192 
PAKTODONTA. 
tliere are no cross-crests representative of those of Ilyrachyus and RJiinocerus, 
the strong anterior crest which appears to be such being homologous with 
the anterior basal cingulum of Ilyrachyus and Palceosyops. The last milk 
molar of Coryphodon resembles the fourth premolar of Palceosyops^ and there 
can be no doubt of the homology of the single inner cusp of this tooth in 
the two genera. In Palceosyops, this represents the anterior cusp of the true 
molars where two are present, so that the same relation can be claimed for 
the same cusp in Coryphodon. As I have heretofore shown,* the crowns of 
the molars in Loxolopliodon support two transverse crests, which unite or 
approach at their interior extremities, forming a V, near to the apex of which, 
posterior to it in the last true molar, is a tubercle. . The anterior transverse 
crest appears to be homologous with that in the molars of Coryphodon, but 
the posterior cannot be homologized with the posterior transverse crest of 
the last molar of that genus, since the tubercle found in Loxolophodon\>Qhva.A. 
this crest must be compared with that part in Coryphodon, or with the 
posterior inner tubercle of Limnohyus. It is thus probable that the posterior 
transverse crests in Loxolopliodon must be compared with one of the two 
transverse crests in the Tapiroid genera and Dinotherium. The homologies 
of the inferior molars of Coryphodon are simple, as they resemble nearly 
the corresponding teeth in Ilyrachyus, an approach in the anterior V to 
Palceosyops. 
Vertebrce, ribs, and sternum. —The least satisfactory part of the descrip¬ 
tion of the skeleton is that relating to the vertebral column. The number 
of dorsal, lumbar, and sacral vertebrae remains unknown, and of the caudal 
vertebrae it can only be said that they were rather numerous, indicating a 
tail of medium proportions. 
The cervical vertebrae resemble those of the Carnivora more than those 
of living Perissodactyla, but resemble most those of Lobasileus. The centra 
are short and wide, and usually depressed, and do not display the ball-and- 
socket articulation, nor are they absolutely flat, as in Elepliantidce. They 
are slightly concave behind and convex before, as in the Carnivora gen¬ 
erally. The vertebral canal is present, including the axis. In the atlas, it 
pierces the transverse process on the superior surface close to the posterior 
* Paleontological Bulletin, No. 7, August. 1872. 
