Content uploaded by Peter Galton
Author content
All content in this area was uploaded by Peter Galton on May 29, 2014
Content may be subject to copyright.
Received October 18,
1973.
LEONARD E.
KELLY
Dinosaur
Monophyly
and
a New
Class
of
Vertebrates
Nature Vol. 248 March 8
1974.
TRADmONALLY
dinosaurs are classified as
tWQ
or three separate,
independent groups
of
reptiles in the Subclass Archosauria.
But evidence from bone histology, locomotor dynamics, and
predator/prey ratios strongly suggest that dinosaurs were
endotherms with high aerobic exercise metabolism, physio-
logically much more like birds and cursorial mammals than
any liying reptiles'<". Recently Ostrom has argued forcefully
that birds are direct descendants of dinosaurs and inherited
high exercise metabolism from dinosaurs--s. Here we present
evidence that dinosaurs are a single, monophyletic group, and
that the key advancements of endothermy and high exercise
metabolism are justification for removing dinosaurs from the
Reptiffa and placing them with birds in a new class, the Dino-
sauria.
The two generally accepted orders of 4jnosaurs, the Saur-
ischia and Ornithischia, are usually interpreted as independent
derivatives of primitive thecodontian reptiles of the Triassic,
but
all known Triassic dinosaurs can be distinguished from
typical thecodontians (Fig. 1)9-23.
Most thecodontians had a wide-track sprawling gait or a
crocodile-like semi-erect gait, both requiring much multi-
directional mobility at the shoulder". The thecodontian
glenoid, or shoulder socket, was lizard or crocodile-like-
being a saddle-shaped. notch facing outward and backward,
permitting humerus rotation, abduction, and backswing.
Dinosaur glenoids were concave sockets facing mostly down
and backwards, little outwards, restricting humerus movement
severely to a fore-and-aft vertical
plane-the
'fully erect gait'",
Also, in thecodontians and reptiles, generally, the delto-
pectoral crest (dp in Fig. 1) is located close to the humerus
head, giving the pectoralis musculature good leverage for
rotating the humerus about its long axis but relatively little
leverage for a vertical humerus backswing, In dinosaurs, the
vertical backswing leverage was increased by moving the
deltopectoral crest down the humerus shaft (Figs. 1 and 2).
In thecodontians and reptiles generally, the fourth trochanter
(4t in Fig. 1) usually was located close to the femoral head,
giving the caudifemoralis brevis musculature good long-axis
rotational leverage, but little leverage for a vertical backswing.
In dinosaurs, the apex of the fourth trochanter was developed
into a distally directed flange, increasing the leverage for a
vertical backswing (Figs 1 and 2).
A sprawling or semi-erect gait directs the thrust of the femur
strongly inwards into the acetabulum, or hip socket. The
thecodontian acetabulumusually was a strong, continuous bony
'Surface
as in most living reptiles. In dinosaurs, femoral action
was more restricted to a vertical plane; the acetabulum was
perforated to allow deeper penetration of the inturned femoral
head, and the weight of the body was transmitted from the
proximal surface of the femur to the strong dorsal rim of the
acetabulum-e.
The thecodontian hand, known in phytosaurs>", aeto-
saurs1 7 •25, rauisuchids-" and 'Cheirotherium'27.28, was croco-
dile-like with five long digits, the inner three subparallel, the
outer two divergent (Fig. 1). The hand of Triassic saurischian
dinosaurs is distinctive (Fig. 1) with a very short, stout meta-
carpal I; a long, strong thumb phalanx 1; a powerful, curved,
trenchant thumb claw; thumb articulations forcing the claw
to diverge and point inwards during extension, converge with
digits II and
III
and point downwards during flexion; long
digits II and III, subequal and subparallel, bearing trenchant
claws; and reduced digits IV and
V29.
The hand
of
the better
known primitive ornithischians (fabrosaurids and hypsilopho-
dontids) differed from thai of Triassic saurischians in lacking
the long thumb and having blunt hooves instead
of
claws on
digits I to
III
(ref. 30). Recently, a complete articulated
skeleton of Heterodontosaurus, a Triassic ornithischian, was
collected by A. W. Crompton, now at Harvard, for the South
African Museum. The hand is virtually identical to that of
8
a
Concentration
of
TTX
(1/g
ml-
')
100
90
8
=
Department of Zoology,
University of British Columbia,
Vancouver, B. C.
1Grigliatti, T. A., Wqliamson, R., and Suzuki, D. T., Genetics,
64, S27 (1970).
