Vet olka iat Ve
BD: Leted Mia hid

enter ne, ota ee heey Saath ehsicthanaton ists -
Tes SOT Rab Meno eT De i
entra Tats Uphearrasicanag tsar Bete 2

PN MTD fll TOME ED fdr Nid a
AAD R SA Pa beide tilly FA bie at fort

1M LAME OU od i
es a a eo
PERI E LE Drie F
Treat “

”
J"

tides

Ree hc Ap eating Sethe bets bah aoe SY
ee teat EY Sdeine onchate iteenietsan or ross
=D ayihy pRicbat hertor bo vaba hy be heey
ns ASN Getty 5 sn te tet
Sees ina Hire Or ee Poe
emyith ataliestbathy Ma thanity My targe hers SahaheT Ray"

mage 3
Speen

710 LVAD eho, .
LEANING MATA 5 cay EADIE hath Re titlne'§ RIS
Pe AHS REDON MR. Peerviche ls

beet eee: SS
tare he Nac his tater

ee

PN LAA MED Oni

Ste hndtinn Pent atieg bean Mr

tee

~
Saat aye tiene enyres lee

- ne j-
teeth INS Nalin dhe

ne Te vf ns
as Le i tamer ewe Ses a tee Ses a
Per Oe Petpet Sa + . avery
a » rer por MP ettor Pe TCS Pe ee trastirtyte .
ee PPAF % sh Sane ers 2

soma eteee
a

fos

HARVARD UNIVERSITY

EE

LIBRARY

OF THE

Museum of Comparative Zoology

Cra el a6 TP RLA as. fp
Ywuia te \ Tas

Dae o> oo oy ee ae a

.

Beer Wik ote Ow -
RA) Sie Asi Ww GR &e —
Nid Le) ee ee Ped

aie type
i 655% ees OO, ce8

ss
=o
2
es

ry ‘iden |
V. mae teed Shas .
oe YA we de
TR PRA ey a teiqe ee! |
Be mg. ay Neto bee | -

;
,

P—K- Ue are Haven]

POSTILLA

PEABODY MUSEUM
YALE UNIVERSITY

NUMBER 131. 15 MAY 1969

THE PELVIC MUSCULATURE
OF THE DINOSAUR HYPSI-
LOPHODON (REPTILIA:
ORNITHISCHIA)

PETER M. GALTON

POSTILLA

Published by the Peabody Museum of Natural History, Yale University

Postilla includes results of original research on systematic, evolution-
ary, morphological, and ecological biology, including paleontology.
Syntheses and other theoretical papers based on research are also
welcomed. Postilla is intended primarily for papers by the staff of
the Peabody Museum or on research using material in this Museum.

Editors: Jeanne E. Remington and Nancy A. Ahlstrom

Postilla is published at frequent but irregular intervals. Manuscripts,

orders for publications, and all correspondence concerning publications
should be directed to:

Publications Office
Peabody Museum of Natural History
New Haven, Conn., 06520, U.S.A.

Lists of the publications of the Museum are available from the above
office. These include Postilla, Bulletin, Discovery, special publications,
and available back numbers of the discontinued journal, Bulletin of the
Bingham Oceanographic Collection. All except Discovery are available
in exchange for relevant publications of other scientific institutions
anywhere in the world.

THE PELVIC MUSCULATURE OF THE DINOSAUR
HYPSILOPHODON (REPTILIA: ORNITHISCHIA)

PETER M. GALTON
Peabody Museum of Natural History, Yale University

ABSTRACT

Evidence is provided to show that: 1) the prepubic process,
diagnostic of an ornithischian, did not provide the main support
for the abdomen; 2) the abdomen was large and extended ventral
to the pelvic girdle with a strong M. rectus abdominis that was
not functionally replaced by the lateral abdominal muscles;
3) either the M. pubo-tibialis or the ventral part of the M. pubo-
ischio-femoralis internus originated on the lateral surface of the
prepubic process. An important femoral protractor, the anterior
part of the M. pubo-ischio-femoralis externus, had to be func-
tionally replaced before the pubis changed position to lie close
to the ischium in ornithischians and birds. This was accomplished
by the development of a long anterior process to the ilium with
the differentiation of a large M. ilio-tibialis 1. The lateral curva-
ture of the anterior process of the ilium enabled it to clear the
adjacent ribs and improved the mechanical position of part of
the M. dorsalis trunci, the M. ilio-tibialis 1 and the dorsal part
of the M. pubo-ischio-femoralis internus. A third dorsal muscle
to the femur, the M. ilio-trochantericus, was differentiated in
Hypsilophodon. The well-developed depression at the base of the
fourth trochanter was the main insertion area of the M. caudi-
femoralis longus. The large size of the fourth trochanter of the
ornithopod femur lengthened the moment arm of the M. caudi-
femoralis brevis during the initial part of femoral retraction; its
pendant form resulted from stresses imposed by part of the M.
gastrocnemius which originated on the tendon connecting the
trochanter to the fibula.

POSTILLA 131: 64 p. 15 MAY 1969.

Abstract

Introduction

POSTILLA

CONTENTS

Pelvic Muscles of the Alligator and Birds ....

The Probable Areas of Origin and Insertion of the Pelvic
Muscles in Hypsilophodon .......

A. Axial Musculature

B. Appendicular Musculature...—.......

1. Muscles that extended the tibia. .
M. ilio-tibialis, M. ambiens, M. femoro-tibialis
and M. pubo-tibialis.
2. Muscles that protracted the femur
M. ilio-femoralis and M. pubo-ischio-femoralis
internus.
3. Muscles that flexed the tibia ....
M. ilio-fibularis, M. flexor-tibialis internus,
M. flexor-tibialis externus and M. pubo-
ischio-tibialis.
4. Muscles that retracted the femur
M. pubo-ischio-femoralis internus, M. ilio-tro-
chantericus, M. caudi-femoralis longus and
brevis, M. obturator internus, M. adductor
femoralis and M. ischio-trochantericus.
Discussion
A. The Pubis
1. The Ornithischian Pubis and Sedona ieapeon
2. The Pubis in Birds ...
3. The Abdominal Muscles
4. The Question of Muscle Attachment to the
Lateral Surface of the eae Process
of Hypsilophodon
5. The Pelvic Girdle of Oonischine and Birds
B. The Anterior Process of the Ilium .

ile)

16

HYPSILOPHODON PELVIC MUSCLES

C. The Femur
1. The Greater Trochanter

2. The Fourth Trochanter
a. Possible Functions for the Depression at the

Base of the Fourth Trochanter ... .
b. The Function of the Fourth Trochanter
S ONIN AY gee, op eee aes LP Oh Ot ee
Acknowledgements

Abbreviations / Soe
Literature Cited us he eee: oe

| STECOLRESY (a EAA Ore eae ree ae Re a en ir We a a Tas

4 POSTILLA

INTRODUCTION

The first study of the pelvic musculature of an ornithopod dinosaur
was by Dollo (1883) who discussed the muscles inserting on the
femoral trochanters of /guanodon. Subsequently Dollo (1888)
suggested that the “trochanter creté” of /guanodon was primitive
for ornithopods. He regarded the pendant trochanter of Hypsilo-
phodon and Camptosaurus as a secondary development but
Nopcsa (1905), on the basis of the primitive nature of Hypsilo-
phodon, argued that the reverse was the case. Gregory (1919 and
in Romer, 1927b) made the first attempt to locate the area of
attachment of all the pelvic muscles. Romer contributed several
papers on the pelvic musculature of dinosaurs and related forms:
1923a (dinosaurs and birds), 1923b (alligator), 1923c (sauris-
chian dinosaurs) and 1927a (chick). These studies were followed
(1927b) by a detailed discussion of the pelvic muscles of ornithis-
chians with a reconstruction of the pelvic musculature of the
hypsilophodont Thescelosaurus. The pelvic musculature of the
ceratopsian Chasmosaurus (Russell, 1935) and the hadrosaur
Anatosaurus (Lull & Wright, 1942) has been briefly described
but in both instances Romer’s interpretations were followed.
Janensch (1955) discussed the attachment areas in the region of
the fourth trochanter of the femur of the hypsilophodont
Dysalotosaurus.

Marsh (1878), using material of Dryosaurus (—Laosaurus)
altus (YPM 1876), first showed that the complete ornithischian
pubis consisted of an anteriorly directed prepubic process and a
postpubic rod that lies close to the ischium. However, the presence
of the latter was originally demonstrated by Huxley (1870) using
a specimen of Hypsilophodon |BM(NH) 28707]. Although Dollo
(1888), Nopcesa (1905), Romer (1927b) and Janensch (1955)
referred to Hypsilophodon, the material has not been studied in
detail with reference to the pelvic musculature. Most of the
material is in the British Museum (Natural History), London and
includes several articulated specimens, details of which are given in
Galton (in press). The good specimens consisted of blocks with.
only some of the bone(s) exposed (see Hulke, 1882). I have
prepared most of these in acetic acid so that the bones are com-
pletely free of matrix. Among these specimens is BM(NH) R193
(Figs. 4, 6-11, 13, 15) consisting of a pelvic region and partial

HYPSILOPHODON PELVIC MUSCLES 5

hind limb. The figure of the pubis and ischium of Hypsilophodon
given by Romer (1927b, fig. 7), was based mainly on the figure
of this specimen given by Hulke (1882, pl. 66). The surface
of the bones of the pelvic girdle and femur of this specimen
is very Well preserved so that the boundaries of several areas of
muscle attachment can be located. These areas were identified
using for the main part the descriptions and the reconstruction
of the pelvic musculature of Thescelosaurus given by Romer
(1927b) and information from living forms. In the process it was
possible to compare the areas in these two genera that both belong
to the family Hypsilophodontidae (for diagnosis see Romer, 1956).
It also became apparent that there were several problems that
needed discussion. These, together with certain functional con-
siderations, are dealt with in the second half of this paper (see
below, p. 21).

PELVIC MUSCLES OF THE ALLIGATOR AND BIRDS

The reconstruction of the musculature of an extinct form should
agree as far as possible with the musculature in the closest living
relatives. Dinosaurs are classified in the subclass Archosauria, the
only living order of which is the Crocodilia. Birds are archosaurian
derivatives and, because the pubis lies close to the ischium, are
particularly relevant to the musculature of any ornithischian. Con-
sequently a brief summary of the pelvic musculature of the alligator
and birds is given below. The following classification of the pelvic
muscles of reptiles and birds has been adapted from Romer
hS23b, 1927a, b, 1962).

Axial muscles

A. Dorsal

M. dorsalis trunci and caudae

B. Ventral
M. obliquus abdominis externus and internus,
M. transversus abdominis, M. rectus abdominis

M. ilio-caudalis and M. ischio-caudalis

6

POSTILLA

Appendicular Muscles
A. Dorsal

1)

ii)

To the lower leg

a) M. triceps femoris —M. ilio-tibialis (including
sartorius of birds), M. ambiens and M. femoro-
tibialis

b) M. ilio-fibularis

To the femur

a) M. pubo-ischio-femoralis internus of reptiles, M.
ilio-femoralis internus of birds

b) M. ilio-femoralis of reptiles, M. ilio-femoralis
externus and M. ilio-trochantericus of birds

B. Ventral

1)

ll)

To the lower leg

a) M. pubo-tibialis

b) M. pubo-ischio-tibialis

c) M. flexor tibialis internus of reptiles, M. ischio-
flexorius of birds

d) M. flexor tibialis externus of reptiles, M. caudo-
ilio-flexorius of birds

To the femur

a) M. caudi-femoralis longus and brevis of reptiles,
M. caudo-ilio-femoralis of birds

b) M. adductor femoralis of reptiles, M. pubo-ischio-
femoralis of birds

c) M. pubo-ischio-femoralis externus of reptiles, M.
obturator internus of birds

d) M. ischio-trochantericus of reptiles, M. ischio-
femoralis of birds

In the following summary, unless indicated otherwise, informa-
tion concerning the pelvic muscles of the alligator was taken from:
Romer (1923b) and for birds from Romer (1923b, 1927a,

b). The
individual

areas of attachment and the lines of action of the
muscles are shown for the alligator in Figures 1 and 2

and for an eight day old chick in Figure 3.

HYPSILOPHODON PELVIC MUSCLES 7

Dorsal Axial Musculature

The dorsal axial muscles in the alligator, lizards and Sphenodon
lie lateral to the neural arches and dorsal to the transverse pro-
cesses of the dorsal and caudal vertebrae. The M. dorsalis trunci
has areas of insertion and the M. dorsalis caudae areas of origin on
the inner surface of the dorsal part of the ilium. Further subdivi-
sion of this musclature is unnecessary in the present connection.
The M. dorsalis trunci and caudae are completely separated by
the ilia in birds.

Ventral Axial Musculature

In the alligator anteriorly the three lateral muscles originate
from the lumbodorsal fascia. This fascia attaches to the surface
of the M. dorsalis trunci, the tips of the transverse processes, the
tip of the anterior process of the ilium and the surface of the M.
pubo-ischio-femoralis internus. The M. obliquus abdominis ex-
ternus passes posteroventrally and it has three insertion areas: by
a tendon to the anterior edge of the acetabulum below the M.
ambiens; on the posterodorsal end of the last abdominal rib,
which is connected to the external edge of the pubis by a tendon;
and on an aponeurosis lying above the main part of the M. rectus
abdominis. The first insertion corresponds to the one on the lateral
process of the pubis and the second to that on the pubo-ischiadic
ligament in lizards (Snyder, 1954, fig. 21) and Sphenodon
(Gregory & Camp, 1918, pl. 45, fig. A*). The M. obliquus
abdominis internus passes anteroventrally to insert on the posterior
long rib and the anterior abdominal ribs. The M. transversus
abdominis is the deepest muscle and it passes ventrally to insert
on the deep surface of the M. rectus abdominis.

The three lateral abdominal muscles are slightly different in
birds (George & Berger, 1966). The M. obliquus abdominis
externus originates on the costal margin of the sternum and a
midline raphe anteriorly or a fleshy contact with the muscle of
the other side. It passes dorsally and slightly anteriorly or poste-
riorly depending on the genus to insert on an aponeurosis on the
lateral surface of the ribs and on the ventral edge of the anterior
part of the pubis. The M. obliquus abdominis internus originates
from the last rib, passes posterodorsally and inserts by an aponeu-
rosis on the distal third to two-thirds of the pubis. The origin of

8 POSTILLA

the M. transversus abdominis forms a ventral midline raphe that
extends from the posterior margin of the sternum to the interpubic
ligament. It inserts on the medial surface of the last two or three
ribs and by an aponeurosis and/or by fleshy fibers on the ventral
margin of the pubis. In ratites all three muscles insert on the
whole of the ventral edge of the pubis (Gadow, 1880).

