 to 5, see above Verner, Chapter II. 4.
137
W. C. Hayes, “The Middle Kingdom in Egypt: Internal History from the Rise
of the Heracleopolitans to the Death of Ammenemes III”, (CAH I.3, 1961), 4–5.—
H. G. Fischer, Inscriptions from the Coptite Nome (Rome, 1964), 42.
138
This yields a total of ca. 183 years for Dyns. 6 and 8 which is reminiscent of
the total of (181 + 6) years + 6 monhts + 3 days in TC.
139
Posener-Kriéger, Archives II, 491.
140
Posener-Kriéger, Archives II, 486–488.
141
As a rule the first count occurred after rnpt zmA tAwj; for an exception see
Baud, “Ménès”, 125. Thus the year before rnpt zp 4 is regnal year 7, rather than 6.
430 rolf krauss

ritual = LD 2.142 The verso is dated to rnpt zp 21 or rather 22; i.e.


regnal year 42 or rather 44, if the count was biennial.143
There are two models for interpreting the lunar dates I and II. In
model A lunar date I belongs to 7 Izezi and II belongs to year 21 of
Izezi. In model B lunar date I belongs to 7 Wenis and II occurred 11,
25, or 36 years earlier, corresponding to year 15, 26 or 40 of Isesi,
provided he ruled for 44 years. If so, then in relative chronological
terms at least ca. 250 years elapsed between his year 7 and the Khozam
lunar date. Only 261 y + 119 d or 286 y + 119 d are proper lunar
distances for model A. Table III. 8.15 contains the possibilities for 1
Isesi between 2393 and 2335 BC, if model A is followed. The only
astronomically viable possibility would be 7 Isesi = 2346 BC and 21
Isesi = 2325 BC, implying 208 years between year 1 of Dyn. 6 and
2078 BC for the Khozam lunar date, or 183 years, if the Khozam
lunar date corresponds to 2103 BC. Both intervals are apparently too
short and thus model A does not seem to work, in harmony with
Posener-Krieger’s slight preference of model B over A,144 which we
share.
For model B only 211 y, 236 y, or 261 y (+ 119 d each) are proper
lunar distances. Table III. 8.16 contains the possibilities for 1 Wenis
between 2382 and 2296 BC and the corresponding earlier years of
Izezi.

Table III. 8.15


Izezi Izezi LD 1 Izezi LD 2
Year 1 Year 7 in IV Shemu Year 21 in II Shemu

2335 2328 19 2314 19


2346 2339 17?18? 2325 18
2357 2350 16 2336 17
2360 2353 18 2339 20
2371 2364 18 2350 18
2382 2375 16 2361 17
2385 2378 19 2364 20
2393 2386 15 2372 16

142
Posener-Kriéger, Archives I, 52–54.
143
See above Verner, Chapter II. 4.
144
Posener-Kriéger, Archives II, 487.
lunar dates 431

Table III. 8.16


Wenis ‘7 Wenis’ ‘40 Izezi’ ‘26 Izezi’ ‘15 Izezi’
Year 1 LD 1 in LD 2 in LD 2 in LD 2 in
IV Shemu II Shemu II Shemu II Shemu

2296 ‘2289’ ‘2300’ ‘2314’ ‘2325’


17 18 19 18
2321 ‘2314’ ‘2325’ ‘2339’ ‘2350’
17 18 20 18
2346 ‘2339’ ‘2350’ ‘2364’ ‘2375’
17? 18? 18 20 19
2371 ‘2364’ ‘2375’ ‘2389’ ‘2400’
18 19 19? 20? 19
2382 ‘2375’ ‘2386’ ‘2400’ ‘2411’
16 17 19 17

Model B allows for three astronomically workable possibilities for year


7 of Wenis. Of them 2289 BC implies unacceptable 200 years between
year 1 of Dyn. 6 and 2078 BC for the Khozam lunar date. The remain-
ing possibilities for 7 Wenis are 2314 or 2339 BC; they imply accept-
able 225 years between year 1 of Dyn. 6 and the Khozam lunar date
of 2078 or 2103 BC. Thus the earliest lunar dates that are available
for chronological analysis yield two possibilities for the end of Dyn. 5
which are 25 Egyptian years or a lunar cycle apart.
III. 9 THE HELIACAL RISING OF SIRIUS

Teije de Jong

The heliacal rising of Sirius, the brightest star in the sky, was used in
antiquity, both in Egypt and in Mesopotamia, to synchronize the cal-
endar to the solar year. On the day of its heliacal rising Sirius is seen
again for the first time in the morning twilight sky after having been
invisible for about 70 days (at the geographical latitude of Memphis).
On that day it appears a few degrees above the Eastern horizon and
disappears again after about 15 minutes due to the brightening of the
sky just before sunrise. The date of heliacal rising depends on the rel-
ative positions of Sirius and the Sun with respect to the horizon and
on atmospheric conditions.
According to Parker in the ancient Egyptian lunar calendar an addi-
tional 13th month was intercalated in the next year whenever the first
visibility of Sirius (associated with the Goddess Sothis) occurred during
the last 11 days of the last month Wep renpet of the lunar year.1 In this
way the Egyptians could make sure that the first month Toth of their
lunar calendar always began shortly (within one lunar month) after the
first visibility of Sirius. The heliacal rising of Sirius plays a crucial role
in Egyptian chronology because it is supposed to fix the zero-point of
the Egyptian civil calendar of 365 days by postulating that at the time
of its installation the first visibility of Sirius occurred on the first day
of the first month.
One of the earliest references to the use of Sirius for intercalation
in the Mesopotamian lunar calendar is found on Tablet II of MUL.APIN
where we are told that if Sirius rises in the month Du"uzu (the 4th
month) this year is normal but if it rises in the month Abu (the 5th)
an extra month has to be intercalated that year.2 On Tablet I the nom-
inal date of the first visibility of Sirius in the lunar calendar is given

1
R. A. Parker, “The Calendars and Chronology”, in: The Legacy of Egypt, J. R.
Harris, ed., (Oxford, 1971), 13–26; cf. also below, Chapter III. 10.
2
H. Hunger & D. Pingree, MUL.APIN: An Astronomical Compendium in Cuneiform, AfO
Beiheft 24 (1989).
the heliacal rising of sirius 433

as “the 15th of Du"uzu” (the 4th month) as part of a list of first visi-
bility dates of about 30 stars. This list was based on observations car-
ried out in Babylon during the 13th century BC.3 Later (during the
5th to 1st centuries BC) dates of the first appearance of Sirius are given
in the Astronomical Diaries.4 Sachs has shown that in the Babylonian
19-year calendar cycle, introduced in the early 5th century BC and
used without any further change during five centuries throughout the
ancient near-East, the intercalation pattern was arranged in such a way
that Sirius always rose in the 4th lunar month Du"uzu.5
Starting with Ptolemy (~130 AD) in his Almagest (Book VIII.6),6
astronomers have discussed the heliacal rising (also referred to as first
visibility or first appearance) of stars and planets in terms of the so-
called arcus visionis, the distance between a star/planet and the sun,
measured in degrees perpendicular to the horizon when it becomes
visible again for the first time in the morning twilight sky. In the sec-
ond book of his Fãseiw éplan«n ést°rvn ka‹ sunagvgØ §pishmais«n.
Ptolemy gives the dates of first and last visibility of some 30 bright
stars at different geographical latitude zones (kl¤mata) from which values
of the arcus visionis can be deduced.7 For Sirius Ptolemy apparently uses
a value of about 11°.
In the 1920’s Carl Schoch was the first to attempt to determine val-
ues of the arcus visionis for Sirius and for the planets from Babylonian
observations.8 He derived a value of 7.7° degrees for Sirius. The recent
edition of the Astronomical Diaries by Sachs and Hunger allows a
reanalysis of the Babylonian observational material. We now know that
the number of genuine observations of the date of first visibility of
Sirius in the Diaries is quite small since Sachs showed that almost all
dates during the Seleucid Era (last three centuries BC) are computed

3
T. de Jong, “A New Attempt to Date the Observations of Rising Stars in
MUL.APIN”, paper presented at the 7th Notre Dame Workshop on the History of
Astronomy ( July 2005).
4
Sachs & Hunger, Diaries.
5
A. Sachs, “Sirius Dates in Babylonian Astronomical Texts of the Seleucid Period”,
JCS 6 (1952), 105–114.
6
See Toomer, Almagest.
7
H. Vogt, “Griechische Kalender V. Der Kalender des Claudius Ptolemäus”, Sitzungs-
berichte der Heidelberger Akademie der Wissenschaften, Phil.-Hist. Klasse, 15. Abh. (1920), 1–61
8
C. Schoch, The Arcus Visionis in the Babylonian Observations (Oxford, 1924); idem,
“The Arcus Visionis of the Planets in the Babylonian Observations”, Monthly Notices of
the Royal Astronomical Society 84 (1924), 731–734.
434 teije de jong

rather than observed.9 This is probably the reason that Schoch’s value
is rather small, because the Babylonian computational scheme is appar-
ently based on dates observed under optimal atmospheric conditions.
It turns out that there are only two genuinely observed dates of the
first appearance of Sirius listed in the Diaries: day 18, month IV, year
12 of Darius III (22 July 325 BC) and day 13, month IV, SE 22 (20
July 290 BC). An analysis of these data results in arcus visionis values
of 11.0° and 8.6°, respectively.
To determine the best value of the arcus visionis of Sirius for the pur-
pose of calibrating the Egyptian civil calendar in studies of Egyptian
chronology L. Borchardt and P. V. Neugebauer in 1926 organized an
observing program of the first visibility of Sirius in Egypt. Results of
this study are summarized and discussed in a recent paper by Pachner.10
He shows that the 1926 program resulted in the determination of arcus
visionis values for Sirius of 8.7°–9.2°.
The problem was discussed again by Schaefer without taking the
results of Pachner into account.11 He suggested a value of the arcus
visionis for Sirius of about 11°, similar to the value derived by Vogt
from Ptolemy’s Phaseis. His analysis was based on a theoretical model
of stellar visibility adopting a visual extinction for the atmosphere near
ancient Memphis of 0.35 magnitudes per air mass (one air mass is a
measure of the thickness of the atmosphere at zenith). This extinction
estimate was based on data for Jerusalem. In his paper he also listed
the expected dates of heliacal rising of Sirius between 3500 BC and
500 AD for his adopted value of the extinction.
To put things into perspective I have computed values of the arcus
visionis and dates of the heliacal rising of Sirius in Egypt using the
model of de Jong and Inklaar,12 which is an improved and updated
version of the earlier model of Inklaar.13 In this model the visibility of
stars and planets is computed based on the brightness of the twilight

9
See n. 5.
10
N. Pachner, “Zur Erfassung der Sichtbarkeitsperioden ekliptikferner Gestirne”,
Ä&L 8 (1998), 125–136.
11
B. E. Schaefer, “The Heliacal Rise of Sirius and Ancient Egyptian Chronology”,
JHA (2000), 149–155.
12
T. de Jong & F. Inklaar, “A New Method to Compute First and Last Visibilities
of Stars and Planets” (2006), in preparation.
13
F. Inklaar, Een Nieuwe Methode voor de Berekening van Heliakische Opkomsten, doctor-
aalscriptie, (Universiteit van Amsterdam, 1989).
the heliacal rising of sirius 435

sky as a function of the depression of the Sun below the horizon, on


the transparency of the atmosphere and on the sensitivity of the human
eye in twilight conditions. The physical principles underlying this model
are similar to those adopted in the earlier models of Bruin14 and of
Schaefer15 but some of the assumptions and parameters used are different.
Using this model I have computed the parameters in Table III. 9.1
characterizing the heliacal rising of Sirius in 1000 BC for a location
at geographical latitude 30° North, representative for ancient Memphis
along the Nile. Visual magnitude, position and proper motion of Sirius
were taken from The Bright Star Catalogue16 and precession was calcu-
lated according to algorithms in the Explanatory Supplement to the Astronomical
Almanac.17 Results are given for three different values of the visual atmos-
pheric extinction k(V) in column (1) by averaging over 4 years around
1000 BC. Values of the apparent altitude of Sirius above the horizon
when it first becomes visible are given in column (2) and of its true
altitude (without atmospheric refraction) in column (3) as well as the
true solar depression below the horizon in column (4). In column (5)
I list the actual average arcus visionis <h>, the distance between Sun
and Sirius perpendicular to the horizon at first visibility (column (4)
subtracted from column (3)). In columns (6) and (7) I list the Julian
date and the local time of the first visibility of Sirius and in column
(8) the duration of its visibility. The averaging takes account of the fact
that the ecliptic longitude of the Sun at sunrise on any Julian date
varies from year to year and returns to the same value every four years
(due to Julian intercalation). Notice that if one wants to compute the
date of heliacal rising based on a value of the arcus visionis one should
use a minimum value h0 which is ~0.5° (one half the daily motion of
the Sun) smaller than the average values listed in Table III. 9.1.
The day on which the heliacal rising of Sirius (Sothis) was observed
in ancient Egypt clearly depends on the prevailing atmospheric con-
ditions. In Table III. 9.2 I show values of the average visual extinc-
tion as measured at different locations on Earth at different epochs.

14
F. Bruin, “The heliacal setting of the stars and planets” I & II, Proceedings of the
Koninklijke Nederlandse Akademie van Wetenschappen, Series B, Vol. 82 (1979), 385–410.
15
B. E. Schaefer, “Predicting Heliacal Risings and Settings”, Sky and Telescope 70
(1985), 261–263; idem, “Heliacal Rise Phenomena”, Archeoastronomy 11 (1987), S19–S33.
16
D. Hoffleit, The Bright Star Catalogue (New Haven, 1982).
17
P. K. Seidelmann, ed., Explanatory Supplement to the Astronomical Almanac (Mill Valley:
University Science Books, 1992).
436 teije de jong

Table III. 9.1. Heliacal rising of Sirius at GB = 30° in 1000 BC


Visual Apparent Real Real Arcus Julian Local Duration
extinction stellar stellar solar visionis date Time of
k(V) altitude altitude altitude <h> visibility

0.20 1,5 1,1 –6,9 8,0 16-jul 4:25 13


0.27 2,3 2,0 –7,3 9,3 17-jul 4:24 14
0.35 3,2 2,9 –7,7 10,7 19-jul 4:22 14
mag/airmass degrees degrees degrees degrees dd-mon hrs:min minutes

Not surprisingly, the data show that by far the clearest skies (lowest
extinction values) are found in dry regions at high altitudes where pre-
sent-day astronomical observatories are located (McDonald Observatory
on Mount Locke, Texas, USA18 and the European Southern Observatory
on La Silla, Chile).19 According to the data in Table III. 9.2 much
poorer atmospheric conditions are found in humid climates at sea level
(Leiden Observatory, the Netherlands,20 no longer in use) and at low
altitudes ( Jena Observatory in Grossschwabhausen, Germany).21 It is
instructive to realize that an increase in extinction of 0.1 magnitudes
corresponds to a small decrease in the intensity of starlight of 10% at
zenith but to a large decrease by a factor 4 at 3° above the horizon.
One important uncertainty that affects estimates of the atmospheric
extinction is the aerosol content of the atmosphere, which over the last
century and a half has been noticeably increasing due to industrial pol-
lution. Therefore I also list in Table III. 9.2 values of the atmospheric
visual extinction in Mesopotamia during the 6th and 7th century BC
derived from an analysis of ancient observations of Saturn from Babylon
and Uruk22 and during the 13th century BC from an analysis of obser-

18
R. J. Angione & G. de Vaucouleurs, “Twenty years of atmospheric extinction at
McDonald observatory”, Publications of the Astronomical Society of the Pacific 98 (1986),
1201–1207.
19
F. Rufener, “The evolution of atmospheric extinction at La Silla”, Astronomy and
Astrophysics 165 (1986), 275–286.
20
K. K. Kwee & A. M. van Genderen, “Photo-electric Observations of 31 and 32
Cygni during November and December 1961”, Bulletin of the Astronomical Institutes of the
Netherlands 17 (1963), 53–55.
21
H.-G. Reimann, V. Ossenkopf & S. Beyersdorfer, “Atmospheric extinction and
meteorological conditions: a long time photometric study”, Astronomy and Astrophysics 265
(1992), 360–369.
22
T. de Jong, “Early Babylonian Observations of Saturn: Astronomical Considerations”,
in: Under One Sky, J. M. Steele & A. Imhausen, eds., (Münster: AOAT 297, 2002),
175–192.
the heliacal rising of sirius 437

Table III. 9.2. Atmospheric visual extinction at different locations and epochs

Location Epoch Season Altitude <k(V)>

Babylon, Mesopotamia ~1300 BC yearly average sea level 0,27


Babylon, Mesopotamia 647–634 BC jul–nov sea level 0,25
Uruk, Mesopotamia 577–575 BC oct–dec sea level 0,34
Leiden, the Netherlands 1961 nov–dec sea level 0,43
Grossschwabhausen, Germany 1968–1991 yearly average 350 m 0,36
Mount Locke, Texas, USA 1960–1980 yearly average 2000 m 0,17
La Silla, Chile 1975–1985 yearly average 2400 m 0,12

vations of about 20 bright stars from Babylon.23 These values are close
to the present-day visual extinction for clear skies at sea level of about
0.28 magnitudes per air mass (0.02 due to Ozone absorption, 0.12 due
to molecular scattering and 0.14 due to aerosols and dust).24 Contrary
to Schaefer25 who adopted 0.35 magnitudes per air mass for the Memphis
area I suggest that 0.27 is a more appropriate choice since the climatic
conditions in the Nile valley near Memphis are probably quite similar
to those in the Euphrates valley near Babylon. This value is consistent
with the 1926 observations summarized by Pachner26 because arcus visio-
nis values of 8.7°–9.2° correspond to a visual extinction of about 0.25
magnitudes per air mass (see Table III. 9.1). Ptolemy’s arcus visionis of
11° (corresponding to a visual extinction of about 0.35 magnitudes per
air mass) is appropriate for the much more humid conditions in
Alexandria at the shore of the Mediterranean.
For chronological purposes the actual date of heliacal rising of Sirius
is the crucial quantity. In Table III. 9.3 I list dates of heliacal rising
of Sirius computed for geographical latitudes 25° North (Edfu) and
30° North (Memphis) for different values of the atmospheric visual
extinction at different epochs in antiquity. These dates are averages
over four consecutive years. As argued above atmospheric conditions
in arid upper Egypt (Elephantine and Memphis) are probably best char-
acterized by a visual extinction k(V) = 0.27 magnitudes per air mass,
while for the more humid climate of lower Egypt (Alexandria) k(V) =
0.35 magnitudes per air mass may be more appropriate.

23
See n. 3.
24
See M. S. Bessell, “UBVRI passbands”, Publications of the Astronomical Society of the
Pacific 102 (1990), 1181–1199.
25
See n. 11.
26
See n. 10.
438 teije de jong

Table III. 9.3. Julian dates of the heliacal rising of Sirius and of summer solstice

k(V) 0.20 0.27 0.35 0.20 0.27 0.35 Summer


Epoch Geographical Latitude = 25° Geographical Latitude = 30° Solstice

3500 BC 8-jul 10-jul 11-jul 14-jul 16-jul 17-jul 22-jul


3000 BC 8-jul 10-jul 11-jul 14-jul 16-jul 17-jul 18-jul
2500 BC 9-jul 11-jul 12-jul 14-jul 16-jul 18-jul 14-jul
2000 BC 9-jul 11-jul 12-jul 15-jul 17-jul 18-jul 10-jul
1500 BC 10-jul 12-jul 13-jul 15-jul 17-jul 18-jul 7-jul
1000 BC 11-jul 12-jul 14-jul 16-jul 17-jul 19-jul 3-jul
500 BC 11-jul 13-jul 14-jul 16-jul 18-jul 19-jul 29-jul
1 AD 12-jul 14-jul 15-jul 17-jul 18-jul 20-jul 25-jun
500 AD 13-jul 14-jul 15-jul 17-jul 19-jul 20-jul 21-jun

The atmospheric extinction varies due to changing weather conditions.


On some dates this will make Sirius unobservable and it may cause
variations in the date of heliacal rising of Sirius of up to about ± 3
days, corresponding to extreme values of k(V) ranging from 0.15 to
0.40 magnitudes per air mass.27
The data in Table III. 9.3 also show that in the course of 4000
years the date of the heliacal rising of Sirius moves forward with respect
to the summer solstice by one day in about 120 years. This is due to
precession but tempered by the fact that Sirius lies far (~40°) south of
the ecliptic. For a star in the ecliptic this forward motion would be
one day in about 75 years, as expected for a rate of precession of 1°
in 72 years. The rather large proper motion of Sirius, which causes a
displacement in the sky of about 1.5° in 4000 years, only causes a
minor shift in the date of heliacal rising of about one day over that
period. Notice that the inaccuracy of the Julian calendar causes the
summer solstice to recede by about one month in 4000 years rather
than two months as expected for an accurate solar (Gregorian) calendar.

Acknowledgements

I am grateful to Rolf Krauss and Peter Huber for several useful com-
ments and suggestions for improvement.

27
See n. 22.
III. 10 EGYPTIAN SIRIUS/SOTHIC DATES,
AND THE QUESTION OF THE SOTHIS-BASED
LUNAR CALENDAR

Rolf Krauss

In addition to the observational conditions discussed by T. de Jong in


the preceding chapter, there are three factors which must be consid-
ered when dealing with Egyptian Sirius (Sothic) dates: 1) the quadren-
nium; 2) geographical reference; 3) regular observation versus schematic
determination.

Quadrennium

Under the same meteorological conditions and at the same site, the
heliacal risings of Sirius occur as a rule thrice at intervals of 365 days,
followed by the fourth rising after 366 days. On each of the these con-
secutive risings the ecliptic longitude of the sun and thus also the dis-
tance between sun and Sirius would have decreased, so that in the
fourth year the distance would have been too small for sighting the
star after 365 days. Only on the 366th day after the last sighting would
the star have been observable.
Because the Egyptian calendar year was only 365 days long, it fell
one day behind the heliacal rising of Sothis every four years when the
rising occurred after 366 days. Accordingly, a specific date in the
Egyptian year correlates to the heliacal risings of 4 consecutive years.
Because Egyptian sources do not specify in which year of a quadren-
nium a given recorded Sothic date fell, there are usually four years in
the Julian calendar that might correlate with it. A specific year can be
determined only when additional data are available, such as the lunar
date associated with the Sothic date, in the Illahun archive.
440 rolf krauss

Geographic Reference and Its Interplay with Astronomically or Schematically


Determined Sothic Dates

Regardless of where the Egyptians observed the rising of Sothis, the


date could not apply elsewhere, because Egypt’s ancient boundaries
extended from 24° (First Cataract, Elephantine) to 31.15° (Diospolis
inferior) north latitude. In the 28th century BC, when the Sothic-linked
365 day calendar will have been introduced, the date when the star
rose differed by more than one day for each degree of latitude, so that
the star rose 8 to 10 days later on the Mediterranean coast than
at Egypt’s southern border. Sighting was effected not only by latitude,
but also by the meteorological differences between Upper and Lower
Egypt: UE is dry and arid, LE is more humid which results in a higher
extinction.
The impracticability of using several different calendars in each zone
evidently led the inventors of the Egyptian calendar to choose one par-
ticular site as the geographical reference for the calendrical heliacal ris-
ings of Sothis. A single source provides information on the geographical
reference: in the 6th century AD Olympiodoros, in his commentary on
Aristotle’s Meteora, remarked that the rising of Sothis was celebrated in
Alexandria when the star rose for the Memphites.1 Until about 1950,
Egyptologists assumed that Memphis was the reference point for all
Sothic dates throughout Egyptian history.2 Presuming schematically
determined rising dates, Eduard Meyer, for example, calculated III
Shemu 9 of 9 Amenhotep I, the Sothic date in Papyrus Ebers, as cor-
responding to the quadrennium 1550/49 to 1547/46 BC; by adding
8 years of Amenhotep I and 26 years of Ahmose he arrived at ca.
1580 BC for the beginning of the NK.
Later Egyptologists abandoned the concept of schematic Sothis ris-
ings, but retained Memphis as the reference point for Sothic dates of
all periods. Using an arcus visionis of 9° the rising of Sothis occurred at
Memphis around 1500 BC on July 17/16; the Ebers Sothic date cor-
responds to these days in 1538–1535 BC.3 It follows that the begin-
ning of the NK would have been ca. 1570 BC.