2Grigliatti, T. A., Hall, L., Rosenbluth, R., and Suzuki, D. T.,
Malec. gen. Genet., 120, 107 (1973).
3Narahashi, T.,
Moore,J.,
and Scott, W. R., J. gen. Pliueiol.,
47,965(1964).
4Hinton, C. W., Genetics, 40, 951 (1955).
16~
with~'l
Ii
min,
while no comparable effect is observed with
Oregoh-R
flies.
The converse is also true; shi'"
flies
which
are debilitated by
TTX
at 17° C are immediately paralysed
and die when shifted to 22° C. This indicates two things;
first, the debilitation induced by
TTX
is temperature rever-
sible in shi"" and secondly, the increased resistance of Ore-
gon-R and, in part, of shi'" at 17°' C is probably due to a
reduced rate of uptake of liquid in both strains. This may be
a consequence of reduced dehydration at the lower tempera-
ture. .
In conclusion, it has been shown
that
the
shi"
alleles con-
fer on Drosophila a resistance to
TTX
which, like the
paralysis of these
flies,
is temperature dependent.
It
has also
been demonstrated
that
this resistance is attributable to an
alteration in the properties of the nervous system. The
TTX
resistance results obtained for the
shi"
and
shi'"
alleles are
interpreted as resulting from alterations of different sites in
the protein molecule encoded by this gene.
It
is suggested
that
the
shi"
locus is involved in the production of some com-
ponent of the regenerative sodium channel and
that
the
presence of the
shi"
allele, as well as producing temperature-
sensitive paralysis, also reduces the affinity of this channel
for TTX.
Ithank Drs David T. Suzuki and Linda Hall for discus-
sions. This research was supported by grants from the Na-
tional Research Council of Canada and the National Cancer
Institute of Canada to Dr Suzuki.
FIG.
3 Effect of increasing concentrations of
TTX
on
u, Oreaon-R, and b, ehi" flies at 17·C. Incubation time was
48 h. Arrows marked a and b indicate LD.. for Oregon-R
and
shi"!
flies respectively.
Nature
VoZ.
248 March 8 1.974 169
Fig. 1
Skulls,
forelimbs and hindlimbs of thecodontians, dino-
saurs and birds. All
views
of right
elements
except femur.
(A)
Euparkeria,
(B)
Anchisaurus,
(G"j
Fabrosaurus, (D) Comp-
sognathus,
(E)
Archaeopteryx,
(F)
Gracilisuschus, (G) Plateo-
saurus,
(H)
Heterodontosaurus,
(I)
Halticosaurus, (J)
Ttcino-
suchus, (K) Efraaskr",
(L)
Thecodontosaurus,
(M)
Syntarsus,
(N)
Deinonychus,
(0)
Ammosaurus,
(P)
Coelophysis, (Q) an
ornithomimid and
(R)
Grus.
Anchisaurus,
Fabrosaurus
(except
skull), Heterodontosaurus, Efraasia and Thecodontosaurus
from originalmaterial; data for the others from refs 20, 21,
26,28,36,44,53-55. (s) Scapula,(c) coracoid,(dp) deltopectoral
crest, (Ic) lunate carpal, (il) ilium, (is)
ischium,
(p)
pubis,
(a) astragalus, (ca) calcaneum,
(ft)
femorotibialis, (4t) fourth
trochanter,
(-
-
-)
long axis of
glenoid.
~C~il
~--....,
<
c~R
&2
ts :PQ
i~
E
_ A 0 p
E,
,.11
II II II
\\
~
V
V'I
~"ffi"
1"
I
III
I IV
III
I::
II QIV
IIi'
R
tv IV JJI M(embryo)
F
BeE
Right humerus
-..l
"l
11
'" 0.14
'"
"
~
:s
.0
20.10
'-
0
><
"
]0.06
a
"~f~'--'
~0.141·~.
//5.M'
L
II
9""'-/
-37'
L
~
3
21·
~~38
33~
.
IV •....31
40
.50.10 \
.11
X·I
:>1
32 .. S
~
~2
'2
~
3\'0
1
.0
18
~~.~
o • • ..
~
"'"
14 18 24 2 ...-
..
'0 0.06
\5
20/
.~,
Right
~
'
••
' femur
~
.
