THE MAMMALIA OF THE UINTA FORMATION. 485 



Leptotmgulus is, therefore, to be regarded either as the direct ancestor of Poehro- 

 therium, or as very closely connected with that ancestor. Unless, however, the doubtful 

 indications of tetradactyl feet in the Uinta genus should be confirmed, it throws no 

 groat amount of light upon the vexed question of the origin of the camels and their 

 relations to other selenodonts. Among the imperfectly known and sparsely represented 

 artiodactyls of the Bridger, the one which most clearly suggests itself as a forei-unner 

 of Leptotragulus is Ilomacodon. This animal is described by Marsh (!N'o. 9, p. 364) 

 as being " very neai'ly allied to Heloliyns, and but a single step away from this genus 

 towards the selenodonts. * * * This piimitive selenodont had forty-four teeth, 

 which formed a nearly continuous scries. The molar teeth are very similar to those 

 of IL'lohijus, but the cones on the crowns have become paitially triangular in outline, 

 so that when worn the selenodont pattern is clearly recognizable. The first and sec- 

 ond upper molai-s, moreover, have three distinct posterior cusps, and two iji front ; a 

 peculiar feature, which is seen also in the European genera, Dicliohune and Caiao- 

 therium. There were four toes on each foot, and the metapodial bones were distinct. 

 The type species of this genus was about as large as a cat. With Helohyus this 

 genus forms a well-marked family, the Ileloh tjidLe." 



The association of these two genera into a single family is (piite inadmissible, 

 as it ignores a fundamental ditference in the construction of the upper molars ; in 

 Hdolnjus the fifth cusp is in the anterior half of the tooth, while in Ilomacodon it is 

 in the posterioi- half This latter genus belongs clearly to the DichobanidiK, and the 

 puljlished description shows no reason for separating it fi'om Dichohune itself. Before 

 identifying it with the European form, however, it seems best to wait for fuller infor 

 mation with regard to the American species. Similarly, in the absence of the upjjcr 

 molars, it would be premature to state that Leptotrar/nlus is to be derived from IIo- 

 inacodon, though there is nothing in what is known of the two genera to forljid such 

 a derivation, which may thci-efore be fairly assumed. If lliis connection should be 

 confirmed, it would necessitate revision of the current views upon the relationships of 

 the Tylopoda. Schlosser (No. 13, p. 42) regards this group as a branch of the same 

 general stem which developed the Ano])lothen'dce, the Anthracolheridit', etc., and espe- 

 cially as being connected with the oreodonts, a suggestion which has also been made 

 by Uutimeyer (Xo. 11, p. 98). On the other hand, Schlosser derives the tragulines 

 and true ruminants from the JJichobunidti', thus making a very radical distinction 

 between the Tylopoda and the other existing selenodonts. If, however, the sugges- 

 tiiui hcie made, that Ilomacoflan and Leptolraqxhis are genetically connected, be con- 

 firmed, it follows that the Tylopoda are also to be derived from the Dicholmiiiddi, and 

 therefore from the same origin as the other recent selenodonts. Such a result would 



A. P. S. — VOL. XVT. .'5.1. 



