556 THE MAMMALIA OF THE UINTA FOUMATIOlsr. 



In a general review of the above series, we observe practically the same type of 

 intercarpal articulations in all tetradact}'! and tridaetyl forms irrespective of the geo- 

 logical 23eriod to which they belong. Thus the tapir manus is a persistent example 

 of a type developed independently, at different periods, in Hyracotheritim, Hyracliyus, 

 Acerathermm and Palmosyops. The tridactjd rhinoceros, with Tr/plopus, Hyraco- 

 don and PalcwtTierium present another type. The significance of this parallelism is 

 discussed in another section. 



In the tarsus we find a striking analogy between the equine and Hyracodon 

 series. In another series may be grouped all the remaining Perissodactyla. 



Thei-e is actual disproof of Kowalevsky's generalization as to the bearing 

 of the spreading of the metapodials upon the survival of the fittest, iu the non- 

 spreading of the metapodials at an important stage in the development of the horse 

 series (see p. 546). In fact we can trace no connection between certain foot struc- 

 tures and survival in perissodactyl phyla. We must either with Schlosser (^op. cit., p. 

 121) regard the inadaptation of the teeth to the changes in the character of the food 

 as the cause of extinction, or rather look at the organization as a whole in its rela- 

 tions to the changes both in food and to the rapid increase of Carnivora and com- 

 peting Herbivora. 



IV. THE CLASSIFICATION OF THE UNGULATES. 



This detailed comparison certainly supports in a measure Riitimeyer's* designa- 

 tion of the foot bones as a " verschiebare mosaik," and raises the question as to how 

 far these articulations can be employed in classification. With all the direct con- 

 tradictions in Cope's system urged by Riitimeyer and others, which are here pointed 

 out, tliore still remains the universal principle that in every step of transition from 

 the primitive plantigrade foot towards digitigi-adism, some form of displacement 

 from the serial order, or interlocking of i)arts, took place. This proceeded not at 

 random, but along certain definite lines, conforming in general to the succession in 

 geological time and large existing subdivisions of the Ungulata, and of great value 

 as expressing their relations to each other and to the primitive stock. "VVe cannot, 

 therefore, as the distinguished Swiss palaiontologist advises us (p. 18), rest content 

 with the lines drawn by Cuvier and Owen, because the derivation from the plaui igradc 

 taxeopod ungulate was not recognized in their system, and the underlying principle 

 of Cope's system, since adopted by Marsh, and many of the younger school of palaa- 

 ontologists, rests upon a secure basis of fact. The real problem is, therefore, to ex- 



* Ut'bcr ciiiiiii' Bczicluiiiiii'ii '/.. d. SiiiiiritliitTs'.aiiimrn ullrr ii ncucr Wult. Zuricli. 18S8. 