1
Sethe, “Zeitrechnung”, 309.
2
Hornung, Untersuchungen, 20.
3
III Shemu 9 corresponds to July 17 from 1541 to 1538 BC and to July 16 from
1537 to 1534 BC.
egyptian sirius/sothic dates 441

Around 1950 new sources relating to the Assyrian-Babylonian kinglists


came to light which shortened Near Eastern and Egyptian chronology
by about 20 years. Shifting the geographical reference point for the
Ebers Sothic date from Memphis (f = 29.9°) to Thebes (f = 25.7°),
compensated for the 20-year chronological difference, provided that the
rising of Sothis mentioned in Papyrus Ebers had actually been observed.
In the 1970s it became clear that the relative and the absolute
chronology of the NK had to be shortened further by at least 10 years,
yielding ca. 1540 BC for the beginning of the NK.4 The only possi-
bility to reconcile this chronology with the Ebers Sothic date was to
propose a geographical reference point further south, namely on the
island of Elephantine (f = 24°). Within Egypt’s traditional borders, the
rising of Sothis could be seen earliest at Elephantine, the southernmost
point. Here lay the mythical source of the Nile. Late Period texts asso-
ciate the rising of Sothis with the beginning of the inundation at
Elephantine. Measurement of the height of the inundation at Elephantine
(as at Memphis) was particularly important for the entire country.5
Because of precession, the heliacal risings of all fixed stars shift within
the tropical year. For Sirius, the large proper motion modifies the effect
of precession. Around 2750 BC Sirius will have risen at latitude 24°
on June 16th Greg., but today it rises there on July 29th Greg. In
pharaonic times the Sothic year was always longer than the Julian year
by about one minute. When, over centuries, the difference added up
to a quarter of a day, the rising of Sothis fell for only three, rather
than for four, years on the same Egyptian calendar day and the shift
to the next day occurred in the third year, not the fourth. If the annual
observation of the rising of Sothis was a feature of the Egyptian cal-
endar, then the occasional shift of the rising should have been accounted
for in triennia. By contrast, it is possible that the calendar dates of
the rising depended upon a series of observations made during the
period before the introduction of the calendar with Sothic dates there-
after schematically shifted one day every four years. Only an evalua-
tion of the Egyptian historical sources—not astronomical or calendrical
theorizing—can determine which alternative is correct.
Hieroglyphic sources preserve two relevant statements. (1) The ris-
ing of Sothis in 7 Senwosret III was not actually observed, instead it

4
Bierbrier, Kingdom, 109ff.
5
Krauss, Sothis, 63–67.
442 rolf krauss

was announced 22 days in advance and might have been ‘known’ ear-
lier. (2) The Canopic Decree of Ptolemy III Euergetes of March 7 in
238 BC asserts that the rising of Sothis shifts forward one day every
four years.6 The purpose of the decree was to introduce a sixth epagom-
enal day every four years,7 so that the rising of Sothis continued to fall
on Payni 1 (II Shemu 1), as it had in the year before the decree (239
BC).8 No mention is made of a possible shift after three years.
According to Clagett there is a third hieroglyphic source. He under-
stands an inscription of Ptolemy IV to contain an assertion about the
regular shift of the rising of Sothis in intervals of four years.9
Other statements with a bearing on the question are to be found in
Latin and Greek sources. In “De die natali” (18, 10), the orator Censo-
rinus commented on Great Years:
The Great Year of the Egyptians has nothing to do with the moon. In
Greek we call it Dog’s Year, in Latin canicularis, because it begins on
the 1st day of the month Thoth, when the Dog star rises. For the civil
calendar of the Egyptians, the year numbers 365 days without any inter-
calation. So it is that a period of 4 years is about one day shorter than
the natural quadrennium. And it follows that New Year’s Day coincides
in both systems only after 1461 [ Julian] years. Some call this Great Year
a Helios year; others, God’s Year.
Censorinus also noted (c. 21, 10–11) that in the year when he was
writing, namely AD 238, the “Egyptian New Year’s day fell on June
25th. But a hundred years earlier, during the second consulate of
Emperor Antoninus Pius and Bruttius Praeses, the Egyptian New Year’s
Day fell on the *20th of July, the day when the Dog star rose.”10 Thus
on the 20th of July in AD 139 a new period of 1460 Egyptian years
had started.11 Borchardt claimed that coins minted at Alexandria in
year 2 of Antoninus Pius showing a haloed phoenix and the legend
AIVN commemorated the Sothic period.12 Even though AIVN may be
a Hellenistic deity that was identified in Alexandria with Osiris embod-

6
Clagett, Science II, 326–331.
7
Presumably at the end of a quadrennium, the first time in 235 BC.
8
For details Krauss, Sothis, 54–57, and Spalinger, Studies, 39–50.
9
Clagett, Science II, 331–333.
10
C. Rapisarda, De die natali liber ad Q. Caerellium (Bologna: ESUFC 47, 1991),
44–45.—Censorinus writes “21st of July”; cf. Meyer, Chronologie, 24, and Rapisarda,
ibidem, 258–259cf.
11
Borchardt, Annalen, 55–56; idem, Mittel, 11; cf. Hornung, Untersuchungen, 17.
12
Borchardt, Annalen, 55–56.
egyptian sirius/sothic dates 443

ied in his turn as Phoenix,13 Borchardt might have been right never-
theless, since Tacitus, Annals VI. 28, refers to one opinion that the
Phoenix lived for 1461 [Egyptian years = 1460 Julian years] and was
thus associated with the Sothic period.14
About AD 400 Chalcidius, in his commentary on Plato’s Timaeus,
called Sothis the Egyptian equivalent of Sirius and stated, like Censorinus,
that the length of a Sothic cycle was 1460 years.15
According to Theon of Alexandria (ca. 335–405 AD),16 the day of
Sothis’s rising would have shifted by 426 days from Thoth 1 in year
1 of the era apo Menofrevw until year 100 of the era of Diocletian
(384 AD), had the fixed Alexandrian year not been introduced.17 Because
of this calendar reform, there were only 324 shifts before the 100th
year of the Diocletian era.18 Thus the era “apo Menophreos”19 would
have begun, in our terms, in 1322 BC and would have ended in AD
138/139, as implied by Censorinus, i.e. for Theon the Sothic cycle
lasted full 1460 years.
Clearly Graeco-Roman sources reckoned with a regular shift of the
rising of Sothis in quadrennia; the later 19th century astronomer
Oppolzer asserted that during the Graeco-Roman period it was unknown
that the Sothic year varied in length.20 Accordingly, Egyptologists tra-
ditionally assumed that the calendric rising of Sothis was shifted regu-
larly and schematically every four years, implying exactly 365.25 days
as the length of a Sothic year. Oppolzer computed the length of the
Sothic year as being longer than the Julian year during the pharaonic
period. For him the question of whether the historical Sothis-bound
calendar relied on annual observation, resulting in triennia, or on a

13
G. W. Bowersock, Hellenism in Late Antiquity (Ann Arbor, 1990), 24–26.
14
Krauss, Sothis, 200.
15
Meyer, Chronologie, 30 Anm. 1.
16
Or perhaps a commentary on his works, cf. L. Depuydt, JARCE 32 (1995), 46–47.
17
C. R. Lepsius, Königsbuch der alten Aegypter (Berlin, 1858), 123–124; Meyer, Chronologie,
29 n. 1.
18
Other details of Theon’s computation remain unclear, see Meyer, Chronologie,
28–29.
19
Menophris is neither a garbled royal name nor a rendering of “Memphis” (Egyptian:
Menfe), but rather the name of a solar god attested in the Hellenistic period, cf. Krauss,
Ende der Amarnazeit (1978), 268–273. Like Censorinus, Theon implied the solar nature
of the Sothic cycle when he spoke of it as the “era of (the solar god) Menophris”.
20
Th.v. Oppolzer, “Über die Länge des Siriusjahres und der Sothisperiode”, Sitzungs-
berichte der mathematisch-naturwissenschaftlichen Classe der Kaiserlichen Akademie
der Wissenschaften, XC. Band, II. Abtheilung, Jahrgang 1884 (Vienna, 1885), 576;
Ginzel, Handbuch I, 191.
444 rolf krauss

cyclic scheme of regular quadrennia remained undecided, although he


showed a bias for yearly observation.21
In 1904 Förster computed the length of the Sothic year at Meyer’s
request. Förster arrived at a smaller difference between the Sothic and
the Julian year than Oppolzer.22 He calculated the rising of Sothis at
Memphis on July 19 ( Jul.) throughout pharaonic history. Later Schoch23
and Neugebauer24 corrected Förster’s figures; they found a difference
between the Sothic and Julian years that caused shifts of the rising day
in triennia. The most recent calculation of the length of the Sothic
year confirmed their result.25 Neugebauer concluded that the recorded
rising dates of Sothis cannot have been schematically determined, but
must have depended on observation. The assumption that the rising
dates were determined schematically results in the following paradox:
if the rising of Sothis be reckoned backwards schematically from the
Censorinus and Canopus dates, then the starting date for the 1460-year
cycles is astronomically incorrect by several days. If, however, an astro-
nomically correct date is calculated for the beginning of the historical
cycles, then the Canopus and Censorinus dates do not match.
The paradox can be resolved by supposing that the Sothis-bound
365 day calendar was schematical from its introduction, in the 28th
century BC, with reference to a site in UE. The shift of the reference
point to Memphis that Egyptologists recognized around 1950 may have
taken place as late as Dyn. 30, i.e. in the 4th century BC. If a schematic
calendar had been used since its introduction down to the 4th century
BC, then the risings of Sothis that were observable at the geographic
reference point occurred 4 days later than those determined schemat-
ically. Presuming the discrepancy were to be overcome in the 4th cen-
tury BC, two workable possibilities offered themselves: a) either the
schematic date of the rising was corrected by the addition of 4 days
without altering the geographic reference, b) or the geographic reference
point was moved from UE to a northern site.
Evidently Memphis was chosen as the new reference point. The cal-
endar day of the rising of Sothis at Memphis was determined by obser-

21
Oppolzer (n. 20), 584.
22
Meyer, Chronologie, 14–15.
23
C. Schoch, Die Länge der Sothisperioden beträgt 1456 Jahre (Steglitz, 1928; Reprint:
Kiel: Astronomische Nachrichten 8, 2; 1930).
24
P. V. Neugebauer, Astronomische Nachrichten No. 6261 (Kiel, 1937), 377–378.
25
M. F. Ingham, JEA 55 (1969), 36–40.
egyptian sirius/sothic dates 445

vation and served as the initial day of a new schematically determined


series, i.e. the calendar remained schematical. With the Gregorian
reform, days had to be eliminated to achieve its goal, whereas there
would have been no need to do so when reforming the Sothic calen-
dar. In 400/399 BC the shift would have caused the rising day to leap
10 days from IV Peret 10 to 20. In other words, the cycle would not
have started in 1322 BC, i.e. 1460 years before 138/139 AD, but 10 ×
4 years = 40 years later. A calendar that was based on regular obser-
vation of Sothis would not have required a shift of its geographical
point of reference; thus the shift that Egyptologists recognized around
1950 implies a schematic calendar.

Comments on Peret Sepdet Dates

I Peret 4–19 Festival calendar of Amenhotep I26


[? Ra'neferef archive]
II Peret 21 coffin of Ashayt (T3C)
III Peret 30 coffin S3P
IV Peret 21 coffin S11C
IV Peret Festival calendar of Amenhotep I26
IV Peret 16 Abydos, Osireion27
IV Peret 16 or *18 year 7 [Senwosret III]
IV Peret year 1 [Amenemhet III]
IV Peret [Amenemhet III]
<II Shemu 20 year **11, Luxor—Farshut road>28
III Shemu 9 year 9 of Amenhotep I
III Shemu 28 Festival calendar of Elephantine, Thutmose III29
(continued on next page)

26
The festival calendars of Amenhotep I cite festival calendars of much earlier peri-
ods without indicating the names of the contemporaneous kings, cf. Spalinger, Studies,
15–17; idem, “Sothis and ‘Official’ Calendar Texts”‚ VA 10 (1995), 175–181, esp. 179.
27
Clagett, Science II, 357–392, at 357; 380.
28
J. C. Darnell, Theban Desert Road Survey in the Egyptian Western Desert I. (Chicago:
OIP 119, 2002).—At the SCIEM Workshop “Egypt & Time” (Vienna June 30 to July
2, 2005), K. Ryholt argued (a) that the year ought to be read 31, and (b) that the
graffito does not mention Sothis at all.
29
The inscription (Urk. IV 827,8) belongs to a list of festivals which is dated epi-
graphically to the end of reign of Thutmose III, cf. Krauss, Ä&L 3 (1994), 90–92.—
The translation of prt spdt m #'.tw as “true rising of Sothis”, in: E. Blumenthal, I. Müller
& W. F. Reineke, eds., Urkunden der 18. Dynastie. Übersetzung zu den Heften 5–16
(Berlin, 1984), 271 is incorrect; read, rather, “Rising of Sothis, one offers”.
446 rolf krauss

Table (cont.)
– Festival calendar of Buto, Thutmose III30
I Akhet Festival calendar of Medinet Habu, Ramesses III31
[Thoth 1, 1322 BC era apo Menofrevw], see above
II Shemu 1, 239 BC Canopus decree, see above
Thoth 1, 139 BC Censorinus date, see above

Peret Sepdet in the Ra'neferef Archive?

The archive functioned at least until the last years of Izezi. It men-
tions the moveable w #g j feast that is also attested in the Illahun archive32
where it was celebrated on LD 17, apparently in the second lunar
month after the rising of Sothis.33 However, the moveable w #g j feast
may have been a seasonal event, perhaps related to the availability of
new wine.34 If so, it might have been celebrated on a LD 17 when
new wine was available, whether the month happened to be the third
or second after the rising of Sothis. Regardless, Luft suggests that the
w #g j date III [Akhet] 28 in the Ra'neferef archive corresponded to LD
18 (sic) in the first lunar month after the rising of Sothis,35 resulting in
a date around ca. 2450 BC which is far too early for the archive,
according to any recent Egyptological chronology.36 If the w #g j date III
[Akhet] 28 referred to the second lunar month after the rising of Sothis,
corresponding to the time when it was celebrated at Illahun, then a
date around 2600 BC would result. If the date is emended to III [Peret]
28,37 then it would have occurred in the third lunar month after the
rising of Sothis, provided Izezi’s reign lay in the second half of the

30
Bedier, Shafia, in: Aspekte spätägyptischer Kultur. Fs Erich Winter (Mainz:
Aegyptiaca Treverensia 7, 1994), 35–50.—The rising of Sothis is noted between festi-
vals of the first month of Shemu, but neither month nor day is specified.
31
MHC 629: “I Akhet, when Sothis goes forth on her day”, and thus on any day
in I Akhet from 1 to 30, cf. Parker, Calendars, 40.
32
P. Posener-Kriéger, “Remarques préliminaires sur les nouveaux papyrus d’Abousir”,
in: Ägypten: Dauer und Wandel (Mainz: DAIS 18, 1985), 35–43.
33
Krauss, Sothis, 86–94; for LD 17 (not LD 18) see Krauss, in: Bietak, SCIEM
Haindorf 2001, 192, with additional literature.
34
P. Posener-Kriéger, “Wag-fest”, in: LÄ VI, 1135–1139; R. Krauss, GM 162 (1998),
57–58.
35
U. Luft, “The Date of the w #g y Feast: Comments on the Chronology of the Old
Kingdom”, in: Spalinger, ed., Revolutions, 39–44.
36
Krauss (n. 34), 53–57; cf. above Verner, Chapter II. 4.
37
L. Depuydt, JARCE 37 (2000), 167–186, at 173.
egyptian sirius/sothic dates 447

24th century BC. Thus it seems impossible to deduce a chronological


result without circular reasoning.

Heliacal Rising Dates of Sothis in the Star Clocks

To date, there are 19 complete and/or fragmentary known coffin lids


decorated with so-called diagonal star clocks or diagrams of certain ris-
ing stars, conventionally called decans.38 The diagrams indicate risings
at 12 intervals (“hours”) during the night, and at 10-day intervals
(“decades”) during the Egyptian year.39 A star that rises in the 12th
hour of the night, i.e. at dawn, on the first day of a decade, rises heli-
acally. According to Neugebauer and Parker the diagonal star clocks
are roughly datable by the position of Sirius/Sothis as 12th hour star.40
They reckoned a margin of ca. 50 years, considering that the extant
star clocks are copies from older ones; their grouping of the extant star
clocks has been superseded by the analysis of J. Kahl.41
The star clocks preserve dates for the rising of Sothis between II
Peret 21 and IV Peret 21, corresponding to an interval of 240 years
from roughly 2060 to 1820 BC. The only coffin with a star clock that
is datable to a specific reign is T3C, the coffin of Ashayt, from the
reign of Nebhepetre' Mentuhotep II.42 T3C is at least once removed
from the Vorlage.43 In T3C, and in its Vorlage, the heliacal rising of
Sothis is dated to II Peret 21. In relation to the Illahun Sothic date
and supposing the same geographic point of reference, II Peret 21 cor-
responds to the early years of Inyotef II, whereas Ashayt lived and died
more than 50 to 75 years later.

38
K. Locher, “Middle Kingdom astronomical coffin lids: extension of the corpus
from 12 to 17 specimens”, in: Eyre, Proceedings, 697–701. For lids 18–19 see S. Symons,
JHA 33 (2002), 257–260.
39
For the decades see Gardiner, Grammar, Excursus C.
40
Neugebauer & Parker, EAT I 29–32; 106.
41
J. Kahl, “Textkritische Bemerkungen zu den Diagonalsternuhren des Mittleren
Reiches” SAK 20 (1993), 95–107.
42
In 2060 BC: II Peret 21 = July 9 ( Jul.).
43
Locher (n. 38), 608, Fig. 1.
448 rolf krauss

Sothic Dates of Senwosret III and Amenemhet III

Bot instances of Peret Sepdet in IV Peret in the Illahun archive—but


without specification of the day—are datable to Amenemhet III.44 The
rising of Sothis on the specific day IV Peret 16 in 7 Senwosret III was
announced in a letter dated to III Peret 25, i.e. 22 days before the
event.45 A fragment of the temple diary, pBerlin 10012 A, contains a
copy of the letter. The rising date IV Peret 16 is ambiguous. Another
fragment of the temple diary states that the offerings for the Sothic fes-
tival in year 7 [of Senwosret III] were received, or entered in the tem-
ple diary, on IV Peret 17.46 As a rule, offerings were delivered to Illahun
one day or even two days before a festival.47 On the basis of the deliv-
ery date alone, one would conclude that the rising of Sothis took place
on IV Peret 18. Luft has pointed out grammatical and syntactical errors
which were made, presumably, by the scribe when he copied the let-
ter with the announcement.48 It is possible that the scribe also made a
mistake when he copied the date, writing IV Peret 16 instead of IV
Peret *18.49
Supposing that the Illahun Sothic date refers geographically to Illahun,
Brix computed the years 1876–1872 BC as corresponding to IV Peret
16 for the rising day in year 7 of Senwosret III.50 Later Edgerton com-
puted 1870 BC ± 6 years, by allowing a range of the arcus visionis
from 9.5° to 8.4° and the possibility that the observations might have
been made somewhere between Illahun and Heliopolis.51
Parker tried to accommodate the Illahun lunar dates to Edgerton’s
interval for the Sothic date and arrived at year 7 Senwosret III = 1872
BC.52 His sequencing of the Illahun lunar dates was wrong, and con-
sequently 7 Senwosret III = 1872 BC was also wrong;53 for the same

44
Luft, Fixierung, 44–47; 123.
45
Luft, Fixierung, 54–58.
46
Luft, Fixierung, 57–58.
47
Krauss, Sothis, 79. For an apparent exception see Krauss (n. 33; SCIEM ), 186
n. 76.
48
Luft, Fixierung, 55–57.
49
If the Sothis-bound lunar year of pBerlin 10056 is projected to later periods, only
peret Sepdet on IV Peret *18 yields the calendric situation in year 9 of Amenhotep I as
represented in the Ebers calendar; cf. below, end of section.
50
W. Brix, “Über das im zweiten Papyrusfund von Kahun enthaltene Sothisdatum
des mittleren Reiches der ägyptischen Geschichte,” ZÄS 41 (1904), 26–33; cf. also ZÄS
37 (1899), 101.
51
W. F. Edgerton, “Chronology of the Twelfth Dynasty”, JNES 1 (1942), 307–314.
52
Parker, Calendars, 63–69.
53
Krauss, Sothis, 73–101; cf. Clagett, Science II, 323.
egyptian sirius/sothic dates 449

reason Barta’s computation of 7 Senwosret III = 1875 BC was wrong.54


On the basis of the correct sequencing of the Illahun lunar dates, the
next possible, but not necessarily astronomically correct equation for 7
Senwosret III before 1872 BC would be 1880 BC, or after 1855 BC.
Luft postulated that the Illahun Sothic date refers to Memphis where
Sothis rose in the 19th century BC on July 17 ( Jul.).55 As Table III.
10.1 shows, IV Peret 16 would have corresponded to July 16/17 ( Jul.)
during the quadrennium 1869 to 1866 BC, if the Egyptian calendar
day began at sunrise as Luft erroneously supposes. Given that the cal-
endar day began at dawn, IV Peret 16 actually corresponded to July
16/17 during the quadrennium 1873 to 1870 BC.
Table III. 10.1 also shows that III Peret 9, a day that is recorded
in the Illahun archive as a lunar day 1 in year 9 of Senwosret III,
corresponded to a lunar day 1 in 1864 BC, provided the calendar day
began at dawn. Luft concluded that 7 Senwosret III, the year of the
Sothic date, corresponded to 1866 BC, combining the lunar event in
1864 and the stellar event in 1866 BC. It is clear, however, that the
rising of Sothis and the rising of old crescent were observed (or the
absence of old crescent confirmed) during the same hour before sun-
rise. It contradicts simple logic and common sense to date the stellar
and the lunar events to different Julian calendar days. Thus Luft’s astro-
nomical dating of the MK is patently wrong.56

Table III. 10.1


year BC III Peret 9 IV Peret 16 IV Peret 16
(dawn) (sunrise) (dawn)

1873 LD 24 July 17/18 July 16/17


1872 6
1871 16
1870 27
1869 8 July 16/17 July 15/16
1868 18
1867 28
1866 9
1865 20 July 15/16 July 14/15
1864 1

54
W. Barta, “Die ägyptischen Sothisdaten und ihre Bezugsorte”, JEOL 26 (1979–80),
26–34.
55
Luft, Fixierung, 224–229.
56
Furthermore, the proportion of 6 correct and 15 delayed lunar dates that Luft’s
date implies, cannot be the result of observation; cf. above Table III. 8.3.
450 rolf krauss

When both possible dates (IV Peret 16 and *18) for the rising of
Sothis in 7 Senwosret III are taken into consideration, together with
the uncertainties of pharaonic Sothic dates, then the Illahun Sothic
date corresponds to any year between ca. 1882 and 1830 BC.57 As
shown in Chapter III. 8 it follows from the astronomically correct com-
putation of the Illahun lunar dates that 1 Senwosret III = 1837/36
BC and thus year 7 of Senwosret III = 1831/30 BC.

Sothic Date of the Ebers Calendar

The so-called Ebers calendar (see Fig. III. 10.1) is written on the recto
of the medical papyrus Ebers. According to Möller the hand that wrote
the calendar is palaeographically a few years younger than the hand
of the medical text.58
For ease of orientation, the columns and lines have been numbered.
According to line 1 the calendar refers to regnal year 9 of King
Amenhotep I. If the regnal year and the period covered by the cal-
endar were exactly concurrent, then the accession day of Amenhotep
I coincided also with III Shemu 9. Whether Col. I lists months or
feasts is contested (see below). Col. II contains civil dates, each 30 days
apart, disregarding the epagomenal days after a well-attested fashion.59
In Col. III the entry peret Sepdet (rising of Sothis) follows the calendar
day III Shemu 9 and refers therefore to that day and only indirectly
to wep-renpet of col. I, i.e. the text asserts directly that peret Sepdet occurred
on III Shemu 9. The Ebers calendar has been studied and commented
on by many authors, those since 1980 are Helck,60 Luft,61 Barta,62
Krauss,63 Leitz,64 Beckerath,65 Depuydt,66 Spalinger67 and Belmonte.68

57
The upper limit is ca. 1882 BC, if the rising date is IV Peret 16 and Diospolis
inferior is the observation site, where an arcus visionis * 9° is to be expected, because
of the high humidity near the Mediterranian coast.—The quadrennium 1833/30 BC
is the lower limit, if Elephantine is the geographical point of reference and if the ris-
ing date IV Peret *18 was determined schematically with the arcus visionis * 8.6°.
58
G. Möller, Hieratische Paläographie I (Leipzig, 1909), 20.
59
Cf. above, Chapter I. 5.
60
W. Helck, GM 67 (1983), 47–49.
61
U. Luft, GM 62 (1986), 69–71.
62
W. Barta, SAK 8 (1980), 43–47.
63
R. Krauss, Ä&L 3 (1992), 75–85.
64
Leitz, Studien, 24–34.
65
J.v. Beckerath, SAK 14 (1987), 28–31.
66
L. Depuydt, “The Function of the Ebers Calendar”, Orientalia 65 (1996), 61–88.
67
A. J. Spalinger, “Notes on the ancient Egyptian Calendars”, Orientalia 64 (1995), 17–32.
68
Belmonte, “Questions”, 50–55.
egyptian sirius/sothic dates 451

line 0

line 1

line 2

line 3

line 4
line 5

line 6
line 7
line 8
line 9

line 10

line 11
line 12

col.III col.II col.I


Fig. III. 10.1

If around 1500 BC the rising of Sothis was observable at the southern


border of Egypt on July 10 (or 11) and at Diospolis inferior on July
19 ( Jul.), then the rising on III Shemu 9 in 9 Amenhotep I referred
to any of the years between ca. 1549 BC ( July 19/18; Diospolis infe-
rior) and ca. 1517 ( July 10/11; Elephantine). If the rising dates were
schematically determined, then III Shemu 9 in 9 Amenhotep I referred
to any of the years from ca. 1541 BC ( July 17; Diospolis inferior) to
1509/1506 BC ( July 9; Elephantine). Thus the Ebers Sothic date can-
not be earlier than ca. 1549 BC and not later than 1506 BC. Recently
Kitchen argued on the basis of historical reasoning that Amenhotep I
reigned from 1515 to 1494 BC,69 implying that year 9 corresponded

69
K. A. Kitchen, Acta Archeologica 67 (1996), 12.
452 rolf krauss

to 1507/06 BC. If so, the Ebers Sothic date must be understood as


schematically determined with geographical reference to Egypt’s south-
ern border.
Provided that there was no calendar reform between the Illahun and
Ebers Sothic dates,70 the relative distance between allows chronologi-
cal conclusions. 333±3 years (Illahun Sothic date = IV peret 16) at
most and 325±3 years (Illahun Sothic date = IV peret *18) at least
elapsed between the two Sothic dates, dependent on the position of
each within its quadrennium.71 If these maximum and minimum figures
are added to Kitchen’s date for 9 Amenhotep I, then the Illahun Sothic
date fell at the earliest in 1840/39 BC ± 3 and at the latest in 1831/30
BC ± 3. It is methodologically preferable to presume the same calen-
drical and geographical conditions for the Illahun Sothic date as for
the Ebers Sothic date. If so, the Illahun Sothic date corresponds to
1830 BC and was schematically determined with the southern border
as geographical reference. Alternatively, the Illahun Sothic date could
correspond to 1840/39 BC. If so, it would have been determined by
actual observation between the latitudes of Coptos and Asyut.

