304 UNGULATA 



previously unsuspected extinct fauna in the American continent of 

 the Tertiary period, as interpreted by Leidy, Cope, Marsh, and 

 others, has thrown a flood of light upon the early history of this 

 family, and upon its relations to other mammals. 



There have been found in these regions many Camel -like 

 animals exhibiting different generic modifications ; and, what is 

 more interesting, a gradual series of changes, coinciding with the 

 antiquity of the deposits in which they are found, have been traced 

 from the thoroughly differentiated species of the modern epoch 

 down through the Pliocene to the early Miocene beds, where, their 

 characters having become by degrees more generalised, they have 

 lost all that specially distinguishes them as Camelidce, and are 

 merged into forms common to the ancestral type of all the other 

 sections of the Artiodactyles. Hitherto "none of these annectant 

 forms have been found in any of the fossiliferous strata of the Old 

 World ; and it may therefore be fairly surmised (according to 

 the evidence at present before us) that America was the original 

 home of the Tylopoda, and that the true Camels have passed over into 

 the Old World, probably by way of the north of Asia, where we 

 have every reason to believe there was formerly a free communica- 

 tion between the continents, and then, gradually driven southward, 

 perhaps by changes of climate, having become isolated, have under- 

 gone some further special modifications ; while those members of 

 the family that remained in their original birthplace have become, 

 through causes not clearly understood, restricted solely to the 

 southern or most distant part of the continent. The occurrence 

 in the dentition of the fossil Siwalik Camels of a feature now 

 found only in Auchenia is especially interesting from this point 

 of view. 



Briefly referring to some of these fossil types, we may note 

 that Pliauchenia, of the Loup Fork beds (Lower Pliocene) of 

 the United States, has three lower premolars, while in Procamelus 

 there were four of these teeth. In Protolabis of the Miocene 

 we have a more generalised form, in which the dental formula 

 is i f , c ^, p %, m % ; and from this type a transition may be 

 traced to Poebrotherinm, which, while having the same dental 

 formula, was no larger than a Fox, and had the third and fourth 

 metacarpals separate, with rudiments of the fourth and fifth. The 

 earliest undoubted representative of the group is Leptotragulus, of 

 the Uinta Eocene, which appears to have been closely allied to 

 Poebrotherium. It is, however, probable that the first lower pre- 

 molar was wanting; while the other premolars of the mandible 

 were much shorter antero-posteriorly than in the last-named genus. 

 The manus, moreover, appears to have been less reduced, the second 

 metacarpal retaining its connection with the magnum. It is 

 suggested that Leptotraguhis may have been derived from the 