The M. rectus abdominis is large in the alligator as is the case in
Sphenodon and lizards. It originates from the posterior edge of
the sternum and the distal part of the long ribs. The more lateral
part is interrupted by the gastralia and inserts on the last gastralia,
which is connected to the pubis by membranous tissue ventrally
and by a strong tendon laterally. The deeper fibers of the more
medial part insert on the posterior edge of the pubis while the rest
insert On the ventral surface of the M. ischio-caudalis. The M.
rectus abdominis in birds originates from the posterior margin of
the sternum and the last sternal rib; it inserts by an aponeurosis
on the posterior part of the pubis and the interpubic ligament
(George & Berger, 1966). In ratites the M. rectus abdominis
inserts on the distal part of the pubis (Gadow, 1880).

Posteriorly the M. ilio-caudalis and M. ischio-caudalis (=M.
ilio-ischio-caudalis; Romer, 1923b) are separated by the M. caudi-
femoralis in the alligator. The more dorsal M. ilio-caudalis orig-
inates on part of the dorsal edge of the ilium and inserts on the
undersides of the transverse processes and centra of the caudal
vertebrae. The M. ischio-caudalis originates on the posteroexternal
angle of the ischium and inserts on the haemal arches of the tail.
There are four posteroventral axial muscles in birds that control
the movements of the very short tail. The details (George &
Berger, 1966, p. 282), are unimportant because they are not
relevant to the muscles in ornithischians.

Appendicular Muscles
The M. triceps femoris consists of three parts:

M. ilio-tibialis. This muscle originates along most of the dorsal
margin of the ilium and is divided into two parts in lizards, three
in crocodiles and five in birds. All these origins are tendinous
except the most anterior one in birds which is fleshy. The inser-
tion in all forms joins the tendon of the M. femoro-tibialis on the
cnemial crest of the tibia.

HYPSILOPHODON PELVIC MUSCLES 9

M. ambiens in reptiles and birds has a tendinous origin just
anterior to the acetabulum close to the ilio-pubic boundary and
inserts on the tendon attaching to the cnemial crest. A second
tendon crosses the knee to the fibular side in crocodiles and birds
to join the external head of the M. gastrocnemius.

M. femoro-tibialis has a fleshy origin from much of the femoral
shaft. In the alligator there are two parts but in birds there are
several additional small heads. The tendinous insertion is on the
cnemial crest of the tibia.

M. ilio-fibularis. This muscle in crocodiles and birds has a
large fleshy origin on the postacetabular part of the ilium and is
overlain by the M. pubo-tibialis. It inserts on the fibula and also
on the external head of the M. gastrocnemius.

There are two dorsal muscles inserting on the femur in modern
reptiles. The M. ilio-femoralis originates on the outer surface of
the ilium and inserts on the posterior surface of the femoral
shaft. The two parts of the M. pubo-ischio-femoralis internus
primitively originate on the medial surface of the pubis. In the
alligator they have a more dorsal origin. The main part originates
on the ventral surfaces of the transverse processes of the last six
dorsal vertebrae; it inserts on the dorsal surface of the proximal
part of the femur. The more ventral part originates on the internal
surface of the ilium and ischium and the ventral surfaces of the
sacral ribs and passes ventrally to insert more anteroventrally
on the femur.

There are three dorsal muscles inserting on the femur in birds
that all originate on the outer surface of the ilium. The homo-
logues of the.two reptilian muscles are the M. ilio-femoralis ex-
ternus and the more anterior M. ilio-femoralis internus. The
muscle in between is the M. ilio-trochantericus that inserts in
part on the greater trochanter.

The M. pubo-tibialis of primitive reptiles is lost. in crocodiles
and birds. The M. pubo-ischio-tibialis of lizards is a large ventral
muscle (Snyder, 1954). It is reduced to a slip that originates on
the lateral surface of the ischium below the! acetabulum in the
alligator. It is lost in birds.

The M. flexor-tibialis internus has three heads in the alligator.
The first originates on the medial surface of the ischium near the
posteroventral corner, the second on the outer surface postero-

10 POSTILLA

ventral to the acetabulum and the third is on the posterior corner
of the ilium. These three heads have a double tendon, one part
inserting on the medial surface of the tibia and the other passing
along the external head of the M. gastrocnemius. The only part
in birds is the M. ischio-flexorius.

The M. flexor-tibialis externus arises from the posterior angle
of the ilium in the alligator and from the tail in birds (M. caudo-
ilio-flexorius). It inserts on the medial aspect of the head of
the tibia.

The M. caudi-femoralis brevis originates on the posteroventral
edge of the ilium and the sides of the centra of the last sacral and
the first caudal vertebra in the alligator. It inserts on the dorsal
surface of the fourth trochanter of the femur. The iliac portion
is more important in birds.

The M. caudi-femoralis longus originates on the sides of the
centra and the ventral surface of the transverse processes of caudal
vertebrae 3 to 15 in the alligator. The thick tendon inserting on
the fourth trochanter is connected to the head of the fibula by
a thin tendon that is also present in lizards and birds.

M. adductor femoris has two heads that originate on the outer
surface of the ischium. These two heads are separated by a portion
of the M. pubo-ischio-femoralis externus in the alligator but not in
birds. The insertion is a long and narrow area on the ventral side
of the femur.

M. pubo-ischio-femoralis externus has three heads in the al-
ligator. The first is on the ventrolateral surface of the pubis, the
second on the dorsomedial surface of the pubis and the third on
the outer surface of the ischium. The common tendon inserts on
the posteroventral edge of the femur close to its head. This muscle
has only one head in birds, the M. obturator internus, which orig-
inates on the ventral edge of the ischium and the dorsal edge of
the pubis.

M. ischio-trochantericus of alligators is a small muscle which
originates on the posterior part of the inner surface of the ischium.
Its tendinous insertion is on the outer dorsal edge of the femur
close to its head. This muscle originates on the outer surface of.
the distal part of the ischium and neighboring pelvic bones in birds.

HYPSILOPHODON PELVIC MUSCLES 1]

THE PROBABLE AREAS OF ORIGIN AND INSERTION OF THE PELVIC
MUSCLES IN HyPSILOPHODON

Romer (1927b) discussed the probable attachment areas of each
pelvic muscle in ornithischians and gave a detailed reconstruction
of these muscles for Thescelosaurus (Figs. 5, 16). His study has
been used to interpret certain well-defined surface markings on
BMIGNED) R193 (Figs. 6-11, 13, 15), a specimen of Hypsilo-
phodon that has an extremely well-preserved ilium, ischium, pubis
and femur. The following section provides a summary of the
evidence shown by this specimen but the reasons for certain of
the interpretations are discussed in later sections. Data for other
ornithischians are taken from Romer (1927b) unless otherwise
indicated. Romer (1927b) applied avian names to several of the
muscles of ornithischians. However, as ornithischians are reptiles
I have not followed this practice unless there is good evidence
that the muscle in Hypsilophodon resembled that of birds. When
Romer (1927b) used the avian term this is given in parenthesis
after the appropriate section heading (see also the classification
of muscles given on p. 5).

A. AXIAL MUSCULATURE

Only the more superficial part of the dorsal axial musculature
attached to the ilium (Figs. 4, 6). Part of the M. dorsalis trunci
probably inserted on the medial surface of the anterior process,
which is strengthened by a couple of longitudinal ridges. The
square and rugose posterior edge of the ilium probably provided
an area of origin for part of the M. dorsalis caudae.

The reconstruction of the ventral axial muscles anterior to
the pelvic girdle must be rather tentative but I consider that the
following is a reasonable approximation. The M. obliquus abdo-
minis internus probably originated from the posterior edge of
the sternum and by a midline raphe or a fleshy contact with the
muscle of the other side. This muscle passed laterally and postero-
dorsally to insert on a dorsal fascia, similar to that in the alligator,
and on the pubis. The M. transversus abdominis originated from a
midline raphe and passed dorsolaterally to insert on the dorsal
fascia and on the pubis. The M. obliquus abdominis externus
originated from a midline raphe and from the lateral edge of the
M. rectus abdominis and passed laterally and anterodorsally to

12 POSTILLA

insert on the pubis and on the dorsal fascia. The M. rectus
abdominis originated from the posterior edge of the sternum
and inserted on the distal part of the pubis and on the expanded
distal end of the ischium.

The three lateral abdominal muscles probably inserted on the
ventral edge of the pubis, as in birds. The insertion of the M.
obliquus abdominis externus may have been restricted to the
more anterior part of the pubis and that of the pars internus to the
more posterior part as is the case in carinate birds (George &
Berger, 1966). It would appear more likely that both these mus-
cles, like the M. transversus abdominus, inserted along the whole
of the ventral edge of the pubis (Fig. 6) as is the case in ratites
(Gadow, 1880). It is possible that more distally the area of
insertion of these three muscles was on the adjacent surface of
the ischium. The problem of the abdominal musculature is dis-
cussed on page 21.

B. APPENDICULAR MUSCULATURE

The functional grouping of the limb muscles of Hypsilophodon is
simplified because the action of the hind limbs was mainly
anteroposterior.

1. Muscles that Extended the Tibia.

The three parts of the M. triceps femoris provided much of the
propulsive force. The more anterior portion of the M. ilio-tibialis
also raised the leg during forward movement and effected minor
adduction and abduction. The insertion of all parts was presumably
a tendinous one onto the cnemial crest of the tibia. |

a) M. ilio-tibialis. This originated on the dorsal margin of the
ilium. The dorsal edge of the ilium in lateral view (Fig. 6) has a
definite bevel running from the posterior corner onto the ante-
rior process. The complete bevel is visible in BM(NH) R196 and
that part of it present in BM(NH) R193 has well-marked muscle
scars. A separate portion, the M. ilio-tibialis 1 (the sartorius) |
was probably differentiated on the anterior process of the ilium
as in birds. The posterior boundary of this portion is indicated
by a rugose line running across the process in hadrosaurs and
Thescelosaurus (Fig. 5); this is not present in Hypsilophodon.

HYPSILOPHODON PELVIC MUSCLES 13

b) M. ambiens. The origin on the pubis seems to be just
anterior to the acetabulum. Romer (1927b) showed that this
region varies in Thescelosaurus because a distinct bump is pre-
served in the type (Gilmore, 1915, fig. 17) that is lacking on
specimens at the American Museum of Natural History. Hypsilo-
phodon specimens BM(NH) R193, R195 and R196 show muscle
scars on the dorsolateral edge of the prepubic process close to the
acetabular region (Figs. 4, 6, 7) in the same position as the
bump in Thescelosaurus. The main insertion was presumably on
the cnemial crest of the tibia but, as in crocodiles and birds,
there may also have been a tendon that crossed to the fibular side
of the leg to insert on the external head of the M. gastrocnemius.

c) M. femoro-tibialis. Unlike the other two divisions of the
M. triceps femoris there was no possibility of the M. femoro-
tibialis functioning to protract the femur. Among archosaurs there
was: a general tendency for the M. femoro-tibialis to enlarge its
area of insertion (Romer, 1923b, c; 1927b) so that it covered
much of the femoral surface. Concerning Thescelosaurus, Romer
(1927b, p. 261) wrote that “the femoro-tibialis origin, as is
usually the case, is not well defined, but apparently includes a
main area lying on the dorsal (anterior) surfaces of the femur,
bounded externally by a ridge extending down from the ‘lesser
trochanter’. Probably a head originated on the ventral (posterior)
surface external to the ridge, while a third (not seen in the figures)
extended around the internal margin towards the ventral surface
interior and anterior to the coccygeo-femoral insertion.” These
three divisions are referred to below as M. femoro-tibialis 1, 3 and
2 respectively and their areas of origin are clearly seen in the
femur of Hypsilophodon, BM(NH) R193 (Figs. 8-10). It should
be noted that the ridge below the lesser trochanter in Thescelo-
saurus is not equivalent to the groove in this position in Hypsilo-
phodon but corresponds to the more posteriorly placed ridge.

M. femoro-tibialis 1. The area of insertion of this muscle is
delimited by two longitudinal features on the femur. At the base
of the lesser trochanter there is a well-defined shallow groove
(Fig. 10C), with surface markings, which becomes wider close
to the broken end of the bone. On the outer surface there is a
strong and curved ridge (Figs. 8A, 9A, 10B) which commences
level with the top of the fourth trochanter. Distally this ridge

14 POSTILLA

curves inward and widens to form a flat area with surface striations
(Figs. 8A, 9A, 10A, B). The area bounded by these two features
is convex and its anterior limit is marked by a series of rugose
markings, which extend from the base of the lesser trochanter to
the flatter area at the end of the ridge. There are few surface mark-
ings on this area but distally there is a diagonal system of low
parallel ridges and grooves (Figs. 8A, 9A).

M. femoro-tibialis 2. The insertion area of this muscle is con-
vex and is mainly bounded by the M. femoro-tibialis 1 and 3.
The proximal extent of this muscle is not certain. There are some
rugose markings, situated at the level of the lower part of the
depression at the base of the fourth trochanter, which may mark
the proximal limit (Figs. 8B, 9B, 10C, D). Alternatively, this
muscle may have extended medial to the depression toward the
base of the lesser trochanter.

M. femoro-tibialis 3. The area of insertion of this muscle is
bounded externally by the sharp ridge mentioned above. Internally
the proximal boundary is a faint line, formed by a series of small
depressions, that extends from the proximal end of the ridge onto
the base of the fourth trochanter. More distally on the trochanter
this boundary continues as a rugose ridge (Figs. 8A, 9A, 10A)
and then by an edge formed by the thickening of the trochanter
(Fig. 10B). Distal to the trochanter the internal boundary is a
line of small depressions extending from the distal end of the
well-developed depression at the base of the fourth trochanter
(Figs. 8B, 9B, 10D). The surface for the origin of this muscle
is mostly concave but internally, in the region distal to the fourth
trochanter, it is convex. A part of this muscle probably originated
on the central area underneath the pendant part of the fourth
trochanter (Figs. 8A, 9A, 10B) where there are surface markings.

d) M. pubo-tibialis. There is a well-defined area on the lateral
surface of the prepubic process of Hypsilophodon on which the
M. pubo-tibialis or the ventral part of the M. pubo-ischio-femoralis
internus may have originated (see below, p. 28). The lateral area
is best preserved in BM(NH) R193 (Figs. 4, 6, 7). The anterior -
part is also shown on BM(NH) R196 and the posterior part on
BM(NH) R195 and R5829. The surface markings cease ante-
riorly close to the end of the process. Ventrally the area is de-
limited by a groove running along the corner of the process (Fig.