The Problem of the Sothis-based Lunar Calendar

From data in the Illahun archive and in the Ebers calendar Borchardt
deduced the existence of a Sothis-based lunar calendar.72 Parker con-
curred in principle although he corrected some of Borchardt’s ideas.73
Spalinger, who recently criticized lunar calendrics in general, never-
theless accepts the Sothis-based lunar calendar74 although he points to
the fact that Egyptological calendrics would be simpler without it.75 In

70
I.e. the geographical point of reference and the determination of the rising dates
remained the same.
71
These figures may be subtracted from 7 Senwosret III = 1830 BC, as determined
by the Illahun lunar dates (see Chapter III. 8). The subtraction yields 1508 to 1494
BC for the Ebers Sothic date. The interval overlaps from 1508 to 1506 BC with 1549
to 1509/06 BC, the period of the Ebers Sothic date (see above). It follows that the
astronomically correct interval for the Ebers Sothic date lies between 1508 and 1506
BC, harmonizing well with Kitchen’s 1507/06 BC for year 9 of Amenhotep I.
72
Borchardt, Mittel, 5ff.
73
Parker, Calendars, 30ff.
74
A. J. Spalinger, “Thot and the Calendars”, 47–48, in: Spalinger, ed., Revolutions.
75
A. J. Spalinger, “Months representations”, CdE 70 (1995), 110–122.
egyptian sirius/sothic dates 453

general a Sothis-based lunar year represents a luni-stellar year, which


is documented in other cultures.76
The Illahun papyrus Berlin 10056 contains a complete account and
two fragmentary ones, each spanning a lunar year.77 The complete
account is headed: “Account of earth almonds and honey over the
course(?) of a year. List over the course (?) of six months of the tem-
ple scribe Hornakhte. Regnal year 31. Offerings from this list.”78 After
the heading, 6 intervals follow:
II Shemu 26 nfrjjt-r [to] III Shemu 25
IV Shemu 25 nfrjjt-r regnal year 31, I Akhet 19
regnal year 31, II Akhet 20 nfrjjt-r III Akhet 19
IV Akhet 19 nfrjjt-r I Peret 18
II Peret 18 nfrjjt-r III Peret 17
IV Peret 17 nfrjjt-r I Shemu 16
Each interval is designated as month (#bd ) of a phyle leader. It was
evident to Borchardt that the six alternate months are lunar and rep-
resent either months 1, 3, 5, 7, 9 and 11 or months 2, 4, 6, 8, 10 and
12 of a lunar year of 354 or 355 days. He further realized that the
rising of Sothis is the likely starting point for this lunar year.79 Parker
agreed in principle, but he ascribed pBerlin 10056 to Amenemhet III,
and not to Senwosret III as Borchardt had done.80 Reckoning 36 years
of rule for Senwosret III, Parker placed peret Sepdet of years 30/31 of
Amenemhet III on I Shemu 1. He further argued that I Shemu 1 is
included within the last month of the account: IV Peret 17 to I Shemu
16. If this were the twelfth month of the lunar year, then the year
would have begun on I Shemu 27 or 26 with the beginning of the
lunar year following after peret Sepdet on I Shemu 1.
Parker was right to attribute pBerlin 10056 to Amenemhet III, but
Senwosret III ruled only for 19 years, not 36. Therefore the correct
date for peret Sepdet in 30 Amenemhet III is at least IV peret 26 or 28,

76
Parker, Calendars, 31.
77
For the fragmentary accounts, see Krauss, Sothis, 83–84.
78
Luft, Fixierung, 74, and Pl. 13 provides a transcription and a photo of the complete
account.
79
Borchardt also noticed a possible relationship to summer solstice. It so happens
that the summer solstice and the rising of Sothis at f = 24° coincided on July 9 ( Jul.)
in the 19th century BC. But since solstices are not attested in Egyptian calendric
sources, by contrast to the rising of Sothis, it is pointless to relate the Illahun lunar
year to summer solstice.
80
Parker, Calendars, 37.
454 rolf krauss

depending on whether Sothis rose on IV Peret 16 or IV Peret *18 in


7 Senwosret III. Peret Sepdet in IV Peret 26 or *28 fell in 31 Amenemhet
III within IV Peret 17 to I Shemu 16, the last month of the account.
Borchardt and Parker interpreted the dates before nfrjjt-r as first days,
and the ones after as last days of an Egyptian lunar month, but Luft
realized that “LD 2 nfrjjt-r LD1” is meant (see above Chapter III. 8).
The Illahun lunar year may be interpreted as standard calendar, used
and modified for a temple roster.

The Ebers Calendar As Another Example of the Sothis-based Lunar Calendar

Borchardt interpreted the Ebers calendar as an example of a Sothis-


based lunar year of the type that he had recognized in the Illahun
archive. According to him col. I (see Fig. III. 10.1) contains a series
of lunar months.81 Parker modified Borchardt’s interpretation:82 the date
III Shemu 9 is common both to peret Sepdet and to the beginning of
the lunar month wep-renpet, i.e. on III Shemu 9 in 9 Amenhotep I, a
LD 1 and peret Sepdet coincided.
Gardiner disagreed,83 but Hornung concurred with Parker,84 utiliz-
ing III Shemu 9 as LD 1 for chronological purposes.85 In the early
1980s Helck denied a correlation between the lunar months of col. I
and the associated civil days in col. II, suggesting instead that the lunar
month wep renpet could have begun on any day between III Shemu 9
and IV Shemu 8 in year 9 of Amenhotep I.86 Furthermore, he pro-
posed that peret Sepdet could have occurred on any day within the lunar
month wep renpet. Helck’s purpose was to eliminate an obstacle to his
ultra short NK chronology characterised by a 15 year reign at most
for Horemhab, and implying that year 9 of Amenhotep I was corre-
spondingly later than 1506 BC, which excluded III Shemu 9 as a Sothic
date.
Subsequently, Luft doubted that col. I refers to lunar months and
explained it as a list of feasts celebrated on unspecified days within the

81
Borchardt, Mittel, 19–29.
82
Parker, Calendars, 37–42.
83
A. H. Gardiner, RdE 10 (1955), 18–22.
84
Hornung, Untersuchungen, 16–17.
85
E. Hornung, ZDMG 117 (1967), 14–15, cf. also W. Barta, ZÄS 110 (1979), 2.
86
Helck (n. 60); idem, SAK 15 (1988), 163.
egyptian sirius/sothic dates 455

civil months of col. II;87 the latter he identified as months of the reg-
nal year, following a suggestion of Barta.88 Beckerath argued against
Luft’s interpretation of col. I; he also rejected the utilization of the non-
existent months of the regnal year.89
Whereas it follows from a straightforward reading of the Ebers cal-
endar that Sothis rose on III Shemu 9, the identification of III Shemu
9 as LD 1 presupposes the identification of the Ebers calendar as
another example of a Sothis-based lunar year of the Illahun type. Parker
discussed other sources which possibly attest a Sothis-based lunar cal-
endar, but these are liable to justifiably different interpretations and
should not be used to elucidate the Ebers calendar. However, two argu-
ments may be adduced to strengthen Parker’s case.
Without any recourse to calendrical theory, Kitchen has concluded
that Amenhotep I reigned between 1515 and 1494 BC, implying as
noted above that year 9 = 1507/06 BC.90 Because III Shemu 9 coin-
cided in 1506 BC with a LD 1, Kitchen confirms, if unintentionally,
Parker’s lunar interpretation of the Ebers date.91
The possibility of projecting the Illahun lunar year backwards and
forwards can also be utilized in support of Parker. If peret Sepdet is
shifted from IV peret 26 or *28 in 30 Amenemhet III to III Shemu
9, keeping in mind that 30 Amenemhet III can be the first or the last
year of a quadrennium, then the shift corresponds to 72 or *70 qua-
drennia ± 3 years + 43 days each or 288 ± 3 years or *280 years ±
3 years. If the LD 1 that fell on I Shemu 26 in 30 Amenemhet III is
projected over the same interval by using mean synodic months,92 it
coincides with IV Shemu 9 as first lunar day after peret Sepdet, as
Tables III. 10. 2a.b shows.
According to the projection there was a coincidence of IV Shemu
9 and a LD 1 after peret Sepdet on III Shemu 9, 282 years after 30
Amenemhet III. It is clear that in the same projected year peret Sepdet
coincided with a LD 1 on III Shemu 9.93 This implies that the Ebers

87
Luft (n. 61), 69–71.
88
Barta (n. 62).
89
Beckerath (n. 65), 28–31.
90
K. A. Kitchen, Acta Archeologica 67 (1996), 12.
91
III Shemu 9 corresponded to LD –1 in 1531, to LD 2 in 1517, to LD 1 in 1506,
to LD –1 in 1495, and to LD 2 in 1492 BC; cf. also Krauss, Sothis, 109.
92
For example, the intervall of 3426 mean lunar months corresponds to 101171.8
days. Thus the day that fell 3426 mean lunar months after I Shemu 26 in 1788 BC
corresponded to IV Shemu 3.
93
There is no such result, if peret Sepdet fell on IV Peret 16 in 7 Senwosret III.
456 rolf krauss

Table III. 10.2a


Years after Lunations since LD 1 after
30 Amenemhet III I Shemu 26 III Shemu 9

277 3426 IV Shemu 3


278 3438 III Shemu 22
279 3450 III Shemu 12
280 3463 III Shemu 30
281 3475 III Shemu 20
282 3488 IV Shemu 9
283 3500 III Shemu 28

Table III. 10.2b


Years after Lunations since LD 1 after
30 Amenemhet III I Shemu 26 III Shemu 9

285 3525 IV Shemu 6


286 3537 III Shemu 25
287 3549 III Shemu 15
288 3562 IV Shemu 4
289 3574 III Shemu 23
290 3586 III Shemu 13
291 3599 IV Shemu 2

calendar represents a Sothis-based lunar year of the Illahun type,94 and


it eliminates the argument that the Ebers calendar is without parallel
and thus only “proves itself ” (so Belmonte).

Unusual Features of the Ebers Calendar

One peculiar feature of the Ebers calendar is the position of the months
wep renpet and techi at first and second position. These two months are
otherwise known as the 12th or last month and first month respec-
tively. According to Borchardt the Sothis-based lunar year, as exemplified
in the Ebers calendar, began with the intercalary lunar month wep
renpet. Parker accepted that the Sothis-based lunar year could begin

94
Following Barta, ZÄS 110 (1983), 19–25, Beckerath, Chronologie, 50, presumes that
the Ebers calendar represents the civil-based lunar calendar.
egyptian sirius/sothic dates 457

with the intercalary month, but he argued against the identification of


the intercalary month as wep renpet. Independent of the Ebers calendar,
he had hypothesized the existence of an intercalary month “≈hwtyt”; he
emphasized that wep renpet is the name of the 12th month, and denied
the possibility that the intercalary 13th month could bear the same
name. But Parker’s intercalary month “≈hwtyt” is fictitious, and it is
quite possible that the name of an intercalary 13th month is the same
as the preceding 12th month.95
Provided that the accession day of Amenhotep I lay on III Shemu
9, it will have been its coincidence with peret Sepdet and the first day
of the lunar month in which peret Sepdet occurred, that prompted a
scribe to record the calendric situation in regnal year 9. The last month
of a Sothis-based lunar year began on the accession day, and a new
Sothis-based lunar year began a lunar month later. Disregarding the
overlapping of two lunar years, the scribe listed a series of 12 months,
corresponding to the standard format of a single year. To each lunar
month he allotted the standard length of 30 days;96 according to stan-
dard procedure he omitted the epagomenai. Thus the Ebers calendar
can be explained as the Sothis-based lunar year of the Illahun type,
adapted to a regnal year in order to make known and commemorate
the very rare coincidence of a royal accession day with the beginning
of an intercalary lunar month in which peret Sepdet occurred.
The coincidence of peret Sepdet and the first lunar day of a 13th lunar
month tended to occur each 19 Egyptian years less 4 to 5 days. The
coincidence occurred because the rising day of Sothis stayed the same
over centuries in the Julian calendar while lunar dates tend to repeat
after 19 Julian years = 19 Egyptian years less 4 to 5 days (Metonic
cycle). The coincidence of an accession day and peret Sepdet was only
possible during the NK and the TIP when regnal years began with
the accession day; in fact, the occurrence in year 9 of Amenhotep I
may have been the only one of its kind.

95
Krauss, Sothis, 112–115; cf. also J. S. Nolan, “The Original Lunar Calendar and
Cattle Counts in Old Kingdom Egypt”, in: AH 17 (2003), 84–88.
96
Cf. A. J. Spalinger, rev. of Leitz, Studien, in: OLZ 87 (1992), 25, pointing out that
the Babylonians had also set up a schematic 360 day lunar year composed of twelve
months each containing 30 days.
III. 11 FOUNDATIONS OF DAY-EXACT CHRONOLOGY:
690 BC–332 BC

Leo Depuydt

Of only three rulers of Egypt before the Roman conquest of 30 BC


do we know the exact day of death in absolute terms, that is, can we
define the distance in time to the present precisely. They are Psamme-
tichus II, Alexander the Great, and Ptolemy VIII Euergetes II. Only
Psammetichus II reigned in the period covered by this handbook. A
stela dedicated to his daughter, the God’s Wife Ankhnesneferibre, states:
rnpt-zp 7 #bd 1 #¢t sw 23 pr nΔr pn . . . psmΔk r pt “Year 7, Month 1 of
the #¢t-season (first month of the year), Day 23, this god . . . Psammetichus
went up to heaven.”1 Year 7 Month 1 Day 23 of Psammetichus II is
most probably 9 Feb 589 BC. The event happened in the period of
Egyptian history for which a widely accepted chronological model is
available that allows dating to the exact day. The period begins with
Taharqa’s reign about 690 BC and spans most of the seventh to fifth
centuries BC. Nowhere else in that period—and not anywhere before
it—can events be dated absolutely to the exact day.2 It is all of his-
tory’s earliest period of day-exact dating. From about 400 BC, day-
exact dating is suspended for a few decades. It resumes with Nectanebo
II’s reign in the 350s and 340s BC. Uncertainty then returns for a few
years due to a lack of sources. With Alexander’s conquest in 332 BC,
day-exact dating resumes once more, never to cease again. Day-exact
dating is a topic with sharply defined contours deserving treatment in

1
G. Maspero, “Deux monuments de la princesse Ankhnasnofiribrî”, ASAE 5 (1904),
84–90.
2
For converting Egyptian dates into Julian dates, that is, into our modern calen-
dar extended backward into the past, a table listing the Julian equivalent for Day 1
of every Egyptian month in the period at hand is found in P. W. Pestman (with
S. P. Vleeming), Les papyrus démotiques de Tsenhor (Leuven: Studia demotica 4, 1994),
167–183. In this table, the years before Taharqa are hypothetical, those of the fourth
century BC are approximate, and the beginning of Nectanebo II’s reign is dated two
years earlier than will be proposed here in section 3 of Chapter II. 11. For tables list-
ing only the Julian equivalent only of Day 1 of the whole year, see section 4 in Chapter
II. 11; also Ginzel, Handbuch II, 576–585; E. J. Bickerman, Chronology of the Ancient World
(Ithaca, 19802), 115–122.
foundations of day-exact chronology: 690 bc‒332 bc 459

its own right. What follows is an outline of its foundations. A com-


prehensive account remains desirable. This Chapter’s focus is on gen-
eral principles. For chronological details about the period’s dynasties
see Chapter II. 11.

1. Concepts of Time-reckoning: 1.1. Calendars

Three calendrical structures are relevant: the Egyptian civil calendar,


Egyptian lunar time-reckoning, and the Babylonian lunisolar calendar.
Throughout Egyptian history, the civil calendar was the dominant cal-
endar of daily life. Details about its structure, its 12 months of 30 days
plus five epagomenal days, and its three seasons of four months, Akhet,
Peret, and Shemu, appear elsewhere in this handbook (see Chapter I. 5).
In what follows, they will be called Months 1 through 12. Month-
names have also been preserved.3 In the period at hand, the names
appear almost exclusively in fifth century BC Aramaic papyri from
Egypt. The Greek variants are now conventionally used in modern his-
toriography. One form for each, along with the Egyptian origin, is as
follows:

1 Thoth ≈˙wty 7 Phamenoth p n jmn ˙tp


2 Phaophi p n jpt 8 Pharmouthi p n rnnwtt
3 Hathyr ˙wt ˙r 9 Pachons p n ¢nsw
4 Choiak k # ˙r k # 10 Payni p n jnt
5 Tybi t# 'bt 11 Epeiph jpjp
6 Mecheir m¢r 12 Mesore mswt r'

The Egyptian 365-day year is almost a quarter day shorter than the
solar year, or cycle of seasons, of about 365.2422 days. The Egyptian
year therefore wanders or shifts in relation to the solar year at a rate
of about one day in four years, and of exactly one day in four years
in relation to the Julian calendar year of 365.25 days (that is, three
years of 365 days followed by one of 366). The Egyptian new year
returns to the same day in the Julian calendar in exactly 1460 (365 ×
4) years. In 664 BC, new year fell on 5 February for the first of four
successive years. By 332 BC, it had receded to 14 November. Two

3
For a survey of the sources for these month-names in hieroglyphic Egyptian, Coptic,
Greek, and Aramaic, see Depuydt, Calendar, 109–136.
460 leo depuydt

new years fall in 521 BC, which is a leap year of 366 days, like all
years BC divisible by 4 after subtracting 1: the first on 1 Jan 521 BC
and the second 365 days later on 31 Dec 521 BC. New year then
shifts to a day earlier in the Julian calendar, to 30 December, already
in 517 BC, not in 516 BC.
Two kinds of lunar time-reckoning are relevant. The native Egyptian
kind lived in the shadow of Egypt’s civil calendar. It was used mainly
for religious purposes. Native lunar dates play a crucial role in the
chronology of Dynasties 26 and 30 (see sections 2 and 3 in Chapter
II. 11. The Babylonian lunisolar calendar enters Egyptian history when
Persia, which had adopted it,4 conquered Egypt in or shortly before
525 BC.
There is not much room for lunar calendars to differ. They all fol-
low the cycle of lunar phases. Astronomical lunar months are on aver-
age about 29.53 days long. Obviously, calendrical lunar months last a
full number of days. The average 29.53 is obtained by alternating
months of 29 days and of 30 days, with slightly more 30-day months
because the average is just above 29.5. Lunar calendars differ mainly
in two respects: (1) when the month begins; (2) how months are orga-
nized into years. As for (1), Babylonian and Egyptian lunar months
began close to new moon or conjunction, the point in time when the
moon is right between the earth and the sun and therefore invisible
from the earth.5 As for (2), 12 months (on average about 354 days) are
shorter than a solar year of about 365.2422 days and 13 months (on
average almost 384 days) are longer. The average of 365.2422 is obtained

4
Persian kings may have used accession dating in their native Persia (see L. Depuydt,
“Evidence for Accession Dating under the Achaemenids”, JAOS 115 (1995), 193–204),
with regnal years lasting from the accession to the successive anniversaries of the
accession.
5
As to how close, see L. Depuydt, “The Date of Death of Jesus of Nazareth”, JAOS
122 (2002), 466–80, at 471–477. It is now often stated, following Parker (Calendars, 10),
that the Egyptian lunar Day 1 was the day that follows the day in the morning of
which the last crescent is last sighted. The last crescent is last seen above the eastern
horizon, rising ahead of the sun, one to two days before conjunction. However, the
evidence, which includes civil-lunar double dates, proves only that the Egyptians did
not wait for first crescent visibility, which occurs one to two days after conjunction.
Lunar months began a little earlier. In fact, nowhere in the ancient world, Classical
or Near Eastern, is there positive evidence that first crescent visibility marked the begin-
nings of lunar months. The role of first crescent visibility is one of the most overrated
assumptions of ancient history. The Muslim calendar is an exception. But it may be
alone. Babylonian astronomical texts, mentioned elsewhere (Depuydt, JAOS 122, 471)
as a second exception, may actually not be one, as I hope to show in another place.
foundations of day-exact chronology: 690 bc‒332 bc 461

by alternating years of 12 months and years of 13 in a 19-year cycle


containing 12 years of 12 months and 7 years of 13. That is because
235 (12 × 12 + 7 × 13) lunar months are roughly as long as 19 solar
years. From the fourth century BC, and already mostly earlier, the
years with 13 months are years 3, 6, 8, 11, 14, 17, and 19 of the
cycle. In heeding both moon and sun, the Babylonian calendar is called
lunisolar. Month 1, Nisanu, begins around a new moon close to the
spring equinox, in later March or in April. The Babylonian month-
names, in one variant each, are as follows. The derived Hebrew set is
also listed.

Babylonian Hebrew
Nisanu Nisan
Aiaru Iyyar
Simanu Sivan
Duzu Tammuz
Abu Ab
Ululu Elul
Tashritu Tishri
Arahsamnu Heshvan
Kislimu Kislev
Tebetu Tebeth
Shabatu Shebat
Addaru Adar

1.2. Year-counting

Calendars account for how days add up to months and months add
up to years. That leaves counting the years. The year-count recom-
menced with each new reign in ancient Egypt. Two systems are rele-
vant here. Predating was used in Dynasty 26, even if a formal proof
is still desirable, and presumably also in Dynasties 28–30.6 Predating
of postdating was used in the Persian Dynasty 27, and presumably also
in the short Persian Dynasty 31, for which evidence is scarce.7 A fictional

6
See Gardiner, “Years”, 11–28.
7
No single principle is more important to the chronological structure of 664 BC–332
BC than predating of postdating. For details, see L. Depuydt, “Regnal Years and Civil
Calendar in Achaemenid Egypt”, JEA 81 (1995), 151–73. For the special case of
Cambyses, first ruler of Dynasty 27, see id., “Egyptian Regnal Dating under Cambyses
and the Date of the Persian Conquest”, in: Studies in Honor of William Kelly Simpson
462 leo depuydt

example involving a king assuming power on 1 Sep AD 2001 may


illustrate both systems. Let us assume that the Egyptian new year falls
on 1 January.
In Egyptian predating, Year 1 lasts four months, from 1 Sep 2001
to 31 Dec 2001. A new regnal year begins on 1 Jan 2002. Regnal
years and calendar years are henceforth coterminous. The old king’s
last regnal year, lasting from 1 Jan 2001 to 31 Aug 2001, and the new
king’s first together form a whole calendar year. This system is called
predating because the beginning of regnal Year 2 on 1 Jan 2002 pre-
cedes the beginning of the full second year of reign on 1 Sep 2002.
In Babylon, new year began around a new moon close to the spring
equinox in late March or in April, let us assume 21 March for argu-
ment’s sake. In Babylonian postdating, the eight months from the reign’s
beginning on 1 Sep 2001 to 20 Mar 2002 are not counted and called
“head (beginning) of the reign” or the like. It is a numberless acces-
sion year. This is postdating because the beginning of regnal Year 2
on the Babylonian new year of 21 Mar 2003 follows the beginning of
the full second year of reign on 1 Sep 2002.
In Egyptian-Babylonian predating of postdating, postdated Babylonian
regnal years of Persian kings are predated in Egypt. The reign’s first
Babylonian new year in the spring was for all practical purposes taken
as the reign’s beginning in Egypt. Persian-Egyptian regnal years are
predated in relation to the postdated beginning of the reign, the Baby-
lonian new year.
In predating, the beginning of regnal Year 2 precedes the beginning
of the full second year of reign. In postdating, it follows. In predating
of postdating, both are possible, following or preceding. All depends
on when the reign begins. A twofold distinction applies. In case one,
the Persian reign begins between the Babylonian and Egyptian new
years. Following applies. Postdating has an effect. In case two, the
Persian reign begins between the Egyptian and Babylonian new years.
Preceding applies, as in pure predating. Postdating has no effect.

(Boston, 1996), 179–190; W. Barta, “Zur Datierungspraxis in Ägypten unter Kambyses


und Dareios I”, ZÄS 119 (1992), 82–90; cf. also D. Devauchelle, “Un problème de
chronologie sous Cambyse”, Transeuphratène 15 (1998), 9–17. F. X. Kugler had earlier
indirectly noted the principle of predating to postdating (Sternkunde und Sterndienst in Babel
(Münster, 1907–24), vol. 2.2.2, 190–1), but in reference to the years of Ptolemy’s Canon
only, not to actual historical regnal years, and also without distinguishing between the
two distinct cases to be described next. In the Canon, the lengths of all the reigns are
converted into multiples of the Egyptian calendar year, that is, into multiples of 365 days.
foundations of day-exact chronology: 690 bc‒332 bc 463

Regnal Year 2 of a king assuming power on 15 Dec 2001, between


the Babylonian and Egyptian new years, begins on 1 Jan 2002, that
is, before the second full year of reign begins. Regnal Year 2 of a king
assuming power a month later, on 15 January, between the Egyptian
and Babylonian new years, begins on 1 Jan 2003, that is after the sec-
ond full year of reign begins.
Xerxes I’s reign illustrates case one. It began in late Nov 486 BC.
The Egyptian new year soon followed on 23 Dec 486 BC. Babylonian
regnal Year 1 began about 4 Apr 485 BC. By about 4 Apr 485 BC,
Egyptian regnal Year 1 must have begun as well, probably not earlier.
Egyptian regnal Year 2 began on 22 Dec 485 BC, on the reign’s sec-
ond Egyptian new year. The reign’s beginning fell between the Babylonian
and Egyptian new years. Egyptian regnal Year 2 therefore began on
the second Egyptian new year of the reign, and after the beginning of
the second full year of reign in late Nov 485 BC.8
Darius II’s reign illustrates case two. It began in all probability in
25 Dec 424 BC–13 Feb 423 BC,9 after the Egyptian new year of 7
Dec 424 BC. The Babylonian regnal Year 1 began on or close to 31
Mar 423 BC. By that day, Egyptian regnal Year 1 must have begun,
probably not earlier. Egyptian regnal Year 2 then began on 7 Dec 423
BC, the reign’s first Egyptian new year. The reign’s beginning fell
between the Egyptian and Babylonian new years. Regnal Year 2 there-
fore began on the first Egyptian new year of the reign, as in pure pre-
dating, and before the beginning of the second full year of reign, which
fell after 24 Dec 423 BC in the Egyptian civil calendar.
A lack of evidence prevents establishing whether Artaxerxes II’s reign
is case one or case two. The reign began between about 17 Sep 405
BC and about 10 Apr 404 BC. The Egyptian new year of 2 Dec 405
BC falls inside this period.

