RECENTLY DISCOVERED TERTIARY VERTEBRATA OF EGYPT. 299 



this resemblance Avith the elephant, in the hind feet at least, is only 

 suj)erficial, the actnal arrangement of the tarsal bones being widely 

 different. As remarked above, the affinities of this remarkable crea- 

 ture are nncertain, and it was considered necessary to establish a new 

 subdivision of the Ungulata for its reception, though at the same time 

 relationships with the Ilyracoidea were pointed out. Winge, on the 

 other hand, in a recently published memoir on the Ungulata, boldly 

 refers it to the Hvracoidea. Probably its real position will remain 

 doubtful till some earlier and less specialized members of the same 

 stock have been discovered. 



Barytherium^ from the Middle Eocene beds, is another large and 

 heavih^ built ungnlate, of which unfortunately very little is yet 

 known. Only the npper and lower jaws, with the cheek teeth, and a 

 few limb bones have yet been found. All are characterized by their 

 immensely massive construction. The teeth have comparativeh^ lov/ 

 crowns, with two transverse ridges. The humerus has all its ridges 

 and processes for the attachment of muscle greatly developed, indi- 

 cating a fore limb of great strength, and, judging from its form, 

 possibly employed in digging. The relationships of this creature 

 are unknown; it is by some regarded as belonging to the Proboscidea, 

 and it has even been suggested that there may l)e some relationship 

 with the South American Pyrotheria. 



Although Aris'moitJieriuin and Barytherimn are interesting for 

 the peculiarities they present, their very isolation detracts consid- 

 erably from their importance, for they throw no light on the earlier 

 history of any of the previously known groups of mammals. From 

 this point of view the remains of primitive Proboscideans from these 

 Eg}'ptian deposits are of vastly greater interest, for the}" at once 

 settle the point of origin of the group and carry back the line to a 

 generalized type of ungulate showing only the beginning of the 

 extraordinary specializations characteristic of the later forms. Pre- 

 vious to the discovery of these Egyptian forms, the earliest Probos- 

 cideans known were species of Tetrabelodon and Dinotherium from 

 the lowest Miocene beds of Europe, where they appear suddenly at 

 this horizon, no trace of any related form being found in the earlier 

 Tertiary deposits of that continent. The sudden apj^earance in the 

 European fauna of these and members of some other groups led to 

 a number of speculations as to where these animals had originated. 

 Osborn, Stehlin, and Tullberg for various reasons all came to the 

 conclusion that the evidence pointed to the existence of an Ethiopian 

 land area in early Tertiary times and they considered that not only 

 the Proboscidea, but several other groups — ^notably the Sirenia, 

 Ilyracoidea, certain Edentates, the Antelopes and Giraffes, the Hip- 

 popotami, several divisions of the Rodentia, and lastly the Anthro- 

 poiclea — originated in that region. Of many of these the early 



