MUSEUM OF COMPARATIVE ZOOLOGY. 83 



lower molars. Cosoryx is very closely allied to Blastomeryx, and is distin- 

 guished from it chiefly by the much more hypodont molars. The bones of 

 the various Loup Fork species of this genera cannot be distinguished apart in 

 the absence of associated teeth, and it is quite probable that the John Day 

 species of Blastomeryx will prove to belong to a different genus from the Loup 

 Fork species. 



The Dentition. 



One undoubted specimen of Cosoryx contained in the Cambridge collection 

 consists of a fragment of the superior maxillary containing one molar, the lower 

 jaw with first and third molars, an antler, the sacrum, all the lumbar and the 

 five posterior dorsal vertebrae in unbroken succession, the scapula, humerus* 

 pelvis, and posterior cannon-bone. The resemblance of these bones to Antilo- 

 capra is veTy striking, and fully justifies what Schlosser has said with regard 

 to the relationships of the two genera. The second upper molar is not much 

 extended in the antero-posterior direction, and has a fairly high crown, though 

 not hypsodont to the same degree as in the prong-buck ; the median fold of 

 enamel on the external wall, or, more properly speaking, the projecting anterior 

 horn of the postero-external crescent is less strongly developed than in the re- 

 cent form, and the corresponding horn of the anterior crescent hardly projects 

 at all. The valleys are shorter and wider than in Antilocapra, and though the 

 tooth is in an advanced state of wear, they are still quite deep, in contrast to 

 what occurs in the lower molars. The lower incisors and canines are all 

 broken away, but from the alveoli and remaining fangs it may be seen that they 

 were of the ordinary ruminant pattern, probably not very long; they decrease 

 in size from the median incisor outwards, and the canine is the smallest of the 

 series. The premolars, three in number, are represented only by their alveoli, 

 which shows them to have been very small. The most anterior is implanted 

 by a single root, the others by two. Leidy's figure (Merycodus necatus) shows 

 them to possess considerable complication, but they are less molariform and 

 more trenchant than in Antilocapra. The true molars are more truly hypso- 

 dont than in the upper jaw ; the first is very small, but the third resembles that 

 of the modern genus exceedingly closely. 



The same may be said as to the form of the mandible itself; the horizontal 

 ramus is very long, compressed, and rather shallow, and with an extremely long 

 diastema between the canine and premolar 3"; the ramus is less rounded on the 

 external side than in Antilocapra, and in that genus there is no such descent of 

 the upper margin in front of the premolars as occurs in Cosoryx. The sym- 

 physis is short (much shorter than in C. trilateralis, Cope) and much contracted, 

 and on a level with its posterior edge is a large single mental foramen. The 

 antler is branched like the one figured by Leidy with the name of Cervus Warrcni, 

 but with a much longer beam, and the tines meeting at a more open angle. 

 The beam is longer and the tines shorter than in any of the antlers figured by 

 Cope, except, perhaps, the imperfect specimen named Cosoryx (Dicrocerus) teres 

 (Wheeler, PI. LXXXII. fig. 6). The antler is composed of dense bone, with a 