2. Four Key Sources

Four sources of wider scope are fundamental to day-exact dating.


Additional sources of narrower scope are adduced in the Chapter on

8
A full year of reign, or the time from accession to the first anniversary of the
accession, is here taken at 365 days. We are after all in Egypt.
9
See L. Depuydt, “The Date of Death of Artaxerxes I”, WdO 26 (1995), 86–96.
464 leo depuydt

the chronology of Dynasties 26 to 31 (see Chapter II. 11). Ptolemy’s


Canon is the anchor source. It fixes Dynasty 27 in time, and hence
indirectly also the other dynasties. The Canon is supported in this
role—and will in the future be entirely replaced in it—by Babylonian
astronomical texts. These texts cement the Canon’s veracity. A second
anchor, dependent on the Canon, is a set of Serapeum stelae. These
stelae mostly fix the reigns of Dynasty 26 in relation to the previously
fixed Dynasty 27.

2.1. Ptolemy’s Royal Canon The Canon remains the foundation of ancient
Near Eastern chronology in the first millennium BC.10 An adaptation
of the relevant portion’s Greek original is as follows.

Ptolemy’s Canon of Kings, Segment Relevant to Saite and Persian Egypt

King’s Name Years Total


Reigned

Nabonassar 14 14
Nabu-nadin-zeri 2 16
Mukin-zeri and Pul 5 21
...
Cyrus 9 218
Cambyses 8 226
Darius I 36 262
Xerxes I 21 283
Artaxerxes I 41 324
Darius II 19 343
Artaxerxes II 46 389
Artaxerxes III 21 410
Arses 2 412
Darius III 4 416
Alexander the Great 8 424

10
For an analysis, see L. Depuydt, “ ‘More Valuable than All Gold’: Ptolemy’s Royal
Canon and Babylonian Chronology”, JCS 47 (1995), 97–117. The foundation is shift-
ing, however; see id., “The Shifting Foundation of Ancient Chronology”, in: Modern
Trends in European Egyptology (British Archaeological Reports S1448; Oxford, 2005),
53–62, and id., “Ancient Chronology’s Alpha and Egyptian Chronology’s Debt to
Babylon” (to appear in a Festschrift).
foundations of day-exact chronology: 690 bc‒332 bc 465

The Canon’s diminutive size belies its importance. Its three columns
provide: (1) the ruler’s name; (2) the length of reign in a number of
full 365-day Egyptian civil years; (3) the cumulative total of years from
Nabonassar, the Canon’s first king. The cumulative count is also known
as the Era of Nabonassar. Year 1 of Cambyses, ruler no. 22, is Year
219 from Nabonassar. 218 years precede. The first day of the first year
of the Canon is 26 Feb 747 BC. The Egyptian years wander in rela-
tion to the Julian calendar (see 1.1 above). The Canon may now be
expanded as follows.

Ptolemy’s Canon of Kings, Segment Relevant to Saite and Persian Egypt

King’s Name Length of Reign in


Full Egyptian Civil Years

Nabonassar 26 Feb 747–22 Feb 733


Nabu-nadin-zeri 23 Feb 733–21 Feb 731
Mukin-zeri and Pul 22 Feb 731–20 Feb 726
....
Cyrus 5 Jan 538–2 Jan 529
Cambyses 3 Jan 529–31 Dec 522
Darius I 1 Jan 521–22 Dec 486
Xerxes I 23 Dec 486–16 Dec 465
Artaxerxes I 17 Dec 465–6 Dec 424
Darius II 7 Dec 424–1 Dec 405
Artaxerxes II 2 Dec 405–20 Nov 359
Artaxerxes III 21 Nov 359–15 Nov 338
Arses 16 Nov 338–14 Nov 336
Darius III 15 Nov 336–13 Nov 332
Alexander the Great 14 Nov 332–11 Nov 324

The two dates in each line are Day 1 and Day 365 of an Egyptian
year. Babylonian regnal Year 1 began on the first Babylonian new year
of the reign in late March or in April, namely in the springs of 747,
733, 731, 538, 529, 521, 485, 464, 423, 404, 358, 337, 335, 331, and
323 BC. The reign of course begins earlier, up to as much as a year
earlier. The Canon’s reigns begin on the Egyptian new year that pre-
cedes the reign’s first Babylonian new year. As a result, the reign’s his-
torical beginning can both precede and follow the Canon’s beginning
(cf. the related distinction between the two cases in 1.2). Xerxes I’s
466 leo depuydt

reign began before the Canon’s beginning of 23 Dec 486 BC. Darius
II’s reign began after the Canon’s beginning of 7 Dec 424 BC.

2.2. Serapeum Stelae Many of the stelae found in the tombs of the Apis
bulls in the Serapeum in Memphis list the exact dates of birth and
death of people and bulls along with their life-span to the day. If the
person or bull was born under one king and died under another, the
first king’s length of reign can be inferred. Such inferences provide the
chronological skeleton of Dynasty 26.11 One example will need to suffice.
According to stela Louvre 193, an Apis was born in Year 53 Month
6 Day 19 of Psammetichus I, died in Year 16 Month 2 Day 6 of
Necho II, and lived 16 years 7 months 17 days, that is, 16 years and
227 (7 × 30 + 17) days. The two fractions of a year that the bull lived
in Year 53 of Psammetichus I and in Year 16 of Necho II add up to
232 days, excluding apparently Month 6 Day 19, the date of birth,
from the count. That is five days less than the 227 days given in the
text. Evidently, the five epagomenal days are erroneously not counted.12
That leaves 16 full 365-day years to be identified. Necho II’s Years
2–15 are 14 of them. That leaves two. According to predating (see
1.1), one must be the 365-day year encompassing Psammetichus I’s last
regnal year and Necho II’s first. One last full year now still needs to
be accounted for. It cannot be but Psammetichus I’s Year 54, a full
365-day year, lasting from new year to new year. Year 55, a regnal
year shorter than 365 days, was then presumably his last.

2.3. The Cuneiform Record From 525 BC or shortly before, Persia ruled
both Egypt and Mesopotamia. Rulers of Egypt were now dated con-
tinuously in cuneiform tablets by the Babylonian calendar. Cuneiform
evidence thus becomes directly relevant for Egyptian chronology. R. A.
Parker and W. H. Dubberstein have collected what the cuneiform
record transmits about three aspects of the Babylonian calendar: (1)
which lunar months had 29 days and which 30 days; (2) which years
had 12 months and which 13 months; (3) on which day daylight of
each lunar Day 1 fell.13 The following line from their tables describes
Darius II’s Babylonian regnal Year 1.

11
For a tabulation of the key evidence, see Kienitz, Ägypten, 154–59. See earlier
Wiedemann, Geschichte, 115–21; see also Gardiner, “Years”, 17–20.
12
So already Gardiner, “Years”, 17.
13
Parker & Dubberstein, Chronology. Evidence that has emerged since would make
foundations of day-exact chronology: 690 bc‒332 bc 467

1 423 4/11 5/11 6/9 7/9 8/7 9/5 10/5 11/3 12/3 422 1/1 1/31
3/1
This line is to be interpreted as follows. Month 1 (Nisanu) of Year “1”
began about 11 Apr 423 BC (“423 4/11”), Month 2 (Aiaru) about 11
May 423 BC (“5/11”), and so on. Month 12 (Addaru) began about 1
Mar 422 BC (“422 . . . 3/1”). Most of the Julian dates for lunar Day
1 in Parker’s and Dubberstein’s tables are obtained by computation.
These computations are based on the assumption that the first cres-
cent could have been seen in the evening preceding daylight of lunar
Day 1, that is, in the evenings on 10 April, 10 May, and 28 February.
The first crescent is first seen in the evening one or two days after
conjunction. The conjunctions in question occurred about 3:56AM on
9 Apr 423 BC, about 6:52PM on 8 May 423 BC, and about 1:22PM
on 27 Feb 422 BC.14 It is important to note that a Julian day date of
an ancient lunar Day 1 may be off by one to two days from the actual
historical beginning. It is certain that lunar Day 1 was always close to
conjunction, but it can no longer be known for most lunar months
exactly how close, hence the description “about 11 Apr 423 BC,” and
not just “11 Apr 423 BC” for lunar dates throughout this Chapter and
Chapter II. 11.
Once the course of the Babylonian calendar is reconstructed, the
earliest and latest attested dates for each reign provide approximate
dates for the reigns’ beginnings (see sections 1 and 4 in Chapter II. 11).

2.4. Aramaic Double Dates Aramaic papyri from Egypt dating to the
fifth century BC have yielded a remarkable set of Egyptian-Babylonian
double dates. Double dates date single documents by both the Egyptian
civil calendar and the Babylonian lunisolar calendar. The fact that the
two independently obtained Julian equivalents of the Egyptian and
Babylonian dates as a rule match serves as an absolute guarantee of
the correctness of our understanding, at least from the fifth century BC
onward, of both calendars, of our understanding of the dating of
Egyptian history, and of our understanding of the Egyptian year’s wan-
dering motion. The persistent match also adds confidence to the estab-
lished understanding of earlier Egyptian chronology. That by itself is

an update desirable. On the cuneiform record, see now also C. O. Johnson, The Gentile
Times Reconsidered: Chronology and Christ’s Return, Third Edition (Atlanta, 1998).
14
Goldstine, Moons.
468 leo depuydt

a paramount contribution of these double dates to Egyptian chronol-


ogy,15 in addition to sundry detail regarding the dating of individual
kings.
For a list of the completely preserved double dates see Table III.
11.1.16 Columns (1) and (2) list what is in the text. The texts give the
month-names in Aramaic garb. The familiar Hebrew and Greek forms
are used here. The equation (1) = (2) found in the texts is reduced to
lunar Day 1 in column (3). On the one hand, the Julian date of the
Egyptian equivalent of lunar Day 1 is known because the course of
the Egyptian wandering year is fully transparent ((4)). On the other
hand, so is the time of conjunction or new moon through computa-
tion using formulas of lunar motion (see (5)).17 A comparison of (4) and
(5) reveals that, strikingly, lunar months begin around conjunction.18

3. The Limits of History’s Earliest Day-exact Chronology

Day-exact chronology sets in with Taharqa’s Year 1, the 365-day year


12 Feb 690 BC–11 Feb 689 BC, or perhaps 12 Feb 691 BC–11 Feb
690 BC (see section 2 in Chapter II. 11). But day-exact is inactive and
merely potential without actual dated events of history. The earliest
day-exact date is I Peret 10 (Month 5 Day 10) of Taharqa’s Year 3,
the date of papyrus Louvre E3228d, which concerns the sale of a slave.19
That would be 20 Jun 688 BC. Considering the uncertainty regarding
Taharqa’s Year 1, a more secure earliest day-exact date is the first
dated event of Psammetichus I’s reign, namely Month 1 (I Akhet) Day

15
Depuydt, “Consistency”, 53–4. On special problems regarding double dates nos.
3 and 10 and on statistical considerations supporting the foundational value of the
Aramaic evidence, see now sections 4.3 and 4.4 in: L. Depuydt, “Calendars and Years
in Ancient Egypt” (forthcoming in the Acts of a session on “Calendars and Years”
held at the 2005 Notre Dame Workshop on the History of Astronomy).
16
The manuscript sigla are those of B. Porten’s and A. Yardeni’s comprehensive
Textbook of Aramaic Documents from Ancient Egypt, 4 vols. ( Jerusalem, 1986–99).
17
The times are Goldstine’s (Moons) for Babylon, minus 53 minutes for Memphis,
or minus 47 minutes for Aswan.
18
Apparently, they began on average a little too early for the first crescent to have
been visible in the evening preceding daylight of Day 1 (see note 5).
19
See Depuydt, “Consistency”, 52. At 52a, line 19, for ‘11 June 688 BCE’ read
‘20 June 688 BCE’. At 52a, note 42, line 3, for ‘Month 10 Day 23’ read ‘Month 11
Day 23’.
foundations of day-exact chronology: 690 bc‒332 bc 469

28 of his Year 9, or 2 Mar 656 BC. On this day, the Nitocris Adoption
stela states, princess Nitocris departed from Sais.20
Day-exact chronology is suspended for several decades when Persian
rule over Egypt ends around 400 BC. The latest exact dates appear
in Aramaic papyri. The latest complete one is Month 1 Day 12 (12
Thoth) of Year 4 of Artaxerxes II, that is, 13 Dec 402 BC.21 If the
Artaxerxes associated with a date of Month 2 Day 18 (18 Phaophi) of
Year 4 is indeed the second king of that name, as he appears to be,22
then 18 Jan 401 BC would be the latest date.

20
Cf. Kienitz, Geschichte, 15.
21
Papyrus “B3.12” in: Porten & Yardeni, Textbook (n. 16), vol. 2.
22
B. Porten, “The Calendar of Aramaic Texts from Achaemenid and Ptolemaic
Egypt”, in: Irano-Judaica, II: Studies Relating to Jewish Contacts with Persian Culture through-
out the Ages, Sh. Shaked & A. Netzer, eds., ( Jerusalem, 1990), 13–32, at 19 (document
“C.43”).
Table III. 11.1. Completely Preserved Double Dates in Aramaic Papyri from Egypt
470

(1) (2) (3) (4) (5)


Text Egyptian Babylonian Equation (1) = (2) Daylight of New Moon
Date Lunar Date Converted for Egyptian Date Closest to (4)
Babylonian Day 1 in (3)

No. 1 C3.8 9 Choiak 24 Adar 16 Hathyr 4 Mar 471 1 Mar 2:41PM


Year 15 Year 14 = 1 Adar
of Xerxes I
No. 2 B2.1 28 Pachons 18 Elul 11 Pachons 26 Aug 471 24 Aug 6:13PM
Year 5 of Xerxes I = 1 Elul
No. 3 B3.1 4 Thoth 7 Kislev 3 epagomenal 12 Dec 457 16 Dec 8:06AM
Year 9 of Artaxerxes I = 1 Kislev
No. 4 B3.2 25 Phamenoth 20 Sivan 6 Phamenoth 17 Jun 451 16 Jun 1:59PM
Year 14 of Artaxerxes I = 1 Sivan
No. 5 B2.7 10 Mesore 2 Kislev 9 Mesore 16 Nov 446 16 Nov 5:17AM
Year 19 of Artaxerxes I = 1 Kislev
No. 6 B2.8 19 Pachons 14 Ab 6 Pachons 13 Aug 440 12 Aug 7:16PM
Year 25 of Artaxerxes I = 1Ab
No. 7 B3.4 9 Payni 7 Elul 3 Payni 8 Sep 437 7 Sep 12:02PM
leo depuydt

Year 28 of Artaxerxes I = 1 Elul


No. 8 B3.5 25 Epeiph 25 Tishri 1 Epeiph 6 Oct 434 4 Oct 8:43AM
Year 31 of Artaxerxes I = 1 Tishri
No. 9 B3.6 7 Phamenoth 20 Sivan 18 Mecheir 24 May 427 22 May 4:58AM
Year 38 of Artaxerxes I = 1 Sivan
No. 10 B3.9 22 Payni 6 Tishri 17 Payni 17 Sep 416 15 Sep 5:32AM
Year 8 of Darius I1 = 1 Tishri
No. 11 B2.10 12 Thoth 3 Kislev 10 Thoth 14 Dec 416 12 Dec 11:52PM
Year 9 of Darius I1 = 1 Kislev
No. 12 B2.11 9 Hathor 24 Shebat 16 Phaophi 18 Jan 410 17 Jan 2:54AM
Year 14 of Darius I1 = 1 Shebat
No. 13 B3.10 29 Mesore 24 Marcheshvan 6 Mesore 2 Nov 404 1 Nov 10:41AM
Year 1 of Artaxerxes I1 = 1 Marcheshvan
No. 14 B3.11 8 Choiak 20 Adar 19 Hathyr 18 Feb 402 16 Feb 8:14PM
Year 3 of Artaxerxes I1 = 1 Adar
PART IV

CONCLUSIONS AND CHRONOLOGICAL TABLES


IV. 1 CONCLUSIONS

Rolf Krauss and David Alan Warburton

From Dyn. 27 Back to Early Dyn. 22

The chronology of Dyn. 27, also known as the First Persian Period, is
fixed by Ptolemy’s Canon. Analysis of a series of Serapeum and bio-
graphical stelae provides the chronological framework for Dyn. 26, with
the exception of the reign of Amasis who ruled for 43 or 44 years.
Parker solved this problem in favour of 44 years by utilizing a lunar
date from year 12 of Amasis, setting year 1 of Psammetichus I in 664
BC. A Serapeum stela linking Psammetichus I to Taharqa and other
dated sources yield 690 BC as the latter’s year 1 (Chapter II. 11).
Dated documents of Taharqa’s predecessor Shebitku are few, but accord-
ing to the Tang-i Var inscription, regnal year 1 of Shebitku corresponded
to 706 BC at the latest (Chapter II. 12). His predecessor Shabaka ruled
at least into a year 15; at the beginning of his reign, he defeated
Bocchoris of Memphis. Using dead reckoning 723/22 BC is the latest
possible date for year 6 of Bocchoris. There is a gap between Bocchoris’s
predecessor Tefnakhte and the last king of Dyn. 22, Shoshenq V. 26
years at most, the lifespan of the Apis buried in year 6 of Bocchoris,
bridges the gap. This Apis was the successor of the bull that died in
year 37 of Shoshenq V.
For dead reckoning back to the middle years of Dyn. 22 we follow
Aston’s reconstruction of the period’s history which differs in certain
points from Kitchen’s valuable analysis of the history of the Third
Intermediate Period. According to Aston, there were parallel Memphite
and Theban lines from Takelot II and Shoshenq III onwards; the
Theban line fought at times with an Upper Egyptian rival (Chapter
II.10). Precise dates for Dyn. 22 can be deduced by following the rea-
soning of Parker, Vernus, and Kruchten in interpreting the Tepi Shemu
dates of the Late Period as lunar dates which most likely yield 1 Takelot
II = 845, 1 Shoshenq III = 841 BC, and 1 Petubaste I = 834 BC
(Chapter III. 8). The history and chronology of the Theban dynastic
line after Petubaste I and Iuput I remain open, and the same applies
to the Lower Egyptian Dyn. 23.
474 rolf krauss and david alan warburton

The chronology of early Dyn. 22 depends on dead reckoning. The


sum of the highest attested regnal dates for Osorkon II, Takelot I,
Osorkon I, and Shoshenq I, added to 841 BC as year 1 of Shoshenq
III, yields 938 BC at the latest for year 1 of Shoshenq I (Chapter II.
10). The large Dakhla stela provides a lunar date in the form of a wr“
feast in year 5 of Shoshenq [I], yielding 943 BC as his year 1 (Chapter
III. 8). The traditional date of 945 BC for Shoshenq I’s accession rests
on a combination of Biblical and Egyptological information. There is,
however, no basis for the projected Biblical dates, as no contemporary
archaeological or epigraphical evidence provides any support for the
generation counts in the Old Testament. It is true that the results of
dead reckoning from the Egyptian sources seem to coincide roughly
with the projected Biblical date for the raid on Jerusalem, but the
Biblical source cannot provide support for the conjecture.

Dyn. 21

Dead reckoning of the highest attested dates for Psusennes II, Siamun,
and 'Akheperre' yields at least 13(?) + 17 + 2 = 42 years for the sec-
ond half of Dyn. 21. Another six or perhaps even a few more years
can be added, if year 19 mentioned on the large Dakhla stela is attrib-
uted to Psusennes II.
Young recognized that a “generation” elapsed between the induc-
tion of a priest in year 2 of King 'Akheperre' and the induction of the
priest’s son in year 17 of Siamun. The distance implies a short reign
for 'Akheperre' and suggests his identification with “Osochor” < Osorkon,
the predecessor of “Psinaches” according to Manetho, who ruled for 6
years. We conflate Young’s thesis with Vernus’s and Kruchten’s inter-
pretation of the inductions as occurring on lunar Tepi Shemu feasts. Thus
2 'Akheperre' = 990 BC and 17 Siamun = 970 BC, implying a reign
of exactly 6 years for 'Akheperre'. Manetho’s figure for Osochor proves
to be correct, but there is no guarantee that his figures for other kings
of Dyn. 21 are also correct. Further confirmation for 2 'Akheperre' =
990 BC is provided by the date of an oracle in year 3 of ['Akheperre']
on the last day of the Tepi Shemu feast, i.e. on a lunar day 5. A gap
of 8 unattested years remains between 17 Siamun = 970 BC and 962
BC as the latest possibility for year 1 of Psusennes II (Chapter III. 8).
Jansen-Winkeln questions whether the contemporaneous regnal years
of the first half of Dyn. 21 refer to the Tanitic kings or to the Theban
conclusions 475

High Priests; at present the problem remains unresolved (Chapter II.


9). The alternatives result in different dates for the first year of Dyn.
21. If the recorded regnal years refer to the Tanitic kings, at least 9 +
48 + 24 = 81 full years must be reckoned for Amenemope, Psusennes
I, and Smendes. Jansen-Winkeln’s model reckons with at least 48 +
24 + 5 = 77 full years for High Priests Menkheperre', Pinudjem, and
Herihor.
If the reconstruction of the inscription on the well-known linen wrap-
ping strip is correct with [year x under] king Amenemope = year 49
[of High Priest Menkheperre'], there was an overlap of unknown length
between Amenemope and Menkheperre'. The degree of uncertainty
can only be somewhat minimized since it is highly probable that year
49 Menkheperre' = year 5 ± 4 of Amenemope, i.e., year 49 of
Menkheperre' could be any year of Amenemope, from year 1 to year 9.
There are no contemporaneous dates for Amenemnisut; thus only
the duration X (less than all of Amenemope’s 10 years) of the pre-
sumed coregency of Psusennes I and Amenemope has to be consid-
ered, resulting in 1073 BC—X years at the latest for the beginning of
Dyn. 21. Should 992/991 BC = 1 'Akheperre' Osorkon, then Dyn. 21
began in 1074 BC ± 4 years.
Although this chronology for Dyn. 21 rests on a methodologically
unsatisfactory admixture of dead reckoning and presumed lunar dates,
we arrive at a reasonably precise point of departure for the end of
Dyn. 20.

From Dyn. 20 Back to Dyn. 19 (Chapter II. 8)

There is a gap between the beginning of Dyn. 21 and the reign of


Ramesses XI, the last king of Dyn. 20. Egyptologists generally concede
that his reign could have ended 1 or 2 years later than year 10 of the
we˙em mesut era = regnal year 28. The reigns of Ramesses IX through
XI add up to 49 y + 8 m, if their highest attested dates and accession
dates are considered. The exact lengths of the reigns of Ramesses VIII
and VII are unclear, but 1 year and 7 years, respectively, are certain.
The sum of the highest undisputed dates of Ramesses III through VI
is 8 + 4 + 7 + 31 = 50 years. Thus year 1 of Ramesses III corre-
sponds to 1181 BC—X years or 1182 BC ± 4 years at the latest.
Between Ramesses III and Ramesses II there are two chronological
problems, viz. the transition from Twosre to Sethnakhte and the position
476 rolf krauss and david alan warburton

of Amenmesses. Both questions are resolvable on the basis of Ramesside


lunar dates (Chapter III. 8). Several lunar feast-of-the-valley dates sup-
plement the Piramesses lunar date of year 56 of Ramesses II. Their
mutual distances imply that Sethnakhte’s reign of 3 years (incomplete)
followed year 8 of Twosre; furthermore, the entire reign of Amenmesses
was subsumed within Sety II’s. Thus the distance between the accessions
of Ramesses III and Sety II amounts to 16 y + 3 m. Since 9 y + ca.
5 m are attested for Merneptah and 66 years + ca. 2 months for
Ramesses II, 1 Ramesses II corresponds to 1273 BC—X years or 1274
BC ± 4 years at the latest. On the basis of the Piramesses lunar date,
Parker preferred 1290 BC as year 1 of Ramesses II, while admitting
1304 and 1279 BC as other possibilities. By contrast, Bierbrier decid-
edly favours 1279 BC, relying on his research into the genealogy of
the late New Kingdom; Kitchen supports Bierbrier’s argument. The
best match for the astronomically possible equivalents for the Piramesses
lunar date and the festival-of-the-valley lunar dates is 1279 BC. The
degree of agreement is significantly less for 1304 BC, which is histor-
ically impossible; the remaining lower possibilities—1268 and 1265
BC—are excluded for astronomical reasons.

