486 THE MAMMALIA OF THE IJTNTA FOEMATION, 



further conclusively demonstrate that the oi-eodonts are but remotely connected with 

 the camels, as is indeed made exceedingly probable by what we already know of the 

 earlier members of the two groups. Protoreodua, which will be fully described here- 

 after, shows the unpaired cusp in the anterior half of the upper molars, just as in 

 Heloliyus, from which, indeed, it seems to have been derived, and its i-esemblance to 

 such types as Anoplutherimn and the (so-called) Eocene species of ILjopotaruus is 

 very marked, and Schlosser is fully justified in associating the oreodonts with 

 these phyla. On the other hand, the resemblances between the camels and the 

 oreodonts are only such as are common to all the more primitive selenodonts, 

 and indeed this peculiar family has decidedly more similarity to the tragulines than 

 to the camels. Lejitotragvlvs is almost as different from its contemporary Protoreo- 

 don, as Oreodon is fi-om Poeirothermm, or Merycliyus from Procamelus. The Bi'idger 

 representatives of these two phyla are apj^arently Ilelohyus and Ilomacodoji, which 

 not imjirobably take their rise from a common starting point in Pantolestes of the 

 Wasatch, which would thus be the progenitor of the two chai'acteristically Amei*- 

 ican groups of selenodonts, as well as through the DicJwhwiidfi} of the European 

 series also. Of course, in the present state of knowledge, these suggestions can 

 only be made tentatively ; they are not yet capable of demonstration. 



Whether we adopt the scheme of genetic relationships propounded by Sclilosser, 

 or the one here suggested, it necessarily follows that the word, selenodont can only be 

 used as a term of description and not of classification ; for it seems clear that this 

 pattern of dentition has been independently assumed by many groups of artiodactyls. 

 Nearly all the types of selenodonts may be traced back step by step to the general- 

 ized buno-selenodonts of the Eocene, and the resemblance between the molar teeth 

 of such forms as Trayulus and Cainotherium, Oreodon and Cervus is much more 

 probably due to mechanical conditions than to genetic relationship. 



Measurements. ^f. M. M. 



* Jlomitrodoii •carjiins. \ Lcjitvtriigiiliis proavus. Poihrothcrium wilsoni. 



Lonsilli lower premolar , 'erics (last three) .021 .030 



Lasl lower premolar antorcpost. diumeter .007 .011 



" " transverse " .003 



First lower molar anlcropost. diameter .008 .011 



transverse " .005 .000 



Length lower molar series 017..') .020f .042 



Last lower molar aiitero-post. diameter 007.3 .012t .01!) 



" transverse " ,001 .OOOf -007 



Height of astragalus 014 .020 



* These measurements from Marsh, Am. .Jour. Sti. and Arts (3). IV. p. 120. 

 f Second specimen. 