.s
0.24 0.32 0.40 0.48
Power index=M/L
0.13 0.21 0.29 0.37 0.45
Power index =M/L
The ankle
of
most thecodontians was complex
and
crocodile-
like: the astragalus articulated movably with the calcaneum
and tibia,
and
the calcaneum bore a 'heel' (calcaneal tuber) for
the gastrocnemius musculature-v-". In advanced thecodontians
the astragalar-calcaneal joint was a deep ball and sockets:'. In
sharp contrast, the dinosaur ankle was stiff, simple and bird-
like: the astragalus and calcaneum were rounded caps fixed
immovably on the ends
of
the tibia
and
fibula; the distal tarsals
were immovable caps on the metatarsal heads, and the tarsal
joint was a simple unidirectional hinge between astragalus-
calcaneum and distant tarsals (Fig.
1)24.28.
Saurischian diphyly
has been suggested because some theropods, such as
Allosaurus,
have deep ilia and a thin, anterior ascending process
of
the
astragalus, while prosauropods have unexpanded ilia, no
ascending process, and a peg-in-notch astragalar-tibial
joint2 2 •3 4 •Triassic forms, however, show how the transition
from prosauropod-grade to advanced theropod grade sauris-
chian occurred. The Triassic theropods Syntarsus and Halti-
cosaurus
had
expanded ilia but retained a pro sauropod-type
astragalar-tibial joint without an anterior ascending pro-
cess3 S •3 6 •In general, the prosauropod grade
of
postcranial
anatomy seems to be the primitive dinosaur pattern. Details
of
tarsal structureare often obscured by incomplete ossification,
but
the pattern in Syntarsus is virtually identical to that
of
Heterodontosaurus
(Fig. 1). Digit I arises two-thirds down the
length of metatarsal
II
and
is reduced and slightly divergent;
digits
II
and IV are subequal; digit
III
is the longest; digit V
is reduced to a splint; distal tarsal IV is cuboid; III is rect-
angular;
II
is a thin plate and the astragalus-calcaneum are
prosauropod-like.
Fig.2
Robustnessand positionof
muscle
attachments,(a) in the
femur and, (b) in the humerus. The data are taken from
speci-
mens and refs 9, 10, 13, 15-20, 26, 28, 30, 36, 38, 39, 42, 43.
Thecodonts, e, are (1) Proterosuchus, (2) Erythrosuchus, (3)
Euparkeria, (4) Shansisuchus, (5) Fenhosuchus, (6) Wangisuchus,
(7) Argentinosuchus, (8) Neoaetosauroides, (9) Stagonolepis,
(10) Palaeorhinus,
(1)
Phytosaurus, (12) Rutiodon, (13) Mystrio-
suchus, (14) Gracilisuchus, (15) Turfanosuchus, (16)
Omitho-
suchus, (17) Riojasuchus, (18) Tictnosuchus, (19) Sphenosuchus,
(20) Pseudhesperosuchus, (21) Pedeticosaurus, (22) Triassolestes,
(23) Alligator, (24) Lagosuchus, and (25) Lagerpeton. Dinosaurs
(~,
are (26) Staurikosaurus, (27) Ischisaurus, (28) Herrera-
saurus, (29) Halticosaurus, (30) Efraasia, (31) Massospon-
dylus, (32) Teratosaurus, (33) Plateosaurus, (34) Riojasaurus,
(35) Melanorosaurus, (36) Anchisaurus, (37) Fabrosaurus, (38)
Heterodontosaurus, (39) Dysalotosaurus, and (40)Hypsilophodon.
(8)
Minimumshaft diameter,
(L)
length of femur or humerus,
(M)
distance from head to
muscle
attachment (deltopectoral
crestin humerus, fourth trochanter in
femur).
E
H
Triassic saurischians in all six
of
the features cited above.
Although specialised craniallys-,
Heterodontosaurus
probably
represents the original ornithischian hand pattern, inherited
from saurischians, where the thumb was specialised for defence
and
digits
II
and
III were used for defence, digging, and support
during slow, quadrupedal locomotion. In more advanced
ornithischian hands, the defensive function was lost, the
thumb was secondarily shortened
and
simplified, and the
claws were replaced by hooves, but in Hypsilophodon, digits
II
and III are still subequal and subparallels".
The origin of the femorotibialis, a knee extensor, in most
thecodontians
and
in reptiles generally, is on the smooth
anteriolateral surface
of
the femoral shaft. In primitive
saurischians and ornithischians, the origin was expanded by
the development of a spike-like ridge
(f.t,
in Fig. 1). In ad-
vanced saurischians
and
ornithischians, and in birds, this
spike becomes a large
crest-the
'lesser trochanter', not
homologous to the lesser trochanter
of
mammals. Increased
muscle power for knee extension probably was related to the
development
of
the fully erect gait. Ornithosuchid thecodonts
have this crest
but
differ from dinosaurs dramatically at the
ankle and forelimb'P-P.