From Early Dyn. 19 Back to the Amarna Kings

Jansen-Winkeln gives cogent reasons why the biography of the High


Priest Bakenkhonsu does not support a 15 year reign for Sety I (Chapter
II. 8). His highest attested date is year 11; for his predecessor Ramesses
I a year 2 is documented. For Haremhab, the series of regnal dates
ends with year 13, but years 26 (oIFAO 1254) and 27 (oIFAO 1254,
Medinet Habu graffito), are probably documented. This provides sup-
port for the interpretation of the regnal year 59 (or 58) mentioned in
the Mes-inscription as 32 years of the Amarna rulers plus 27 years of
Haremhab’s own, as done traditionally. With 27 years, dead reckon-
ing results in 1319 BC as year 1 of Haremhab. The comparative dearth
of archaeological documentation would appear to suggest a shorter
reign. However, such criteria must be weighed carefully. Helck exper-
imented with 15 years for Haremhab; Beckerath among others does
not follow him. The twin anchors of the known royal synchronisms for
Dyns. 18 and 19 (see below), combined with the astronomical dates,
would present substantial chronological anomalies if Haremhab’s reign
were shorter than 27 years. Only a long reign for Haremhab is com-
conclusions 477

patible with these figures and with 1279 BC as year 1 of Ramesses II.
Furthermore, without a long reign, the distance between 1508/03 BC
as the latest possibility for the Ebers Sothic date and 1 Ramesses II =
1279 BC could not be covered.
The chronology of the period when Amarna was the residence seems
clear (Chapter II. 8). Sixteen successive wine vintages, each represent-
ing a calendar year, occurred during the occupation of the site; 13 vin-
tages correspond to years 5 through 17 of Akhenaten, whereas 3 vintages
occurred under his successors. The first 2 of them are evidently to be
attributed to King 'Ankhkheprure' Smenkhkare'. Since Nebkheprure'
Tut'ankhamun is attested at Amarna, the last vintage would seem to
be datable to his regnal year 1. However, the last vintage is possibly
attributable to Queen 'Ankhetkheprure', successor of 'Ankhkheprure'
Smenkhkare'. From year 1 of Akhenaten until the royal court left
Amarna, 20 full years elapsed. Tut'ankhaten/Tut'ankhamun ruled at
least 8 or 9 years (depending on the attribution of the last vintage on
record at Amarna), and Aya 4 years which amounts to 33 or 32 years.
Thus dead reckoning yields 1352/51 BC at the latest for 1 Akhenaten.

Near Eastern and Egyptian Synchronisms (Chapter II. 13)

To some extent, the synchronisms of Near Eastern and Egyptian rulers


provide a basis for postulating absolute dates for this period. The pre-
served correspondence demonstrates that the reign of Ramesses II was
contemporary with the reigns of the Hittite kings Muwattali and Hattushili;
and likewise the reign of Akhenaten was contemporary with Shuppiluliuma
I of Hatti, Tushratta of Mitanni, Ashshur-Uballit I of Assyria, and
Burnaburiash II of Babylon. Whereas the reigns of the Hittite and
Mitanni kings cannot be dated directly, the Assyrians and Babylonians
can be dated with near precision, and linked to Hatti and Mitanni and
thus also to the Egyptian historical relations with these countries. The
proposal of Gasche et al. (Dating) is the only astronomically and archae-
ologically supported chronology available today. Although some debate
persists about the first half of the second millennium, there is general
agreement on rough dates for the reigns of Adad-nerari I (1300–1270
BC), Shalmaneser I (1269–1241 BC), Tukulti-Ninurta I (1240–1205
BC), and Ashshur-Uballit I (1356–1322 BC) of Assyria as well as for
Kadashman-Enlil (1369–1355 BC), Burnaburiash II (1354–1328 BC)
and Kadashman-Turgu (1276–1259 BC) of Babylon.
478 rolf krauss and david alan warburton

Correspondence preserved in the archives at Hattusha links Hattushili


II of Hatti not only to Ramesses II, but likewise to the reign of
Kadashman-Turgu and the period of Egyptian—Hittite conflict before
the treaty of peace (year 21 Ramesses II) and after the accession of
Hattushili, years after the battle of Kadesh (year 5). The same corre-
spondence also allows a synchronism of Muwattali with Adad-nerari I
immediately after the battle of Kadesh.
Amarna letter EA 7 was addressed to Akhenaten and written by
Burnaburiash II. EA 6 was written by the same Burnaburiash II but
addressed to an Egyptian king with a different name, and the text
specifies that the father of Burnaburish maintained friendly relations
with the recipient. EA 5 confirms that Kadashman-Enlil was in con-
tact with Amenhotep III. Furthermore, EA 9, written by Burnaburiash
II to Akhenaten, refers to the Assyrian messengers in Egypt mentioned
by Ashshur-Uballit I in EA 16. The Amarna correspondence thus
confirms that although Burnaburiash and Akhenaten were contempo-
raries, Burnaburiash came to the throne before the death of Amenhotep
III; the synchronism of Ashshur-Uballit I and Burnaburiash II known
from Mesopotamian sources can thus also be dated to the reign of
Akhenaten, with an overlap for both rulers into the reigns of Akhenaten’s
immediate successors.
The Hattusha and Amarna correspondence document synchronisms
between the first decades of Ramesses II’s reign and the reigns of
Kadashman-Turgu and Adad-nerari I, and that the accession of
Akhenaten coincided with the reign of Burnaburiash. Therefore it is
clear that a shift of any reign necessitates a commensurate change in
the others. Since the anchors for the astronomical absolute dates and
the historically relative dates for Mesopotamia and Egypt are funda-
mentally different, these synchronisms can be viewed as virtually certain.

From Amenhotep III Back to Thutmose III (Chapter II. 8)

The highest attested full years are 37 for Amenhotep III, 7 for Thutmose
IV, 22 (or 25) for Amenhotep II, 53 for Thutmose III. Parker’s argu-
ments for a coregency of Thutmose III and Amenhotep II of more
than 2 years are invalid. If there was any coregency at all, it amounted
to 4 months at most. Adding half a year to each of the full years yields
1472 (or 1475) BC at the latest for year 1 of Thutmose III.
conclusions 479

There are two lunar dates of Thutmose III, one referring to the
foundation date of a temple at Karnak in year 24 and the other to
the battle of Megiddo in year 23. Both dates have a complicated his-
tory of interpretation; whether the recorded dates ought to be emended
or not is moot. Utilizing both possibilities results in 1479 BC = 1
Thutmose III as the only astronomically viable alternative.

Astronomical Determination of the Period between Thutmose III and Senwosret III
Based on the Dates in the Illahun Archive and the Ebers Calendar

Dead reckoning and generation-counts do not provide a sufficient basis


for a chronology of the Second Intermediate Period. However, the
Ebers calendar and the Illahun archive furnish astronomical data enabling
the limits of the period to be defined. It is a fact that the combined
reigns of Thutmose II and I amounted to a minimum of 4 years. A
comparison of the archaeological documentation for Thutmose I and
II with Hatshepsut’s yields 4 years for Thutmose II and 11 years for
Thutmose I. But we shall not use these figures, because exact reign-
lengths cannot be based upon archaeologically recoverable remains.
The traditional argument that Manetho’s Chebron is 'Akheperenre'
Thutmose II and that Chebron’s 13 regnal years are to be emended
to *3 years, is speculative. The two predecessors of Thutmose I—
Amenhotep I and 'Ahmose—ruled, respectively, into a year 21 and at
least into year 22. Thus on the basis of dead reckoning, year 1 'Ahmose =
1526 BC at the latest.
The Ebers calendar refers to regnal year 9 of Amenhotep I. The
discussion about the proper interpretation of the astronomical data in
the Ebers calendar had achieved an initial consensus following Parker’s
study of 1950 and Hornung’s of 1964. Accordingly, the rising of Sothis
and a first day of lunar month wep renpet coincided on III Shemu 9 in
regnal year 9 of Amenhotep I. Although there were methodological
shortcomings associated with a challenge to this consensus mounted in
the 1980’s, it must be stressed that the discussion did produce the
insight that the period covered by the Ebers calendar is concurrent
with regnal year 9 of Amenhotep I, i.e. the accession day of the king
very probably coincided with III Shemu 9. When taken at face value,
the text of the Ebers calendar reports that the rising of Sothis occurred
on III Shemu 9 in year 9 of Amenhotep I. Given that equivalents for
pharaonic Sothic dates are subject to various conditions, the Ebers
480 rolf krauss and david alan warburton

Sothic date could correspond to any year between 1546 BC and 1503
BC (Chapter III. 10).
The Illahun archive contains a Sothic date and 21 lunar dates from
the reigns of Senwosret III and Amenemhet III. The Sothic date is
problematic since IV Peret 16 in year 7 of Senwosret III in the temple
diary could be a mistake for IV Peret *18. When both possible dates,
along with the uncertainties of pharaonic Sothic dates are considered,
the Illahun Sothic date corresponds to any year between ca. 1882 and
1827 BC. By contrast, there is only one astronomically correct solu-
tion for the Illahun lunar dates, because lunar dates do not repeat reg-
ularly in 25 year cycles (Chapter III. 8). The result can be expressed
as 1 Senwosret III = 1837/36 BC, corresponding to 1831/30 BC =
year 7 of Senwosret III, the year of the recorded Sothic date.
Provided there was no calendar reform between year 7 of Senwosret III
and year 9 of Amenhotep I, a maximum of 333 ± 3 years (Illahun
Sothic date = IV peret 16) and a minimum of 325 ± 3 years (Illahun
Sothic date = IV Peret *18) elapsed between the Illahun and Ebers
Sothic dates. Subtraction of these figures from 7 Senwosret III = 1830
BC yields 1508 to 1494 BC, overlapping 1549 to 1503, the interval of
the Ebers Sothic date. The overlap implies that the astronomically
correct interval for the Ebers Sothic date is the period from 1508 to
1503 BC.
It follows from the Illahun lunar dates that 1 Amenemhet III =
1818/17 BC. The Nile level inscription referring to year 44 + x of
Amenemhet III and year 1 of Amenemhet IV may indicate a core-
gency. The highest attested dates of Amenemhet IV and Nefrusobk,
which compare favourably with the figures in the TC, yield ca. 1760
BC as last year of Dyn. 12. On this basis, year 3 of the immediate
predecessor of Khendjer can be fixed, provided the Monthu feast
recorded in pBoulaq 18 was celebrated on lunar day 2, like the Monthu
feast that Luft recognized in the Illahun archive. If so, year 3 of
Khendjer’s predecessor, whether he was Sobekhotep II or not, was
1734 BC.

The Second Intermediate Period (Chapter II. 7)

The changes in the sequence of Second Intermediate Period kings and


dynasties proposed by Ryholt are not necessarily compelling, and in
any case, they have no chronological significance since dead reckoning
conclusions 481

depends upon a fixed sequence of kings—and this does not exist. Rather,
we must concentrate on dynasties. Ryholt’s proposal of an overlap for
Dyns. 12 and 13 is speculative; neither Ryholt nor Beckerath suggests
that Dyn. 13 overlaps 15. By contrast, the concurrence of Dyn. 15
with 17 and 18 is certain.
There are very few contemporaneous dates for Dyn. 13: Ryholt
(Situation, 193–194) lists less than two decades for Sobekotep I, Sonbef,
Khendjer, and Sobekhotep IV, with another three decades documented
for unknown rulers who should be assigned to this period. The remain-
ing reigns must be assigned arbitrary dates based on averages, and ref-
erences to Manetho or the TC which cannot be used to base an
argument. Ryholt documents roughly half a century for Dyn. 13; he
then extrapolates 154 years for all of Dyn. 13; Beckerath (Untersuchungen,
220) allotts only 133 years; Franke is closer to Beckerath, and Kitchen
closer to Ryholt.
The overlap with Dyn. 18 can be estimated at less than two decades,
with the conquest of Avaris not later than ca. year 18 of the reign of
'Ahmose, followed by the siege of Sharuhen before year 1 of Amenhotep
I = 1514 BC. Based on attested dates, only year 33 of pRhind can be
attributed to Dyn. 15. The stelae of Kamose provide some additional
decades. Regardless of the paucity of contemporary sources, Dyn. 15
will have lasted more than a century, and Dyn. 13 at least 133 years.
Thus the interval between the end of Dyn. 12 and the fall of Avaris
would be z = (x + 100) + (133 + y). This would place the end of
Dyn. 12 before 1751 BC. The length of the Second Intermediate Period
cannot be reduced further than the minimal figures provided here.

The Beginning of Dyn. XII

In Chapter II. 7, Schneider accepts, as have others, Jansen-Winkeln’s


arguments in favour of coregencies. Moreover, double-dating implies
that coregencies have chronological consequences. But arguments in
favour of coregencies are not compelling and if they existed in prac-
tice, there is no reason to presume they played a chronological role.
The Illahun papyrus Berlin 10055 states that year 19 of [Senwosret
III] was followed by year 1 of [Amenemhet III], while the Korosko-
inscription implies that year 29 of Amenemhet I was valid as a date.
If there were coregencies, possibly the 30-reign of Amenemhet I was
followed by a 45–year reign of Senwosret I, whereas year 19 of Senwosret
482 rolf krauss and david alan warburton

III was followed by the 46–year reign of Amenemhet III. This has
chronological implications since the Illahun lunar dates suggest that
1818/17 BC was the year of the accession of Amenenhat III, and this
is the latest date compatible with our dead-reckoning for Dyns. 13 and
15, thus constituting the link between the length of the Second Inter-
mediate Period and the Sothic dates for Senwosret III (year 7 = 1830
BC) and Amenhotep I (year 9 = 1508–1503 BC). We thus project back
the dates for the earlier part of Dyn. 12, based on dead-reckoning, to
arrive at year 1 of Amenemhet I as either ca. 1939/40 (with possible
16 years of coregencies) or ca. 1955/56 (without coregencies).

From the Early Middle Kingdom Back to Early Dyn. 11

The highest attested dates for Dyn. 11 add up to 106 years. Arguably
the highest attested dates are not last regnal years; furthermore, no
contemporaneous dates exist for three kings. The TC gives a total of
143 years for Dyn. 11 which seems to be of correct magnitude, but
might be incorrect in detail. If the uncertainties of MK coregencies
are considered, dead reckoning implies a date for the beginning of
Dyn. 11 well before 2046 (with MK coregencies) or 2062 BC (no MK
coregencies).
These limits must be reconciled with the Khozam lunar date of a
regnal year [1] from the last years of Herakleopolitan hegemony in the
Coptite nome, corresponding to the early years of Dyn. 11. Astronomically
possible years for the Khozam lunar date are 2053, 2078, and 2103
BC. 2053 BC would appear to be too low, making it likely that Dyn.
11 began shortly before 2078 or 2103 BC.

From Early Dyn. 11 Back to Late Dyn. 5

For the First Intermediate Period we rely on the judicious estimate of


Hayes and Fischer of not more than 40 years for the pre-Theban
period, and we follow Baud in allowing Dyn. 8 to span a “generation”
or about 30 years (Chapter II. 5).
Only 10 m-¢t zp years and 12 zp years are documented for Dyn. 6.
At least 72 zp years actually elapsed, i.e. 83.3% of the zp years are
missing from the archaeological record. Presuming that the percentage
of missing m-¢t zp years is the same as for zp years, it is likely that at
conclusions 483

most 60% of the regnal years of Dyn. 6 were zp years and at least ca.
40% (m-)¢t zp years. Thus Dyn. 6 lasted between 120 and 144 years.
Conversely, one can argue that the autobiography of Weni mandates
that he served in a minor capacity during the reign of Teti, entering
royal service under Pepy I, and continuing to serve energetically well
into the reign of Merenre'. Since the highest known date for Pepy II
would be m-¢t zp 31, one can assume at least 63 years for that reign,
and less than a century for those preceding.
By adding the minimum figures for the Second Intermediate Period
and Dyns. 6 and 8 to ca. 2078/2103 BC for the very early years of
Dyn. 11, we arrive at ca. 2268/2292 BC and 2293/2317 BC for year
1 of Dyn. 6 at the latest.
Posener-Kriéger found two lunar dates in the Neferirkare' archive;
both could be regnal years of Izezi, or only one date his and the other
Wenis’s. We follow Posener-Kriéger who preferred the latter alterna-
tive, ascribing one lunar date to regnal year 7 of Wenis and the other
to a regnal year of Izezi that lay 11 or 25 years earlier. It is gener-
ally agreed that the TC’s 30 years, and Manetho’s 33 years for Onnos
(Wenis) are far too high. The highest attested date of Wenis is zp 8,
corresponding, at most, to regnal year 16. Provided the count was bien-
nial, then 7 Wenis is at least (or exactly) 10 years earlier than the
beginning of Dyn. 6, corresponding to ca. 2278/2302 BC or 2303/2327
BC, with 200/224 or 225/249 years thus lying between 7 Wenis and
2078, or 2089, or 2103 BC as year of the Khozam lunar date. Within
this minimum range the appropriate lunar distance between year 7 of
Wenis and the Khozam date amounts to at least 211 years + 119 days;
other astronomically feasible possibilities are 236 y + 119 d or 261 y +
119 d. The corresponding possibilities for year 1 of Wenis can be
expressed as 2321 BC ± 25 years. Taking into account that 2321 BC—
25 years presupposes minimum figures for Dyn. 6, the preferable figure
for year 1 of Wenis is 2321 BC or 2321 BC + 25 years.
Izezi ruled at least 33 years, provided 68% of his m-¢t zp years are
missing, as in the case of his zp years. If the count was regular, the
reign lasted for at least 42, or 44 years, depending on the reading of
the highest zp year; the year of his accession would be 2365/64 BC
or 2365/64 BC + 25 years. Manetho’s 44 years for Tancheres (<
Djedkare') would be correct, if only coincidentally, whereas the TC’s
28 years are inacceptable regardless.
484 rolf krauss and david alan warburton

Early and Mid-Dyn. 5

There are equal numbers of contemporaneous zp and m-¢t zp years for


Userkaf and Sahure', implying a regular biennial count. For these kings
the Royal Annals confirm a biennial count, whether the reconstruction
of the stone’s verso is exact or only approximate. The Annals should
be chronographically reliable at least for early Dyn. 5, since the orig-
inal presentation of the data (i.e., even if the Palermo stone itself is a
copy) must have followed shortly after the actual compilation of the
data, towards the end of the dynasty. The reconstructed Annals imply
8 year compartments for Userkaf, corresponding to at least 7 full reg-
nal years. Sahure', who reigned into m-¢t zp 6, evidently occupied the
throne for 6 zp years, 6 m-¢t zp years, and a year zm #-t #wj, altogether
ca. 13 years.
For Neferirkare' rnpt zp 5 is the only securely attested contempora-
neous date, whereas the Annals record the year zm #-t #wj and the year
after the 5th count. The count was evidently biennial, although the
first count might have occurred only in the second full calendar year
of the reign. Neferirkare' did not complete his funerary complex which
suggests a shorter reign than that of Sahure'; we follow Verner in
assigning him not more than 11 years. On the basis of discoveries at
Abusir, Verner attributed 2 years at most to Ra'neferef, and identified
Shepseskare' as his probable successor who ruled for only a fraction of
a year. The few secure dates for Neuserre', taken together with his
building achievements, allow an estimate of at least 15 years, but pos-
sibly we err by a decade in attributing about 30 years to him. There
are no contemporaneous dates for Menkauhor; on the basis of his
archaeological record, it seems that his reign was not brief. The TC
and Manetho possibly share the same tradition, the TC rounding to
full 8 years, Manetho to a full 9. In conformity with common prac-
tice and fully aware of the risk, we attribute 8 full years to Menkauhor.
If an uncertainty of ± 5 years is allowed for Neuserre', we arrive at
ca. 2440 BC ± 5 years or 2465 BC ± 5 years for year 1 of Userkaf.

End of Dyn. 4 and Row 1, Verso of the Annals Stone

For Shepseskaf, rnpt m-¢t zp jpt 1 is attested; the Annals record his rnpt
zm# t#wy. According to Beckerath’s recent reconstruction, row 1 on the
verso contained 5 or 6 year compartments. These ca. 6 regnal years
conclusions 485

at most can be allowed for Shepseskaf alone or for him and Thamphthis
combined, if the latter existed. The TC with 4 + 2 years for the two
rulers before [Wsr-]k #-[r'] does not contradict this arrangement.
There are no securely attested contemporaneous dates for Mycerinus.
The amount of work that would have been required to complete his
funerary complex as planned, amounted to only a tenth of that required
for those of Khufu or Khephren. The fact that the project was largely
unfinished, indicates a short reign. Nevertheless, Beckerath’s recon-
struction of the Annals favours the TC’s *28 years for Mycerinus, over
alternative *18; his proposal shows 10 year compartments in row 1 on
the verso, and he postulates 19 more above in a row “0”. However,
if the compartments in row 1 were of uniform width, then about 6
compartments result for Mycerinus. A short reign of ca. 6 years is com-
patible with the archaeological record. Coincidentally Herodotus appears
to report a reign of 6 years for Mycerinus.
The remainder of Dyn. 4.—The masons’ marks from the Giza necrop-
olis documenting rnpt zp 4 through 13 can be attributed to Khephren
and Cheops. These record only zp years, presumably a shortened nota-
tion that dispensed with references to m-¢t zp years. In view of Khephren’s
considerable building activity, it is to be expected that he ruled more
than 20 years, corresponding to the TC’s 20 + × years.
The TC and Manetho seem to concur in placing Bikheris after
Khephren. Regardless, the king who excavated the Great Pit at Zawyet
el-Aryan was more likely the successor of Ra'djedef than Khephren’s.
Analysis of the pyramid’s architecture reveals similarities to Ra'djedef ’s
monument as planned; the discovery of a palette with Ra'djedef ’s name
at the site likewise supports the contention. The reign must have been
short; Manetho’s 22 years can be tentatively explained as inflated from
an original *2 years; the figure in the TC is lost. For Ra'djedef, rnpt
zp 1 is attested; the TC’s 8 years are compatible with his building
achievement.
The masons’ marks indicate a reign of more than 20 years for Cheops,
again commensurate with his enormous building activity. An expedi-
tion graffito attests rnpt m-¢t zp 13, corresponding to regnal year 27,
provided the count was biennial, and contradicting the 23 years in the
TC. Manetho’s 63 or 66 years for Suphis (Cheops) could be secondarily
influenced by Herodotus’s 50 years for Cheops and 56 for Khephren,
but could have originated in *23 or *26 (or *33 or *36) years.
The highest attested date for Snofru is zp 24, corresponding to ca.
year 48, presuming again a biennial count. However, graffiti from the
486 rolf krauss and david alan warburton

Maidum pyramid site indicate that the count was irregular, and this is
confirmed by the Annals that record the 7th and 8th count without
an intervening year. The construction of three pyramids, one at Maidum
and two at Dahshur, could have been achieved within three decades.
Biographical data allow a limited check on these results. The life of
Netjeripunisut extended from the reign of Ra'djedef through Sahure';
Sekhemre' lived under Khephren through Sahure', and Ptahshepses
under Mycerinus to Neuserre'. Thus the overall length of our relative
chronology seems correct. But since each of these men may have lived
for 50, 60 or even 70 years, further refinement cannot be expected.
Thus we arrive at ca. 2548 BC ± 5 years or 2573 BC ± 5 years for
the beginning of Dyn. 4.

Dyns. 1–3

The contemporaneous dates available today do not suffice to establish


a coherent chronology for Dyns. 1–3. Furthermore, the history of the
second half of Dyn. 2 is unclear, insofar as kings might have ruled
simultaneously and not successively. Thus we are dependent on the
fragments of the Annals stone (Chapter I. 1). The uncertainty inher-
ent in utilizing them is twofold: the dates are not contemporaneous
with the stone and any reconstruction of the fragments will contain
mistakes. By correcting the errors of recent attempts, we arrive at a
total of 355 years for Dyns. 1–3. This implies that the first year com-
partment of the Annals corresponded to ca. 2900 BC or 2925 BC, be
this a year of 'Aha or Na'rmer.
The range of the C-14 dates for Abydos tomb U-j imply that it was
probably roughly contemporary with the tomb of 'Aha (Görsdorf et al.,
“Results”, 173), which it must have preceded. The archaeological mate-
rial likewise confirms that Naqada IIc–d cannot have been significantly
earlier than Naqada IIIa (Braun & van den Brink, “Comments”), as
indicated by the C-14 dates from Abydos U-224, U-287, which means
that Naqada IId almost overlaps with Naqada IIIa2. The archaeolog-
ical material suggests that Naqada IIIa swiftly followed on Naqada IId
and from the C-14 samples, it follows that Naqada IIIa was quickly
succeeded by Dyn. 1. As Braun and van den Brink note, the time
elapsed from final Naqada IId to early Dyn. 1 cannot have been
significantly more than a century.
conclusions 487

For absolute dating, the C-14 method can only offer a limited degree
of precision. The C-14 dates for Uruk and Abydos demonstrate that
the median dates for the relevant levels of the Late Uruk Temple C
at Uruk (± 3450 cal. BC) were invariably more than two centuries
older than those for Naqada IIIa2 and Dyn. 1 (ca. 3200 BC: Boehmer
et al. “Datierungen”). The relative position of Late Uruk centuries
before the end of the Predynastic sequence must be valid, but precise
dates for the end of the Predynastic cannot be taken at face value. As
published, these are impossible with the C-14 dates for Abydos Dyn.
1 tomb B-19 earlier than the Jemdet Nasr period (which should begin
ca. 3100 BC: Porada et al., “Mesopotamia”), since the iconography of
the palettes and sealings from the later Predynastic and earliest Dynastic
Periods cannot antedate the Jemder Nasr period; in fact, Boehmer
(1991: 60) argues that Naqada IIIa2 = Late Jemder Nasr, and thus
the Abydos Naqada IIIa2 tomb U-j should be dated to around or after
3000 BC. Either the Mesopotamian chronology must be revised, or the
end of Naqada IIIa and the beginning of Dyn. 1 must be assigned a
date ca. 3000 or 2950 BC.
Dreyer’s reconstruction of a seal from impressions recovered from
the tomb of Qa-'a listing all kings of Dyn. 1, should leave no doubt
that Nar-mer was viewed by the Egyptians as the first king of Dyn. 1;
and thus his reign will have begun the dynastic period around 2950
BC (Dreyer et al., “Umm el-Qaab”, 77–167).

