1910.] Perissodactyls; Chalicotheres. 397 



strengthened the eonnection but Hniited the niobilitv, between the astrairahis 

 and the ealeanenni. The sustentaeular and eetal facets at the same time 

 formed the peeuhar J -shaped facet ch\scribed above (p. 395). The narrow- 

 ing of the sustentaeular facet is foreshadowed to some extent in Phenacodus 

 wortmani. After examining considerable material, the writer ventures to 

 assert that in the characters of the astragalus the Phenacodont-Condylarthra 

 approach the Perissodactyl type somewhat more nearly than do the known 

 Eocene representatives of any other order. 



To conclude, the derivation of the Perissodactyl order from the general 

 Insectivore-Creodont-Condylarth group of Placentals seems fairly well 

 established. There is no reason to suspect direct derivation from either of 

 the two lower orders of this group (Insectivora, Carnivora), because a com- 

 prehensive study of the osteology reveals nothing in support of the hypothesis, 

 but shows on the contrary that the stem Perissodactyl had already attained 

 relatively advanced cursorial and herbivorous adaptations which are not 

 foreshadowed in those orders. On the other hand, genetic derivation from 

 any well known Condylarths {Phenacodus, Euprotogonia, Meniscotherium) 

 is almost erpially improbable. But with regard to many important dental 

 and osteological features it is obvious also that Euprotogonia and Phena- 

 codus bridge over the structural gap between the Perissodactyls and the lower 

 unguiculate orders, and in brief, that the stem of the Perissodactyls would 

 very likely fall under the Condylarthra, as re-defined by Matthew (1895). 



The Ancyhpoda (Chalicotheroidea). 



The deeply fissured imgues impressed all earlier writers on the "clawed" 

 ungulate Chalicotherium of the European Miocene and its allies and led Cope 

 to disregard the numerous ungulate features in the rest of the skeleton in 

 favor of the view that Chalicotherimn must have been derived from some 

 primitive unguiculate. He therefore (1889, p. 153) made the Chalicotheri- 

 idfc the type of a distinct order, the Ancylopoda. 



Deperet (1892) saw its prevailing resemblance to "les Pachydermes " ; 

 but the inclusion of Anoplotherium in that group somewhat confused the 

 issue. 



Osborn (1893, pp. 118-133) came to the conclusion that the Ancylopoda 

 were intermediate between Meniscotherium of the Condylarthra on the one 

 hand and the Perissodactyls on the other. He later (1898) included the 

 Chalicotheroidea as an aberrant superfamily of the Perissodactyls. This 

 view has recently been supported with fresh evidence by Peterson (1907, pp. 

 733-752) who, in describing the skeleton of Moropus, points out many peculiar 



