232 BULLETIN UNITED STATES GEOLOGICAL SURVEY. [FoZ.V. 



not difficult to trace, and I give the following diagram in explanation of 



it. 



Ccclodonta. 



/ 

 Ehinocerus. Atelodus. 



\ / 

 CeratorMnus. 



Aplielops. 



( 

 Zalabis. Aceratlierium. 



\ / 



It is evident that the descent diverged at a comparatively late period 

 of geological time into two lines, which are represented at the present 

 day by the African and li^dian species respectively. The earliest species 

 of the toothless or African series is the Atelodus pacliygnatlius of Wag- 

 ner, whose characters have been so well worked out by Gaudry in his 

 great work on the Fossil Fauna of Attica. That species sometimes pre- 

 sents a single small incisor or canine tooth in the mandible. From what 

 has preceded it is also apparent that the generally most specialized type 

 of rhinoceros, the genus Ccelodonta, has become entirely extinct. Its 

 species yet known, were confined to Europe and Northern Asia, and 

 the most formidable of them extended its range with the hairy mam- 

 moth within the Arctic circle. The Coelodonta antiquitatis was evidently 

 the most effectively armed of the family, as it had two horns, which, 

 judging from the character of the surface of the skull to which they 

 were attached, must have been of unusual size. To jjrovide further 

 against the shocks incitlent to their use in combat, the nareal septum 

 was ossified, thus becoming a solid support to the nasal bones, etc., on 

 which they stood. 



It remains to look backwards, and to discover, if possible, the probable 

 origin of the family in that of its earliest known genus, Aceratlierium. 

 A late survivor of this ancestral type is seen in the genus Zalabis Cope, 

 of which one species, the Z. sivale^isis, has been discovered by Cautley 

 and Falconer in the late Tertiary of Hiudostan. In this form, according 

 to Falconer, there are f incisors and ^ canines. The early type, which 

 corresponds most nearly with this genus, and which preceded the Acera- 

 tlieria in time, is the genus Amynodon Marsh, which has left a species 

 in the Uinta or Upper Eocene of Utah. Here the incisors are f and the 

 canines \. This formula is intermediate between that of Aceratlierium 

 and that of the Eocene tapirs, where the normal numbers f \ prevail. 

 According to Marsh, Amynodon further differs in the primitive condition 

 of the premolars above, which, as in the LopModontida', differ from the 

 molars in their greater simplicity. Thus it is probable that tapiroid 

 animals, probably Lopliiodontidce, gave origin to the RMnoceridcc, as 

 Marsh has suggested. And it is further altogether probable that the 

 general type of dentition presented by the Rhinoceridw, LojjModontidcv, 