Postscript by Stan Hendrickx

When this chapter was in press, Christiana Köhler published an article


of great importance for the relative chronology of the Naqada III period,
based on her excavations at Helwan.1 Helwan is by far the largest
Naqada III cemetery and offers great possibilities for chronological stud-
ies. This is especially relevant for the Naqada IIID period, which at

1
E. C. Köhler, “On the origins of Memphis—The new excavations in the Early
Dynastic necropolis at Helwan”, in: S. Hendrickx, R. F. Friedman, K. M. Cialowicz
& M. Chlodnicki (eds.), Egypt at its Origins. Studies in Memory of Barbara Adams. Proceedings
of the International Conference “Origin of the State. Predynastic and Early Dynastic Egypt”, Krakow,
28th August–1st September 2002 (Leuven-Paris-Dudley: OLA 138, 2004), 295–315.
488 rolf krauss and david alan warburton

present has only been rather vaguely defined. Köhler proposes an addi-
tional phase for the Naqada III period, Naqada IIIC3, and a triple
subdivision of Naqada IIID, i.e., IIID1–IIID3, after Helwan tomb groups
distinguished according to the presence and absence of diagnostic finds
such as cylindrical vessels, “wine jars” and “beer jars” with their respec-
tive typological sequence.2 Naqada IIIC3 would encompass the reigns
of Qa-'a and Semerk-het, the last two kings of 1st dynasty, and be
characterised by very late and rudimentary variants of the cylindrical
jars. Eventually, Naqada IIIC3 might continue into the 2nd dynasty.
Cylindrical jars no longer occur during the Naqada IIID period, while
there are also changes in tomb architecture. During Naqada IIID1, the
“wine jars” become more elongated in shape and early “beer jars”
occur at the same time. Naqada IIID2 is not well defined and could
be contemporaneous with IIID1, from which it mainly differs through
the architecture of the tombs. During Naqada IIID3, a new type of
“beer jars”, with direct rim, occurs together with the old type with
shoulder and lip rim. Another important characteristic are the earliest
examples of bowls with internal rim, that will become one of the main
characteristics of the early OK. Naqada IIID3 as defined by Köhler
dates to the end of the 2nd dynasty and is to be compared with the
late 2nd dynasty assemblage from Elephantine.3 As is stressed by Köhler
herself, the distinction between the groups and their sequence is only
preliminary and will have to be confirmed or adapted following further
excavations.

Résumé

The chronology presented here is predicated on historical methodol-


ogy. It agrees with Beckerath and Kitchen on 1279 and 1479 BC as
the accession years of Ramesses II and Thutmose III. A noteworthy
discrepancy is our acceptance of 19 years for Psusennes II which reduces
the gap in absolute chronology between Dyns. 21 and 20.
For the Middle Kingdom, the difference between our figures and
those of Beckerath and Kitchen in particular amounts to three and a

2
Köhler (n. 2), 299–306.
3
D. Raue, ‘Ägyptische und nubische Keramik der 1.–4. Dynastie’, in: W. Kaiser
et al., “Stadt und Tempel von Elephantine 25./26./27. Grabungsbericht”, MDAIK 55
(1999), 180–182.
conclusions 489

half decades; the discrepancy results from our consistent exploitation


of two different kinds of astronomical data, lunar and Sothic. There
are further cases of widely differing regnal year figures in the Old
Kingdom; compare, for example, our allocation of 2 years for Ra'neferef
and Shepseskare' combined, to Beckerath’s 18 and Kitchen’s 14 years.
Another example is our attribution of 6 years to Mycerinus, as com-
pared to 28 years in Beckerath’s chronology and Kitchen’s 18. In view
of the uncertainties in any reconstruction of the Annals, the discrep-
ancies for Dyns. 1 to 3 should not be stressed. However, both the typo-
logical analysis and the comparative material from Mesopotamia do
not leave much leeway for significant changes.
A problem that appears to be resolvable concerns the absolute dates
for the Middle Kingdom or, formulated differently, the duration of the
Second Intermediate Period. When Borchardt presented the newly dis-
covered Illahun Sothic date in 1899, he wrote (ZÄS 37, 1899, 102):
Es ist also das 7. Jahr Usertesen’s III in die Jahre von 1876–1873 v. Chr.
anzusetzen, d.h. immer noch etwa 100 Jahre später als es der am niedrig-
sten greifende Historiker Ägyptens, Eduard Meyer, in seinen Minimaldaten
annahm. Es bleiben uns also für die Zeit zwischen Usertesen III und
Amenophis I, dessen 9. Jahr durch die Sothisangabe des Papyrus Ebers
auf die Jahre 1545–1542 v. Chr. bestimmt ist, nur rund 330 Jahre und
für die Zeit vom Ende der 12. bis zum Anfang Dynastie gar nur 200 bis
210 Jahre übrig. Ist das mit den sonst bekannten historischen Angaben
vereinbar?
Borchardt answered his own rhetorical question in the affirmative, but
decades later, when Gardiner commented in Excursus C of his Grammar
on the Sothic dates of Thutmose III, Amenhotep I (year 9) and Senwosret
III (year 7), he wrote that “the first two dates fit admirably with other
considerations, but the third has been thought by some to allow too
small an interval between the Twelfth and the Eighteenth Dynasty.”
Today several Egyptologists contend that Borchardt’s interval is too
short. By contrast, we share Borchardt’s opinion and adduce further
arguments in its favour. Although our absolute dates for Senwosret III
and Amenhotep I are lower than Borchardt’s, the relative distance
between 7 Senwosret III and 9 Amenhotep I remains more or less
the same. Whereas Borchardt had only the Sothic dates to compute the
distance between Senwosret III and Amenhotep I, we also utilize the
Illahun lunar dates. The only astronomically workable distance between
year 7 of Senwosret III and 9 of Amenhotep I is 324 years—regard-
less of whether those who think this interval too short like it or not.
IV. 2 CHRONOLOGICAL TABLE FOR THE
DYNASTIC PERIOD

The selection of names and their transcription follows the lead of Baines
& Malek, Atlas, 36–37 for the most part except in those cases where
authors insisted on their own alternatives. Throne names are rendered
in italics. For the Early Dynastic Period chronology is based on the Annals
stone (Chapter I. 1); for the period between ca. 2350 and 750 BC on
lunar dates (Chapter III. 8), and therefore differs slightly in some cases
from the information found in the chapters on relative chronology.

Early Dynastic Period ca. 2900–2545+25


Dyn. 1 ca. 2900–2730+25
Nar-mer ca. 2900–?+25
'Aha ?–2870+25
Djer 2870–2823+25
“Serpent” 2822–2815+25
Den 2814–2772+25
'Adj-ib/Anedjib 2771–2764+25
Semer-khet 2763–2756+25
Qa-'a 2755–2732+25
Dyn. 2 ca. 2730–2590+25
Hetep-sekhemwy 2730–?+25
Ra'-neb ?–2700+25
Ny-netjer 2700–2660+25
Per-ibsen 2660–2650+25
Sekhem-ib 2650–?+25
Sened ?–2610+25
Kha-sekhemwy 2610–2593+25
Dyn. 3 ca. 2592–2544+25
Djoser (Netjery-khet) 2592–2566+25
Sekhem-khet 2565–2559+25
Kha'ba 2559–?+25
Nebka ?–?+25
Huni ?–2544+25
Old Kingdom ca. 2543–2120+25
Dyn. 4 ca. 2543–2436+25
Snofru 2543–2510+25
(continued on next page)
chronological table for the dynastic period 491

Table (cont.)
Khufu (Cheops) 2509–2483+25
Ra'djedef/Djedefre' 2482–2475+25
Bikheris 2474–2473+25
Khephren (Ra'kha'ef ) 2472–2448+25
Menkaure' (Mycerinus) 2447–2442+25
Shepseskaf 2441–2436+25
Dyn. 5 ca. 2435–2306+25
Userkaf 2435–2429+25
Sahure' 2428–2416+25
Neferirkare' Kakai 2415–2405+25
Ra'neferef/Neferefre' 2404+25
Shepseskare' Izi 2403+25
Neuserre' Ini 2402–2374+25
Menkauhor 2373–2366+25
Djedkare' Izezi 2365–2322+25
Wenis 2321–2306+25
Dyn. 6 ca. 2305–2118+25
Teti 2305–2279+25
Userkare' ?–?+25
Pepy I Meryre' 2276–2228+25
Nemtyemzaef Merenre' ca. 2227–2217+25
Pepy II Neferkare' 2216–2153+25
Nemtyemzaef II 2152+25
Dyn. 8 ca. 2150–2118+25
Neferkaure' 2126–2113+25
Neferkauhor 2122–2120+25
Neferirkare' 2119–2218+25
First Intermediate Period ca. 2118–1980+25
(Herakleopolitan) Dyns. 9 and 10 ca. 2118–1980+25
Middle Kingdom ca. 1980+16–1760
(Theban) Dyn. 11 ca. 2080–1940+16
Mentuhotep I (Tepi'a) ca. 1980–?+16
Inyotef I (Sehertawy) ca. ?–2067+16
Inyotef II (Wah'ankh) 2066–2017+16
Inyotef III (Nakhtnebtepnufer) 2016–2009+16
Mentuhotep II Nebhepetre' 2009–1959+16
Mentuhotep III S'ankhkare' 1958–1947+16
Mentuhotep IV Nebtawyre' 1947–1940+16
Dyn. 12 1939+16–1760
Amenemhet I Sehetepibre' 1939–1910+16
Senwosret I Kheperkare' 1920–1875+6
Amenemhet II Nebukaure' 1878–1843+3
(continued on next page)
492 chronological table for the dynastic period

Table (cont.)
Senwosret II Kha'kheperre' 1845–1837
Senwosret III Kha'kaure' 1837–1819
Amenemhet III Nima'atre' 1818–1773
Amenemhet IV Ma'kherure' 1772–1764
Nefrusobk Sebekkare' 1763–1760
Second Intermediate Period 1759–ca. 1539
Dyn. 13 1759–ca. 1630
Wegaf Khutawyre' 1759–1757
Amenemhet VII Sedjefakare' ca. 1753–1748
Sobekhotep II Sekhemre'-khutawy 1737–1733
Khendjer Userkare' ca. 1732–1728
Sobekhotep III Sekhemre'-swadjtawy ca. 1725–1722
Neferhotep I Kha'sekhemre' ca. 1721–1710
Sobekhotep IV Kha'neferre' ca. 1709–1701
Sobekhotep V Khahotepre' ca. 1700–1695
Ibiau Wahibre' ca. 1695–1685
Aya Merneferre' ca. 1684–1661
Ini Merhetepre' ca. 1660–1659
Swadjtu, Ined, Hori, Dedumose
Dyn. 14 ?
Dyn. 15 (Hyksos) ca. ?–ca. 1530
Khian Swoserenre'
Apophis 'Awoserre' ca. 1575–1540
Khamudi
Dyns. 16 and 17 ca. ?–1540
Sobekhotep VIII, Nebiriau, Rahotep, Sobekemzaf I & II, Bebiankh
Inyotef Nebukheperre' ?–?
Ta'o Senakthenre' ?–?
Ta'o Seqenenre' ?–?
Kamose Wadjkheperre' ca. ?–1540
New Kingdom ca. 1539–1077
Dyn. 18 ca. 1539–1292
'Ahmose Nebpehtire' ca. 1539–1515
Amenhotep I Djeserkare' 1514–1494
Thutmose I 'Akheperkare' 1493–1483
Thutmose II 'Akheperenre' 1482–1480
Thutmose III Menkheperre' 1479–1425
Hatshepsut Ma'atkare' 1479–1458
Amenhotep II 'Akheperure' 1425–1400
Thutmose IV Menkheprure' 1400–1390
Amenhotep III Nebma'atre' 1390–1353
Amenhotep IV/Akhenaten Neferkheprure' 1353–1336
(continued on next page)
chronological table for the dynastic period 493

Table (cont.)
Smenkhkare'/Nefernefruaten 'Ankhkheprure' 1336–1334
Nefernefruaten 'Ankhetkheprure' 1334–?
Tut'ankhaten/amun Nebkheprure' ?–1324
Itnetjer Aya Kheperkheprure' 1323–1320
Haremhab Djeserkheprure' 1319–1292
Dyn. 19 1292–1191
Ramesses I Nebpehtire' 1292–1291
Sety I Menma'atre' 1290–1279
Ramesses II Userma'atre' setepenre' 1279–1213
Merneptah Baenre' 1213–1203
Sety II Userkheprure' 1202–1198
Amenmesses Menmire' 1202–1200
Siptah Akhenre' 1197–1193
Towsre Sitre' meritamun 1192–1191
Dyn. 20 1190–1077
Sethnakhte Userkha'ure' 1190–1188
Ramesses III Userma'atre' meriamun 1187–1157
Ramesses IV Heqama'atre' setepenamun 1156–1150
Ramesses V Userma'atre' Sekheperenre' 1149–1146
Ramesses VI Nebma'atre' meryamun 1145–1139
Ramesses VII Userma'atre' setepenre' meryamun 1138–1131
Ramesses VIII Userma'atre' akhenamun 1130
Ramesses IX Neferkare' setepenre' ca. 1129–1111
Ramesses X Kheperma'atre' setepenre' ca. 1110–1107
Ramesses XI Menma'atre' setepenptah ca. 1106–1077
Third Intermediate Period ca. 1076–723
Dyn. 21 ca. 1076–944
Smendes Hedjkheperre' setepenre' ca. 1076–1052
Psusennes I 'Akheperre' setepenamun ca. 1051–1006
Amenemnisut Neferkare' ca. 1005–1002
Amenemope Userma'atre' setepenamun ca. 1002–993
Osorkon 'Akheperre' setepenre' 992–987
Siamun Netjerkheperre' setepenamun 986–ca. 968
Psusennes II Titkheprure' ca. 967–944
Dyn. 22 943–ca. 746
Shoshenq I Hedjkheperre' setepenre' 943–923
Osorkon I Sekhemkheperre'setepenre' 922–ca. 888
Takelot I Userma'atre' setepenamun ca. 887–874
Shoshenq II Heqakheperre' setepenre' ca. 873
Osorkon II Userma'atre' setepenamun ca. 872–842
Shoshenq III Userma'atre' setepenre'/amun 841–803
Shoshenq IIIa Hedjkheperre' ?–790
(continued on next page)
494 chronological table for the dynastic period

Table (cont.)
Pami Userma'atre' setepenre'/amun 789–784
Shoshenq V 'Akheperre' 783–ca. 746
Dyn. 23 (UE) and Rival Kings
Takelot II 845–821
Iuput I 820–809–
Osorkon III, Takelot III ca. 780 ± 20
Petubaste I 834–812–
Shoshenq IV, Rudamun, Iny
Dyn. 23 (LE) ca. 730
Petubaste II (?), Osorkon IV
Dyn. 24 ca. 736–723
Tefnakhte Shepsesre' ca. 736–729
Bocchoris Wahkare' 728–723
Late Period ca. 722–332
Dyn. 25 ca. 722–ca. 655
Piye/Pi'ankhy ca. 753–723
Shabaka Neferkare' ca. 722–707
Shebitku Djedkaure' ca. 706–690
Taharqa Khure'nefertem 690–664
Tantamani Bakare' 664–ca. 655
Dyn. 26 664–525
Psammetichus I Wahibre' 664–610
Necho II Wehemibre' 610–595
Psammetichus II Neferibre' 595–589
Apries Ha'aibre' 589–570
Amasis Khnemibre' 570–526
Psammetichus III 'Ankhkaenre' 526–525
Dyn. 27 (Persian) 525–404
Cambyses 525–522
Darius I 521–486
Xerxes 486–466
Artaxerxes I 465–424
Darius II 424–404
Dyn. 28 404–399
Amyrtaios 404–399
Dyn. 29 399–380
Nepherites Baenre' merynetjeru 399–393
Psammuthis Userre' setepenptah 393
Hakoris Khnemma'atre' 393–380
Nepherites II 380
(continued on next page)
chronological table for the dynastic period 495

Table (cont.)
Dyn. 30 380–343
Nectanebo I Kheperkare' 380–362
Teos Irma'atenre' 365–360
Nectanebo II Senedjemibre' setepenanhur 360–343
2nd Persian Period 343–332
Artaxerxes III Ochus 343–338
Arses 338–336
Darius III Codoman 335–332
Alexander the Great 332–323
IV. 3 KUSHITE RULERS OF PRE-25TH DYNASTY,
25TH DYNASTY, NAPATAN AND MEROITIC TIMES
INCLUDING THEIR BURIALS AND PRESUMED DATES
FOR REIGNS1

Generation Name of Ruler Burial2 Dates

Gen. A Ruler A Ku. Tum. 1 c. 885–835 BC3


Gen. B Ruler B Ku. Tum. 2 c. 865–825 BC
Gen. C Ruler C Ku. Tum. 6 c. 845–815 BC
Gen. D Ruler D Ku. 14 c. 825–805 BC
Gen. E Ruler E Ku. 11 c. 805–795 BC
Gen. F Alara (Ku. 9) c. 785–765 BC
Kings of Dyn. 254
Gen. 1 Kashta (Ku. 8) c. 765–753 BC
Gen. 2 Piankhi [Piye] Ku. 17 c. 753–722 BC
Gen. 3 Shabaka [Shabaqo] Ku. 15 c. 722–707 BC
Gen. 4 Shebitku Ku. 18 c. 707–690 BC
Gen. 5 Taharqa [Taharqo] Nu. 1 690–664 BC
Gen. 6 Tanwetamani Ku. 16 664–655/53 BC (in Egypt)
Napatan Kings
Gen. 7 Atlanersa (Nu. 20) 2nd half of 7th cent. BC5
Gen. 8 Senkamanisken Nu. 3 2nd half of 7th cent. BC
Gen. 9 Anlamani Nu. 6 late 7th cent. BC
Gen. 10 Aspalta Nu. 8 early 6th cent. BC
(campain Psammetik’s II
against Nubia in 593 BC)
Gen. 11 Aramatelqo Nu. 9 2nd quarter of 6th cent. BC
Gen. 12 Malonaqen Nu. 5 1st half of 6th cent. BC
Gen. 13 Analmakheye6 [Analma'aye] Nu. 18 middle of 6th cent. BC
Gen. 14 Amani-nataki-lebte Nu. 10 2nd half of 6th cent. BC
Gen. 15 Karkamani Nu. 7 2nd half of 6th cent. BC
Gen. 16 Amaniastabarqo Nu. 2 late 6th cent. BC
(continued on next page)

1
Questionable rulers in parentheses; alternative or conventional readings of names
in brackets; completely hypothetical rulers excluded, e.g., Pisekara, who has been thought
to have been king solely because he fathered a presumed king.
2
Ascribed burials in parentheses, controversial burials excluded.
3
Dates for Generations A–E after Kendall, Meroitica 15, 97.
4
Calculations based on conventional dates, antedated using the Tang-i Var inscription.
5
Napatan and Meroitic dates—except for the few fixed points—according to gen-
erally accepted hypothetical reign lengths.
6
After Priese, Sprachmaterial, 105–106 (95).
tables for kushite rulers 497

Table (cont.)

Generation Name of Ruler Burial Dates

Gen. 17 Sikhespiqo7 [Si'aspiqo] Nu. 4 early 5th cent. BC


Gen. 18 Nasakhma Nu. 19 1st half of 5th cent. BC
Gen. 19 Malowi-Amani Nu. 11 middle of 5th cent. BC
[Malowiebamani]
Gen. 20 Talakhamani Nu. 16 2nd half of 5th cent. BC
Gen. 21 Arike-Amanote Nu. 12 2nd half of 5th cent. BC
Gen. 22 Baskakeren Nu. 17 late 5th cent. BC
Gen. 23 Harsijotef Nu. 13 early 4th cent. BC
Gen. 24 King, name unknown Ku. 1 middle of 4th cent. BC
Gen. 25 Akhratan Nu. 14 2nd half of 4th cent. BC
Gen. 26 Amanibakhi ? 2nd half of 4th cent. BC
Gen. 27 Nastasen Nu. 15 last 3rd of 4th cent. BC
Gen. 28–32 Neo-Ramesside Rulers late 4th cent. to 2nd third
of 3rd cent. BC
Aktisanes [Gtsn] (Bar. 11)
Ary(amani) (Bar. 14)
Kash . . . amani (Bar. 15)
Arike-Pi(ankhi)-qo ?
Sabrakamani ?
Meroitic Kings and
ruling Queens8
Gen. 33 *Arkamanis9 Beg. S. 6 2nd quarter of 3rd cent. BC
[Ergamenes I.,
Arqamani-qo,
Arakakamani]
Gen. 34 *Amanisaraw10 [Amanislo] Beg. S. 5
Gen. 35 *Amani-tekha Beg. N. 4
Gen. 36 *Arnekhamani (Beg. N. 53) last 3rd of 3rd cent. BC
Gen. 37 *Arqamani Beg. N. 7 end of 3rd cent. BC
[Ergamenes II]
Gen. 38 *Tabirqo [Adikhalamani ?] Beg. N. 9 1st third of 2nd cent. BC
Gen. 39 King, name unknown Beg. N. 10
Gen. 40 King, name unknown Beg. N. 8
Gen. 41 Königin Sanakadakhete (Beg. N. 11) 2nd half of 2nd cent. BC
Gen. 41 King, name unknown Beg. N. 12
Gen. 43 *Naqyrinsan (Beg. N. 13)
Gen. 44 Taneyidamani (Beg. N. 20) early 1st cent. BC
Gen. 45 King, name unknown Bar. 2
Gen. 46 Queen, name unknown Bar. 4
Gen. 47 Queen Nawidemaka Bar. 6
(continued on next page)

7
After Hintze, Fs Dunham, 93.
8
Asterisks designate Meroitic kings whose names are known only in Egyptian hiero-
glyphic transcriptions.
9
Improved reading after Hallof, International Conferenc for Meroitic Studies, 2004, Paris, in press.
10
According to the improved reading Amanisaraw (see note 9), the name could pos-
sibly be understood as Meroitic Amani-Sedewa or Amani-Salawa.
498 tables for kushite rulers

Table (cont.)

Generation Name of Ruler Burial Dates

Gen. 48 Amanikhabale (Beg. N. 2)


Gen. 49 Teriteqase (Beg. N. 14) last 3rd of 1st cent. BC
Gen. 50 Queen Amanirenase (Beg. N. 21) late 1st cent. BC
Gen. 51 Queen Amanisakheto Beg. N. 6
Gen. 52 Natakamani Beg. N. 22 2nd half of 1st cent. AD
Gen. 53 (Sorakarora) ?
Gen. ? (Amanitaraqide) (Beg. N. 16)
Gen. ? Amanikhareqerema ? end of 1st cent. AD11
Gen. ? Amanitenamomide Beg. N. 17 2nd half/end of 1st cent. AD
Gen. ? Queen *Amanikhatashan Beg. N. 18 to middle of 2nd cent. AD
Gen. ? Tarekeniwala Beg. N. 19
Gen. ? Ariteneyesebokhe Beg. N. 3412
Gen. ? Takideamani (Beg. N. 29)
Gen. ? (Arayesebokhe) (Beg. N. 36)
Gen. ? Teqorideamani Beg. N. 28 accession 248/249 AD
Gen. ? (Tamelordeamani) ?
Gen. ? Yesebokheamani ? end of 3rd cent. AD
Gen. ? Queen, name unknown Beg. N. 26
Gen. ? Queen, name unknown Beg. N. 25 End of kingdom, about
330–370 AD
Known rulers not
sequenced
Gen. ? Karaki13 ?
Gen. ? ”sp-'n¢-n-Jmn stp.n R' 14 ?
Gen. ? Aqrakamani15 ?
Burials in Beg. N.
not ascribed or
sequenced in table
Gen. ? Beg. N. 41
Gen. ? Beg. N. 30
Gen. ? Beg. N. 38
Gen. ? Beg. N. 51
Gen. ? Beg. N. 24
Gen. ? Beg. N. 27

11
According to Wenig, Meroitica 15, 681, he should be dated close to Taneyidamani,
ª
L is
since nswt-bjt occurs in the king’s title. However, cf. Naqa, Inschr. 18. 21, where l
used as title (Török, FHN III, 937–38).
12
This formally disputed location of burial is now ascertained (Hinkel, Suppl. BIFAO
81 (Bulletin du Centenaire), 1981, 379–88.
13
A king Karaki to be listed before Sabrakamani is mentioned in Beg. S. 10. See
Macadam, Kawa I, 74 (Inscr. XIII); Hallof, International Conference for Meroitic Studies,
2004, Paris, in press.
14
Early Ptolemaic; see FHN II, 571 f. (118. 119); for the epithet stp.n R' see Jansen-
Winkeln, BSEG 23, 1999, 51–61, esp. 54.
15
FHN II, 686 (161): about 29–25 BC; Burckhardt, Meroitica 8, 1985, 76 and note
12: late 1st to early 2nd cent. AD.
ABBREVIATIONS AND BIBLIOGRAPHY
LIST OF ABBREVIATIONS

Cf. also the abbreviations in Lexikon der Ägyptologie VII (1992), XIII–XXXVIII

AD Anno Domini
BC Before Christ
BP Before present
Dak. Dakke
FIP First Intermediate Period
Fs Festschrift
HP High Priest
HPA High Priest of Amun
LD Lunar Day
LE Lower Egypt
LM Lunar month
LN Lower Nubia
LP Late Period
MK Middle Kingdom
NK New Kingdom
OK Old Kingdom
Ph. Philae
SIP Second Intermediate Period
TC Turin Canon
TIP Third Intermediate Period
TL Thermoluminescence
TT Theban Tomb
UE Upper Egypt
WB Wörterbuch

Periodicals and Series

AAR African Archaeological Review


ÄA Ägyptologische Abhandlungen
ÄAT Ägypten und Altes Testament
ÄgFo Ägyptologische Forschungen
AfO Archiv für Orientforschung
AH Aegyptiaca Helvetica
AHw Akkadisches Handwörterbuch
Ä&L Ägypten und Levante
AnBib Analecta Biblica
ANM Archéologie du Nil Moyen
ANRW Aufstieg und Niedergang des Römischen Reiches, W. Haase, ed.
AJA American Journal of Archaeology
AJSL American Journal of Semitic Languages and Literatures
AoF Altorientalische Forschungen
APAW Abhandlungen der Preussischen Akademie der Wissenschaften
APF Archiv für Papyrusforschung und verwandte Gebiete
ASAE Annales du Service des Antiquités de l’Égypte
502 abbreviations and bibliography

