494 THE MAMMALIA OF THE UINTA FORMATIOjISr. 



but with widely extonded transverse processes, which are perforated by the vertebrar- 

 terial canal. The anterioi- cotylus for the occipital condyles is cpiite shallow, the 

 posterior faces foi- the axis are quite flat and with j^reater vertical than transverse 

 diameter. The axis is likewise similar to that of Oreodon, though with differences. 

 The atlanteal faces are narrower than in that genus, but higher, and form a consider- 

 able part of the side walls of the neural canal; the odontoid process is narrower and 

 more peg-shaped, and the articular surface on its lower side is not continuous with 

 the facets for the atlas, but separated by a faint ridge; the upper surface of the pro- 

 cess is marked by a quite high and strong i-idge, which loses itself posteriorly in the 

 floor of the neural canal. In Oreodon the upper surface of the odontoid is either flat 

 or, as is the case in \mmj specimens, it shows an approximation to the spout-like 

 form of the ruminants in the elevation of the edges, so that it becomes somewhat 

 concave from side to side. The shape of the process is thus seen to be quite different 

 in Protoreodoit. The centrum of the axis is keeled and quite strongly opisthoccelus ; 

 neither neural spine nor ti-ansverse process is pi-eserved in any of the specimens. 



The other cervical vertebra^ ai'e i-ather longer in proportion than those of Oreodon 

 and have somewhat more markedly opisthocoelous centra ; the anterior ones at least 

 have either obsolete or very low spines, and apparently all except the seventh exhibit 

 the vertebrarterial canal. 



Except for their light and slender construction, neither the dorsal nor the lumbar 

 vertebrae present any special peculiarities; the posterior dorsals are long and com- 

 pressed, the lumbai-s become broad and depressed in the hinder part of the region. 

 The artiodactyl characteristic of cylindrical and interlocking zygapophyses in the 

 posterior dorsal and lumbal* vertebr;e is quite as well developed as in the White 

 Kiver genera of the group. Several caudal vertebrae indicate that the animal pos- 

 sessed an unusually long and stout tail, even more so than in Oreodon, as is made 

 probable by the size of the transverse process upon the more anterior vertebrje. 



Of the scapula only the distal portion is preserved, which is, however, sufficient 

 to show its more essential characters. The glenoid cavity is subcircular in outline, 

 much as in Oreodon and Hyopotamus ; the coracoid process is recurved and promi- 

 nent, and the spine rises abruptly from the neck, as in Oreodon and the ruminants, 

 and having a very different shape from that which occurs in the pigs. As far as can 

 be judged from the fractured condition of the specimens, the spine may be said to 

 divide the blade into nearly equal pre- and ])ostscapulai- fossae. Thjs position of the 

 spine is rendered the more probable from the iact that it occurs in Oreodon and 

 Hi/ojJotamtis, with the scapula of which the jrortions of the shoulder blade of 7Vo- 

 toreodon which are preserved in the collection closely agiee. 