~
-e
§
:I: IV II I
III
J
170
The radiations
of
thecodontians
and
dinosaurs were based
on different trends in locomotor anatomy, behaviour,
and
exercisemetabolism. The earliest thecodontians, the protero-
suchians!",
had
short limbs
and
a sprawling posture with a
wide range
of
movement at each joint". Stride length and power
in the sprawling gait are increased by vertebral column undula-
tions, and thecodontians generally lack adaptations for
inhibiting lateral undulation. The complex ankle
of
advanced
thecodontians was probably aresponse to selection for finer
-control of supination-pronation and greater leverage in
sprawling locomotion. Thecodontian footprlnts"? indicate
that
digit V in the
hand
and foot were strongly divergent
and
powerfully muscled, like digit 1
of
climbing mammals. Flexi-
bility at the major limb joints plus divergent digit V probably
gave thecodontians a wide locomotor repertoire, including'
some digging, climbing,
and
movement over uneven ground.
Some
of
the later Triassic thecodontians (such as aetosaurs
and
phytosaurs") retained the sprawling gait and others, such as
rauisuchids
and
ornithosuchids'v-!", developed a crocodile-
like locomotor system which make possible both sprawling
and
semierect gaits
li
, 3 7 . Inturned femoral heads indicate that a few
thecodontians such as Hal/opus and some advanced raui-
suchids and ornithosuchids!", may have
had
narrow-tracked,
fully erect hindlimbs, but even these forms retained the complex
ankle.
The first well-known dinosaur is Staurikosaurus,
of
late
Middle or early Late Triassic3s.39, a long-jawed predator
with very long, gracile hindlimbs and postcranial anatomy
of
prosauropod grade (the femoral head is not as sharply in-
turned as in later dinosaurs
and
there is an oval acetabulum).
More advanced theropods, such as Ischisaurus and Herrera-
saurus,and ornithischians, such as Pisanosaurus, appear shortly
after Staurikosaurus", The original dinosaur stock were
probably predators with higher running speeds and exercise
metabolism than in thecodontians
of
comparable size. Early
dinosaurs may have been dipedal at top speed,
but
the strong
forelimb modifications for the fully erect gait and the prosauro-
pod-like
hand
indicate that quadrupedal progression was very
important. Even at the prosauropod grade, dinosaur bone
histology was identical to that
of
endothermic mammals",
Restriction
of
joint movement to a fore and aft, vertical plane
and the stiff, bird-like ankle would have made climbing difficult
for any dinosaurs". The vertebral column
of
early dinosaurs
was stiffened to inhibit lateral undulations by the development
of
extra vertebral articulations, the hyposphene-hypantra in
saurischians,
and
long ossified tendons appressed against the
neural spines in ornithischians. The locomotor repertoire
of
dinosaurs was narrow, rather restricted to moving over fairly
even terrain. Recent experimentsf! show that erect locomotion
does not require less energy for a given speed than sprawling, as
has usually been assumed",
but
erect locomotion probably
does increase manoeuvreability at very high speeds. The dino-
saur
radiation was based on a concentration
of
behavioural,
physiological, and anatomical adaptations for high, sustained
running speeds which made them irresistible predatorsand com-
petitors to contemporary thecodontians
and
the larger mammal-
like reptiles.
Could the similar postcranial adaptations
of
Triassic dino-
saurs reflect merely convergent evolution
of
the erect gait from
different thecodontians? We believe that the answer to this
question is almost certainly no. Evolution
of
erect posture
does not invariably lead to an avian-dinosaur type joint
pattern. Some thecodonts
and
many mammals, such as carni-
vores and ungulates, have a fully erect gait
but
retain complex
ankles with astragalar-calcaneal mobility
and
calcaneal 'heels'.
The very detailed similarity in all the majorjoints, especially the
hands
and
feet,
of
Triassic ornithischians
and
saurischians
makes dinosaur polyphyly exceedingly unlikely. Moreover,
the pattern
of
increasing dinosaur diversity in the Triassic
upwards from the stratigraphic level
of
Staurikosaurussuggests
one monophyletic radiation. Some
of
the little 'rabbit theco-
dontians' such as Lagosuchus
and
Lagerpeton4 2 ,4 3 , have im-
Nature Vol. 248 March
81974
mobile astragalar-calcaneal joints
and
lack calcaneal 'heels',
and
may be related in some way to dinosaur origins.
