between Europe and North America. 69 



group of admittedly American origin does not appear in Europe 

 until the Middle Eocene, that there must have been a connection 

 in the Middle Eocene. But, unless we find also groups of 

 European (Old World) origin which first appear in America in 

 the Middle Eocene, we should look into the matter and test the 

 possibility of the first group having migrated from America to 

 Europe in the Lower Eocene, but not being as yet reported from 

 the Lower Eocene formations of Europe. The test would of 

 course be identity of the Middle Eocene genus in Europe and 

 America. If divergent, or merely paralleling in certain 

 respects the evolution of the North American series as observed 

 in Lower and Middle Eocene stages, I think its occurrence may 

 be explained as a consequence of Lower Eocene rather than of 

 Middle Eocene migration. 



In illustration of these statements, the following example 

 will explain my meaning : Sinopa and Tritemnodon occur in 

 Lower and Middle Eocene in North America, — fairly distinct 

 genera in the Bridger ; almost indistinguishable in Lower Eocene. 

 American species have peculiarly elongate premolars, throwing 

 them a little out of the direct ancestry of later Hysenodonts. 

 Sinopa occurs in the Middle Eocene of Switzerland. " Sinopa " 

 ethiopica occurs in the Upper Eocene of Africa. The Euro- 

 pean and African species apparently do not have the elongate 

 premolars ; they are known only from jaws but apparently 

 correspond with Sinopa and Tritemnodon, respectively, in stage 

 of evolution towards Hycenodon. Oynohymnodon of European 

 Oligocene is very little advanced over " S." ethiopica, and 

 might be directly descended from the European or African 

 species. 



Now if we simply take the recorded occurrence of genera as 

 a basis, we must conclude that Sinopa originated in North 

 America in Lower Eocene, crossed over to Europe in the Middle 

 Eocene and thence to Africa in the Upper Eocene. A more 

 detailed study brings out the following points : 



(1) The American species of Sinopa and Tritemnodon 

 paralleled the Old World series, but developed independently 

 from their sudden appearance in the Lower Eocene (Wasatch) 

 to their disappearance in the Middle Eocene (Washakie). They 

 cannot have been directly ancestral to the later Hysenodonts, 

 but represent side branches which left no descendants. 



(2) The European and African species are more nearly in the 

 line of ancestry of the later Hysenodouts, but are too conserva- 

 tive to be ancestral to the large and highly specialized Hyaino- 

 don and Pterodon. 



(3) The latter genera appear in the Upper Eocene and Oligo- 

 cene of Europe and Africa, and Oligocene of North America, 

 accompanying and soon displacing smaller and more conserva- 



