The Representation of Lower Egypt

by Herbert Donner

Considering the Delta of the Nile (1) we have to make a primary observation: the representation of Lower Egypt totally differs from the other sections of the mosaic map - a matter which scholars usually have not paid attention to. (2) A clear description of the differences is the first step to a pertinent interpretation of this neglected part of the mosaic.

1. The mosaic map on the whole is an illustration of God's salvation history according to the Holy Bible, Old and New Testament. The representation of the Nile Delta, however, does not fit into this principle. It can be recognized by the desiderata, i.e. by the lack of important biblical themes which are to be expected on the map. The story of Joseph (Gen. 37: 39-50), e.g., is missing: that story according to which Joseph settled his father's family in the land of Goshen in the eastern part of the Delta, the modern Wadi at-Tumeilat. Nothing is reported of Israel's stay in Egypt (Ex. 1-12), of their building the store cities of Pithom and Raamses (Ex. 1:11), of the Exodus from Egypt (Ex. 13-15). The figure of Moses is totally absent, and the crossing of the sea of reeds is not mentioned - the latter, probably, because Byzantine tradition localized this event at the northern point of the gulf of Suez. (3) Furthermore, we find no trace of the Prophet Jeremiah's stay in Egypt (Jer. 42-44) nor of the escape of the Holy Family to Egypt (Mat. 2) which is said to have reached Heliopolis or even Memphis. (4) Inscriptions in the style of the mosaic map referring to these biblical themes can easily be invented, e.g.
a) "Tanis, whence came Saint Moses (or: where Saint Moses was born)"; (5)
b) "Goshen (land of Egypt), where Jacob dwelt together with his sons". (6) Such proposals and other ones eliminate the problem for the mosaicist. The question arises: why din't he do his duty?

2. The main literary source of the mosaic map is Eusebius' Onomasticon of Biblical Place Names. But the mosaicist who used the Onomasticon exhaustively tor the Palestinian sections of his map, did not use it for Lower Egypt. Eusebius mentions 10 or 11 items in the Delta of the Nile, the mosaicist 14, but only two of hem can be found here and there namely Sais and Tanis. The same is true with regard to other suggested sources the mosaicist took as a basis: the books of Flavius Josephus, the Bible commentaries of Origenes and St. Jerome, the so-called Diamerismos tes ges ("the distribution of the earth"), (7) a.o. We get the following impression: the mosaicist used a small library of important ancient books, but when he began to prepare the mosaic of Lower Egypt, he closed the door of this library and did not use it anymore. Why did he do so, and which sources did he really use for the Delta of the Nile?
3. The mosaic map of Madaba is the most exact example of cartography before the beginning of modern cartography in the nineteenth century. Naturally, there are some more or less important geographical and topographical mistakes, but they cannot diminish the value and exactness of the map. In the relatively small section of the Nile Delta, however, the accumulation of mistakes and inaccuracies is striking. The sites of both cities Xois and Sais have changed: Xois was situated near modern Saha, about twenty-four kilometres southeast of Tall Faraoun (or Fara'in), i.e., east of the Sebennitic arm; (8) and Sais was located near Kafr az-Zayat north of Sa'al-Hagar at the arm of Rosette, i.e., west of the Sebennitic arm. (9) Consequently, the Saitic arm is misrepresented on the Madaba map: it did not branch off to the right, i.e., to the east, but to the left, i.e. to the west, seen from the Sebennitic arm. The village HNIKIOY is unlocalized; it must have been situated east of the Canopic arm in the Prosopitic district. (10) On the map it is represented too far to the east. The position of Pelusium (Tall al-Farama) was mainly on the east side of the Pelusiac arm; (11) on the Madaba map we find it on the west side. Finally, the Nile Delta is drawn in a false geographical relation to Palestine. The coastline of the Mediterranean Sea really turns west south of Gaza, but on the Madaba map it turns east. (12) This is the same incorrectness as on the so-called map of St. Jerome, a twelfth century copy of a Roman world map originating from the seventh or eighth century. The reason is quite clear. If the mosaic artist would have represented the coast-line correctly, he would have had to abandon the rectangular size of his map: the coast-line going from top to bottom, and the Nile with its arms coming from the right side - totally impossible on a church floor. Moreover, a conflict would have been produced between real geography and religious geography. (13) For in the ancient Christian tradition the Nile was one of the rivers of Paradise, and the Paradise was situated in the east, according to Gen. 2: therefore the Nile had to run from the east to the west without any regard to the geographical facts, even though people may have known the geographical truth. (14) But the other mistakes cannot be explained in this way. The question arises: did the mosaicist ever see Lower Egypt? And once more: which sources did he use?

