47 
principal extensor of the knee joint and apparently a very powerful muscle. 
It was reinforced by the femoro-tibialis, corresponding to the mammalian 
vasti. 
Ilio-fibulans (Figure 9, if.) . This is the functional equivalent, at least, 
of the mammalian biceps femoris, although the origin is different. In cera- 
topsians the origin probably was from the dorsolateral border of the ilium, 
posterior to the antitrochanter. Such an origin implies a muscle almost as 
large as the ilio-tibialis. The insertion was on the external side of the 
fibula, near the proximal end. This was one of the principal flexors of the 
knee joint, but also served to rotate the lower part of the limb somewhat 
outward. 
Flexor tibialis (ilio-flexorius, ischio-flexorius) (Figure 9, f.t .) . This is a 
single muscle in Sphenodon, incompletely separate from tne pubo-ischio- 
tibialis (gracilis). In the crocodiles (6) it is split into a confusing series 
of slips. Romer (7) recognizes two branches in the omithopods, one aris- 
ing from the posterior extremity of the ilium and one from the pubis. The 
insertion was on the inner side, proximal end, of the tibia. In the Cera- 
topsia these muscles would function as flexors and adductors; the limited 
areas of origin suggest only moderate power. 
Adductor femoris . This probably originated from the pubo-ischiadic 
junction. As this area is restricted in ceratopsians, the adductor must 
have been relatively smaller than in omithopods. The insertion would be 
on the posteromedial surface of the femur near midlength or below. 
Coccygeo-femoralis. There are two muscles of this name in reptiles, 
both originating from the anterior caudal vertebras. It is generally agreed 
that in dinosaurs they were inserted in the fourth trochanter of the femur, 
although it is possible that the anterior or brevis member of the group had 
a somewhat more dorsal attachment. As the fourth trochanter in most 
ceratopsians is relatively reduced, these muscles may be assumed to have 
been less powerful than in omithopods. In bipedal dinosaurs the coccygeo- 
femoralis muscles would be very important, as they would serve to depress 
the tail, and thereby elevate the anterior part of the body. In the quad- 
rupedal ceratopsians they naturally would be less important. 
GENERAL BODILY FUNCTIONS 
Locomotion . The most striking feature of the ceratopsian limbs is 
the marked difference in pose and mechanical arrangement between the 
anterior and posterior pair. This fits in well with the theory of bipedal 
ancestry for the ceratopsians. The fore limbs are typically reptilian, pro- 
jecting out from the body, and suggesting that the thoracic region was 
supported free only by considerable muscular effort. The hind limbs, in 
contrast, have the mammalian posture, operating principally in the fore 
and aft plane. A certain amount of rotation was possible, but very little 
adduction, for the great trochanter is situated uncomfortably close to 
the overhanging margin of the ilium. Locomotion in the ceratopsians must 
have been very peculiar. We must visualize the fore limb, partly extended, 
being swung through an arc in the forward stroke. As it approached the 
forward position it would be flexed. The back stroke would be accom- 
plished by extension of the limb, accompanied by adduction of the 
humerus. At the same time the hind limb would be going through a 