The prosauropod saurischians have a mixture
of
primitive
and
advanced features which show how ornithischians evolved
from saurischians (Fig. 1).
In
the primitive ornithischian
dentition, represented by
Fabrosaurus,
the anterior teeth were
non-serrated, pointed and slightly incurved with swollen bases
and
the cheek teeth were serrated with triangular crownst".
In some prosauropods, such as Massospondylus and probably
Anchisaurus, the basic pattern was similar: anterior, simple
teeth
and
triangular, serrated cheek teeth. In advanced pro-
sauropods, such as Plateosaurus
and
in most ornithischians,
the jaw articulation was depressed below the
tooth
row level,
but
in Fabrosaurus
and
Anchisaurus the articulation was still
on the same level as the
tooth
row4 4 •The lower
jaw
of
most
ornithischians was deep and massive,
but
the fabrosaur
jaw
was slender like that
of
Anchisaurus. In early ornithischians the
origin
of
iliotibialis, a knee extensor co-inserting with the
femorotibialis, was expanded by the development
of
a long
anterioriliac prong. Such a prongis unknown in thecodontians
and
most prosauropods,
but
is present in very frabrosaur-like
configuration in Anchisaurus
and
Ammosaurus'", In ornithis-
chians the primitively broad, plate-like ventral surface
of
the
pubis and ischium were deeply excavated
and
reduced; such
excavations
had
begun in the prosauropods Anchisaurus and
Ammosaurus'",
The prosauropod grade is quite primitive,
and
herbivorous
prosauropods must have diverged very early from the basal,
predatory dinosaur stock.
It
should be noted that no theco-
dontians show trends away from carnivory towards herbivory,
except the aetosaurs, sprawling, armoured types totally unlike
ornithischians
17.
Small prosauropods with Anchisaurus-type
jaw
articulations, ilia and pubes were probably
the
immediate
ancestors
of
ornithischians. The adaptive shift probablywas an
emphasis on ultra-high speed bipedality plus adaptations for
prehension
and
cutting
of
plant fibres. Excellent fabrosaur
material collected for the South African Museum
and
now at
Harvard, shows
that
the development
of
an upper beak
had
just barely begun in
Fabrosaurus.
The fabrosaur predentary
was small, sharply pointed,
and
movably articulated with the
dentary; the dentary symphysis was mobile. The predentary
beak could serve as a gouge for digging into plant stems and
as a cutting edge occluding with the arcade
of
premaxillary
teeth, much like the incisors
of
kangarooss", while the mobile
symphysis permitted the retention
of
independent control
of
each
jaw
ramus, important for tooth-to-tooth shearing on
one side
of
the
mouth
at a time. Excavations in the maxilla
and
dentary, for cheek pouches to retain the cut ends
of
plant
material?",
had
begun in
Fabrosaurus.
All known Triassic
ornithischians
had
very long hind limbs with elongate distal
extremities indicatingvery high speeds. A short
trunk
facilitates
the maintainance
of
balance in a highly bipedal animal,
but
herbivory demands a long, bulky gut. Ornithischians solved
the balance problem by rotating the pubis posteriorly so that
the distal end, marking the end
of
the gut, lay far posterior
to the acetabulum. Thus, the gut was extended between the
hindlimbs
and
the pre-acetabular
trunk
section could be
abbreviateds",
Although Walker49has cited some resemblances between
living birds
and
crocodile-like thecodontians (sphenosuchids),
we believe that Ostrom has shown that the similarities between
Archaeopteryx
and
small theropods are so detailed
and
comprehensive
that
the immediate ancestor
of
birds must have
been a saurischian dinosaurl.s,5o. Some
of
the features cited
by Walker as evidence
of
crocodile-bird relationships are
actually widespread among tetrapods: cranial salt glands are
developed in lizards, especially in Amblyrhynchus51 ;forearm
articulations which force the radius distally during forearm
flexion are present in salamanders, Sphenodon,
and
lizards.
Although details are obscure, the Archaeopteryx skull seems
much more like the light, kinetic skulls
of
small theropods than
the solidly braced sphenosuchid structure'
2.
Archaeopteryx
'
..
Nature Vol. 248 March
81974
has the complete suite of seven features listed above as diag-
nostic of the primitive dinosaur erect gait. _In flying birds, bats
and pterosaurs the glenoid faces strongly outwards to make
possible the abduction of the humerus necessary for gliding
and powered flight. Significantly,
Archaeopteryx
retained the
typical, downward-facing dinosaur glenoid; it is difficult to
imagine how this joint would have made either flight or gliding
possible. The initial evolution of feathers probably was not
associated with flight, a- hypothesis developed at length by
Ostrom'.