The Nile Delta:
civil provinces, main monasteries (hook) and bishoprics (crosses)
(after F. Van der Meer - Ch. Mohrmann,
Atlas of the Early Christian World, London 1966, Map 18)

As far as the representation of the arms of the Nile is concerned, the second question can easily be answered: it is based on the oldest description of the Delta we know, namely Herodotus, Hist. II, 17:3-6. The text runs as follows: "The Nile intersects Egypt in two, from the cataracts unto its mouth. Until the present city of Kerkasoros it is running as one Nile; after this city it is split into three arms. And the arm going to the east is called the Pelusiac arm; the other one is going to the west and is called the Canopic arm. The arm going straight ahead, however, runs as follows: coming from above it reaches the top of the Delta; from this point it intersects the Delta, flows into the Sea and keeps a quantity of water which is neither insignificant nor unknown. It is named the Sebennitic arm. There are still two other arms, branching off from the Sebennitic arm and running into the Sea: their names are Saitic and Mendesic. The Bolbyticon and the Bucolicon are no real arms, but artificial canals."
The representation on the mosaic map corresponds exactly with this description. There are three small differences only:
1. The name Bulbyticon on the Madaba map instead of Bolbitinon in Herodotus' description is without parallel. Probably, it is nothing but an error, or the mosaicist used another text of Herodotus than we have.
2. The artist did not distinguish between the "canals" and the real arms, the reason of which is clear: he wanted to draw the arms in the Delta symmetrically. There was no need to differ from Herodotus. He only had to interpret him, because Herodotus does not describe how the arms are running.
3. The Mendesic arm, mentioned by Herodotus, seems to be absent. But it can easily be demonstrated that it originally was represented on the map. The inscription BOYLBY[TIKON] is completely preserved until Y. Of the following letter T two white cubes are still existing, forming part of a horizontal line, the cross-beam of the T. If this cross-beam is lengthened to the left, trying to restore the whole letter T, it becomes clear that the inscription together with the Sebennitic arm slightly deviated to the right. On the other hand, however, the black left limitation-line is slightly moving to the left. From these observations we have to conclude: another arm of the Nile which is not preserved branched off from the Sebennitic arm to the east, i.e., Herodotus' Mendesic arm.

So far, things are clear. According to what principles, however, the mosaic artist chose the cities and villages to be represented in the Nile Delta? Neither according to biblical traditions nor to pilgrims' requirements. Or did he want to portray the Christian Lower Egypt in Byzantine times by giving the ecclesiastical centres and bishops' residences? If it be the case, one could compare the Madaba map with the lists of Byzantine bishoprics, the most important and most complete of which is the Descriptio Orbis Romani, written by Georgius Cyprius during the reign of the emperor Phokas (602-610). (15) The Descriptio mentions fifty metropolis cities, the Madaba map fourteen only. Moreover, five of these fourteen cities are not mentioned in the Descriptio, and on the other hand, significant bishops' sees are lacking on the Madaba map: e.g., Bubastis, Leontopolis, Naucratis, Taua, Cleopatris, Busiris a.o. In a word, this is certainly no representation of ecclesiastical Lower Egypt in the sixth century.
Consequently, there is no other possibility but to examine the relations of the represented cities to the road system in the Nile Delta. Briefly, there are three main roads, running approximately along the collateral lines of the triangle. We know these roads from the written itineraries, e.g., from the collection of Itinerarium Antonini (16) and from the Tabula Peutingeriana, a Roman road map originating from the third or fourth century and preserved in a medieval copy.(17) The first main road runs from Pelusium to Memphis, the second one from Alexandria to Memphis, the third one from Pelusium to Alexandria. Nobody will be surprised hearing that all cities and villages represented on the Madaba map were situated at one of these main roads, except Thennesos. Thennesos, known from Byzantine and early Arabic sources, was a commercial town and a seaport upon a small island within the lagoon region of Birkat or Bahr Manzala, nowadays Tall or Kom Tannis. This town, of course, was connected with the inland: there must have been roads, although we do not know them, probably to Heracleopolis parve, i.e. Sethroites, or to Tanis, or to both of them. The strange rhombus near Thennesos seems to be a hint of the lagoon region in the northwestern part of the Delta.
To summarize: the mosaic artist used the classical description of the Delta, written by Herodotus, and a profane Roman-Byzantine itinerary, the latter not being identical with the Itinerarium Antonini, but similar to it. One cannot exclude that he also used a Roman-Byzantine road map. But it seems better to assume that he did not, for the inaccuracies in localizing the places and minor arms of the Nile are more intelligible provided that he did not make use of a map which could have corrected him. It was sufficient to know where the big cities of Pelusium, Alexandria and Memphis were situated; after that, he could complete the representation according to his itinerary.
Finally, the capacity of the mosaic artist is to be admired: using only a few and poor literary sources, and probably without having seen the land, he created the oldest map of Lower Egypt we know, a map which is more reliable than all its successors until the beginning of modern cartography.