HYPSILOPHODON PELVIC MUSCLES 15

7). Posteriorly the surface of the prepubic process is concave, so
that the area projects laterally (Fig. 11). Dorsally the markings
continue to the rounded edge of the process. The surface markings
are best developed posteriorly, where the area is wider, making an
angle of about 45° to the long axis of the process (Figs. 6, 7).
These same markings are present in BM(NH) R5829 but the
area does not project laterally. Anteriorly, the area is narrower
and the markings have no apparent direction. This part is the
same in BM(NH) R196 where the markings stop a short dis-
tance from the end. If the M. pubo-tibialis was present, it would
presumably have inserted on the medial surface of the tibial head.

The prepubic process of Dryosaurus altus (YPM 1876) has an
area on the lateral surface similar to that in Hypsilophodon. Pos-
teriorly the well-defined muscle scars are diagonally inclined and
the area continues onto the dorsolaterally facing surface of the
slightly flattened and expanded anterior part where the markings
are longitudinal. The prepubic process of Thescelosaurus (AMNH
117) is similar to that of Dryosaurus but the posterior region is
less well-defined and the markings less diagonally inclined. The
markings on the prepubic process of most other ornithischians
examined are, as Romer (1927b) noted, longitudinal stria-
tions. This is true for Thescelosaurus neglectus (NMC 8537,
USNM 2210 and AMNH 6120), J/guanodon atherfieldensis
[BM(NH) R5764] and the mounted skeletons of hadrosaurs and
ceratopsians in the Peabody Museum of Yale University, United
States National Museum, National Museum of Canada and the
Royal Ontario Museum, Toronto. In some of the mounted hadro-
saurs at the American Museum of Natural History (Procheneo-
saurus, AMNH 5461; Saurolophus, AMNH 5220; and Cory-
thosaurus, AMNH 5240) the pattern is slightly different. In these
specimens the striations radiate from the narrow part into the
dorsoventrally expanded anterior part of the process. As a result
the most dorsal and ventral striations are parallel to the ap-
propriate edge of the process.

2. Muscles that Protracted the Femur.

a) M. ilio-femoralis(—=M. _ ilio-femoralis externus; Romer,
1927b). Romer (1927b) concluded that the M. ilio-femoralis
Originated on the antitrochanter of the ilium of hadrosaurs. In

16 POSTILLA

Thescelosaurus (Fig. 5) and Hypsilophodon (Fig. 6), in which
the area of origin cannot be distinguished, the area shown on the
ilium is comparable in position to that'of the hadrosaurian anti-
trochanter. The insertion of the M. ilio-femoralis completely
covered the lesser trochanter (Figs. 4, 10); it probably extended
to the depression at the base on the anterior edge (Fig. 10C).
This muscle helped to prevent disarticulation of the femur.

b) M. pubo-ischio-femoralis internus(—=M. ilio-femoralis in-
ternus; Romer, 1927b). The ventral part of this muscle, rather
than the M. pubo-tibialis, may have originated on the well-defined
area on the lateral surface of the prepubic process (see p. 14, 28).
A possible area of insertion for this part was on the femoral shaft
proximal to the depression at the base of the fourth trochanter
(Figs. 8B, 9B, 10D) and possibly also proximal to M. femoro-
tibialis 2. On the ventral surface of the head and neck there is an
area, with very strongly developed markings (Figs. 8B, 9B, 10C),
which probably was not for this ventral part but for ligaments that
held the head of the femur in the acetabulum.

3. Muscles that Flexed the Tibia.

a) M. ilio-fibularis. There is no indication of the area of
origin of this muscle on the ilium in either Thescelosaurus or
Hypsilophodon. Romer (1927b) concluded that it originated on
the ilium posterior to the antitrochanter in hadrosaurs. The M.
ilio-fibularis is shown in a comparable position in Thescelosaurus
(Fig. 5) and Hypsilophodon (Fig. 6). The insertion was on the
lateral surface of the fibula head. In BM(NH) RS5830 there is a
well-defined area with muscle scars on the posterior part of this
surface which was probably for the M. ilio-fibularis (Fig. 4). This
muscle, together with the other flexors, aided in retracting the
femur.

b) M. flexor-tibialis internus(—=M._ ischio-flexorius; Romer,
1927b). Romer (1927b) showed that in Thescelosaurus the.
origin of this muscle was on a projection on the dorsal edge of
the ischium just distal to the end of the obturator process (Fig.
5S). There is no comparable projection in Hypsilophodon so its
position in Thescelosaurus has been used in the reconstruction

HYPSILOPHODON PELVIC MUSCLES 17

(Fig. 6). A second head of the M. flexor-tibialis internus may
have originated from the postacetabular part of the ilium as in
the alligator but there is no evidence that this slip was present.
The main insertion was presumably a tendinous one on the poste-
rior part of the tibial head. A second tendon may have inserted
on the external head of the M. gastrocnemius as in the alligator.

c) M. flexor-tibialis externus(—M. ilio-flexorius, Romer, 1927b).
This muscle was probably present, but its area of origin has not
been observed on the posterior part of the ilium of any ornithis-
chian. The insertion was presumably on the posterior part of
the tibial head.

d) M. pubo-ischio-tibialis. This muscle is reduced to a slip
in the alligator and is lost in birds as was probably also the case in
ornithischians.

4. Muscles that Retracted the Femur.

a) M. pubo-ischio-femoralis internus(—=M. ilio-femoralis in-
ternus; Romer, 1927b). The more dorsal part of this muscle
Originated from the ventral surface of the transverse processes
of the more posterior dorsal vertebrae. Romer (1927b) deduced
that this muscle inserted on the greater trochanter in Thescelo-
saurus (Fig. 16). On the outer surface at the base of the greater
trochanter in Hypsilophodon, there is a very prominent ridge that
continues distally (Figs. 8A, 9A, 10B). Proximally, this ridge
is sharp edged because it is the line of contact between two con-
cave surfaces. More distally the anterior surface becomes convex
and the posterior concave area cuts a distinct edge into this. The
anterior concave area has strong longitudinal striations. The pos-
terior area has only faint striations proximally but distally there
are strongly developed and irregular markings on BM(NH) R193.
In BM(NH) R2487 the posterior area is covered by longitudinal
striations.

The two surfaces evidently formed an extension onto the shaft
of the insertion area of two muscles that were attached to the
greater trochanter. The M. pubo-ischio-femoralis internus probably
inserted on the posterior area and the M. ilio-trochantericus
on the anterior area (see p. 36).

18 POSTILLA

b) M. ilio-trochantericus. The area of insertion of this muscle
has just been discussed. Romer (1927b) deduced that in hadro-
saurs the M. ilio-trochantericus originated from the lateral surface
of the ilium anterior to the antitrochanter. The origin of this
muscle is shown in a comparable position in Thescelosaurus
(Fig. 5) and Hypsilophodon (Fig. 6). The M. ilio-trochantericus
helped to prevent disarticulation of the femur.

c) M. caudi-femoralis longus and brevis(—=M. _ coccygeo-
femoralis longus and brevis; Romer, 1927b). The M. caudi-
femoralis longus probably originated on the sides of the centra
and the ventral surface of the transverse processes of the anterior
third of the tail. Romer (1927b) pointed out that in many primi-
tive ornithischians, including Thescelosaurus, the ilium has a
broad brevis shelf from which the M. caudi-femoralis brevis
originated. This muscle probably originated on the broad brevis
shelf in Hypsilophodon (Figs. 4, 6) and, in addition, may have
originated from the sides of the centra of the last sacral and the
first caudal vertebra as in the alligator. Both of these muscles
inserted on the fourth trochanter of the femur.

The outer surface of the fourth trochanter of Hypsilophodon
continues the concave curve of the adjacent shaft (Fig. 10A).
More distally on the outer surface there is a ridge extending to
the tip of the trochanter (Figs. 8A, 9A, 10B). The position of
this ridge relative to the distal edge of the trochanter varies—
it is progressively farther away in the series BM(NH) R193,
R195 and R196. Between this ridge and the dorsal edge there
is a concave depression that proximally becomes indistinct as the
ridge disappears. In BM(NH) R193 (Figs. 8A, 9A, 10A, B)
this depression is rugose; it can be followed proximally where its
boundary is formed by two series of small ridges that are parallel
to the dorsal edge. The line of these ridges is continued onto the
shaft of the femur as a faint line formed by a series of small
depressions (Figs. 8A, 9A, 1OA). In BM(NH) R195 there is a
well-defined series of fine ridges running across this concave
depression. In BM(NH) R196 the depression is less well defined
and there are only faint insertion markings.

Distally the dorsal edge of the fourth trochanter is sharp because
there is another concave area, with strongly developed insertion

HYPSILOPHODON PELVIC MUSCLES 19

markings, on the inner side of the pendant portion of the tro-
chanter (Figs. 8B, 9B, 10D). Proximally the dorsal edge is
thickened on this side to form a strong ridge with very well-
developed but irregular insertion markings (Figs. 8B, 9B, 10D).
Internal to this ridge there is a smooth concave area that widens
out distally between the strongly concave area on the pendant
part of the trochanter and the depression on the shaft of the
femur (Figs. 8B, 9B, 10D).

In BM(NH) R193 (Figs. 8B, 9B, 10D), R195 and R2477a
the depression at the base of the fourth trochanter is deep with a
well-defined edge anteriorly and distally. Proximally the depression
merges w:th the shaft surface and posteriorly it merges with the
smooth concave area below the dorsal ridge. The whole of the
surface of this depression is rough with extremely well-developed
insertion markings. In BM(NH) R196, R5829 and R5830 there
is a well-defined area, with irregular and strong insertion mark-
ings, in the same position on the shaft but forming only a slight
depression. Distally this area is slightly higher than the adjacent
part of the femoral shaft. The remaining boundaries of this area
can only be distinguished by the presence of insertion markings.

I consider that the main insertion of the M. caudi-femoralis longus
was on the medial depression and that the pars brevis inserted on
the lateral surface of the fourth trochanter. The anterior limit of
M. caudi-femoralis brevis is indicated by the line approximately
parallel to the edge of the trochanter (Figs. 8A, 9A, 10A). Con-
sequently the pars brevis inserted on a marginal zone on the
lateral surface that extends the complete length of the trochanter.
Proximally the brevis may have extended slightly onto the shaft
but no boundary is visible. The brevis probably wrapped round
the edge of the trochanter to insert on the medial surface of the
strong ridge (Figs. 8B, 9B, 10D). The medial depression on the
pendant part of the fourth trochanter was for a tendon which
extended to the head of the fibula (Figs. 4, 8B, 9B, 10D). The
lateral surface of the base of the M. caudi-femoralis longus was
closely applied to the adjacent structures (Figs. 8B, 9B, 10D):
the unoccupied medial surface of the trochanter, the surface of
the tendon to the knee and the medial surface of the brevis. The
fourth trochanter and the depression at its base is discussed on
page 38.

20 POSTILLA

d) M. obturator internus(—M. pubo-ischio-femoralis exter-
nus, anterior part). Romer (1927b) thought that the posterior
part of the M. pubo-ischio-femoralis externus had been lost in
ornithischians and that it was replaced by the anterior part, which
originated from ligaments connecting the postpubic rod and the
ischium. There is no evidence to show whether or not the poste-
rior part was lost in ornithischians but it is reasonable to conclude
that it was as is the case in birds. In Hypsilophodon the postpubic
rod has a sharp upper edge on which the puboischiadic ligament
could have attached. Specimen BM(NH) R193 has surface mark-
ings running below the dorsal edge and passing below the obtura-
tor foramen (Figs. 6, 7) that were probably for the puboischiadic
ligaments. Romer (1927b) stated that there is no sure indication
of the area of origin of the M. obturator internus on the ischium
of Thescelosaurus but that in hadrosaurs there is a ridge running
lengthwise that marks its dorsal border. In Hypsilophodon the
ischium of BM(NH) R193 has on its outer surface a definite
ridge running along just below the middle of the shaft which
marks the dorsal limit of the M. obturator internus (Figs. 6, 7).
The lateral surface of the obturator process probably formed
another surface for the origin of this muscle but no definite
boundaries can be seen in this region or along the length of the
blade. Romer (1927b) did not find any indication of the inser-
tion area; he noted that it is on the ventral surface of the head
in living forms. The femur of BM(NH) R193 of Hypsilophodon
(Fig. 1OA) has a small, faint circular rugose area that was pos-
sibly the area of insertion for the tendon of this muscle.

e) M. adductor femoralis. Romer (1927b) placed the origin
of the two heads of this muscle on the shaft region of the ischium
in Thescelosaurus (Fig. 5) where there is a series of rugose mark-
ings between the origin areas of the M. obturator internus and the
M. ischio-trochantericus. In Hypsilophodon there is a longitudinal
depression on the shaft of the ischium [BM(NH) R193, Figs.
6, 7 and BM(NH) R195] bounded by a slight ridge dorsally and
ventally. Romer (1927b) noted that the insertion of the M. -
adductor femoralis is sometimes shown by a slight indication
on a line from the fourth trochanter towards the outer condyle.
This has not been located on the femora of Hypsilophodon unless
it is represented by an area on the posterolateral part of the shaft

HYPSILOPHODON PELVIC MUSCLES 21

in BM(NH) R193 (Figs. 8A, 9A, 1OA, B). This area bears well-
developed longitudinal striations and is separated by a sharp edge
from the insertion area of M. femoro-tibialis 3.

f) M. ischio-trochantericus. In Thescelosaurus (Fig. 5) this
muscle probably originated on the flattened upper surface internal
to the dorsal edge of the ischium, which forms a prominent shelf
below this muscle (Romer, 1927b). Romer noted that in all forms
in which the position is determinable except Camptosaurus and
Protoiguanodon there is no prominent shelf and the muscle
originated from the outer surface. A rugose or ridged line running
along most of the length of the ischium marks the ventral limit
of this muscle in these forms.

In Hypsilophodon (Figs. 6, 7) the posterior part of the M.
ischio-trochantericus probably originated on the flattened inner
surface of the blade of the ischium as in Thescelosaurus. The
dorsal margin of the shaft region does not form a distinct ledge
and the muscle probably originated on its rounded margin,
extending slightly onto the outer surface as far as the dorsal limit
of the adductor muscles (Fig. 6). This dorsal margin shows very
strongly developed markings in BM(NH) R195 and R196.