AV Archäologische Veröffentlichungen des DAI


BÄBA Beiträge zur ägyptischen Bauforschung und Altertumskunde
BAR Breasted, Ancient Records
BCE Bulletin de liaison du groupe international d’étude de la céramique égyptienne
BdE Bibliothèque d’Études
BIFAO Bulletin de l’Institut français d’archéologie orientale
BiAe Bibliotheca Aegyptiaca
BiOr Bibliotheca orientalis
BS Bollingen Series
BMFA Bulletin of the Museum of Fine Art
BSFFT Bulletin de la Société Française des Fouilles de Tanis
BzS Beiträge zur Sudanforschung
CAA Corpus Antiquitatum Aegyptiacarum
CAH The Cambridge Ancient History
CdE Chronique d’Égypte
CG Catalogue Général des Antiquités Égyptiennes du Musée du Caire
CRAIBL Comptes rendus de l’Académie des Inscriptions et Belles-lettres
CRIPEL Cahiers de Recherches de l’Institut de Papyrologie et d’Égyptologie de
Lille
CSSM Malinine, Catalogue des stèles du Sérapeum
DE Discussions in Egyptology
EAT Neugebauer & Parker, Egyptian Astronomical Texts
FHN Eide, Fontes Historiae Nubiorum
Gifts Friedman, Gifts
GLR Gauthier, Livre des rois
GM Göttinger Miszellen
GOF Göttinger Orientforschungen
GRL Geophysical Research Letters
HAS Harvard African Studies
HPBM Hieratic Papyri in the British Museum
HESPOK Smith, History
HO Cernÿ, Hieratic Ostraca
JACF Journal of the Ancient Chronology Forum
JAnAr Journal of Anthropological Archaeology
JARCE Journal of the American Research Center in Egypt
JAS Journal of Archaeological Science
JCS Journal of Cuneiform Studies
JEA Journal of Egyptian Archaeology
JEOL Jaarbericht van het Vooraziatisch-Egyptisch Genootschaap Ex Oriente Lux
JHA Journal for the History of Astronomy
JNES Journal of Near Eastern Studies
JRAI Journal of the Royal Anthropological Institute
JSSEA Journal of the Society for the Study of Egyptian Antiquities
Kawa Macadam, Kawa
KRI Kitchen, Ramesside Inscriptions
LD Lepsius, Denkmäler
LD Text Lepsius, Denkmäler Text
LEM Caminos, Late Egyptian Miscellanies
LRL Wente, Late Ramesside Letters
MDAIK Mitteilungen des Deutschen Archäologischen Instituts Abteilung Kairo
MAe Monumenta Aegyptiaca
MHC Medinet Habu Calendar
MIO Mitteilungen des Instituts für Orientforschung
abbreviations and bibliography 503

MNL Meroitic Newsletter


MonAeg Monumenta aegyptiaca
NABU Nouvelles assyriologiques brèves et utiles
NAWG Nachrichten der Akademie der Wissenschaften in Göttingen
OLA Orientalia Lovaniensia Analecta
OLP Orientalia Lovaniensia Periodica
OLZ Orientalistische Literatur-Zeitung
OMRO Oudheidkundige mededelingen uit het Rijksmuseum van Oudheden
te Leiden
OrAnt Oriens Antiquus
PÄ Probleme der Ägyptologie
PM Porter & Moss, Topographical Bibliography
PPS Proceedings of the Prehistoric Society
PSBA Proceedings of the Society of Biblical Archaeology
QuF Quellen und Forschungen zur Zeitbestimmung der ägyptischen
Geschichte
RAD Gardiner, Ramesside Administrative Documents
RCK Dunham, Royal Cemeteries of Kush
RE Pauly-Wissowa, Real-Encyclopädie der classischen Altertumswissenschaften
RlA Reallexikon der Assyriologie
RT Petrie, Royal Tombs
RdE Revue d’Égyptologie
RecTrav Recueil des travaux relatifs à la philologie et à l’archéologie égypti-
enne et assyriennes
Reliefs The Bubastite Portal
REM Repertoire d’Epigraphie Méroitique, J. Leclant, ed.
RITA Kitchen, RITA
RSO Rivista degli studi orientali
SAe Studia Aegyptiaca
SAK Studien zur altägyptischen Kultur
SARS The Sudan Archaeological Research Society
SCCNH Studies on the Civilization and Culture of Nuzi and the Hurrians
SCIEM The Synchronisation of Civilizations in the Eastern Mediterranean in
the Second Millennium B.C.
SIM Serapeum Inventaire Mariette Louvre
SNR Sudan Notes and Records
Studies Kakosy Luft, ed., Studies Kakosy
TAVO Tübinger Atlas des Vorderen Orients
TUAT Texte aus der Umwelt des Alten Testaments
UGAÄ Untersuchungen zur Geschichte und Altertumskunde Ägyptens
Urk. Urkunden des ägyptischen Altertums
VA Varia Aegyptiaca
WdO Welt des Orients
WZKM Wiener Zeitschrift für die Kunde des Morgenlandes
ZA Zeitschrift für Assyriologie
ZÄS Zeitschrift für Ägyptische Sprache und Altertumskunde
ZDPV Zeitschrift des Deutschen Palästina-Vereins
ZPE Zeitschrift für Papyrologie und Epigraphik
504 abbreviations and bibliography

Abbreviated Titles of Books and Articles

Allam, Shafik, Hieratische Ostraka und Papyri aus der Ramessidenzeit (Tübingen, 1973)
Åström, Paul, High, Middle or Low?: Acts of an International Colloquium on Absolute Chronology
held at the University of Gothenburg, 20th–22nd August 1987, vol. 1 (Gothenburg, 1987)
Baines, John & Málek, Jaromir, Atlas of ancient Egypt (Oxford, 1980)
Baud, Michel, “Ménès, la mémoire monarchique et la chronologie du IIIe millénaire”,
Archéo-Nil 9 (1999), 103–141
Baud, Michel & Dobrev, Vassili, “De nouvelles annales de l’Ancien Empire égyptien.
Une ‘Pierre de Palerme’ pour la VIe dynastie”, BIFAO 95 (1995), 23–63
Beckerath, Jürgen von, Untersuchungen zur politischen Geschichte der Zweiten Zwischenzeit in
Ägypten (Glückstadt: ÄgFo 23, 1964)
——, Chronologie des Neuen Reiches (Hildesheim: HÄB 39, 1994)
——, Chronologie des pharaonischen Ägypten. Die Zeitbestimmung der ägyptischen Geschichte von
der Vorzeit bis 332 v.Chr. (Mainz: MÄS 46, 1997)
——, Handbuch der ägyptischen Königsnamen2 (Mainz: MÄS 49, 1999)
Belmonte, Juán A., “Some open Questions on the Egyptian Calendar: An Astronomer’s
View”, Trabajos de Egiptología 2 (2003), 7–56
Bierbrier, Morris L. The Late New Kingdom in Egypt (c. 1300–664 BC). A Genealogical and
Chronological Investigation (Warminster, 1975)
——, The Tomb-builders of the Pharaohs (London, 1982)
Bietak, Manfred, ed., The Synchronisation of Civilisations in the Eastern Mediterranean in the
Second Millennium B.C. Proceedings of an International Symposium at Schloß Haindorf, 15th–17th
of November 1996, and at the Austrian Academy, Vienna, 11th–12th of May 1998 (Vienna:
ÖAW, Denkschriften der Gesamtakademie, vol. 19, 2000)
——, Proceedings of the SCIEM 2000–EuroConference, Haindorf May 2–7, 2001 (Vienna:
ÖAW, Denkschriften der Gesamtakademie, vol. 29, 2003)
——, Proceedings of the SCIEM 2000–2nd EuroConference, Vienna May 28 to July 1, 2003,
in press
Blackman, Aylman, The Rock Tombs of Meir III (London, 1915)
Boehmer, Rainer M., “Gebel-el-Arak- und Gebel-el-Tarif Griff ”, MDAIK 47 (1991),
51–60.
Boehmer, Rainer M., G. Dreyer & B. Kromer, “Einige frühzeitliche 14C-Datierungen
aus Abydos und Uruk”, MDAIK 49 (1993), 63–68
Borchardt, Ludwig, Die Annalen und die zeitliche Festlegung des Alten Reiches der ägyptischen
Geschichte (Berlin: QuF 1, 1917)
——, Die Mittel zur zeitlichen Festlegung von Punkten der ägyptischen Geschichte und ihre Anwendung
(Kairo: QuF 2, 1935)
——, Das Grabdenkmal des Königs Ne-user-Re (Leipzig: WVDOG 7, 1907)
Braun, E. & E. C. M. van den Brink, “Some Comments on the Late EB I Sequence
of Canaan”, Ä&L 7 (1998), 71–94
Brink, E. C. M. van den & T. E. Levy, T. E., eds., Egypt and the Levant. Interrelations
from the 4th through the Early 3rd Millennium BCE (London-New York, 2002)
The Bubastite Portal by the Epigraphic Survey. Reliefs and Inscriptions at Karnak III (Chicago:
OIP LXXIV, 1954)
Caminos, Ricardo A., The Chronicle of Prince Osorkon (Rome: AnOr 37, 1958)
Cernÿ, Jaroslav & Gardiner, Alan H., Hieratic Ostraca (Oxford, 1957)
Clagett, Marshall, Ancient Egyptian Science II. Calendars, Clocks, and Astronomy (Philadelphia,
1955)
Clère, Jacques J. & Vandier, Jacques, Textes de la première période intermédiaire et de la XI ème
dynastie (Brussels: BiAe X, 1982)
Couyat, Jules & Montet, Pierre, Les inscriptions hiéroglyphiques et hiératiques du Ouâdi Hammâmât
(Cairo: MIFAO 34, 1912)
Daressy, Georges, Ostraca. Catalogue général des antiquités égyptiennes du Musée du Caire Nos.
25001–25385 (Cairo, 1901)
abbreviations and bibliography 505

Depuydt, Leo, “On the Consistency of the Wandering Year as Backbone of Egyptian
Chronology”, JARCE 32 (1995), 43–58
——, Civil Calendar and Lunar Calendar in Ancient Egypt (Leuven: OLA 77, 1997)
Dreyer, Günther et al., “Umm el-Qaab: Nachuntersuchungen im frühzeitlichen Königs-
friedhof 9./10. Vorbericht”, MDAIK 54 (1996), 77–167
Dunham, Dows, The Barkal temples, excavated by G. A. Reisner (Boston, 1970)
——, The Royal Cemeteries of Kush, vol. 1–5 (Cambridge, Mass., 1950–1963)
Edel, Elmar, Altägyptische Grammatik I. II (Rome, 1955/1964)
Eide, Tormod, Hägg, T., Pierce, R. H. & Török, L., eds., Fontes Historiae Nubiorum:
Textual Sources for the History of the Middle Nile Region between the Eighth Century BC and
the Sixth Century AD. Vol. I: From the Eighth to the Mid-Fifth Century BC (Bergen, 1994)
Emery, Walter B., Great Tombs of the First Dynasty II, III (Oxford, 1954, 1958)
Eyre, Chris J., ed., Proceedings of the Seventh International Congress of Egyptologists (Leuven:
OLA 82, 1998)
Firth, Cecil M. & Quibell, James E., Excavations at Saqqara: The Step Pyramid I (Cairo,
1935)
Fischer, Henry G., Dendera in the Third Millennium (Locust Valley, 1968)
Franke, Detlef, Personendaten aus dem Mittleren Reich (20.–16. Jahrhundert v. Chr.) (Wiesbaden,
1984)
——, Das Heiligtum des Heqaib auf Elephantine. Geschichte eines Provinzheiligtums im Mittleren
Reich (Heidelberg: SAGA 9, 1994)
——, Zur Chronologie des Mittleren Reiches, Teil I, Or 57 (1988), 113–138; Teil II:
Die sogenannte ‘Zweite Zwischenzeit’ Altägyptens, Or 57 (1988), 245–274
Friedman, Renée F., ed., Egypt and Nubia: Gifts of the Desert (London, 2002)
Gabolde, Marc, D’Akhenaton à Toutânkhamon (Paris, 1998)
Gardiner, Alan H., “Regnal Years and Civil Calendar in Pharaonic Egypt”, JEA 31
(1945), 11–28
——, The Royal Canon of Turin (Oxford, 1959)
——, Egypt of the Pharaohs (Oxford, 1961)
Gasche, Hermann, Armstrong, J. A., Cole, S. W. & Gurzadyan, V. G., Dating the Fall
of Babylon: a reappraisal of second-millenium chronology (Ghent, 1998)
Gauthier, Henri, “Quatre nouveaux fragments de la pierre de Palerme”, in: G. Maspero,
Le Musée égyptien III (Cairo, 1915), 29–53
Gestermann, Louise, Kontinuität und Wandel in Politik und Verwaltung des frühen Mittleren
Reiches in Ägypten (Wiesbaden: GOF IV.18, 1987)
Ginzel, Friedrich Karl, Handbuch der mathematischen und technischen Chronologie, vol. I–III
(Leipzig, 1906–14)
Goedicke, Hans, Königliche Dokumente aus dem Alten Reich (Wiesbaden: ÄA 14, 1976)
Görsdorf, J., G. Dreyer & U. Hartung, “14C Dating Results of the Archaic Royal
Necropolis”, MDAIK 54 (1998), 169–175
Goldstine, Herman H., New and Full Moons 1001 BC to AD 1651 (Philadelphia: MAPS
94, 1973)
Harris, James E. & Wente, Edward F., An X-Ray Atlas of the Royal Mummies (Chicago,
1980)
Helck, Wolfgang, Untersuchungen zu Manetho und den ägyptischen Königslisten (Berlin: UGAÄ
18, 1956)
——, Untersuchungen zur Thinitenzeit (Wiesbaden: ÄA 45, 1987)
——, Die datierten und datierbaren Ostraka, Papyri und Graffiti von Deir el-Medineh, A. Schlott,
ed. (Wiesbaden: ÄA 63, 2002)
——, “Gedanken zum Mord an König Teti”, Essays in Egyptology in honor of Hans Goedicke,
Bryan, B. & Lorton, D., eds., (San Antonio, Texas, 1994)
Hintze, Fritz, Studien zur meroitischen Chronologie und zu den Opfertafeln aus den Pyramiden von
Meroe, Abhandlungen der deutschen Akademie der Wissenschaften Berlin, Klasse für
Sprachen, Literatur und Kunst Nr. 2 (Berlin, 1959)
—— & Reineke, Walter F., Felsinschriften aus dem sudanesischen Nubien (Berlin, 1989)
506 abbreviations and bibliography

——, Ägypten und Kusch: Fritz Hintze zum 60. Geburtstag gewidmet, Endesfelder, E., ed.
(Berlin, 1977)
Hofmann, Inge, Beiträge zur meroitischen Chronologie (St. Augustin, 1978)
Hornung, Erik, Untersuchungen zur Chronologie und Geschichte des Neuen Reiches (Wiesbaden:
ÄA 11, 1964)
—— & Staehelin, Elisabeth, Studien zum Sedfest (Basel and Geneva: AH 1, 1974)
—— & Staehelin, Elisabeth, Skarabäen und andere Siegelamulette aus Basler Sammlungen (Mainz,
1976).
Israelit-Groll, Sarah, ed., Studies in Egyptology: Presented to Miriam Lichtheim I–II ( Jerusalem,
1990)
Jansen, Jacques J., Village Varia. Ten Studies on the History and Administration of Deir el-
Medina (Leiden: EU 11, 1997)
Jacquet-Gordon, Helène, The Graffiti on the Khonsu Temple Roof at Karnak (OIP 123:
Chicago, 2003)
Jéquier, Gustave, Fouilles à Saqqarah. Deux pyramides du Moyen Empire (Cairo, 1933)
Junker, Hermann, Bericht über die Grabungen der Kaiserlichen Akadamie der Wissenschaften in
Wien auf dem Friedhof in Turah, Winter 1909–1910 (Vienna: DAWW 56, 1912)
——, Giza I–XII (Wien & Leipzig, 1929–1955)
Kahl, Jochem, Das System der ägyptischen Hieroglyphenschrift in der 0.–3. Dynastie (Wiesbaden:
GOF IV/29, 1994)
Kaplony, Peter, Die Inschriften der ägyptischen Frühzeit I–III (Wiesbaden: ÄA 8, 1963)
——, Steingefässe mit Inschriften der Frühzeit und des Alten Reiche (Brussels: MonAeg 1, 1968)
Kienitz, Friedrich Karl, Die politische Geschichte Ägyptens vom 7. bis zum 4. Jahrhundert vor
der Zeitwende (Berlin, 1953)
Kitchen, Kenneth A., The Third Intermediate Period in Egypt (1100–650 BC) (Warminster,
1973, 19862, 19953)
——, Ramesside Inscriptions, Translated and Annotated. Notes & Comments II (Oxford, 1999)
——, “Regnal and Genealogical data of Ancient Egypt (Absolute Chronology I). The
Historical Chronology of Ancient Egypt, A Current Assessment”, in: M. Bietak, ed.,
SCIEM Haindorf 1996/98, 39–52
Krauss, Rolf, Sothis- und Monddaten. Studien zur astronomischen und technischen Chronologie
Altägyptens (Hildesheim: HÄB 20, 1985)
Kruchten, Jean-Marie, Les annales des prêtres de Karnak (XXI–XXIII mes dynasties) et autres
textes contemporains relatifs à l’initiation des prêtres d’Amon (Leuven: OLA 32, 1989)
Lacau, Pierre & Lauer, Jean-Philippe, La pyramide à degrès IV. Inscriptions gravées sur les
vases (Cairo, 1959)
——, La pyramide à degrès V. Inscriptions à l’encre sur les vases (Cairo, 1965)
Leahy, Anthony, ed., Libya and Egypt c. 1300–750 BC (London, 1990)
Leclant, Jean, Enquêtes sur les sacerdotes et les sanctuaires égyptiens à l’époque dite ‘Éthiopienne’
(XXV e dynastie) (Cairo: BdE 17, 1954)
——, ed., Repertoire d’Epigraphie Méroitique (Paris, 1976; 1982)
Leitz, Christian, Studien zur Ägyptischen Astronomie (Wiesbaden: ÄA 49, 1991)
Luft, Ulrich, Die chronologische Fixierung des ägyptischen Mittleren Reiches nach dem Tempelarchiv
von Illahun (Vienna, 1992)
——, ed., The intellectual heritage of Egypt. Studies presented to L. Kakosy (Budapest, 1992)
Macadam, Miles F., The Temples of Kawa I–II (Oxford, 1949–1955)
Malinine, M., Posener, G., Vercouttter, J., Catalogue des stèles du Sérapeum de Memphis,
Vol. 1–2 (Paris, 1968)
Manuelian, Peter Der, ed., Studies in Honor of William Kelly Simpson I. II (Boston, 1996)
Marciniak, Marek, Deir el-Bahari I: Les inscriptions hiératiques (Warsaw, 1974)
Maystre, Charles, Les grands prêtres de Ptah de Memphis (Freiburg, 1992)
Meeks, Dimitri, “Les donations aux temples dans l’Egypte du Ier millénaire avant
J.-C.,” in: Lipi…ski, Edward, ed., State and Temple Economy in the Ancient Near East II
(Louvain: OLA 6, 1979), 605–687
abbreviations and bibliography 507

Meyer, Eduard, Ägyptische Chronologie (Berlin, 1904)


——, “Nachträge zur aegyptischen Chronologie”, in: APAW 1907 (Berlin, 1908), 1–46
Montet, Pierre, Les constructions et le tombeau de Psousennès à Tanis. Le nécropole royale de
Tanis II (Paris, 1951)
Morkot, Robert, “Kingship and Kinship in the Empire of Kush”, in: S. Wenig, ed.,
Studien zum antiken Sudan (Akten der 7. Internationalen Tagung für meroitische Forschungen,
14–19 Sep. 1992 in Gosen/bei Berlin), 179–229
Murnane, William J., Texts from the Amarna Period in Egypt (Atlanta, 1995)
Murnane, William J. & Siclen III, Charles C. van, The Boundary Stelae of Akhenaten
(London & New York: Studies in Egyptology, 1993)
Neugebauer, Otto & Parker, Richard A., Egyptian Astronomical Texts I–III (Providence,
Rhode Island, 1960–1969)
Parker, Richard A., The Calendars of Ancient Egypt (Chicago: SAOC 26, 1950)
——, A Saite Oracle Papyrus from Thebes in the Brooklyn Museum (Providence, 1962)
——, & Dubberstein, Waldo H., Babylonian Chronology 626 BC–AD 75 (Providence: BUS
19, 1956)
Peden, A. J., The Graffiti of Pharaonic Egypt (Leiden: PÄ 17, 2001)
Petrie, William M. Flinders, The Royal Tombs of the First Dynasty I (London: MEEF 18,
1900)
——, The Royal Tombs of the Earliest Dynasties II (London: MEEF 21, 1901)
——, Diospolis Parva. The Cemeteries of Abadiyeh and Hu. 1898–1899 (London, 1901)
——, Abydos I (London, 1902)
——, Tarkhan and Memphis V (London, 1913)
Porada, E., D. P. Hansen, S. Durham & S. H. Babcock, “Mesopotamia,” in: R. W.
Ehrich, ed., Chronologies in Old World Archaeology (Chicago, 1992), I: 77–121, II: 90–124,
Addendum
Posener-Kriéger, Paule & de Cenival, Jean-Louis, The Abusir Papyri (London: HPBM
5, 1968)
Posener-Kriéger, Paule, Les archives du temple funéraire de Néferirkarê-Kakaï (Cairo: BdE 65,
1976)
——, Graffiti in the revetment blocks of the Pyramid, in: Ali el-Khouli, Maidum (ACE
Reports 3, 1991)
Quibell, James E., Hierakonpolis I (London, 1900)
Redford, Donald B., Pharaonic King-Lists, Annals and Day-Books (Mississauga: SSEAP 4,
1986)
Reisner, George A., A History of the Giza Necropolis I (Cambridge, Mass., 1942)
Ryholt, Kim S. B., The Political Situation in Egypt During the Second Intermediate Period,
c. 1800–1550 BC (Copenhagen: CNIP 20, 1997)
Sachs, Abraham & Hunger, Hermann, Astronomical Diaries and Related Texts from Babylon,
vol. 1–3 (Vienna, 1988–1996)
Schäfer, Heinrich, Ein Bruchstück altägyptischer Annalen (Berlin: APAW philos.-hist. Kl.,
1902)
Schenkel, Wolfgang, Frühmittelägyptische Studien (Bonn: BOS NS 13, 1962)
——, Memphis, Herakleopolis, Theben (Wiesbaden: ÄA 12, 1965)
Schmitz, Franz-Jürgen, Amenophis I. Versuch einer Darstellung der Regierungszeit eines ägypti-
schen Herrschers der frühen 18. Dynastie (Hildesheim: HÄB 6, 1978)
Schneider, Thomas, Lexikon der Pharaonen (Munich, 1996)
——, Ausländer in Ägypten während des Mittleren Reiches und der Hyksoszeit Teil 1: Die aus-
ländischen Könige (Wiesbaden: ÄAT 42, 1998)
Schott, Siegfried, Altägyptische Festdaten, Akademie der Wissenschaften und der Literatur
Mainz. Abhandlungen der Geistes- und Sozialwissenschaftlichen Klasse, No. 10 (1950),
1–70
Simpson, William K., Studies in ancient Egypt, the Aegean, and the Sudan: essays in honor of
Dows Dunham on the occasion of his 90th birthday, June, 1980 (Boston, 1981)
508 abbreviations and bibliography

Smith, William S., “Inscriptional Evidence for the History of the Fourth Dynasty”,
JNES 11 (1952), 113–128
——, A History of Egyptian Sculpture and Painting in the Old Kingdom1 (London, 1946)
Spalinger, Anthony J., Three Studies on Egyptian Feasts and their Chronological Implications
(Baltimore, 1992)
——, ed., Revolutions in Time (San Antonio: VAS 7, 1994)
——, “Dated Texts of the Old Kingdom”, SAK 21 (1994), 275–319
——, The Private Feast lists of Ancient Egypt (Wiesbaden: ÄA 57, 1996)
——, “Egyptian Festival Dating and the Moon”, in: Steele, John M. & Imhausen,
Anette, Under one Sky. Astronomy and mathematics in the Ancient Near East (Münster: AOAR
297, 2002), 379–403
Spencer, Alan J., Early Dynastic Objects (London, 1980)
Stationen. Beiträge zur Kulturgeschichte Ägyptens, Rainer Stadelmann gewidmet, Guksch, H. &
Polz, D., eds. (Mainz, 1998)
Stephenson, F. R., Historical Eclipses and Earth’s Rotation (Cambridge, 1997)
Swelim, Nabil, Some Problems on the History of the Third Dynasty (Alexandria: AHS 7, 1983)
Toomer, Gerald J., Ptolemy’s Almagest. Transl. and annotated by G. J. Toomer (Princeton,
1998)
Török, Laszlo, The Kingdom of Kush: Handbook of the Napatan-Meroitic civilization (Leiden,
1997)
Valbelle, Dominique, Les Ouvriers de la tombe: Deir el-Médineh à l’époque ramesside (Cairo:
BdE 96, 1985)
Verner, Miroslav, Forgotten Pharaohs, Lost Pyramids. Abusir (Prague, 1994)
——, “Archaeological Remarks on the 4th and 5th Dynasty Chronology”, Archiv Orientální
69 (2001), 363–418
Waddell, William G., Manetho (Cambridge, Mass.: The Loeb Classical Library 350,
1940)
Welsby, Derek A., ed., Recent research in Kushite history and archaeology: Proceedings of the 8th
International Conference for Meroitic Studies (London: OP 131, 1999)
Wente, Edward F., Late Ramesside Letters (SAOC 33: Chicago 1967)
——, & Siclen III, Charles C. van, “A Chronology of the New Kingdom,” in: Studies
in Honor of George R. Hughes (Chicago: SAOC 39, 1976)
Wiedemann, Alfred, Ägyptische Geschichte (Gotha, 1884)
Wildung, Dietrich, Die Rolle ägyptischer Könige im Bewusstsein ihrer Nachwelt (Berlin: MÄS
17, 1969)
Wilkinson, Toby A. H., Royal Annals of Ancient Egypt (London, 2000)
——, Early Dynastic Egypt (London, 1999)
Ziegler, Christiane, Catalogue des stèles, peintures et reliefs égyptiens de l’Ancien Empire et de la
Première Période Intermediaire (Paris, 1990)
INDEX OF PERSONAL NAMES