(1) Cf. H. Donner, "Das Nildelta auf der Mosaikkarte von Madeba", Fontes atque Pontes. Aegypten und Altes Testament, 5 (1983) p. 75-89.

(2) Except A. Shulten, "Die Mosaikkartee von Madabe und ihr Verhaeltnis zu den aestesten Karten und Beschreibungen des Hl. Landes", Abhandlungen d. Koenigl. Geselschaft d. Wissenschaften, Goettingen, Phil. Hist. Kl. IV, 2 (1900) p. 30-33, 103. 115-121; A. Jacoby, Das Geographische Mosaic von Madaba. Die aelteste Karte des hl. Landes, ein Beitrag zu ihrer Erklärung, Studien über christliche Denkmaler, 3 (1905) p. 35-43; R.T. O'Callaghan, loc. cit (note 11), p. 696-702.

(3) Cf. the pilgrim's report of the nun Etheria or Egeria (around 400 A.D.), chapter 7: translated into German and explained by H. Donner, Pilgerfahrt (note 24), p. 95-99; from the century of the Madaba map: the report of an anonymous pilgrim from Piacentza (around 570), chapter 41 (H. Donner, loc. cit., p. 304-306).

(4) H. Donner, Pilgerfahrt, p. 309, note 20s.

(5) According to Etheria 9,5: H. Donner, Pilgerfahrt, p. 102, note 80.

(6) According to Eusebius, Onom. 62: 10-11.

(7) Cf A Jacoby, loc. cit. (note 28), p. 34.

(8) W. Helck, RE II, 9 (1967) p. 2152-2155.

(9) H. Kees, RE II, 1 (1920) p. 1758-1759.

(10) Perhaps identical with Ibshadi, about 8 km. west of Sersena=Arsinoë/Cleopatris, or Kom Razin, about 9 km. south-west of Menuf. see H. Kees, RE, XVII, I (1936) p. 342-344. The basic note can be found in Ptolemaios, Geogr. IV, 5 (§ 49 Nobbe): ap'anatolon pros to megalo potamo Prosopites, nomos, kai metropolites Nikiou.

(11) H. Kees, RE, XIX, 1 (1937) 407-415.

(12) Donner-Cüppers, pl. 38, 84, 122, 124.

(13) Fine examples of religious geography in the pilgrim's report of Etheria: H. Donner. Pilgerfahrt, p. 84 (Note 12), 87s, (note 23), 109 (note 97).

(14) Fl. Josephus, Ant. 1, 1,3: "Finally, the Geon is running through Egypt and is called 'streaming towards us from the east'; the Greeks call it Nile." Eusebius, Onom. 60, 3-4.

(15) Ed. By O Cuntz, Itineraria Romana, I, 1929. For general information on the itineraria see W. Kubischek, RE, XI, 2 (1916) p. 2308-2663.

(16) Ed. by H. Gelzer, Leipzig, 1890.

(17) Ed. By K. Miller, Die Peutingersche Tafel, 1887/88, reprinted 1962. Cf. K. Miller, Itineraria Romana. Roemische Reisewege an der Hand der Tabula Peutingeriana dargestellt, 1916, reprinted 1964.

This article was first published in: "Transjordan and Egypt in the Mosaic Map of Madaba", Annual of the Department of Antiquities of Jordan 28 (1984) 254-257.

logo logo

Created Saturday, December 16, 2000 at 11:13:20
by Eugenio Alliata ofm in collaboration with Stefano de Luca ofm
Webmaster: John Abela ofm - Space by courtesy of Christus Rex
copyright - Studium Biblicum Franciscanum - Jerusalem 2000