Romer (1927b) noted that the area of insertion of the M.
ischio-trochantericus near the head of the femur is not visible.
In his reconstruction of Thescelosaurus the insertion is shown very
close to the head of the femur, behind the greater trochanter.
In Hypsilophodon it probably inserted in the same region (Figs.
4, 8A, 9A, 10A).

DISCUSSION

Several problems of an interpretational or functional nature be-
came apparent while trying to identify the areas of muscular
attachment described in the above section. The conclusions rele-
vant to the identification of attachment areas were included but
the reasons still need to be considered.

A. THE PUBIS
1. The Ornithischian Pubis and Abdominal Support.

The ornithischian pelvic girdle has a characteristic tetraradiate
form (Figs. 5, 6) with the pubis forming an anteriorly directed
“prepubic process” and a posterior “postpubic rod”. The prepubic

22 POSTILLA

process is diagnostic of the order Ornithischia, one of the two
orders of dinosaurs which were so. successful during the Mesozoic
era. The presence of a rod-like pubis close to the ischium occurs
in only one other group, the Class Aves (Fig. 14D, E). Con-
sequently it is important that the functional significance of the
similarities and differences of the form of the pelvic girdle of
ornithischians and birds be understood.

The prepubic process is generally regarded as a new structure
with the postpubic rod as part of the normal reptilian pubis that,
early in the history of the Ornithischia, came to lie close to the
ischium. However, the postpubic rod had originally been regarded
as a new development (Marsh, 1878, p. 451; more recently,
Lebedinsky, 1913) but it is difficult to account for its development
in terms of a muscle shifting its attachment area posteriorly.
As Romer (1927b) maintained, a muscle involved in such a
shift would probably have passed onto the adjacent anterior edge
of the ischium. He thought it unlikely that a thin rod, lying close
to the ischium, would have been tormed because this is mechani-
cally weaker. In addition, I consider that the implied evolutionary
history of this “new” structure would be rather improbable be-
cause, after the lengthening of the rod until it was as long as the
ischium in hypsilophodonts and camptosaurs, which on other
characters are regarded as primitive Ornithopoda, it was then
considerably shortened in the advanced Ornithopoda iguanodonts
and hadrosaurs as well as in psittacosaurs and ceratopsians.

The pubis of birds has a postpubic rod similar to that of orni-
thischians and an anteriorly directly pectineal process (Figs. 3,
14D) for the M. ambiens. The pectineal process is best developed
in ratites (Lebedinsky, 1914) but proportionally it is always
minute in comparison with any prepubic process. The postpubic
red of birds clearly represents the original reptilian pubis because
in the earliest bird known, the Jurassic Archaeopteryx (Fig. 14E),
there is no pectineal process. In modern birds during embryonic
development the pubis swings back to lie close to the ischium
(Romer, 19274). It seems likely; as Romer” (1927b,, 1956)
believed, that the same thing occurred during the embryonic devel-
opment of ornithischians.

In most reptiles the abdominal muscles insert on the anterior
margin of the pubis and, if it is present, on the pubo-ischiadic
ligament. As noted the postpubic rod probably represents the

HYPSILOPHODON PELVIC MUSCLES 23

original reptilian pubis that, during the evolution of the Ornithis-
chia, changed position to lie close to the ischium. Romer (1927b)
pointed out that this change would have resulted in longer
abdominal muscles under heavy pressure and without adequate
support. He accounted for the development of the prepubic
process of ornithischians as an adaptation to offset this weakness.
Romer considered that a structure comparable to the processus
lateralis of the Lacertilia lengthened and took over most of the
obliquii muscles and possibly part of the rectus abdominis
muscle. According to Romer the main part of the pubis could
change position only after it had been released from its role of
supporting the abdominal muscles. Romer (1927b, p. 246) sug-
gested that the rotation provided the obturator muscle with a
more advantageous position. Romer cited several points that he
thought tended to confirm that the ornithischian prepubic process
provided the main support for the abdomen: the direction of
the prepubic process; its form, which in many types is a long
thin blade; and when present, the muscle markings, which are
longitudinal striations. Further evidence was obtained from a
mounted skeleton of Triceratops in which the prepubic process
was orientated so that its anterior end was in the plane of the
ribs (Fig. 12). Romer (1927b, p. 244) pointed out that on the
posterior border of the adjacent long dorsal rib there are “indica-
tions of a muscular or ligamentous attachment of considerable
strength which matched in size and position the end of the prepubic
process. This quite definitely seems to show that abdominal sup-
port was a major, although perhaps not the sole function of the
process.” Later, in discussing the musculature of Thescelosaurus,
he argued (p. 260) that the M. rectus abdominis was probably not
of great strength, “as suggested by the lack of any well-marked
area for its insertion”, and that the obliquii muscles had taken
over the great part of the abdominal support. Romer followed
W. K. Gregory in considering that there was a considerable
constriction of the waist region, with littlke more than a flap of
skin below the girdle, so that the thighs were not restricted.

2. The Pubis in Birds.

The pubis of birds lies close to the ischium and there is no
prepubic process to provide support of the abdomen. Romer
(1927b, 1956) considered that a prepubic process was unneces-

24 POSTILLA

sary because of the shortness of the abdomen and the large
sternum, which extends almost to the pelvic region. As a result the
abdominal muscles have only a comparatively small space to cross.
This is certainly the case in carinate birds that are extremely
specialized for flying, but not in the earliest known bird, the
Jurassic Archaeopteryx. In Archaeopteryx the pubis is directed
backward and there is neither an anterior prepubic process nor a
pectineal process (Fig. 14E). The abdominal area with 12 dorsal
vertebrae (Heilmann, 1926, fig. 23; and in Romer, 1966, fig. 253)
is proportionally almost as extensive as in ornithischians, which
have 15 dorsal vertebrae. The sternum of Archaeopteryx could not
have provided much support for the abdomen because, although
broad, it is very short (Beer, 1954). It is therefore apparent
that the short abdomen and the large sternum of birds developed
after the pubis had changed position and, as a result, these fea-
tures cannot be cited to explain the absence of a prepubic process
in birds. Archaeopteryx, however, had at least 9 to 10 pairs of
gastralia (Heilmann, 1926, fig. 7) which would have provided
accessory support for the abdominal muscles.

Archaeopteryx was about the size of a large pigeon, which
may reduce its relevance to the present inquiry, but the size of
the early ornithischians in which the pubis shift occurred is not
known. Gastralia are primitively present in reptiles of all groups
and their occurrence in Sphenodon, Archaeopteryx, crocodiles,
phytosaurs, thecodonts and some saurischians strongly sug-
gests that they were present in primitive ornithischians. They
have been reported only from Stegoceras (Gilmore, 1924a), but
these may have belonged to a saurischian dinosaur. It must be
concluded that gastralia were either cartilaginous or absent in most
ornithischians. The possible absence of gastralia in most ornithis-
chians does not necessarily mean that the prepubic process was
essential for abdominal support because gastralia are also absent
in ratites. The sternum of ratites is small and the abdominal mus-
cles span a distance that is comparable to that in an ornithischian
of the same height (see figure Struthio in Gadow & Selenka, 1891,
pl. 5, fig. 1 and Dinornis in Romer, 1966, fig. 257). There is no
accessory support provided by gastralia, a large sternum or a
prepubic process yet some moas reached a height of 10-11 feet
(Romer, 1966). An ornithopod of this height would have been
at least 20 feet long. In addition there is no accessory support

HYPSILOPHODON PELVIC MUSCLES 25

in any large mammal in which the ventral abdominal muscles
are very long (see figures in Romer, 1966).

To summarize, Archaeopteryx shows that the pubis can lie
close to the ischium without the development of a prepubic process,
a large sternum or a short abdomen. Although gastralia were
probably present in early ornithischians, their absence would be
no problem because the ventral abdominal muscles of ratites and
large mammals can span long distances without any accessory
support. Consequently I consider that the basic assumption made
by Romer (1927b) is incorrect. The arguments used to show that
the obliquii muscles had functionally replaced the M. rectus
abdominis as the principal supporter of the abdomen. will now
be considered.

3. The Abdominal Muscles.

The mounted skeleton of Triceratops cited by Romer (1927b)
is shown in Figure 12 (see also Osborn, 1933, for other views;
for another mount see Erickson, 1966). The presence of a strongly
developed ligamentous or muscular connection between the last
dorsal rib and the prepubic process would certainly have braced
and strengthened the side wall of the abdominal cavity. The high
position of this process in Triceratops, like that in other ornithis-
chians must, however, have limited its effectiveness in providing
the main supporting point for the abdomen. According to Romer
(1927b, fig. 17) for Thescelosaurus the M. obliquus abdominis
passed posterodorsally to insert on the anterior end of the
prepubic process (Fig. 5). Consequently in Triceratops the poste-
rior limit of the effective abdominal cavity is marked by the last
long dorsal rib (Fig. 12). This would give a short abdominal
cavity and, as Romer noted, little more than a flap of skin
below the pelvic girdle. This is a rather strange condition for
herbivorous animals which in many cases reached a considerable
size. Herbivorous mammals are characterized by a barrel-like rib
cage and a large abdominal cavity to contain the greatly elongated
intestine required to digest plants. The rib cage of Triceratops is
certainly barrel-like as shown in Figure 12, in Osborn (1933)
and even more clearly in Erickson (1966, pls. 1, 2) where the
ventral part of the rib cage has been restored. It would seem more
logical and provide a larger abdominal cavity if the principal

26 POSTILLA

muscles that supported the abdomen continued posteriorly and
attached to the distal end of the ischium.

In marsupials there is a separate anterior part of the pubis,
the epipubic bone, that is definitely important for abdominal sup-
port with the M. obliquus abdominis internus and externus insert-
ing along its length (Elftman, 1929). In some marsupials, such
as the wombat Phascolomys and especially the koala Phascolartos
(Elftman, 1929, pls. 13, 14), the epipubic bone is very well devel-
oped and the abdomen large. In both these genera the epipubic
bone is more ventrally directed so that the anterior end is propor-
tionally much lower than that of the prepubic process of most
ornithischians, e. g. Hypsilophodon (Fig. 4), Thescelosaurus (Fig.
5) and Triceratops (Fig. 12; for figures of other ornithischians
see Colbert, 1961; Romer, 1966).

Romer (1927b) considered that the M. rectus abdominis was
weak because there was no well-defined area for its insertion. In
his reconstruction of Thescelosaurus Romer (1927b, fig. 16)
showed the M. rectus abdominis attached only to the slender post-
pubic rod (Fig. 5). Although not specifically stated, it is apparent
that Romer considered that the slender nature of the postpubic rod
resulted from the reduction of the M. rectus abdominis. However,
the slenderness may indicate that this muscle had transferred at
least in part onto the ischium, a possibility that Romer (1927b,
fig. 18) accepted for the M. obturator internus. Romer (1927b,
p. 260) also noted that when the distal part of the pubis is lost,
as in iguanodonts, hadrosaurs and ceratopsians, the M. rectus
abdominis would have attached to the anterior border of the
ischium below the termination of the pubis. Obviously such a
transfer must have occurred before the postpubic rod could be lost
so I consider that the M. rectus abdominis probably inserted on
the ischium in hypsilophodonts and camptosaurs. The distal part
of the ischium of Hypsilophodon (Fig. 13) and Thescelosaurus is
expanded transversely. The anterior surface of the swollen distal
end would have provided an adequate insertion area for a strong
M. rectus abdominis. The adjacent part of the blade region
probably formed an additional insertion area. This surface is .
covered by longitudinal striations in BM(NH) R193 and R5829
and it is unlikely that the M. obturator internus occupied the whole
of this surface. I consider that the M. rectus abdominis of Thes-
celosaurus was not reduced and that it was a strong muscle that

HYPSILOPHODON PELVIC MUSCLES 27,

provided an important role in supporting the large abdomen.
In addition it would have provided a ventral tie that would have
helped in maintaining the curvature of the dorsal vertebral series.

It should be noted that Romer (1927b, figs. 16, 17) showed
only two abdominal muscles in Thescelosaurus: the M. rectus
abdominis passing posteriorly to insert on the whole of the ventral
edge of the pubis and the M. obliquus abdominis passing postero-
dorsally to insert on the anterior end of the prepubic process
(Fig. 5). He stressed that the former muscle had been functionally
replaced by the latter but ventral to the pelvic girdle the reverse
is the case with the M. rectus abdominis as the only muscle.
The abdominal wall as reconstructed by Romer is rather unique
and ill-adapted to support any viscera. In living amphibians, rep-
tiles (except Chelonia), birds and mammals there is a ventral
M. rectus abdominis and three lateral abdominal muscles that,
though each is only a thin sheet, together form a strong abdominal
wall as the fiber directions form a lattice work. Details of these
muscles in the alligator and birds are given above (p. 7) and
the reconstruction of these muscles in Hypsilophodon on page 11
and Figures 4 and 6.

From the similarity of the form of the postpubic rod the
abdominal musculature of Hypsilophodon probably resembled
that of birds rather than that of living reptiles. If the musculature
resembled that of carinates (see p. 8 and George & Berger,
1966), then the M. obliquus abdominis externus probably
inserted on the ventral surface of the prepubic process and the
anterior part of the postpubic rod, the pars internus on the distal
half of the postpubic rod and the M. transversus abdominis on all
of the ventral edge of the pubis. With such an extensive abdomen
it is more likely that the lateral abdominal muscles extended their
area of insertion on the pubis so that they all inserted along the
whole of the ventral edge of the pubis as in ratites (Gadow, 1880).
In addition, it is possible that part of each muscle inserted on the
laterally expanded distal part of the ischium. Consequently, I
consider that the region below the pelvic girdle did not consist
of “little more than a flap of skin.” Even so, there was no problem
as regards the free movement of the legs because the width of the
sacrum, combined with the offset heads of the femora, provided
plenty of space for a large and tapering abdominal cavity between
the thighs. This is clearly shown in the posterior view of a mounted

28 POSTILLA

skeleton of Camptosaurus (Gilmore, 1912, pl. 60) and in figures
of Triceratops mounts (Osborn, 1933; Erickson, 1966). The
potential strain on the abdominal muscles, due to the posterior
position of the pubis, was probably compensated for by an in-
crease in the strength of all the abdominal muscles including the
M. rectus abdominis. Apparently in the ornithopods part of this
muscle and possibly the lateral abdominal muscles transferred to
the adjacent surface of the ischium and, as a result, the postpubic
rod became reduced in thickness (hypsilophodonts, camptosaurs).
Subsequently this transfer was completed and the distal part was
lost in iguanodonts, hadrosaurs, psittacosaurs and ceratopsians.
It should be noted that the postpubic rod is thick and about the
same length as the ischium in stegosaurs (Gilmore, 1914). Romer
(1927b, p. 251) stated that he could not explain ‘on functional
grounds the peculiar construction of the ischium and postpubis in
this group. The broad surface presented by the two bones affords
a large external area of origin for the obturator. But this can be
but a minor factor.” The position in stegosaurs probably indicates
that the abdominal muscles had not shifted their insertion to the
ischium to such an extent as had occurred in ornithopods.