Achoris 270, 280 Amenhotep IV, see Akhenaten


Adad-nerari I 321–322, 477–478 Ameni Qemau 177
Adikhalamani 297 Amenirdis I 253–254, 257–258, 284
'Adj-ib 23, 24, 90, 96–98, 100–101, Amenmesses 42, 197, 210, 212–213,
107–112 416–417, 476
'Aha 4, 23, 39, 45, 89–91, 97, 101, Amenophis, see Amenhotep
486 Ammisaduqa 304
'Ahmose 34, 41, 191, 195–196, Amyrtaios 270
198–199, 423, 440, 479, 481 'Ankhkheprure' 207–208, 477
'Ahmose Penekheb 41 'Ankhetkheprure' 36, 207–208
Akhenaten 2, 10, 206, 313–315, 317, 'Ankhew, vizier 422–423
335, 375, 419, 477–478 'Ankhtifi, nomarch 1, 165–166
'Akheperre', see Osochor Anlamani 292–293
Akhet'a, official 120 Apries 15, 268, 282
Aktisanes 295 Aqrakamani 299
Alara 284, 289 Arayesebokhe 301
Alexander the Great 4, 35, 265, Arike-Amanote 293
268–269, 458, 464–465 Arike-Pi(ankhi)-qo 295–296
Amanibakhi 293 Ariteneyesebokhe 302
Amanikhareqerema 301–302 Arkamaniqo, see Ergamenes
Amanikhatashan 302 Arnekhamani 295, 297
Amanirenase 298–299 Arnuwanda II 319
Amanisakheto 299–300 Arqamani 297
Amanislo 295–297 Arses 271, 464–465
Amanitaraqide 301 Artaxerxes I 266, 283, 464–465
Amani-tekha 295, 297 Artaxerxes II 266, 283, 463–465, 469
Amanitenamomide 302 Artaxerxes III Ochos 268, 271,
Amasis 267–268, 282, 390, 473 464–465
Amenemhet, astronomer 199 Ary(amani) 295
Amenemhet I 170–171, 428, 481–482 Ashshur-bel-nisheshu 313
Amenemhet II 172 Ashshur-nerari II 310
Amenemhet III 21, 46, 172–173, 341, Ashshur-uballit 313, 477–478
402, 406–407, 423–427, 480–482 Aspalta 288–289, 292–293
Amenemhet IV 173–174, 177, 480 Atlanersa 292
Amenemhet (V) 177 Aya Kheperkheprure' 208–209, 317,
Amenemhet (VII) 175, 178–179 376
Amenemhetsenbef 177 Aya Merneferre' 180–181, 196
Amenemnisut, see Nephercheres Awibre' 31, 178, 367
Amenemope 218, 224–232, 412–414,
475 Bakenchonsu, HP 476
Amenhotep, son of Hapu 205 Bay, chancellor 213
Amenophthis 218 Bebiankh 186
Amenhotep I 11, 199, 440, 479–482, Bicheris 25, 134, 485
490 Bocchoris 261–263, 408, 411, 473
Amenhotep II 10–11, 200, 203–204, Burnaburiash II 314–315, 477–478
374–375, 419–420, 478
Amenhotep III 10–11, 204–206, Cambyses 266–267, 271, 282, 294,
313–315, 317, 478 464–465
510 index of personal names

Chebron 200, 479 Iuput II 246–248


Cheops 25, 131–132, 339–341, 379, Iuwelot 239
485 Izezi, see Djedkare'

Darius I 266, 282–283, 294, 464–465 Jantin'ammu (14), 179


Darius II 266, 269, 283, 463–466
Darius III 271, 464–465 Kadashman-Enlil I 314, 477
Dedumose 187 Kadashman-Turgu I 324, 477–478
Den 23–24, 39, 90, 96–98, 100–101, Kamose 186–187, 189–192, 481
107, 109–111 Kara-indash 314
Djedefre' 25, 40, 127–128, 132, 346, Kash . . . amani 296
379, 485 Kash . . . meri-Amun 295
Djedkare' 3, 126, 139–143, 429–431, Kashta 254, 284
483 Kenherkhepeshef, scribe 42
Djedkhonsuiuefankh, HP 224–225, 228 Khababash 294
Djehuti 184–186 Kha'ba 117, 120–121
Djer 23, 25, 39, 45, 90, 97, 101, 108 Khamudi 192–195, 198–199
Djoser 21–22, 24, 45, 47, 117–120, Kha-sekhem(wy) 21, 23–24, 45, 90,
122, 346, 350 104–107, 114–115, 118, 122
Djoserti/Djoserteti 118 Khendjer 181, 422–424, 480–83
Khephren 25, 40, 133–134, 339–341,
Ergamenes I 285, 296 346, 485–486
Esarhaddon 289 Kheruef, official 205
Khety 164, 166
Gilukhepa, princess 204 Khufu, see Cheops
Kurigalzu I 314
Hammurabi 8, 179, 312
Har-nedjheritef 177 Malonaqen 292
Harsiese A 241, 247–248 Malowi-Amani 284, 294
Harsiese B 240, 248–251 Masaharta, HP 224–225, 228, 231
Harsijotef 294 Menes 3–5, 25, 29, 35
Hatshepsut 5, 10–11, 15, 33, 200–201, Menkauhor 127, 139, 484
479 Menkaure', see Mycerinus
Hattushili II 320–324, 478 Menkheperre', HP 218, 221, 224–225,
Herihor, HP 218, 221, 225–226, 228–232, 475
228–229, 475 Mentuhotep I 30, 160
Hetep-sekhemwy 21, 24, 96, 99, 102, Mentuhotep II 159–160, 162, 165, 447
106, 108, 112–113 Mentuhotep III 159–160
Haremhab 13, 41, 46, 209, 476–477 Mentuhotep IV 30, 159–161
Hosea 259 Mentuhotpe, see Mentuhotep
Huni 117, 119–123 Merenre' 21, 125, 151–157, 370, 483
Hyksos 29, 192–195, 198–199 Meret-neit 39
Merhotepre' Ini 180, 423–424
Imiramesha 423 Merneptah 15, 42, 198, 214, 476
Iny 255–256 Merykare' 1, 165
Inyotef, count 160–161 Merytaten 2, 207
Inyotef I 160–161 Merytneith, see Meret-neit
Inyotef II 160–161, 428, 447 Metjen, official 120
Inyotef III 160–161 Murshili I 305, 311–312, 316
Inyotef V 423 Murshili II 318, 319–320, 377
Inyotef Nebukheperre' 183, 187–189, Murshili III 320–323
191 Muwatalli II 320–322
Iry-Hor 88–89, 95 Mycerinus 40, 127, 134–136, 339–341,
Iuput I 247, 249–252, 473 346, 485–486, 489
index of personal names 511

Nabonassar 49, 464–465 Osorkon III 43, 241, 243, 246, 248,
Nar-mer 4, 23, 39, 90, 94–97, 101, 252–258, 284, 372–373
487 Osorkon IV 246, 260
Nastasen 293–295
Natakamani 300, 302 Pami 21, 244–245, 263–264, 408
Nawidemaka 298 Pasenhor, priest 39, 220, 235–236, 242
Nebiriau 186, 423 Payankh, HP 224–226, 228–229
Nebka 21, 23, 117–123 Peftjau'awybast 247–248, 256–257
Nebsenre' 31 Pentjini 177–178
Nebukheperre', see Inyotef N. Pepy I 21, 146, 148, 153–156, 483
Necho II 15, 268, 282, 466 Pepy II 11, 40, 145, 152–157, 483
Nectanebo I (Nectanebes) 15, 269 Per-ibsen 21, 24, 90, 102, 105–107
Nectanebo II (Nectanebos) 4, 265, Petubaste I 246–251, 408–411, 473
268–270, 284, 458 Piankhi, see Piye
Neferefre' 40, 138–142, 484, 490 Pinudjem I, HP 218, 221, 224–226,
Neferhotep I 14, 179 230–231, 475
Neferirkare' 29, 127, 138, 140–143, Pinudjem II, HP 221–222, 224–225,
145, 484 228–230
Neferkare' 165 Piye 245–247, 253–257, 261–263, 291,
Neferkauhor 157 408
Nehesi 180 Psammetichus I 268, 282, 284, 290,
Neitiqerti Siptah 156 292, 294, 377–378, 408–409, 466,
Nemtyemzaf I 40 468, 473
Nephercheres 35–36, 218–219, 224, Psammetichus II 11, 265, 268, 282,
227, 475 293, 458
Nepherites I 270, 280 Psammetichus III 268, 282
Nepherites II 270, 280 Psammuthis 270, 280
Nerikare' 31 Psinaches 218, 220–221, 224, 227–228,
Netjery-khet 21, 106, 114–115, 231, 474
117–118 Psusennes I 36, 218–219, 224,
Netjerypunisut, official 486 227–229, 237, 412, 475
Neuserre' 40, 139, 484, 486 Psusennes II 218, 220–225, 227–228,
Nibkhururiya 318–319, 377 412, 474, 489
Nimlot C 240, 242 Ptahshepses, vizier 40, 486
Nimlot D 247, 255, 257–258 Ptolemy II 34, 296
Ninurta-apil-Ekur 310, 323 Ptolemy III 48, 297, 442
Nitocris 28, 33, 156 Ptolemy IV 297
Nitocris, God’s Wife 292 Ptolemy VIII 298, 458
N(e)ofrusobk 173, 480
Nub-nefer 103–104 Qa-'a 23, 90, 96–102, 109–112, 489
Nykau-Izezi, official 146, 154–155 Qahedjet 117, 121–122
Ny-maat-Hep 106, 115
Ny-netjer 21, 24, 45, 102–104, Ra'djedef see Djedefre'
106–107, 112–114 Ra'hotep 183–184, 187–189
Ra'khaef, see Khephren
Osochor (Osorkon) 218–220, 224, 227, Ra'neb 21, 24, 102–104, 106, 112–113
229–230, 236, 242–243, 412–414, Ra'neferef, see Neferefre'
474–475 Ramesses I 210, 476
Osorkon B 234, 242–243, 248–249, Ramesses II 2, 4, 10–11, 13, 15, 26,
252 33, 36, 42, 211–212, 320, 324,
Osorkon I 43, 220, 224, 234, 236, 414–418, 476–478, 489
238–241, 264, 474 Ramesses III 10, 15, 42–43, 214–215,
Osorkon II 234, 236–244, 248, 415–417, 475–476
263–264, 474 Ramesses IV 43, 215, 417
512 index of personal names

Ramesses V 42, 215, 417 Shoshenq I 11, 35, 220, 232–233,


Ramesses VI 43, 215–216, 416–417, 236–238, 241, 255–256, 263–264,
475 350–352, 411–413, 474
Ramesses VII (Itiamun) 216, 416–417, Shoshenq II 222–223
475 Shoshenq III 235, 240–245, 248–250,
Ramesses VIII 216, 417, 475 252, 255–256, 263, 408–412, 473
Ramesses IX 43, 216–217, 417, 475 Shoshenq IIIa 244
Ramesses X 210, 214, 216–217, 417 Shoshenq IV 249–250, 252
Ramesses XI 43, 217, 226, 414, 417, Shoshenq V 219, 235–236, 242–247,
475 254, 257, 262–264, 408, 473
Rehabeam 232, 264, 350 Shoshenq VII 255–256
Rudamun 247–248, 254–257 Shuppiluliuma I 313, 316–319, 477
Shuppiluliuma II 323
Sabrakamani 295–296 Shuttarna II 313
Sahure' 127, 137–138, 346, 484, 486 Siamun 218–221, 224, 227–230,
Sanakadakhete 285, 298 412–414, 474
Sargon II 13, 259–260, 290 Siptah 33, 42, 213
Scorpion 91, 96 Smendes I 43, 218, 225, 227–228,
Sehetepibre' 29, 177–178 230, 414, 475
Sekhem-ib 105–106 Smendes II, HP 224–225, 228, 231
Sekhem-khet 21, 24, 117–119, 346 Smendes III, HP 239
Semer-khet 23, 45, 90, 97–98, Smenkhkare', see 'Ankhkheprure'
100–101, 110–112, 489 Snofru 23–25, 45–46, 119–120,
Senakhtenre' 186, 189–191 124–125, 128–131, 339–341,
Sened 104 346–347, 485–486
Senkamanisken 292–293 Sobekemsaf I 184, 186, 190
Sennacherib 260, 291 Sobekhotep I 175, 481
Senwosret I 35, 171–172, 375–376, 481 Sobekhotep II 175, 178–179, 422–423,
Senwosret II 172, 341 480
Senwosret III 9–10, 46, 121, 172–174, Sobekhotep III 180, 422–424
370, 423–428, 479–482, 490 Sobekhotep IV 179, 181, 424, 481
Seqenenre' Tao 186–187, 191 Sobekhotep V 179
“Serpent” 23–24, 39, 45, 90, 97, 108 Sobekhotep VI 179
Seth 423 Sobekhotep VII 180
Sethnakhte 42, 214, 475–476 Sobekhotep VIII 186, 372
Sety I 5, 33, 41, 210–211, 476 Sonbef 31, 481
Sety II 42, 197, 210, 212–213, 416, Sorakarora 300, 302
476
Shamshi-Adad I 307, 311–312 Taharqa 13, 258, 261, 265, 268,
Shabaka 258–262, 288, 290, 408, 473 288–291, 293, 408, 458, 468, 473
Shabataka & Shebitqo, see Shebitku Takelot E 248–251
Shalmaneser I 310–311, 321–323, 477 Takelot F 240
Shebitku 13, 258–261, 289–291, Takelot I 236, 238–239, 242, 474
372–373, 408, 473 Takelot II 242–243, 247–253, 377,
Shepenupet I 253, 257–258, 284 408–411, 473
Shepseskaf 127, 134–136, 346, 484 Takelot III 220, 252–254, 257
Shepseskare' 127–128, 134, 138, 484, Takideamani (?) 302
490 Talakhamani 284, 294
Sheshi 192 Taneyidamani 298
Shishak, see Shoshenq I Tanwetamani 284, 288–289, 291–293
Shoshenq A 230, 237 Tarekeniwala 302
Shoshenq D 239–240 Tefnakhte 246, 257, 262–263, 408, 473
Shoshenq Hekakheperre' 238 Teos/Tachos 270, 280
Shoshenq Tutkheperre' 237–238 Teqorideamani 289, 300, 302–303
index of personal names 513

Teriteqase 298–299 Ukhhotep, nomarch 40


Teti 21, 145–146, 154–156, 346, 483 Urkhi-Teshshup, see Murshili III
Thamphtis 136, 485 Userkaf 136–137, 484
Thutmose I 10, 199–200, 479 Userkare' 21, 146, 151, 156
Thutmose II 10, 198, 200, 479 Usertesen, see Senwosret
Thutmose III 4, 11, 21, 201–203, 374,
378, 419–422, 478, 489–490 Wadj, see “Serpent”
Thutmose IV 10–11, 203–204, Wegaf 175, 179, 424
313–314, 375, 478 Weneg 102–104, 106
Tiglathpileser I 310 Weni, official 154–155, 370, 483
Tjuloy, priest 26, 33 Wenis 29, 142–143, 429–431, 483
Tukulti-Ninurta I 322–323, 477 Wennefer, HP 41
Tushratta 313, 477
Tut'ankhaten/amun 41, 207–208, Xerxes 265–266, 283, 294, 463–465
318–319, 376–377, 477
Tutkhaliya I 316–317 Yamani 259–260, 290–291
Tutkhaliya IV 322–323 Yesebokheamani 302–303
Twosre 42, 210, 212–214, 415–417,
475–476 Zanakht 117, 120, 122
INDEX OF PLACE NAMES

Abydos 26, 57, 62, 74, 83, 88, 90, Hermopolis 247, 253, 255–257
95–96, 105, 108, 114, 172, 222, 237, Hierakonpolis 78, 81, 83, 95, 105
359, 486–487
Alexandria 437, 440, 442–443 Illahun 46, 173, 341, 400–402, 446, 448
Amarna 13, 205–209, 335–338, 375,
419, 477 Jerusalem 13, 232, 434, 474
Armant 59, 64–65, 67–70, 75–77,
80–84, 93, 201 Karnak 160–161, 189, 198, 200–203,
Asyut 1, 376, 452 209, 219, 232, 234, 236, 359, 370,
Aswan 71, 200, 203–204, 211, 299, 372–373, 384–385, 387, 409,
369–372 419–421, 479
Athens 270, 395 Kawa 261, 293, 299
Avaris 34, 195, 199, 481 Konosso 172, 203

Babylon 8, 13, 304–305, 310, 312, Leontopolis 246, 248


436–437, 462, 477 Luxor 371–373
Ballas 61, 63
Barkal, see Gebel Barkal Maadi 56, 93
Begerawiya 285–287, 295, 298, Maidum 46, 119, 128–131, 339–347,
300–303 486
Bubastis 234–235, 246, 250, 263 Malqata 205, 375
Buhen 204, 206, 210, 213 Matmar 58, 74–77
Buto 56, 79, 93, 96, 246 Medamud 175, 184, 423
Megiddo 202, 421, 479
Coptos 96, 184, 187–188, 452 Memphis 5, 19, 31, 203, 205, 210,
234, 239, 244, 246–247, 250, 252,
Dahshur 121, 129–130, 148, 172, 208, 260, 263, 408, 432, 434–435, 437,
341, 486 440–444, 449, 473
Deir el-Bahri 181, 201, 214, 414–415 Meroe 286–287, 295–296, 302–303
Deir el-Medina 37, 42–44, 204–205, Musawwarat es-Sufra 295, 297
208, 212–213, 215, 375, 379
Diospolis Inferior 440, 451 Napata 291, 293, 301
Diospolis Parva 60–63 Naqa 298, 300–301
Naqa (Naga) ed-Deir 68, 74, 78
Elephantine 121, 161, 291–292, 437, Naqada 55, 60–61, 63, 68, 70, 76–78,
440–441, 451, 489 90
el-Kurru 284–285, 293 Nuri 285, 293, 295–296
Eltekeh 260–261, 291
Philae 299, 303, 385
Gebel Barkal 202, 211, 285–286, 292, Piramesses 414
295–298
Giza 113, 153, 155, 200, 339–348, Qasr Ibrim 299, 303
378, 485
Roda 370–372
Heliopolis 21, 199, 245, 291, 386, 448
Helwan 87, 95, 133–134, 487–489 Sais 257, 468
Herakleopolis 31, 167, 218, 234, 247, Saqqara, see also Step Pyramid 26, 57,
252, 256 88, 90–91, 105, 146–148, 152–153,
index of place names 515

157, 199, 201–204, 206, 208–209, 229, 234, 242–243, 247, 249–250,
346, 418 253–255, 284, 291–292, 408
Semna 173, 177, 199, 201, 203, 370 Turah 85, 88, 198–199, 203
Step Pyramid 102–103, 105, 107, 109,
111–114, 117–121 Umm el-Qaab 95–97, 99, 101–102,
Syene, see Aswan 106–108, 110–111, 113, 115
Uronarti 178, 199, 370
Tanis 218, 221–225, 229, 231, 234, Uruk 487
237, 242, 246–247, 263
Tarkhan 62, 81, 83–85, 90, 93 Wadi Maghara 121, 201
Tell el-Dab(')a, see also Avaris 192, 223
Thebes 12, 162, 166, 181–184, 187, Zaw(i)yet el-Aryan 118–122, 485
212–213, 215, 218, 221, 224, 226,
GENERAL INDEX

Africanus 5, 35–36, 116–117, 163–164, pontifical year 47, 474–475


182, 193–194, 198, 210–211, 214, postdating 461–462
227–228, 231, 235–236, 261, 270, predating 461–462
291 regnal year 46, 235
annals 19–25, 45–46, 94, 100–101, dendrochronology 361–368
107, 122, 125, 128, 131, 136, Distanzangaben 310–311
144–147, 151, 154, 156, 202,
484–486, 490 eclipses 307, 377–378, 394
archives era
Amarna 206, 309, 313–315, 478 apo Menophreos 443
Hattusha 309, 320, 322–323, 478 w˙m mswt 46, 217, 475
Neferirkare' 140–141, 388, 429–431, Seth Nubti 46
483 Eusebius 4, 35–36, 116–117, 163, 175,
Ra'neferef 140–142, 446 193–194, 198, 210, 214, 227, 233,
Apis stelae, see Serapeum stelae 291
expedition dates 12, 369
Babylon, fall of 305–306, 310
Badarian 58–60, 71 genealogies 37–44, 219, 235
Bible, chronology of 3–4, 7, 14–15,
232, 264, 350–352, 474 harvest, dates of 374–376, 419–420
heb-sed, see sed-festival
calendar Herodotus 3–4, 33–34, 282, 485
Alexandrian 48 Horus, following of 20, 22, 24, 45
Babylonian 399–400, 459–460,
467–468 Josephus 35, 36, 116–117, 122,
day 49–51 197–201, 203–204, 207, 209–212,
Egyptian 47–48, 391, 459 375
epagomenai 47, 442, 450, 459, 466
Gregorian 49 King-list 5, 14, 26–32, 33–35, 40,
intercalation 48, 456–457 116–118, 121, 156, 160, 166, 269,
Julian 48, 459 282
lunar calendar 48, 460
lunar calendar, civil-based 389–391 Thermoluminescence dating 356–360
lunar calendar, Sothis-based lunar dates
452–456 computation of 395–397, 407
lunar days 386 feast of Ptah 418
lunar months 387–389, 411, 461, feast-of-the-valley dates 414–417,
479 476
census, biennial 45–46, 124–126, foundation dates 384, 419–421, 479
144–158, 429–430, 483–485 Illahun dates 402, 406–407,
Coptos decrees 150, 152–153, 157 424–427, 480–482, 490
coregencies 30, 170–173, 197, 205, Khozam date 427–429, 482–483
258, 374, 419–420, 426, 478, 481 Megiddo date 421–422, 479
Monthu feast 422–424, 480
Dakhla stela 230, 411–412, 474 observational errors 398–404
dating system Piramesses date 414–418, 476
antedating 46 Tepi Shemu feast 408–414, 473–474
general index 517

wagi feast 423, 425 rnpt zm¡-t¡wy 23, 45, 136, 139–140,
weresh (wr“ ) feast 411–412, 474 151, 157, 484
lunar phenomena rnpt zp 23, 46, 124–143, 144–158,
conjunction 387, 395, 460, 467–468 429, 482–485
25 year cycle 389, 404–407, 431,
480 sed-festival 2, 10–11, 100, 107,
variations in lunar motion 393 149–151, 172, 199–202, 204–205
visibility, last/first 387–389, Sequence Dating, Petrie’s 7, 55, 60–64
398–400, 460 Serapeum stelae 234–235, 244,
267–268, 290, 408, 466, 473
Maadi-Buto culture 56, 79, 92–93 seriation, computer 66–67
Manetho 3, 14, 31–36, 128, 131, Sirius
134–139, 142, 162–165, 167, 168, in Babylonian calendar 432–434
172–175, 181, 192–194, 197–200, in Egyptian calendar, see Sothis
204–207, 214, 218–221, 224, Sothic cycle 47, 442–445
227–228, 230–232, 235–239, Sothic dates
246–248, 261, 269–270, 284, 375, Censorinus date 442, 444, 446
412–413, 428, 474, 479, 481, Canopus date 444, 444, 457
483–485 Ebers calendar date 450–452, 477,
moon, see lunar phenomena 479–480
mummies, royal 197, 207 Illahun date(s) 7, 9, 441, 448–450,
454–455, 480, 489
Naqada I (A-C) 71–75 star clock dates 447
Naqada II (A-D2) 75–81 Sothis
Naqada III (A1-D) 81–88, 487–489 arcus visionis 433–434, 440
Nile flood dates 369–374 geographical reference 440–445
heliacal rising 432–438
orientation, astronomical 340–341, quadrennium 439
378–379, 380–385 triennium 441, 443
Stufen chronology, Kaiser’s 64–68,
papyrus, papyri 70–71, 76, 80, 82–84, 86
Carlsberg 9 389–390 synchronisms 8, 12–14, 304–324,
Gebelein 135 476–478
Heqanakht papyri 391 Synkellos, Georgios 4, 33, 35
Illahun 7, 424–427
Rhind 199 Tang-i Var inscription 13, 258,
Westcar 116, 120 289–290, 473
planets 50, 378, 383, 436 Tasian 57–58
precession 380, 435, 438, 441, 457 tomb of
Ptolemy, Claudius Psusennes I 223, 237
Canon 49, 464–465, 473 Ramesses IV 381
Almagest 50–51, 433, 438 Senenmut 379
pyramids, orientation of 378–379 Sety II 213
Tut'ankhamun 208, 376–377
radiocarbon dating 11, 90–92, Turin Canon 3, 23, 25, 26–32, 33,
327–355, 486–487 116–119, 121–123, 126, 128,
Amarna 335–338 131–139, 142, 144, 156–157, 168,
history 327–332 170–196, 422–424, 428, 480–485
MK pyramids 341–343
OK pyramids 338–350 Venus dates 9, 304–305
wiggle matching 333, 339, 354
rotation, slowing of earth’s 392–393, wine delivery, dates of 375
396, 402–404
rnpt m-¢t zp 23, 124–158, 429, 482–485 year labels 45

Related Interests