502 THE MAMMALIA OF THE UINTA FORMATIOX. 



open orbit and probable absence of the lachrymal depression ; (3) the position and 

 shape of the posterior nares ; (4) the position of the infraorbital foramen; (5) the 

 character of the anterior premolars ; (G) the character of the lower molars and of the 

 internal part of the upper. These difterences of im[)ortance separate Agrioclicerus 

 from Protoreodon, namely, that in the former there are considerable diastemata in both 

 upper and lower jaws, that the last pi-emolar has assumed more or less completely the 

 pattern of the molars, and that the outer crescents of the upper molars are over- 

 hanging. 



Can Protoreodon then be regarded as the common ancestor of both sections of 

 the Miocene family, the Oreodontmce and the Agriochoerinoe f As to the derivation of 

 the former subfomily from this genus there seems to be no I'easonable doubt; its rela- 

 tions to the second are more obscure. If we accept Schlosser's view that a closed 

 dental series is always secondary, and a sign " dass der betrefiende Stamm am End- 

 ziel seiner Entwicklung angelangt ist, wenigstens finden wir diesen Zustand, nur bei 

 solchen Formenreihen, welche gerade im Aussterben begriffen sind " (No. 13, p. 40), 

 it is clear that Agn'ocJmrus, with its diastemata, cannot be derived from Protoreodon. 

 But even if this principle be generally true, it is not without exceptions, as the long 

 persistence of the oreodonts themselves demonstrates, and the diastemata are, there- 

 fore, of themselves insufficient to decide the question. A more important difficulty 

 is the constitution of the outer crescents of the upper molars, Avhich in the known 

 species of Protoreodon have commenced to assume the flattened shape characteristic 

 of Oreodon. The Uinta genus therefore stands very near to the common ancestor of 

 Oreodon and Agriochoerus, but seems itself not to be that ancestor, at least so far as 

 we may judge from the teeth of P. parmis ; it is quite possible that another species 

 of the same genus may stand in the same relation to Agrioclicerus. 



The forerunner in the Bridger fauna of Protoreodon would seem to be Helohyus 

 Marsh, known as yet only from the teeth. In this genus the upper molars have five 

 cusps, the unpaired one in the anterior half of the tooth, which are of a pyramidal 

 shape, and refer the genus to the generalized group of buno-selenodonts, which is 

 represented by the Eocene iiyopotamids of Europe. If more perfect specimens shall 

 confirm this supposed connection between Helohyus and Protoreodon, it will demon- 

 strate the connection of the Oreodunti'd'n with the Anthracotheridce, which has been 

 already often surmised. With regard to the connection of the former group with the 

 Tylopoda, we have seen in considering Leptotragulus that the relationship is proba- 

 bly a very remote one, and that the Tylopoda seem to pass backwards into the Dich- 

 ohunidtp, from which Schlosser derives the typical ruminants and the Tragulina. 



In his later papers (No. 5, p. 384) I*rof. Cope has included Protoreodon among 