Ostrom has shown that the hands and feet of
Archaeopteryx
are identical in detail to certain Jurassic/Cretaceous theropods,
and differ only in being larger relative to the body size (Fig.
3)1,50,53-55.
Feathers may have been widespread in bird-like
theropods.
Archaeopteryx
strongly resemblessmall theropods
in general body plan with a short trunk, thin, flexible neck,
very long hindlimbs and long fingers bearing trenchant claws
for snagging prey. Sphenosuchid thecodontians were quite
different (Fig. 3) with a long trunk, thick neck and long fore-
limbswith short digits adapted for running and not prehension.
Near bird-lower cretaceous theropod Microvenator
Archaeopteryx
~
Crocodyloid thecodont (sphenosuchid) Pseudhesperosuchus
Fig.3 Bodyform in a small theropod dinosaur, the firstknown
bird and a crocodyloid thecodontian. Microvenator hand and
skull restored after Ornitholestes. Data from refs 10, 53-55.
Endothermy and high aerobic exercisemetabolism are
suffi-
cient justification for separating birds into a class distinct from
other living sauropsid tetrapods. But endothermy and high
exercise metabolism were probably already present in the
dinosaur ancestors of birds and are the key features differen-
tiating dinosaurs from crocodilians and the other extinct
archosaurs. The bird radiation has produced many species,
but the structural diversity is not more striking than that
within the Ornithischia or Saurischia. Bird physiology is also
rather stereotyped. The avian radiationisan aerial exploitation
of basic dinosaur physiology and structure, much as the bat
radiation is an aerial exploitation of basic, primitive mammal
171
physiology. Bats are not separated into an independent class
merely because they fly. We believe that neither flight nor the
species, diversity of birds merits separation from dinosaurs on
a class level. Among all amniotes, the most profound adaptive
shift was from ectothermy to endothermy, which occurred
during the origin of mammals and dinosaurs. Therefore we
propose the erection of a Class Dinosauria, to include as
subclasses the Saurischia, Aves and Ornithischia. The cur-
rently recognised suborders of dinosaurs would be elevated to
orders (Fig. 4). Thecodontians, crocodilians, and pterosaurids
Cu
~oou~
S~bc""
omitbitcbia:
.~~,
r
Subcu
lauriacbia
r;,:;::
Order
On!«
Ordet.l:1lliial
Order '/ Order I
S~~u
I~
otDIllio-
An~YJo,
-.
Sau"ropodOmorpba
Theropoda Aves
~.
f
...../
..... / .....
/'-1/'-"
I
If
If
~
I
Fig. 4 Suggested phylogeny and classification of dinosaurs.
Position of Lagosuchus and Lagerpeton is very uncertain.
Within orders, ancestor-descendentrelationship is not necessarily
indicated by stratigraphic levelof known specimens,for example,
Pisanosaurus
is earlier than, but in some ways more advanced
than the earliest known Fabrosaurus. (1) Proterosuchus, (2)
Euparkeria, (3) Ticinosuchus, (4) Gracilisuchus, (5) Spheno-
suchus, (6) Pseudhesperosuchus, (7) Lagerpeton, (8) Lagosuchus,
(9) Staurikosaurus, (10) Ischisaurus, (11) Herrerasaurus, (12)
Coelophysis, (13) Syntarsus, (14) Halticosaurus, (15) Coelurus,
(16) Ornitholestes, (17) Compsognathus, (18) Archaeopteryx,
(19) Microvenator, (20) Deinonychus, (21) Thecodontosaurus,
(22) Efraasia, (23) Plateosaurus, (24) Ammosaurus, (25)
Anchisaurus, (26) Pisanosaurus, (27) Fabrosaurus and (28)
Heterodontosaurus.
would remain in the reptilian subclass Archosauria, which
would stand to Dinosauria much as the reptilian subclass
Synapsida (the mammal-like reptiles) stands to Mammalia.
This new classification, we believe, reflects more faithfully the
major evolutionary steps. Ectotherms and forms transitional
to endotherms are retained in the Reptilia and the two highly
successful endothermic groups, mammals and dinosaurs, are
given separate class status.
172
We
thank
J. Ostrom for
data
on
Archaeopteryx
and
A. W.
Crompton for access to fabrosaurid
and
heterodontosaurid
material. Supported in
part
by the National Science Founda-
tion. ROBERT T. BAKKER
Department
of
Vertebrate Paleontology,
Museum
of
Comparative
Zoology,
Harvard
University,
Cambridge,
Massachussetts
02138
PETER
M.
