228 MAMMALIAN DENTITION 



tooth overlaps the one in front laterally as in the ruminant 

 Artiodactyla. The lower cheek teeth present anterior and 

 posterior crescents concave lingually bnt since the paraconid 

 itself is absent the ridge passing forward and inward from the 

 protoconid fades away at the anterior margin of the tooth. The 

 hinder ridge of the anterior crescent passes from protoconid to 

 the curiously reduplicated metaconid. The two ridges of the 

 posterior crescent run from the hypoconid which forms the 

 apex to the metaconid and entoconid respectively. 



One peculiar and primitive feature of the skull is the exten- 

 sion of the malar into the glenoid fossa as in Marsupials. In 

 spite of this Hyrax is obviously an Ungulate and its closest 

 relationships might be suspected to lie with the Artiodactyls 

 but this conception is immediately contradicted by many ana- 

 tomical features. There is one of these which may be mentioned 

 since emphasis has already been laid upon the essential and 

 primitive differences between the premolars of Artiodactyla 

 and those of Perissodactyla, The premolars are already very 

 advanced in type and molariform in Eocene representatives of 

 the order Hyracoidea. It may also be noted in passing that 

 these early forms show an enlarged, pointed and triangular 

 upper median incisor as in the existing Hyraces. 



In many other features Hyrax resembles the remote Peris- 

 sodactyla ; in the reduplicated metaconid of the lower molars 

 for example, a character exhibited also by the Eocene Megalo- 

 hyrax. Nevertheless the resemblances of this animal to the 

 Proboscidea in parts of the skeleton other than the teeth, con- 

 siderably outweigh the significance of the Perissodactyl-like 

 molars and premolars. 



In short, as Dr. Gregory says, "the existence of so many 

 'cross resemblances' between the Proboscidea and Perissodac- 

 tyla by way of Hyrax seems more consistent with the hypoth- 

 esis that all these now very divergent orders have been de- 

 rived from a common protungulate stock, than with the hypoth- 

 esis that all the resemblances are due to convergent evolution. ' ' 



Fascinating though the Hyrax may be in its anatomical fea- 



