496 THE MAMMALIA OF THE UIXTA FORMATION. 



tei'istic of the Oreodontidoi and is cori-elated with the equally characteristic shape of 

 the trochlea of the humerus. The ulnar facet on the posterior side of the head of 

 the radius is much less conspicuous than in Oreodon. The shaft has a somewhat flat- 

 tened, ti-ansversely oval shape, which does not tend to become cylindrical, as is the 

 case in Oreodon. Unfortunately we have not been able to identify the distal end of 

 the I'adius. 



The carpus. As one Avould naturally expect, the structure of the carpus of 

 Protm-eodon is very much like that of the Miocene members of the family, though 

 some variations of importance may be observed, esi3ecially that the various elements 

 are less cuboidal in outline and of lighter construction. 



The scaphoid is more extended transversely and less antero-posteriorly than in 

 Oreodon; the proximal surflice is more deeply concave, the rising in front more 

 abrupt and the radial facet descends further on the anterior side; on the distal sur- 

 face the facet for the trapezium is distinctly larger than in the Miocene t3'pe, where 

 it is exceedingly minute ; the trapezoid facet is also larger in proportion, while that 

 for the magnum is correspondingly smallei', nor are there any such anterior ridge and 

 posterior concavity as are to be seen in the later representatives of the family. In 

 Oreodon the lunai', as Cope has pointed out, has the remarkable peculiai'ity of resting 

 almost entirely upon the unciform, while the magnum has moved almost completely 

 under the scaphoid, a tendency which I'caches its maximum in 3ferycoc7iosrus and 

 Merychyus, when there is only a lateral contact between the lunar and the magnum. 

 Among recent artiodactyls, the only group where such a displacement of the carpal 

 elements is to be found is that of the Tragulina. 



In Protoreodoii the same peculiar consti'uction of the carpus appears, but it has 

 not been carried to quite the same extent. The lunar has a gi-eater antero-posterior 

 diameter, the radial surRice is more extended in the same direction and is not nearly 

 so strongly convex. The distal beak-shaped prolongation, which passes between the 

 magnum and the unciform, has not moved quite so far toward the radial side, and 

 even in front the lunar rests somewhat, though but slightly, upon the magnum, while, 

 behind, the magnum is proportionately much larger and presents more directly down- 

 wards, instead of being rather more lateral than distal, as is found to be the case in 

 Oreodon. The unciform facet of the lunar is of course smaller than in the last-named 

 Ibrm. The Uinta genus thus presents a transitional stage between the carpus of the 

 more typical artiodactyls and that characteristic of the Orcodontida;, though the 

 tendency towards the latter is already very distinct. 



The cuneiform dilfers more from the cori-espouding bone in Oreodon than does 

 any of the other carpal elements. Indeed, the specimen which we have regarded as 