4. The Question of Muscle Attachment to the Lateral Surface of
the Prepubic Process of Hypsilophodon.

Romer (1923b, 1927b) concluded that the abdominal muscles,
the M. ambiens and the anterior part of the M. pubo-ischio-
femoralis externus were the only muscles on the pubis of sauris-
chian and ornithischian dinosaurs. These muscles, together with
the M. pubo-tibialis and the M. pubo-ischio-femoralis internus,
will be considered in this section in an attempt to determine which
muscle was attached to the well-defined area on the lateral surface
of the prepubic process of Hypsilophodon (see above p. 14 and
Figs. 4, 6, 7).

a) An abdominal muscle.

As suggested above (p. 27) all three lateral abdominal muscles -
inserted on the ventral surface of the prepubic process. The orien-
tation of the muscle scars on the posterior part of the lateral area
of BM(NH) R193 and R5829 are at about 45° to the long axis
of the prepubic process. This indicates that the muscle concerned

HYPSILOPHODON PELVIC MUSCLES 29

probably passed posteroventrally or anterodorsally. The M.
obliquus abdominis externus may have had a_ posteroventral
orientation so it could have occupied the lateral area. However,
I think that this muscle was probably restricted to the ventral sur-
face of the pubis. The extra insertion area for abdominal muscles
on the prepubic process was probably a secondary result; certainly
ratites and large herbivorous mammals manage to support their
abdomen adequately without such a process.

In hypsilophodonts (Figs. 5, 6), Psittacosaurus, Protoceratops
and Leptoceratops the prepubic process is rod-like (for figures of
these and following genera, see Romer, 1927b, 1966; Colbert,
1961). In Camptosaurus and Stegosaurus the process is slightly
expanded dorsoventrally. In /guanodon, hadrosaurs and ceratop-
sians the anterior end is considerably expanded dorsoventrally,
reaching its greatest extent in hadrosaurs and advanced ceratop-
sians (Fig. 12). Romer (1927b) correlated this with increased
size and the resulting heavier strains on the obliquii muscles that
supported the abdomen, because the expansion is in the plane
of these muscles. However, this expansion may represent the in-
creased size of a limb muscle, with particular emphasis on the
longer fibers which originated on the more anterior part of the
prepubic process. The ventral and the medial surfaces would still
provide an insertion area for abdominal muscles. In most ornithis-
chians where they are present the surface markings on the lateral
surface of the prepubic process are longitudinal striations (see p.
15). Romer (1927b) cited this as evidence that the obliquii mus-
cles inserted on this surface. However, in most of the ornithischians
mentioned on page 15 the lateral surface of the anterior process
of the ilium also has longitudinal striations. These were caused
by the M. ilio-tibialis 1 and in hadrosaurs (Lull & Wright, 1942)
and advanced ceratopsians (Fig. 12 and Lull, 1933) the prepubic
process is similar but much larger than the anterior process of the
ilium. The possibility that a limb muscle originated from the
prepubic process must therefore be reconsidered.

b) M. pubo-ischio-femoralis externus.

In the alligator two heads of this muscle originate on the pubis
and one on the ischium (Romer, 1923b); in saurischians probably
one head was on each bone (Romer, 1923c). In birds there is no

30 POSTILLA

posterior part but the anterior part of the M. obturator internus
swings back with the pubis during embryonic development to
become secondarily associated with the ischium (Romer, 1927a,
b). This probably occurred in Hypsilophodon so that the M. pubo-
ischio-femoralis externus originated from the postpubic rod and
ischium as a M. obturator internus. As Romer (1927b) pointed
out, this muscle would have had to follow a devious route from
the prepubic process to reach its normal area of insertion on the
posterior surface of the femur.

c) M. ambiens.

There are insertion markings on the dorsolateral edge of the
prepubic process in Hypsilophodon close to the acetabular region
in BM( NH) R193 (Fics, 4, 6, 7), RI9S and RI96 Mhese
markings, which are completely separate from the well-defined
lateral area (Figs. 4, 6, 7), correspond quite closely to the posi-
tion of the bump for the M. ambiens in Thescelosaurus (see p.
13). It is considered unlikely that the M. ambiens also originated
from the well-defined area on the lateral surface of the prepubic
process of Hypsilophodon. Romer (1927b) noted that such an
anterior origin would give the M. ambiens a very open angle quite
unlike that found in any other group; its usual archosaurian course
across the knee would appear to be impractical without interference
with the tendon of the M. triceps femoris.

d) M. pubo-tibialis.

This muscle is absent in crocodiles and birds so Romer (1927b)
concluded that it was also absent in ornithischians but the presence
of the prepubic process itself could also be denied on these same
grounds. In lizards the M. pubo-tibialis originates between the
M. ambiens and the M. obliquii abdominis which is on the proces-
sus lateralis(Snyder, 1954). The origin of the M. pubo-tibialis is
also adjacent to that of the M. ambiens in Sphenodon (Gregory &
Camp, 1918). If the M. pubo-tibialis originated on the lateral area
of the prepubic process of Hypsilophodon, it would have had the
same spatial relationships with the M. ambiens and M. obliquii
abdominis. The M. pubo-tibialis of Sphenodon (and _ lizards)
extends to the medioproximal prominence of the tibia (Gregory &
Camp, 1918). In hadrosaurs and ceratopsians especially, the form

HYPSILOPHODON PELVIC MUSCLES 31

of the prepubic process resembles that of the anterior process of
the ilium. It is possible that the M. pubo-tibialis originated on the
prepubic process and inserted on the medial part of the tibial head
while the M. ilio-tibialis, which originated more dorsally on the
anterior process of the ilium, inserted on the cnemial crest (Fig.
4). Both muscles would have acted to raise the leg. This inter-
pretation involves no change in muscle locations and, when the
femur was no longer held laterally, the prepubic process was
developed to give the M. pubo-tibialis a better leverage with an
anteroposterior line of action. If the M. pubo-tibialis originated
on the prepubic process of Hypsilophodon, then the absence of this
process in birds could be related to the absence of a M. pubo-
tibialis. However, the retention in Hypsilophodon of a muscle that
has also been lost in crocodiles does pose certain problems. I con-
sider that either this muscle or the one considered below probably
originated on the lateral area of the prepubic process.

e) M. pubo-ischio-femoralis internus.

In its course from an origin on the centra of the posterior dorsal
vertebrae to the greater trochanter of the femur this muscle passed
dorsally, close to the prepubic process. Romer (1927b) stated
that, despite this, it is improbable that it had any strong attachment
to that element. There is a part of the M. pubo-ischio-femoralis
internus, regarded by Romer (1927b) as a slip, that may have
originated from the lateral area of the prepubic process. Romer
(1927b, p. 255) noted that a part of the “‘ilio-femoralis internus
may have followed the course normally taken by a portion of the
primitive pubo-ischio-femoralis internus (and the ilio-femoralis of
birds) and inserted farther down the femur towards the inner
margin in the general position of the mammalian lesser trochanter.
Rugosities for such an insertion have been located in specimens of
Corythosaurus, Triceratops, and a few other forms, but they
are generally rare and uniformly weak-appearing.” The abdominal
muscles inserted on the ventral surface of the prepubic process so
in Hypsilophodon this ventral part would have passed laterally,
from the dorsal centra over the prepubic process, and then ven-
trally to insert proximally on the shaft of the femur. This pos-
tulated course means that the ventral part of the M. pubo-ischio-
femoralis internus wrapped around the prepubic process. It seems

32 POSTILLA

more likely that it actually originated on the lateral area on the
prepubic process (Figs. 4, 6, 7). The line of action of this ventral
part would have been more anteroposterior if it originated on
the prepubic process rather than on the centra of the posterior
dorsal vertebrae. Such an origin might account for the orientation
of the muscle scars on the posterior part of this lateral area.
According to Romer (1927b) the muscle scars on the femur for
the insertion of this ventral part are generally rare and uniformly
weak. This does not necessarily prove that the muscle itself was
weak as several undoubtedly powerful muscles originated from
the ilium in Hypsilophodon but, apart from the M. ilio-tibialis, the
areas on which these muscles originated cannot be identified.
There is a large area on the anteromedial surface of the femur
above the area for M. femoro-tibialis 2 (Figs. 8B, 9B, 10E) on
which this ventral portion could have inserted by a large but
fleshy attachment. Consequently, it is possible that the ventral
part of the M. pubo-ischio-femoralis internus originated on the
lateral area on the prepubic process of Hypsilophodon.

5. The Pelvic Girdle of Ornithischians and Birds.

The anterior process of the ilium in Archaeopteryx (Fig. 14 E)
and primitive ornithischians (Fig. 14C) is proportionally much
longer than it is in primitive saurischians (Figs. 14A, B) while in
croccdiles it is practically nonexistent (Fig. 1). The M. ilio-tibialis
1 (sartorius) probably originated from the complete length of this
process and was the principal long femoral protractor that inserted
on the tibia and extended the knee (Fig. 4). As a result of the
much longer anterior process of the ilium in ornithischians and
birds, this muscle was much larger and had a better mechanical
position than in saurischians. In Archaeopteryx the anterior
process of the ilium is deep (Fig. 14E) and covers the centra of
the adjacent lumber vertebrae (Romer, 1966, fig. 253). As a
result the M. pubo-ischio-femoralis internus of Archaeopteryx
originated on the ilium as in modern birds. The ventral part of this
muscle may have originated on the prepubic process of ornithis-
chians, and, if this was the case, it would account for the absence
of this process in birds (see above p. 16). This femoral protractor
occupied a more lateral position in both Archaeopteryx and
primitive ornithischians so that the line of action in both was fore

HYPSILOPHODON PELVIC MUSCLES 33

and aft rather than more obliquely as in crocodiles and sauris-
chians. However, it is possible that the prepubic process was for
the M. pubo-tibialis, a muscle lost in birds, which would have
supplemented the action of the M. ilio-tibialis 1 (see p. 14).
This would also have been the case if the muscle on the prepubic
process were the M. ambiens (see p. 13). Consequently the two
femoral protractors in Archaeopteryx and primitive ornithischians
probably occupied a position of better leverage than in sauris-
chians and crocodiles. Saurischians probably possessed another
anteriorly placed protractor, the M. pubo-ischio-femoralis externus
(see Romer, 1923b, c), which originated from the pubis that was
anteroventrally directed as in other reptiles (Fig. 14B). This
muscle was probably large as in crocodiles (Figs. 1, 2) in which
it is an important protractor of the femur.

It should be noted that when the pubis rotated backwards to-
wards the ischium in birds and ornithischians the M. pubo-ischio-
femoralis externus would have changed from a femoral protractor
to a retractor. I suggest that the presumed greater development
and improved mechanical position of the more anterodorsally
situated protractors (the sartorius and possibly the ventral part
of the M. pubo-ischio-femoralis internus) in Archaeopteryx and
primitive ornithischians was necessary before the pubis could
change position. Only when the M. pubo-ischio-femoralis externus
was no longer essential as a femoral protractor could the pubis
have come to lie close to the ischium.

As Romer (1927b) noted the pubis is directed ventrally rather
than anteriorly in some primitive archosaurs. Thus in the theco-
donts Euparkeria (Ewer, 1965), Stagonolepis (Walker, 1961)
and Saltoposuchus the more distal part of the pubis is rod-like,
ventrally directed and more or less perpendicular to the adjacent
part of the vertebral column so the ventral part of the M. pubo-
ischio-femoralis internus must have passed very close to the
proximal part of the pubis to reach the femoral shaft. In the line
of thecodonts leading to ornithischians, fibers of the ventral part of
the M. pubo-ischio-femoralis internus probably attached to the
proximal part of the pubis. The increased importance of these
fibers with a more fore and aft action would have resulted in the
formation of the prepubic process. When the M. ilio-tibialis 1
and possibly the ventral part of the M. pubo-ischio-femoralis in-
ternus were adequately developed the M. pubo-ischio-femoralis

34 POSTILLA

externus would no longer have been essential for protracting the
femur. When this stage was reached, the pubis continued rotating
backward to lie next to the ischium. In birds both muscles were
on the ilium, but the result was the same with the pubis lying close
to the ischium.

Depending on the degree of development of the M. ilio-tibialis
1 (as shown by the size of the anterior process of the ilium)
it is possible that the prepubic process was quite small or even
nonexistent when the pubis changed its position. In the latter case
the attachment of the M. pubo-ischio-femoralis internus to the
pubis would have occurred after this change occurred. I consider
that the anterior process was probably quite long when the pubis
rotated backward with quite a short prepubic process. If the muscle
concerned was the M. pubo-tibialis then the prepubic process was
probably moderately developed before the pubis changed position.
It is probable that in ornithischians, as was certainly the case in
birds, the development of a large anterior process of the ilium
was the critical factor that enabled the pubis to lie close to the
ischium.

The posterior position of the pubis in the ornithischian line
would have provided a larger abdominal cavity which, as Nopcsa
(1917) suggested, would be advantageous to an herbivorous ani-
mal. Romer (1927b) rejected this idea because it failed to explain
why the same change occurred in birds. He related the backward
shift of the pubis to a change in the position of the M. pubo-
ischio-femoralis externus to form the M. obturator internus. How-
ever, this must have been a result rather than a cause. It is hard
to visualize a selective force whose primary result was to convert a
femoral protractor into a retractor. There would be a half-way
stage when the muscle was neither but this would be no problem
if other selective forces were acting. A backward extension of the
effective abdominal cavity would also result in a posterior shift
of the center of gravity nearer the acetabulum. This is advanta-
geous to a bipedal animal and its importance was increased in birds
in which, with the exception of Archaeopteryx, the tail is very
short. The explanation may be even more general than this since
in eutherian mammals, where the anterior part of the ilium is
elongated, the anteroventral part of the pubis has been lost al-
though it was present in earlier mammal-like reptiles. Conse-
quently in these groups in which the femur moves anteroposteriorly

HYPSILOPHODON PELVIC MUSCLES 35

in a vertical plane the emphasis has been on the more antero-
dorsally situated protractors. Ornithischians achieved a less perfect
mechanical system than birds or mammals because one of the
protractors was still on the pubis. However, the prepubic process
is more anterodorsal than the pubis of other reptiles.

