ITS ANCESTORS AND RELATIONS. 17 



tial characters of the feet can be readily appreciated, 

 even by those who have little anatomical knowledge, 

 and suffice to show the fundamental distinction be- 

 tween them. 



Having now eliminated from consideration all 

 mammals but ungulates, and all ungulates but peris- 

 sodactyles, we may henceforth confine our attention 

 solely to this group, as it is the one which contains 

 the horse and all its nearest relatives, and we must 

 first endeavor to trace its history back in geological 

 time as far as our available records will take us. 



It is now well known that mammals existed far 

 back into the secondary or mesozoic age, as far back 

 as the Khaetic or uppermost beds of the Triassic 

 system j but these had none of the characters of 

 ungulates. They were all very small in size, and 

 apparently more nearly allied to the Mar supi alia 

 and Insectivora than to any other existing orders. 

 Until quite recently not a trace of any mammal had 

 been found in any of the strata attributed to the great 

 Cretaceous epoch. The blank has, however, been 

 partially filled up by the discoveries in North Amer- 

 ica announced by Professor Marsh ; but we know 

 as yet too little of these to be able to form any satis- 

 factory opinion as to their affinities, or to pronounce 

 with any certainty whether they carry back the pedi- 

 gree of the perissodactyle group beyond the com- 