GALTON
Department
of
Biology,
University
of
Bridgeport,
Bridgepor~
Connecticut 06602
Received
June 25;
revised
October8, 1973.
1Ostrom,J. H., Q. Rev. Bioi. (in the press, 1974).
2. Bakker, R. T., in Perspectives
of
Biophysical Ecology (edit. by
Gates, D., and
Schmerl,
R. B.) (New York, in the press).
3Enlow,D. H., and Brown,S.
0.,
Tex. J.
s«,
2, 186 (1957).
4Currey, J. D., Palaeontology, 5, 238 (1962).
5Ricqles,
A,
Annis. Paleont., 54, 9 (1968).
6Bakker, R. T., Evolution, 25, 363 (1971).
7Bakker, R. T., Nature, 238, 81 (1972).
8Ostrom,
I.
0.,
Nature, 242, 136(1973).
1I Bonaparte, I. F., Proc. Pap. 2nd Gond. Symp, S. Afr., 665
(Council for
Scientific
and Industrial Research, Pretoria, 1972).
10
Bonaparte, I. F., Op, lilloana, 22, 183 (1971).
11 Casamiquela, R. M., Ameghiniana, 4, 47 (1967).
12. Reig, O. A., Ameghiniana, 3, 3 (1963).
13
Romer, A. S.,
Breviora,
389,1
(1972).
14
Thulbom, R.
A,
Nature, 234, 75 (1971).
15
Cruickshank, A. R., in Studies in Vertebrate Evolution (edit. by
Joysey, K.
A,
and Kemp, T. S.)(Oliverand Boyd,Edinburgh,
1972).
16
Young,C.C., Mem. Inst,
vertebr,
Paleont,Peking,
10,15
(1973).
17
Walker,A. D., Phil. Trans. R. Soc., B,
244,
103 (1961).
18
Walker, A. D., Phil. Trans. R. Soc., B, 248, 53 (1964).
ill
Walker, A. D., Phil. Trans. R. Soc., B, 257, 323 (1970).
2.0
Kuhn,
0.,
in Die Saurier der Deutschen Trias (Verglag Oeben,
Munchen, 1971).
2.1 Romer, A. S., in Vertebrate Paleontology (University of Chicago
Press,Chicago, 1967).
2.2. Charig, A. I., Attridge,
I.,
and Crompton, A. W., Proc. Linn.
Soc. Lond.,
176,197
(1965).
23
Romer,
AS.,
Breviora;395, 1 (1972).
24
Charig, A J., in Studies in Vertebrate Evolution (edit. by Joysey,
K., and Kemp, T.) (Oliverand Boyd, Edinburgh,1972).
25
Sawin,H. I.,J. Paleont.,21, 201 (1947).
26
Huene, F., von, in Die fossilen Reptilien des Siidamerikanischen
Gondwanalandes
(C. H. Beck'sche VerlagsbuchhandIang,
Milnchen,1935-1942).
27
Soergel,
W., in Die Fahrten der Chirotheria (lena, 1926).
28
Krebs, B., Schweiz. Palaeont. Abh., 81, 1 (1965).
211 Galton, P. M., Arnoldia, 15, 1(1971).
30
Galton, P. M., Lethaia, 4, 453 (1971).
31
Charig, A.
I.,
and Crompton, A. W., Boonstra Commemorative
Vol. (CapeTown, in thepress).
32. Hughes,B., S. Afr. J. Sci., 59, 221 (1963).
33
Sill,W., Bull. Mus.
Compo
Zool. Harv, (in the press).
34
Bonaparte,I. F., Zoo. J. Linn. Soc., 48, 471 (1969).
35 Huene, F. von, Paldont, Z., 16, 145 (1942).
36
Raath, M.
A,
Amoldia, 28, 1 (1969).
37
Cott, H. B., Trans. zool. Soc. Lond., 29,211 (1961).
38
Colbert, E. H., Amer. Mus. Novitates,2405, 1 (1970).
311 Galton, P. M., Paldont,
Z.,
47, 229 (1973).
40
Galton, P. M., Bull. Br. Mus. nat. HiM., A, 25, 1 (1974).
41
Bakker,R. T., Physiologist, Lond., 15, 3 (1972).
42. Romer, A. S., Breviora, 378, 1 (1971).
43
Romer, A. S., Breviora, 394, 1
(972).
44
Thulbom, R.
A,
Palaeontology, 13, 414 (1970).