In ornithischians the more posterior position of part of the
pubis (postpubic rod) probably resulted in the transformation of
the anterior part of the M. pubo-ischio-femoralis externus from
a protractor to a retractor. This anterior part became a M.
obturator internus and replaced the posterior part as Romer
(1927b) noted. This change in position of the pubis also affected
the abdominal muscles. The lateral abdominal muscles extended
their insertion along the complete ventral edge of the pubis.
The lengthened M. rectus abdominis became stronger and, as in the
example of living reptiles (except chelonians), remained the
principal supporting muscle of the abdomen. When the pubis
reached a position close to the ischium the M. rectus abdominis
and the other abdominal muscles probably transferred to the
adjacent surface of the ischium to a variable degree (see above
p28).

B. THE ANTERIOR PROCESS OF THE ILIUM

The anterior process of the ilium of Hypsilophodon curves out
laterally (Fig. 15). This outward curvature is a general feature
of the ilium of ornithischians, as can be seen from the compara-
tive illustrations given by Romer (1927b, figs. 2-5). There are
several functional reasons for this lateral curvature: a) the out-
ward curvature resulted in the anterior process clearing the ribs
of the adjacent dorsal vertebrae; b) by curving outward the
anterior process presented a larger surface, set at an angle to the
longitudinal axis, for the insertion of part of the M. dorsalis
trunci, so that the lateral extent of this muscle was increased;
c) the outward curvature would have given the M. ilio-tibialis
(sartorius, Fig. 4) a slightly improved mechanical position, as
its line of action would be more anteroposterior (Fig. 15); and
d) the anterior process overhung the dorsal surface of the M.
pubo-ischio-femoralis internus that passed from the centra of
the lumbar vertebrae to the greater trochanter of the femur
(Fig. 4). As a result the ventral surface of the anterior process

36 POSTILLA

would have provided a potential surface of origin for this muscle.
Such an origin probably did not occur in primitive ornithischians
but, as Romer (1927b) suggested, it may have been important
in many ceratopsians, stegosaurs and ankylosaurs.

C. THE FEMUR

1. The Greater Trochanter.

Romer correctly pointed out that the greater trochanter of
saurischian (1923c) and ornithischian (1927b) femora is found
almost precisely where the posteriorly inserting portion of the
M. pubo-ischio-femoralis internus attached in primitive reptiles
and crocodiles. Romer (1927b) concluded that in ornithischians
the M. pubo-ischio-femoralis internus inserted on the greater
trochanter, but he argued that the development of the antitro-
chanter on the ilium in hadrosaurs indicated that a second anterior
femoral muscle had developed. This was the M. ilio-trochantericus,
which probably originated on the surface of the ilium anterior
to the antitrochanter in hadrosaurs. This origin closely resembles
that of the M. ilio-trochantericus in birds, which inserts on the
femur in a position similar to that of the greater trochanter of
ornithischians. Thus it is reasonable to conclude that the hadro-
saurian muscle inserted in the same region. Romer also (1927a,
b) pointed out that the derivation and hence homology of the
avian M. ilio-trochantericus is uncertain, because the evidence
from development and comparative anatomy is inconclusive.
The hadrosaurian muscle may have been derived from the M.
pubo-ischio-femoralis internus, in which case an insertion on the
greater trochanter would be expected. On the other hand, it
might have been derived from the primitive M. ilio-femoralis,
in which case an insertion on the lesser trochanter would be
expected.

The ilium of hypsilophodonts does not provide any evidence
concerning the differentiation of a M. ilio-trochantericus. Romer
(1927b, p. 233) wrote that ‘from the point of view of the femur,
however, a division into a more posterior ilio-femoralis and a
more anterior ilio-trochantericus seems required (see Part V).”
In Part V, when discussing the insertion of the M. ilio-trochanter-
icus, Romer (1927b, p. 254) only said that the lesser trochanter,

HYPSILOPHODON PELVIC MUSCLES 37

“when well developed, points upwards and, in most positions
of the limb, markedly backward, very nearly at right angles with
the direction necessarily taken by most fibers of the ilio-tro-
chantericus. It seems improbable that the ilio-trochantericus
inserted on the lesser trochanter.” In his reconstruction of the
musculature of Thescelosaurus, he showed the two muscles insert-
ing on the greater trochanter (Fig. 16). The more posterior M.
ilio-trochantericus is shown inserting on the fascia of the M.
pubo-ischio-femoralis internus, the insertion of which covers the
top of the greater trochanter. Romer (1927b) did not mention
the presence of any insertion marks on the greater trochanter or
in its immediate vicinity. Consequently, Romer did not provide
any evidence to show that a M. ilio-trochantericus was differ-
entiated in hypsilophodonts; the presence of this muscle was
assumed by extrapolation from the ilium of hadrosaurs.

The lateral surface of the femoral shaft adjacent to the greater
trochanter has two well-developed insertion areas that are sep-
arated by an “S” shaped ridge (Figs. 8A, 9A, 10B, see above,
p. 17). This ridge was first figured by Hulke (1882, pl. 77) for
BM(NH) R193 but its significance has not been commented
upon. The ridge is present on all femora of Hypsilophodon in
which this region is not damaged. This ridge is also visible on
the femur of Laosaurus minimus (Gilmore, 1924b, pl. 2, fig. 4)
and there are traces of it on the femur of Dryosaurus altus
(YPM 1876). The origin of the M. pubo-ischio-femoralis internus
was more anteriorly placed so it probably inserted on the ante-
rior area while the M. ilio-trochantericus inserted on the posterior
area (Figs. 4, 10B). The cleft between the proximal ends of the
lesser and greater trochanters enabled the fascia of the M. pubo-
ischio-femoralis internus to reach the inner area with less effect
on the M. ilio-femoralis. This ridge at the base of the greater
trochanter in Hypsilophodon supports Romer’s assumption that
a M. ilio-trochantericus inserted on the greater trochanter of
hypsilophodonts. It also supports his suggestion that this muscle
in ornithischians was derived from the M. pubo-ischio-femoralis
internus rather than from the primitive M. ilio-femoralis. Romer
(1962, table 2) lists the avian M. ilio-trochantericus and M.
ilio-femoralis internus as homologous to the reptilian M. ilio-
femoralis. If this is correct then the M. ilio-trochantericus of
ornithischians and birds may not be homologous.

38 POSTILLA

2. The Fourth Trochanter.

Dollo (1883, 1888) first pointed out that the fourth trochanter
was the area of insertion for the well-developed caudi-femoralis
muscles. Romer (1927b) merely noted that the size of the fourth
trochanter furnished a rough guide to the size of these muscles.
In his reconstruction of Thescelosaurus he showed the M. caudi-
femoralis brevis inserting on the proximal part of the trochanter
(Fig. /16). The M. caudi-femoralis longus is shown inserting
more distally with a tendon passing from the pendant end of the
trochanter. The muscle to which this tendon runs is not labeled
in Romer’s figure but from his text it is clear that it was the M.
gastrocnemius (1927b, p. 257). Romer labeled a second muscle
as the M. gastrocnemius so he recognized two divisions of the
M. gastrocnemius. This omission of a label in Romer’s figure
appears to have misled Janensch (1955) who discussed this
“problem”. Janensch postulated a tendon in Dysalotosaurus that
extended from one division of the M. gastrocnemius to a distinct
facet on the lateral face of the pendant end of the fourth tro-
chanter. He reconstructed the M. caudi-femoralis longus insertion
on both sides of the fourth trochanter, but noted that the more
proximal insertion of the M. caudi-femoralis brevis could not
be definitely placed. He also drew attention to the well-developed
depression near the base of the fourth trochanter in Dysalotosaurus
and Hypsilophodon.

a) Possible functions for the depression at the base of the
fourth trochanter.

1) To house a mucous gland.

Janensch (1955) thought that the depression near the base of
the fourth trochanter in Dysalotosaurus and Hypsilophodon was
too deep and extensive to have been the insertion area of a muscle.
He postulated that the depression was for a mucous gland though,
as he noted, the presence of such a structure is difficult to prove
in fossil material. Janensch did not suggest why a mucous gland
should have been housed in a depression on the femoral shaft
or what function it would have served. However, such a position
is much too deep for a characteristically epidermal structure.
The depression is quite deep in some femora of Hypsilophodon

HYPSILOPHODON PELVIC MUSCLES 39

but in others it is not (p. 19). In both types of femora an
equivalent area is covered with surface markings of muscle attach-
ments. Consequently this depression did not house a mucous
gland but was the area of insertion of a muscle.

ii) Insertion area for part of the M. pubo-ischio-femoralis
internus.

Janensch (1955) noted that on the femur of Crocodilus
niloticus, dorsal to the wide depression on the fourth trochanter,
there is another smaller depression that corresponds in position
to the deep depression in Dysalotosaurus. He cited the figure of
the alligator femur given by Romer (1923b, pl. 25) in which
part of the M. pubo-ischio-femoralis internus is shown inserting
on this region (Fig. 17). Romer (1927b) discussed this muscie
(see above, p. 31) but, as Janensch (1955) noted, did not show it
in the reconstruction of Thescelosaurus. Janensch considered that
this muscle might have inserted on the deep depression on the
femur of Dysalotosaurus.

ili) Insertion area for the M. caudi-femoralis.

The depression in Hypsilophodon, Thescelosaurus (Gilmore,
1915), Camptosaurus leedsi (Gilmore, 1909) and Camptosaurus
(Gilmore, 1909), is partly on or close to the base of the fourth
trochanter. This depression is further from the base in Dryosaurus
altus (YPM 1876) and more markedly so in Dysalotosaurus
where the depression is found quite anteriorly on the femoral
shaft (see Janensch, 1955, pl. 15, fig. 2). Janensch (1955) recog-
nized that the same muscle inserted on the depression is Hypsilo-
phodon, Camptosaurus and Dysalotosaurus but, because it is sep-
arated by a wide space from the fourth trochanter in Dysaloto-
saurus, he considered that the depression was not for the M.
caudi-femoralis. This depression is close to the fourth trochanter
in [guanodon (see Casier, 1960, pl. 10) and hadrosaurs (see
Sternberg, 1924, pl. 4). I consider that the position of this depres-
sion in Dryosaurus and Dysalotosaurus is secondary and that it
corresponds to the large depression at the base of the fourth
trochanter on the alligator femur. Romer (1923b) showed the M.
caudi-femoralis longus inserting on the depression at the base
of the fourth trochanter in the alligator (Fig. 18). I checked the

40 POSTILLA

insertion areas of the M. caudi-femoralis longus and brevis in a
Gavialis gangeticus, the only crocodilian that was available. The
adjacent parts of the M. caudi-femoralis longus and brevis are
closely spaced and when they are parted the fourth trochanter
is seen between them. The M. caudi-femoralis longus inserts on
the anterior surface of the trochanter, with an especially strong
insertion on the depression at its base. The M. caudi-femoralis
brevis inserts on the posterior surface of the trochanter. Because
of the different postures of these reptiles the medial view (Fig.
10D) of the femur corresponds to the anterior view in the gavial
and alligator (Fig. 17). I consider that the M. caudi-femoralis
longus inserted into this depression in Aypsilophodon and
Dysalotosaurus as it does in modern Crocodilia. The areas of the
femur of Hypsilophodon |BM(NH) R193, Figs. 4, 8, 9, 10]
on which the M. caudi-femoralis brevis and longus and the tendon
from the knee region attached are described on page 31.

b) The function of the fourth trochanter.

A large fourth trochanter is characteristic of many dinosaurs,
especially those that are bipedal, and its size is generally con-
sidered to reflect the degree of development of the caudi-femoralis
muscles (Romer, 1927b; Lull & Wright, 1942; Ostrom, 1964).
In crocodiles the caudi-femoralis muscle, which is the principal
retractor of the femur (Snyder, 1962), is a very large muscle
(see Haughton, 1865, fig. 20) but the fourth trochanter is small
(Fig. 17). However, the pars longus, which forms the major part
of the caudi-femoralis muscle, inserts mainly on the depression
at the base of the fourth trochanter. It is possible that the size
of the fourth trochanter of dinosaurs reflects the degree of devel-
opment of the pars brevis rather than that of the longus or the
caudi-femoralis musculature as a whole. The areas of origin
and insertion of the M. caudi-femoralis brevis and longus were
presumably similar in all bipedal ornithischians and like those
of the alligator. The area of origin of the pars brevis was close
to the head of the femur while that of the pars longus extended
some way down the tail. The pars brevis inserted at the same
level as the longus so its line of action was at a smaller angle to
the femoral shaft (Fig. 4) and, as a result, its moment arm rela-
tive to the head of the femur was much shorter. In addition,

HYPSILOPHODON PELVIC MUSCLES 4)

when the femur was swung forward the moment arm of the pars
brevis was shortened to a greater extent than that of the pars
longus so it would have been reduced to zero sooner. Conse-
quently the presence of a large fourth trochanter at right angles
to the femoral shaft would have had a proportionally greater effect
on the moment arm of the pars brevis. In normal positions of
the femur the trochanter projected towards the area of origin
of the pars longus and resulted in only a small increase in the
moment arm of this part. Certainly an insertion on the depres-
sion on the femoral shaft would have been stronger than one on a
projecting process. The moment arm of the pars longus of
ornithopods is increased phylogenetically because the fourth
trochanter is above the middle of the femur in hypsilophodonts,
just below mid-femur length in camptosaurs and about two-
thirds femur-length in iguanodonts and hadrosaurs. I consider
that the function of the large fourth trochanter of dinosaurs was
to increase the moment arm of the M. caudi-femoralis brevis
during the initial part of femoral retraction. As the femur ap-
proached a vertical position the pars brevis became progressively
less important while the moment arm of the pars brevis reached its
maximum and its line of action became more perpendicular to the
femoral shaft.

