44 
and exoccipital which he ascribed to the litissimus dorsi. Professor Lull 1 
now agrees that this muscle did not extend to the head, but he still considers 
the attachment area as belonging to a neck muscle. The writer, as noted 
above, regards it as the probable origin of the parieto-mandibularis. 
The pronounced attachment areas on the back of the ceratopsian skull 
indicate powerful neck muscles. Their function has been discussed by Lull 
(2) and by Tait and Brown (10). The latter authors conclude that the 
skull was capable of rotation on the longitudinal axis through angles up to 
90 degrees from the vertical. Both Nopesa (4, page 69) and Lull (3, 
page 22) reject this view, and to the present writer also it appears some- 
what extreme. The ceratopsian neck is rather short to allow the necessary 
play of the cervico-cranial muscles. No doubt some rotation of the skull 
was possible, as well as up and down, and side to side, movement. Never- 
theless, the functional adaptation of the jaws and crest can be explained 
without postulating marked rotation at the atlanto-occipital joint. 
It should be noted that muscular attachments are not recognized on 
the posterior part of the ventral surface of the crest. Some restorations 
of ceratopsians show the lower side of the crest attached for its entire 
length to the neck. There is no direct evidence for this interpretation, and 
many skulls show subcutaneous vascular grooves in this region of the crest. 
Such an attachment would interfere seriously with the free movement of 
the head and crest. 
THE TRUNK 
There is not much direct evidence on the muscles of this region in 
ceratopsians. The longissimus dorsi (Figure 9, Ig. d.) must have been large 
and powerful, reinforced by ossified tendons in the lumbar region. The 
rectus abdominis (Figure 9, r.) must have been rather attenuate posteriorly, 
as it had to stretch from the posterior ribs to the distal portion of the 
ischium. 
FORE LIMB 
It is not proposed here to attempt the description of all of the limb 
muscles, but only those having a direct bearing on the functional adapta- 
tions of the limbs. 
Latissimus dorsi (Figures 8, 9, Id.) . This was a large, triangular sheet, 
originating from the spines of the anterior dorsal vertebrae about as far 
back as the twelfth. It converged downward and was inserted in the ulnar 
ridge and tuberosity of the humerus. Owing to the peculiar position of the 
humerus, this muscle served to rotate the limb, as well as to flex the 
shoulder joint. 
u Levator scapulce >> (Figure 8, l. s.) (not the levator scapulae or serratus 
cervicus of mammals) . An elongate muscle that extended from the, side 
of one or more cervical vertebrae to the coracoid border of the scapula. 
There may have been a ventral branch, corresponding to the levator 
claviculse. Owing to the marked backward inclination of the scapula, the 
levator action of this muscle must have been rather ineffective. 
i Personal communication. 
