488 THE MAMMALIA OF THE UINTA FORMATIOlSr. 



the camels and llamas, separates from the primitive selenndont line in the Eocene, 

 probably through the genus Paranisryx." Whether any of these genera includes the 

 previously described Agriochoerus ■pumiliis, we are not informed, but the reference to 

 the i-esemblance between the dentition of Eomeryx and that of ffyojwtcnmcs and 

 AgriocJmrus may perhaps be taken to imply that the former is founded upoH the 

 superior dentition of A. pumilas, and may therefore be the same as Protoreodon. 

 That Parameryx may be identical with Leptotragulus is suggested by the reference 

 of the former to the camel line. However, these meagre accounts, with no species 

 descriptions, and no accompanying figures, are insufficient for identification. 



The osteology of Protoreodon is in most res[)eets vei-y similar to that of its 

 White River successor, Oreodon, though in several respects it is more primitive and 

 presents many interesting indications which serve to point out the line of descent 

 thi'ough which this remarkable family has passed, as well as its relationships with 

 other groups of selenodonts, both American and Eui-opean. 



The sJcull, which fortunately is in a fairly good state of preservation, presents 

 some features of great interest. It resembles in general construction that of the 

 Oreodontidce, being in some respects like that of Oreodon, in others more like the 

 skull of Agrioclicerus, while, in several ways, it is more primitive than either. As a 

 whole the skvdl is longer and narrower than in the Miocene forms, a statement which 

 applies especially to the cranium and more particularly to the posterior region of it 

 behind the post glenoid processes, the length of which region is to a remarkable 

 degree greater than in the later genera of the group. The face, on the other hand, is 

 rather short, the anterior edge of the orbit being above the first molar. The cranial 

 and facial axes are in the same straight line, and the upper contour of the skull is 

 also nearly straight, as it is not complicated b^^ the development of large air sinuses 

 in the frontal region. The sagittal crest is very long and prominent, the occiput low 

 and narrow. 



As is general among the Oreodontidx the parietals have a very great antero- 

 posterior extent and form with the squamosals almost the entire cranium ; posteriorly 

 they are very narrow, but become much broader in front of the squamosals, where 

 they send down great processes to meet the alisphenoids. The high sagittal crest 

 extends the entire length of the pai'ietals. The occiput is rather strongly convex, 

 but is low and narrow, as are also the exoceipitals, which do not extend upon the 

 side walls of the cranium. The paroccipital processes are long and slender, and are 

 much like those in Agrioclicerm, but are furthei- in advance of the occipital condjdes 

 and even more widely separated from tlic postgleuoid processes than in that genus, 

 the elongation of this region of the cranium being especially characteristic of Pro- 



