134 The American Naturalist. 



[F,.i 



form. As the Phosphorites are considered to represent the top of the 

 Eocene, we should certainly expect to find in any species of Pachynolo- 

 phus from this formation the last premolar as complex in structure as 

 the molars. In this jaw, however, the last premolar is not molariform 

 and the collection contains a crushed skull of the same species of 

 Paehynolophus in which all the superior premolars are simpler in 

 structure than the true molars. 



I can find no good generic differences based on tooth structure sepa- 

 rating PropalKotherium from Paehynolophus, and shall consider the 

 former genus as a synonym of the latter in this paper. In Propalceo- 

 therium the species are much larger than in those of Paehynolophus, 

 but certainly size alone can not be considered as of value in generic 

 definitions. In all of the species included in Propalceotherivm, the pre- 

 molars are simpler in structure than the true molars. 



I believe, however, if we divide the various known species of the 

 Hyracotheriince into genera according to the complication of the pre- 

 molars, that we shall be adopting an artificial character, and as shown 

 above. Of al 1 the known form s of Paehynolophus, P. siderolithicus is the 

 only one in which the last upper premolars is molariform and even this 

 species shows considerable variation in this respect. I conclude then 

 that the only natural classification of these forms is a careful analysis 

 of the form of the molar cusps, and to group the species into genera 

 according to the development of the same ; I refer here especially to 

 the European forms of ITyracotherium and Paehynolophus. 



Having thus attempted to show that in nearly all the European 

 species of Paehynolophus the last premolar is simpler in structure than 

 any of the true molars, I come to consider what are the generic differ- 

 ences separating Hyracotherivm from Paehynolophus. Kowalevsky 

 studied Paehynolophus siderolithicus, and if we compare the molars of 

 this species with those of Hyracotherium angustidens from the Wasatch 

 of America, we observe at once that the external cusp of the 



nd they i 



upp,,r 



molars in the latter species are nearly round 



scarcely at all flattened. There is no mesostyle and the height of the 

 crown is very low or strongly brachydont. In P. siderolithicus the 

 ectoloph is considerably lengthened from above downwards, and the 

 external cusps are strongly flattened, with a prominent mesostyle. In 

 all the species of Paehynolophus which I have studied the molar crowns 

 are higher than those of Hyracotherium, and in all the mesostyle is 

 strongly developed. The latter characters demonstrate that the molars 

 of Paehynolophus have reached a higher stage of evolution than those of 

 Hyracotherium, and this transformation in the form of the molar cusps 



