146 GRADUAL MODIFICATION OF ANIMAL FORMS xi 



living state, as it brings back, in the most striking way, the 

 general facies of the fauna of those ancient times. In one 

 respect the tapir is remarkable among Perissodactyles, as it has 

 on its fore-feet as many as four toes, thus retaining a primitive 

 or generalised character. The other two existing forms, the 

 rhinoceros and the horse, appear to be more direct modifications 

 of the Palseotherium type, though in different directions. The 

 existing rhinoceros closely resembles the Palaeotherium in the 

 general structure of its skeleton, limbs, number of toes, etc., 

 and in the general pattern of the molar teeth ; it differs, 

 however, in the greatly reduced number of front teeth, incisors 

 and canines, which in the African two-horned species are often 

 absolutely wanting ; and also in the possession of those 

 singular epidermal appendages to the face, the well-known 

 horns, either one or two in number. Now palaeontology 

 points out with tolerable precision the intermediate steps by 

 which these modifications have been brought about. A small 

 ancient rhinoceros has been found in the early rniocene of 

 North America, to which Leidy has given the name of 

 Hyracodon, which had no horn, and had the complete number 

 of incisor and canine teeth, and was in many ways, at least as 

 far as the skull and teeth are concerned, intermediate between 

 Palceotherium and Rhinoceros proper. The earlier known 

 European rhinoceroses have had the name AceratJierium given 

 to them, the small size of the nasal bones being apparently 

 quite unfitted to support such a weapon as a horn. The 

 resemblance of their skull to Pcdceotherium has been pointed 

 out by H. v. Meyer. 



The more recent fossil rhinoceroses present wonderfully 

 intermediate forms between some of the existing species, as 

 R. pachygnathus of Pikermi, as Gaudry has shown, is about 

 equally related to the two species of modern African rhinoce- 

 ros, and might have been (upon the derivative hypothesis) 

 the ancestor of both. In the same way the Himalayan R. 

 sivalensis appears to be related to the modern R. indicus and 

 sondiacus, and the R. schleirmacheri to the Asiatic two-horned 

 species. One special line of variation indicated chiefly by the 

 ossification of the nasal septum culminated in the R. ticlio- 

 rliinus, which became extinct only in the most recent geological 



