484 THE MAMMALIA OP THE UINTA FORMATION. 



in the direct line of descent, though in any event it is but slightly removed from 

 that line. 



The inferior premolars are composed of the same elements as in Poebrotherium, 

 but have quite a different appearance, owing to their moderate antero-posterior extent, 

 while in the latter genus they are very much elongated ; the difference between the 

 two being least marked in the ease of the last premolar. The most important dis- 

 tinction in the molars of Leptotragulus from those of the White River genus consists 

 in their completely brachyodont chaiacter; they are also not so far removed from the 

 bunodont pattern ; the inner cusps are stout and conical, and the outer distinctly tri- 

 hedral, while in Poehr other turn the crescents, and especially the internal ones, are 

 compressed into thin laminae. The crescents are not so sharply separated by deep 

 constrictions, and the valleys are more nearly confluent in the Uinta genus ; the heel 

 of the last molar is also pro])oitionally larger and contains a deeper valley. Another 

 difference consists in the absence fi-ora the molars of Poebrotherium of the small col- 

 umns developed from the cinguUim between the outer crescents. The elongated 

 slender shape of the mandible recalls that seen in Poebrotherium. 



The atlas is almost a reduced copy of that of Poehrotherhim, but the notch at 

 the side of the anterior cotylus is not nearly so deep, and there is no articular surface 

 developed on the hinder edge of the inferior arch to connect with the centrum of the 

 axis below the odontoid process. 



The scapula has a more oval glenoid cavity, a less prominent coracoid, and the 

 spine is placed more nearly in the median line ; the intercondylar ridge of the distal 

 end of the humerus is narrower and less prominent. 



If we may trust the account given above, the manus is very different in the two 

 genera in that the manus of Leptotragidus is but slightly reduced, and that the sec- 

 ond metacarpal retains its connection with the magnum. The fibula seems also to be 

 less reduced, though the process has obviously begun. The tarsus, metatarsus and 

 phalanges are much like those of Poebrotherium, with only slight difierences, such as 

 the lai'ger size of the cuboid hook and the relative narrowness of that bone. The 

 unguals are very closely alike in the two genera, and indicate that the peculiar foot 

 structure which characterizes the camel and llama could not have occurred in the 

 Uinta and White Kiver representatives of the group, which probably had feet with 

 the general appeai'ance of those of the ordinary ruminants. The pad or cushion 

 developed, no doubt, 'i)ari-2KiM^a with the formation of the peculiar cannon-bone with 

 its diverging distal ends. The ungual phalanges of Protolabis and Procamelus have 

 not as yet been identified, but there is every reason to suppose that they exhibit suc- 

 cessive stages of reduction. 