45
Galton, P. M.,J. Paleont., 45, 781 (1971).
46
Ride,W. D., Syst, Ass. Pub., 3, 33 (1959).
47
Galton, P. M., Lethaia, 6, 67 (1973).
48
Galton, P. M., Evolution, 24, 448 (1970).
411 Walker,A D., Nature, 237, 257 (1972).
50
Ostrom, I. H., K. ned.
Akad.
v. Wet., B, 75, 289 (1972).
51 Dunson, W. A., Am. J. Physiol., 216, 995 (1965).
52. De Beer, G., in Archaeopteryx lithographica (British Museum
(Natural History), London, 1954).
53 Heilmann,G., in The Origin
of
Birds (D.Appleton and Co.,New
York, 1927).
54
Ostrom,I. H., Bull. Peabody Mus. nat.
Hlst.,
35 (1970).
55
Ostrom, I. H., Bull. Peabody Mus. nat.
si«,
30 (1969).
Nature Vol. 248 March
81974
Viroids and viral hepatitis in
marmosets
DIENER
land Zuckerman- have recently speculated
that
hepatitis A and B, respectively, may be caused by infective
naked nucleic acids or viroids.
In
response to their hy-
potheses, we felt it would be of interest to investigate
the
nature of an agent known to cause hepatitis in marmoset
monkeys. We reasoned
that
if the aetiologic agent of
hepatitis found in serum is a free, or naked, nucleic acid in
this system, then its infectivity should be destroyed by the
action of a specific endonuclease.
For
example, a free viral
DNA
would be rendered uninfective after treatment with
pancreatic DNase.
The
virus used in our
study
was the
'Barker' agent, originally recovered by Deinhardt et ol.
from acute phase sera of marmosets inoculated with serum
from ahuman case of viral hepatitis", This virus consistently
induces hepatitis in such animals, although its pedigree as
ahuman hepatitis agent has been asubject of controversys.
As many animal sera contain RNase activity,
but
not
DNase activity, we assumed the postulated viroid for Barker-
induced marmoset hepatitis, if it existed, would most likely
be anaked DNA. To test this assumption, we divided a
pool of infective acute-phase marmoset serum into two
equal aliquots. One aliquot was treated with pancreatic
DNase
(EC
3.1.4.5) at 20
p.g
ml" for 1h at 37° C; the
other aliquot was heated to 37° C for 1 h without any other
treatment. Three S. nigricoUis marmosets were inoculated
intravenously (i.v.) with treated serum (0.25 ml
each);
three other S. nigricollis marmosets were inoculated i.v. with
untreated serum. The course of infection in each marmoset
was monitored by measuring serum glutamic pyruvic trans-
aminase (SOPT) and serum isocitric dehydrogenase (SICD)
activities as indicator enzymes for liver damage.
All six marmosets demonstrated elevated
SOPT
and
SICD
activities by. the fourth or fifth week after inoculation, in-
dicating liver damages had been sustained in each animal.
Clearly, then,
the
marmoset hepatitis agent cannot be a
DNA viroid as all three animals receiving DNase treated
serum showed convincing enzymatic evidence of viral hep-
atitis.
The
results of our preliminary study
are
consistent
with the tentative finding by Deinhardt et al.5
that
the
Barker hepatitis agent bands in CsCI
at
abuoyant density
of 1.2. Viruses usually band around p=1.2-1.4, whereas
free viral
DNA
bands at about p=1.7.
While the possibility still exists
that
the
marmoset hep-
atitis agent is another
type
of viroid,
that
is, a unique
double-stranded
RNA
(or
RNA:DNA
hybridr-protein
con-
jugate which bands in CsCI
at
p=1.2, our experimental
evidence at least rules out the presence of a simple DNA
viroid in serum containing the marmoset agent.
D. W. BRADLEY
D. H.
KRUSHAK
J. E.
MAYNARD
Phoenix Laboratories,
Bureau of Epidemiology,
Center for Disease Control,
4402
North
Seventh Street,
Phoenix, Arizona 85014
Received November 29,1973.
1Diener, T. 0., Agric. Res. USDA, February (1972).
2Zuckerman,A.J., Lancet, 1, 1468 (1973).
3Deinhardt, F., Holmes, A., Capps, R., and Popper, H., J.
expo Med., 125, 673(1967).
• Parks, W., and Melnick, J., J. inject.
DiB.,
120,539
(1969).
5Deinhardt, F., Wolfe, L., Junge, U., and Holmes, A., C.M.A.
J., 106, 468
(1972).