Nopcsa (1905) argued that the pendant fourth trochanter of
hypsilophodonts and camptosaurs is primitive for ornithischians
and that the “trochanter creté”, considered by Dollo (1888) to
be primitive, was a later development. However, Ostrom (personal
communication) considers that the fourth trochanter of /guanodon
and hadrosaurs was also probably pendant. The functional sig-
nificance of the pendant fourth trochanter is not certain. The
space below the pendant part might have allowed a nerve, blood
vessel or tendon to cross this part of the femur but there is no
such structure in this region in crocodiles. In Hypsilophodon BM
(NH) R193 (Figs. 8A, 9A, 10B) there are insertion markings
on the distal surface, formed by the thick basal part of the
trochanter, which were probably for part of M. femoro-tibialis
3 (see p. 14). Romer (1927b) noted that it is a general archo-
saurian tendency for the M. femoro-tibialis to increase in size.
This origin on the fourth trochanter would have increased the
moment arm of the fibers concerned, but I have not been able to
find similar muscle scars on other ornithischian femora. I consider

42 POSTILLA

it unlikely that the origin of part of the M. pubo-tibialis 3 was
the primary reason for the pendant form of the trochanter. A
tendon of the M. caudi-femoralis longus muscle probably orig-
inated from the pendant part of the trochanter (Fig. 4) and
passed to the knee region as Dollo (1888) first suggested. This
tendon presumably inserted on the posterior aspect of the fibula
head as it does in crocodiles, lizards and Sphenodon (Romer,
1923b; Snyder, 1954). The pendant form of the fourth trochanter
may reflect stresses transmitted by this tendon if, as was pos-
sibly the case, a head of the M. gastrocnemius originated from
the distal part (Figs. 4, 16). The moment arm of this head, or
rather the part that originated from the tendon, would have been
increased slightly. Consequently the pendant nature of the fourth
trochanter may reflect stresses from the distal part of the leg rather
than from the caudi-femoralis muscles.

SUMMARY

The reconstruction of the pelvic musculature of Thescelosaurus
by Romer (1927b), with certain modifications, adequately ac-
counts for the surface markings on the very well preserved pelvic
girdle and femur of Hypsilophodon BM(NH) R193. Using data
from living forms and functional considerations it is possible to
place different degrees of confidence on the various possibilities
that have been discussed.

Because the prepubic process is characteristic of ornithis-
chian dinosaurs, it is important that its functional signifi-
cance be understood. The suggestion by Romer (1927a, b,
1956) that this process provided a base for the obliquii muscles,
which he considered had taken over the support of the abdomen
from the M. rectus abdominis, was very neat. This theory
accounted for a unique structure in terms of a new function and,
apparently, explained why this process was not developed in birds.
Unfortunately Archaeopteryx and ratites show that it was possible
to have a backwardly directed pubis with long ventral abdominal
muscles not supported by either a large sternum or a prepubic
process. The pubis of birds had changed position to lie close to
the ischium before the very large sternum and the short abdomen
were evolved. The position in birds, together with the dorsal
position of the prepubic process and the probable presence of

HYPSILOPHODON PELVIC MUSCLES 43

gastralia in early ornithischians at least, indicates that the primary
function of the prepubic process was not abdominal support. Any
potential weakening of the abdomen that resulted from the
changed position of the ornithischian pubis was probably com-
pensated for by an increase in size of all the abdominal muscles.
The three lateral abdominal muscles, the fiber directions of which
formed a grid-like arrangement, probably extended their area
of insertion along the ventral surface of the whole length of the
pubis and, in addition, possibly on to the blade-like distal part
of the ischium. The latter also provided an additional insertion
area for the strong M. rectus abdominis. The degree of transfer
of the abdominal muscles on to the ischium occurred to a vary-
ing extent in different groups of ornithischians (see above, p. 28).
The abdomen of ornithischians was probably large as in ratites
and herbivorous mammals and was not just a flap of skin below
the pelvic girdle as Romer (1927b) suggested.

I consider it unlikely that the muscle on the lateral surface of
the prepubic process was an abdominal muscle or the M. ambiens
(p. 28). The similarity in form of the prepubic process to that
of the anterior process of the ornithischian ilium and the position
of the M. pubo-tibialis in Sphenodon suggest that this might be
the muscle concerned. Birds lack this muscle which would account
for the absence of a prepubic process in that group. The presence
of the ventral part of the M. pubo-ischio-femoralis internus on
the prepubic process would also explain the absence of this
process in birds where this muscle is on the ilium (M. ilio-
femoralis internus). The ventral part of this muscle probably
originated on the prepubic process and did not wrap around the
process as it would, if, as Romer (1927b) suggested, it originated
on the centra of the dorsal vertebrae.

Although the postpubic rod of ornithischians probably repre-
sents the original reptilian pubis, as must be the case in birds, the
functional reason for the change in position is not clear. Romer
(1927b) suggested that it was related to a change in position of
the anterior part of the M. pubo-ischio-femoralis which became
the M. obturator internus. However, during the change in position
of the pubis this muscle was converted from a femoral protractor
into a retractor. I consider that the key factor in birds and ornithis-
chians was the development of a very long anterior process to the
ilium to provide a large M. ilio-tibialis 1. When the M. ilio-

44 POSTILLA

tibialis 1 (possibly with the ventral part of the M. pubo-ischio-
femoralis externus or the M. pubo-tibialis in ornithischians) was
adequate for femoral protraction, then the anterior part of the
M. pubo-ischio-femoralis externus would not have been required
for protraction so the pubis could change its position. In ornithis-
chians, birds and mammals emphasis was on the more antero-
dorsally situated femoral protractor with a more fore and aft line
of action. The change in position of the ornithischian pubis would
have enlarged the abdominal cavity which, as Nopcsa (1917)
suggested, would have been an advantage to an herbivorous group.
In addition, it would have shifted the center of gravity more
posteriorly and this would have been an advantage to a bipedal
animal.

The anterior process of the ilium curved laterally so that it
cleared the adjacent ribs, provided both a larger area of insertion
for part of the M. dorsalis trunci and a more fore and aft direc-
tion for the fibers of the M. ilio-tibialis 1. In addition the process
overhung the M. pubo-ischio-femoralis internus and the ventral
surface was an important area of origin for this muscle in
stegosaurs, ankylosaurs and ceratopsians.

There are two distinct areas of muscle insertion near the base of
the greater trochanter of the femora of Hypsilophodon. These
indicate that a third dorsal muscle, analagous to the M. ilio-
trochantericus of birds, was differentiated in hypsilophodonts as
Romer (1927b) had postulated by extrapolation from the struc-
ture of the ilium in hadrosaurs. In some femora of Hypsilophodon
the depression near the base of the fourth trochanter is well
developed while in others it is not. In both cases the equivalent
area has muscle insertion markings and, because the position of
this depression is too deep for an epidermal structure, it is con-
sidered extremely unlikely that this depression housed a mucous
gland as Janensch (1955) suggested. His second suggestion, that
the depression was for the ventral part of the M. pubo-ischio-
femoralis internus, is considered unlikely because this muscle in
the alligator inserts on the shaft next to a comparable large
depression. The M. caudi-femoralis longus inserts on the large
depression in an alligator so it is reasonable to assume that this
was also the case in Hypsilophodon. The degree of development
of the fourth trochanter is probably not related to the size of the
caudi-femoralis musculature as a whole. The function of the large

HYPSILOPHODON PELVIC MUSCLES 45

fourth trochanter of bipedal dinosaurs was to lengthen the moment
arm of the M. caudi-femoralis brevis during the first part of
femoral retraction. The pendant form of the fourth trochanter
of ornithopods may reflect the stresses from part of the M. gas-
trocnemius that attached to the tendon connecting the trochanter
to the fibula.

ACKNOWLEDGEMENTS

Much of this paper is based on work done at the Zoology Depart-
ment, King’s College, University of London, England, made
possible by a three year Research Studentship from the Depart-
ment of Scientific and Industrial Research (subsequently the
Natural Environment Research Council). I am especially grateful
to Dr. C. B. Cox for all his advice and interest during the course
of this work and to the British Museum (Natural History) for
permission to borrow their unique material of Hypsilophodon and
prepare it in acetic acid. The final paper has been much improved
thanks to the criticisms and suggestions of Dr. J. H. Ostrom of
Yale University. The manuscript was also read by Dr. A. W.
Crompton, Mr. F. A. Jenkins, Jr. and Mr. R. T. B. Bakker, all
of Yale University; Dr. G. E. Erickson of Brown University,
Providence, R.I., and Dr. D. A. Russell of the National Museum
of Canada, Ottawa; all of whose comments were much appreciated.

I am very grateful to people in the various museums for their
help while studying the material under their care as cited in the
text: to Dr. A. J. Charig of the British Museum (Natural His-
tory); Dr. E. H. Colbert of the American Museum of Natural
History, who also kindly supplied the original for Figure 12;
Drs. A. W. Crompton and J. H. Ostrom of the Peabody Museum
of Yale University; Dr. A. G. Edmund of the Royal Ontario
Museum, Ontario and Dr. D. A. Russell of the National Museum
of Canada. Last but not least, my thanks must go to Dr. A. S.
Romer of the Museum of Comparative Zoology, Harvard Univer-
sity who. as must be rather obvious by now, provided most of
the information upon which this paper is based.

46

POSTILLA

ABBREVIATIONS

Unless indicated to the contrary all figures show bones from
the left side; Figures 4, 6, 7-11 and 13 have been printed in

reverse.

The museum names

have been abbreviated as follows:

AMNH American Museum of Natural History, New York.
BM(NH) British Museum (Natural History), London.

NMC National Museum of Canada, Ottawa.
USNM United States National Museum, Washington, D.C.
YPM Peabody Museum, Yale University, New Haven.

The structures indicated in Figures, 5, 10, 13 and 15 are as

follows:

acet — acetabulum

obt proc — obturator process

ant proc — anterior process po rod — postpubic rod

br sh — brevis shelf

pre proc — prepubic process

gr troc — greater trochanter 4th troc — fourth trochanter
les tro — lesser trochanter

The muscles have been abbreviated as follows:

ACC — M.
ADD —M.
AMB — M.
CA-FEM BR — M.
CA-FEM L —M.
CA-IL-F —M.
DOR CA — M.
DOR T —M.
FEM-T 1, 2&3 —M.
| ed Bl — M.
eed ia — M.
G —M.
IL-CAUD —M.
IL-COST — M.
IL-FEM — M.
IL-FEMEXT —M.
IL-FEM INT —M.

accessory obturator
adductor femoralis
ambiens
caudi-femoralis brevis
caudi-femoralis longus
caudi-ilio-fibularis
dorsalis caudae
dorsalis trunci
femoro-tibialis 1, 2 & 3
flexor tibialis externus
flexor tibialis internus
gastrocnemius
ilio-caudalis
ilio-costalis
ilio-femoralis
ilio-femoralis externus
ilio-femoralis internus

HYPSILOPHODON PELVIC MUSCLES 47

IL-FIB — M. ilio-fibularis

IL-TIB 1 & 2 — M. ilio-tibialis 1 (sartorius) & 2

IL-TROC — M. ilio-trochantericus

IS-CAUD — M. ischio-caudalis

IS-FLEX — M. ischio-flexorius

IS-TROC — M. ischio-trochantericus

LIG — ligaments for holding head in acetabulum

OA EXT — M. obliquus abdominis externus

OA INT — M. obliquus abdominis internus

OBL — M. obliquus abdominis

OBT — M. obturator internus (anterior part of
P-I-F E)

P-I-F — M. pubo-ischio-femoralis

P-I-FE — M. pubo-ischio-femoralis externus

P-I-F INT 1 — dorsal part of M. pubo-ischio-femoralis
internus

P-I-F INT 2 — ventral part of M. pubo-ischio-femoralis
internus

P-TIB — M. pubo-tibialis

R ABD — M. rectus abdominis

TND — tendon inserting on fibula

TRA — M. transversus abdominis

TR-P — M. transversus perinei

LITERATURE CITED

Beer, G. de, 1954, Archaeopteryx lithographica. A study based on the
British Museum specimen: London, British Museum (Natural His-
tory), 68 p.

Casier, E., 1960, Les iguanodons de Bernissart: Brussels, Inst. Roy. Sci.
Nat. Belg., 134 p.

Colbert, E. H., 1961, Dinosaurs. Their discovery and their world: New
York, Dutton, 300 p.

Dollo, L., 1883, Note sur la presence chez les oiseaux du “troisieme
trochanter” des dinosauriens et sur la fonction de celui-ci: Mus.
Roy. d’Hist. Nat. de Belg. Bull., v. 2, p. 13-18.

1888, Sur la signification du “Trochanter pendant” des Dino-
sauriens: Sci. Fr. Belg. Bull., v. 19, p. 215-224.

Elftman, H. O., 1929, Functional adaptations of the pelvis in marsupials:
Amer. Mus. Nat. Hist. Bull., v. 58, p. 189-232.

Erickson, B. R., 1966, Mounted skeleton of Triceratops prorsus in the
Science Museum: Sci. Mus. St. Paul, Minn., Sci. PublIn., (n.s.) v. 1,
no: 15) ps 1-16:

48 POSTILLA

Ewer, E. F., 1965, The anatomy of the Thecodont reptile Euparkia capensis
Broom: Phil. Trans. (B), v. 248, p. 379-435.

Gadow, H., 1880, Zur vergleichenden Anatomie der Muskulatur des Beckens
und der hinteren Gliedmasse der Ratiten: Jena, Fischer, 56 p.

Gadow, H. & Selenka, E., 1891, Vogel I. Anatomischer Theil. Jn Bronn,
H. G., Klassen und Ordnugen des Thier-Reichs, in Wort und Bild.
v. 6. Leipzig, Winter.

Galton, P. M., In press, On the anatomy of the ornithischischian dinosaur
Hypsilophodon foxii from the Wealden (Lower Cretaceous) of the
Isle of Wight, England: Brit. Mus. (Nat. Hist.) Bull., Geol.

George, J. C. & Berger, A. J., 1966, Avian myology: London, Academic
Press, 500) p.

Gilmore, C. W., 1909, Osteology of the Jurassic reptile Camptosaurus
with a revision of the species of the genus, and descriptions of two
new species: U.S. Natl. Mus. Proc., v. 36, p. 197-332.

1912, The mounted skeletons of Camptosaurus in the United
States National Museum: U.S. Natl. Mus. Proc., v. 41, p. 587-696.

1914, Osteology of the armoured Dinosauria in the United
States National Museum, with special reference to the genus Stego-
saurus: U.S. Natl. Mus. Bull., v. 89, p. 1-136.

1915, Osteology of Thescelosaurus, an orthopodous dinosaur
from the Lance Formation of Wyoming: U.S. Natl. Mus. Proc., v.
49, p. 591-616.

1924a, On Troodon validus an orthopodous dinosaur from the
Belly River Cretaceous of Alberta, Canada: Dep. Geol. Univ. Alberta
Bull., v. 1, p. 1-43.

1924b, A new species of Laosaurus an ornithischian dinosaur
from the Cretaceous of Alberta: Roy. Soc. Can. Trans. (3), v. 18,
sect. 4, p: 1-7.

