182 THE TRANSFORMATIONS OF THE ANIMAL WORLD 



as an almost perfect resemblance with the mandible 

 of the Elephants, to which the Mastodon arvernensis 

 is yet a stranger so far as ancestral affiliation is 

 concerned. 



The branches of the Proboscidians, which we 

 have just been studying in Europe, have always 

 had a remarkable geological longevity. But this 

 longevity has recently been singularly extended 

 by the discoveries made a few years ago in the Oli- 

 gocene and Eocene soils of the Libyan desert. 

 In the portion of that desert which adjoins the 

 pleasant and cultivated valley of the Fayum, the 

 researches of the English geologists, Lyons, 

 Beadnell, and Andrews, have made us acquainted 

 with the ancestors of our Proboscidians, the origin 

 of which had till then remained an insoluble 

 problem to palaeontologists. In a mass of Oligocene, 

 and, probably, Stampian strata, Andrews has de- 

 scribed under the name of Palceomastodon, an 

 Ungulate which is proved by evidence to be the 

 ancestor of our Mastodons with conical teeth - 

 mounds. It is inferior in size to the smallest forms 

 of the Mastodon angustidens (the pygmceus mutation 

 of the Burdigalian stage), and the skull is long, as 

 in this last species ; the upper tusks are much 

 shorter and slightly cast downwards ; the symphysis 

 of the mandible is long, but less in proportion than 

 in the Mastodon angustidens, and it carries two 

 tusks likewise rather small. Finally, the molars, to 

 the number of six all in their places at the same 

 time, thus come near to the normal type of the 

 Ungulates, instead of showing the progressive re- 

 duction of the pre-molars which characterizes the 