Gregory, W. K., 1919, The pelvis of dinosaurs; a study of the relations
between muscular stresses and skeletal form: Copeia, v. 69, p.
18-20.

Gregory, W. K. & Camp, C. L., 1918, Studies in comparative myology and
osteology. No. Ill: Amer. Mus. Nat. Hist. Bull., v. 38, p. 447-563.
Haughton, S., 1865, On the muscular anatomy of the leg of the crocodile:

Ann. Mag. Nat. Hist. (3), v. 16, p. 326-331.

Heilmann, G., 1926, The origin of birds: London, Witherby, 208 p.

Huene, F. von, 1926, Vollstandige Osteologie eines Plateosauriden aus
dem Schwabischen Keuper: Geol. Palaont. Abhl., N.F., v. 15, p.
139-180.

Hulke, J. W., 1882, An attempt at a complete osteology of Hypsilophodon
foxii, a British Wealden dinosaur: Phil. Trans. (B), v. 172, p.
1035-1052.

Huxley, T. H., 1870, On Hypsilophodon foxii, a new dinosaurian from
the Wealden of the Isle of Wight. Geol. Soc. London Quart. Jour., v.
IA; Jo SON (72,

Janensch, W., 1955, Der Ornithopode Dysalotosaurus der Tendaguru-
schichten: Palaeontographica Suppl., v. 7 (3), p. 105-176.

Lebedinsky, N. G., 1913, Beitrage zur Morphologie und Entwicklungs-
geschichte des Vogel Beckens: Zeit. Natur. Jena, v. 50, p. 647-774.

1914, Uber der Processus Pectinialis der Straussenbeckens
und seine phylogenetische Bedeutung: Anat. Anz., v. 46, p. 84-89.

HYPSILOPHODON PELVIC MUSCLES 49

Lull, R. S., 1933, A revision of the Ceratopsia or horned dinosaurs: Pea-
body Museum, Yale Univ., Memoir, v. 3 (3), 135 p.
Lull, R. S. & Wright, N. E., 1942, Hadrosaurian dinosaurs of North
America: Geol. Soc. Amer. Spec. Pap., v. 40, p. 1-242.
Marsh, O. C., 1878, Principal characters of American Jurassic dinosaurs:
Am. Jour. Sci. (3), v. 16, p. 411-416.
Nopcsa, F. B., 1905, Notes on British dinosaurs I — Hypsilophodon: Geol.
Mag. n.s. (5), v. 2, p. 203-208.
1917, Uber Dinosaurier 3. Uber die Pubis der Orthopoden:
Zentbl. Min. Geol. Palaont., v. 1917, p. 348-351.
Osborn, H. F., 1933, Mounted skeleton of Triceratops elatus: Amer. Mus.
Novitates, no. 654, 14 p.
Ostrom, J. H., 1964, A reconsideration of the paleoecology of hadro-
saurian dinosaurs: Am. Jour. Sci., v. 262, p. 975-997.
Romer, A. S., 1923a, The ilium in dinosaurs and birds: Amer. Mus. Nat.
Hist. Bull., v. 48, p. 141-145.
1923b, Crocodilian pelvic muscles and their avian and rep-
tilian homologues: Amer. Mus. Nat. Hist. Bull., v. 48, p. 533-552.
1923c, The pelvic musclature of saurischian dinosaurs:
Amer. Mus. Nat. Hist. Bull., v. 48, p. 605-617.
1927a, The development of the thigh musculature of the
chick: Jour. Morph., v. 43, p. 347-385.
1927b, The pelvic musculature of ornithischian dinosaurs:
Acta. Zool., Stockh., v. 8, p. 225-275.
1956, Osteology of the reptiles: Chicago, Univ. of Chicago

Press, 772 p.
1962, The vertebrate body (3rd edn.): Philadelphia, Saun-
ders, p. 627.

1966, Vertebrate paleontology (3rd edn.): Chicago, Univ.
of Chicago Press, 468 p.

Russell, L. S., 1935, Musculature and functions in the Ceratopsia: Natl.
Mus. Can. Bull., v. 77, p. 39-48.

Sternberg, C. M., 1924, A new species of hadrosaurian dinosaur from the
Edmonton formation (Creataceous) of Alberta: Can. Dept. Mines
Geol. Surv. v. 38, p. 13-26.

Synder, R. C., 1954, The anatomy and function of the pelvic girdle and
hindlimb in lizard locomotion: Amer. Jour. Anat., v. 95, p. 1-46.

1962, Adaptations for bipedal locomotion of lizards: Amer.
Zool., v. 2, p. 191-203.

Walker, A. D., 1961, Triassic reptiles from the Elgin area: Stagonolepis,

Dasynathus and their allies: Phil. Trans. (B), v. 244, p. 103-204.
1964, Triassic reptiles from the Elgin area: Ornithosuchus
and the origin of the carnosaurs: Phil. Trans. (B), v. 248, p. 53-134.

HYPSILOPHODON PELVIC MUSCLES 5)

IL-CAUD

R ABD origins

FIG. 1. Pelvic region of Alligator mississipiensis showing the areas of
muscle attachment visible in lateral view and the lines of action of most
of the muscles. Femur shown under the body in a position comparable

to that of birds and dinosaurs. Data from Romer (1923b, pls. 19-25).
Abbreviations on p. 46.

a2 POSTILLA

FIG. 2. Pelvic region of the alligator showing the line of action of the
muscles whose area of attachment is visible in the medial view of the
right side. Compare with Figure 1. Data from Romer (1923b, pls. 19-25).
Abbreviations on p. 46.

HYPSILOPHODON PELVIC MUSCLES 33

FIG. 3. Pelvic region of an eight day old chick to show attachment areas.
Data from Romer (1927a, figs. 5, 5a, 5b; 1927b, fig. 1). Abbreviations
on p. 46.

POSTILLA

54

‘9p ‘d uo suonerasiqgy “LOL87 pue OL8cd ‘O61 ‘C6lN (HN)WA Wor
eye ‘sepsnul yeNprarpul oy} Jo wuoNse JO soul] OU} Suimoys uopoydopisd& yy JO UorIso1 a1ajed ay} JO UONSNNsUO.Iy “Pp “Old

gav a

vVul .LNIVO

55

HYPSILOPHODON PELVIC MUSCLES

‘op ‘d uo suoneiasiqqy ‘(8I-rl

‘s3U ‘QLZ61) JOWOY Woy RIeq ‘saposnuT SIAjed Jo JUsWYSe}e JO sRolR BZUIMOYS SNAnvSOjsISayT JO [PIS IAJeq “¢ “Old

901d aid
_ aavy / > al
a ido
21 LNI4-I-d NZ
. air 8 ‘

POSTILLA

56

‘Op ‘d uo suonviasiqqy “LOL8T PUR 96IU (HN) WA Woy osye
eye, “SajOsNW JENPIAIPUL oy} JO JUdWYSe}e JO svaIe 9Yy} SUIMOYS “C6IM (HNIWA ‘vopoydojisdd] JO apsls dIAJaq “9 “OIA

snmvanvo /
-olHos!t.|

ai LNI d-I-d

sluvinals 50uL Q
£\-on waa al Lsriviairom 7 |

Ca ey Z Shiviall-onl ea) 1 ¥od

HYPSILOPHODON PELVIC MUSCLES 57

FIG. 7. Pubis and ischium of Hypsilophodon, BM(NH) R193. Compare
with Figure 6. The scale is 10 cm.

58 POSTILLA

FIG. 8. Femur of Hypsilophodon, BM(NH) R193. A. Anteromedial view.
B. Posterolateral view. Compare with Figure 9. The scale is 10 cm.

HYPSILOPHODON PELVIC MUSCLES 59

A B
Pie /, Y\IL-Troc
INT 1 /\
Lie P-I-F INT 2
IL- bed
CA-FEM BR CA-FEML
FEM-T 1
FEM-T 3 FEM-T 2

l

FIG. 9. Diagrams showing the attachment areas in the stereo-photographs
of Figure 8, A and B. Abbreviations on p. 46.

POSTILLA

60

‘gp ‘d UO SUONRIAZIQGY “MOIA [RIPIW “C “MIIA JOWa}UY *)

‘MAIA [RIBVT “A “MAIA JOII9}SOd “V “OESSU PUB 961Y (HN) WE

WOlJ OSTR BEC ‘sapsnW qu] oy} JO JUSUTYDE}e JO SvaIe 94} Zuimoyus ‘E6lU (HNIWA ‘vopoydojisdky jo inwaj ‘O] “old

Zt INI
d-I-d

904} Sa|
\

Wa4-1l

wool

1g0
area 90u 1-11
nA
{
9013 416 .
L INI 4I-d Wada

HYPSILOPHODON PELVIC MUSCLES 61

pre proc

FIG. 11. Pubis of Hvypsilophodon in dorsal view, BM(NH) R193.

1ocm

FIG. 12. Skeleton of Triceratops cf. elatus, AMNH 5116, 5033, 5039,
5095. Original about 20’ or 6.2 meters. Courtesy of the American Mu-
seum of Natural History.

FIG. 13. Ischium of Hypsilophodon in ventral view. BM(NH) R193.
Abbreviations on p. 46.

62

POSTILLA

FIG. 14. Diagrams showing attachment areas of certain pelvic muscles in
archosaurs.

A

Ornithosuchus — Saurischia — Carnosauria (after Walker, 1964).
The dotted part of the illum represents a cartilaginous extension
whose precise size and outline cannot be determined. As shown, it
resembles Tyrannosaurus.
Plateosaurus — Saurischia — Prosauropoda (after von MHuene,
1926).

Hypsilophodon — Ornithischia — Ornithopoda.

Struthio — Aves —ratite (after Gregory & Camp, 1918, and
Romer, 1923a).

Archaeopteryx — Aves (after Heilmann, 1926) The question mark
indicates the possibility of a cartilaginous extension to the ischium
but its form is not known.

Abbreviations used in this figure:

a. dorsal part of M. pubo-ischio-femoralis internus. b. ventral part of
the M. pubo-ischio-femoralis internus. In D and E these two parts form

the M.

ilio-femoralis internus. c. M. ilio-tibialis 1 (sartorius). d. anterior

part of M. pubo-ischio-femoralis externus. In C-E this is the M. obturator
internus that replaced the posterior part.

HYPSILOPHODON PELVIC MUSCLES

63

64 POSTILLA

ant proc

10cm br sh

FIG. 15. Right ilium of Hypsilophodon, BM(NH) R193. Dorsal view.
Abbreviations on p. 46.

ij A-FEM BR
A

CA-FEM \~

FIG. 16. Femur of Thescelosaurus showing areas of insertion of the pelvic
muscles. Data from Romer (1927b, figs. 16-18). Abbreviations on p. 46.

P-I-F INT 2

CA-FEML CA-FEM BR

FIG. 17. Femur of Alligator showing areas of muscle attachment. Ventral
view based on Romer (1923b, pl. 25). Abbreviations on p. 46.

REVIEW

STYLE

FORM

RITLE

ABSTRACT

IOMENCLATURE

ILLUSTRATIONS

FOOTNOTES

TABLES

REFERENCES

THOR'S COPIES

PROOF

COPYRIGHT

INFORMATION FOR AUTHORS

The Publications Committee of the Peabody Museum of Natural
History reviews and approves manuscripts for publication. Papers
will be published in approximately the order in which they are
accepted; delays may result if manuscript or illustrations are not in
proper form. To facilitate review, the original and one carbon or
xerox copy of the typescript and figures should be submitted. The
author should keep a copy.

Authors of biological papers should follow the Style Manual for
Biological Journals, Second Edition (Amer. Inst. Biol. Sci.). Authors
of paleontological manuscripts may choose to follow the Sugges-
tions to Authors of the Reports of the U.S. Geological Survey,
Fifth Edition (U.S. Govt. Printing Office).

Maximum size is 80 printed pages including illustrations (= about
100 manuscript pages including illustrations). Manuscripts must be
typewritten, with wide margins, on one side of good quality
8% x 11” paper. Double space everything. Do not underline any-
thing except genera and species. The editors reserve the right to
adjust style and form for conformity.

Should be precise and short. Title should include pertinent key
words which will facilitate computerized listings. Names of new
taxa are not to be given in the title.

The paper must begin with an abstract. Authors must submit com-
pleted BioAbstract forms; these can be obtained from the Postilla
editors in advance of submission of the manuscripts.

Follow the International Codes of Zoological and Botanical Nomen-
clature.

Must be planned for reduction to 4 x 642” (to allow for running
head and two-line caption). If illustration must go sideways on
page, reduction should be to 334 x 634”. All illustrations should be
called “Figures” and numbered in arabic, with letters for parts
within one page. It is the author’s responsibility to see that illustra-
tions are properly lettered and mounted. Captions should be typed
double-spaced on a separate page.

Should not be used, with rare exceptions. If unavoidable, type
double-spaced on a separate page.

Should be numbered in arabic. Each must be typed on a separate
page. Horizontal rules should be drawn lightly in pencil; vertical
rules must not be used. Tables are expensive to set and correct;
cost may be lowered and errors prevented if author submits tables
typed with electric typewriter for photographic reproduction.

The style manuals mentioned above must be followed for form and
for abbreviations of periodicals. Double space.

Each author receives 50 free copies of his Postilla. Additional copies
may be ordered at cost by author when he returns galley proof.
All copies have covers.

Author receives galley proof and manuscript for checking printer’s
errors, but extensive revision cannot be made on the galley proof.
Corrected galley proof and manuscript must be returned to editors
within seven days.

Any issue of Postilla will be copyrighted by Peabody Museum of
Natural History only if its author specifically requests it.

wi

066 3

BOUND DEC 1972

~~ atid >
~ = 565 im ia ——

ha Pia FS FS

Le eee

FE LAPS APES TEI ALS
neal + PORE PP PSP OTL:
PP RAL IIE EIS PES PPI ae
Pw ame = PLO POD POPE HT Fe
2 noe POPP PFO FF Te ~s

PF PS OPCS
ir oa -

Re oe Pe Rett te OD ¢ 7 a : ¥ TS FF OPS
Se Piarn F Shs CO be Sy ATE DAE LSE ok Pair : 2 Re A oa 42
- PN - TET, AAS

tee ree
ag

PAIL ALF IG
ras Fb PK

MeO OO EE SE AP tnt PEEL APU I OE Be ee:
a eee th i had coe asad kar et cp PAI efit apn Dirt
s ee ¥ = 2

Te ae
wy ++ rs ed

us eee
hath tila de meted

hee a aie mbetach nied eee AP ee ra
~ yn Oe Pe