670 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1908. 



ing those of the lower jaw. This genus is reported from the Pliocene 

 (Blanco) of Texas and Mexico and ranges as far south as Buenos 

 Aires in the southern hemisphere. Two South American species are 

 known to us, one D. andium^ following the chain of the Andes as far 

 south as Chile. This tyjDe is often found at great altitudes, a speci- 

 men from the Quito Valley in Ecuador, now in the Yale collection, 

 having been found 10,000 feet above the level of the sea. 



Dibelodon hiimholdii was a dweller on the plains, being found in 

 the pampas formation near Buenos Aires, while Darwin records it 

 along the banks of the Parana River in Argentina, and Wallace 

 reports the same species among other remains in a limestone cavern 

 near the headwaters of the San Francisco Eiver in southern Brazil. 

 D. Jmmholdii^ like D. andium, has its origin in the Texas Pliocene, 

 the line of migrations nearly paralleling, the one along the tropical 

 plains, the other along the Andine jDlateau as far south as northern 

 Chile. With the exception of a lone specimen of Elephas reported 

 from French Guiana and the mastodon of Honduras, Dibelodon is 

 the only proboscidian of the Neotropical realm. The migration of 

 these great forms occurred in the late Pliocene, and for some reason, 

 evidently climatic and vegetative, the route has been closed ever since. 

 Otherwise it is reasonable to suppose that the elephants and masto- 

 dons of the Pleistocene would have spread into South America as 

 well. 



In Europe Tetrcibelodon angustidens had successors in T. longi- 

 rostris and arvernensis, the latter ranging over western Europe into 

 England. It did not, however, cross the Pyrenees into Spain. T. 

 longii'OstHs and a late mutation of T. angustidens^ palmiidicus, made 

 the long journey to the Orient, transferring the evolution from 

 Europe to India. The path of this migration is as yet unknown, as 

 little or no paleontological exploration has been made in the region 

 lying between Armenia on the west, across Persia, Afghanistan, and 

 Beluchistan to the Indus River. This oriental migration must have 

 occurred during the Upper Miocene and was followed by a relatively 

 rapid evolution involving a number of species of mastodons and ele- 

 phants. Tetrahelodon longirostris seems to have given rise to Mam- 

 mut'^ cautleyi with a shortened lower jaw, thence through J/, latidens 

 to Stegodon clifti^ the transitional form between the mastodons and 

 the elephants. S. clifti was followed in succession by Stegodon hom- 

 hifi'ons and S. insignis and finally by the genus Elephas itself. 

 Elephas proved to be a great migrant, although the stegodont species 

 had spread from their original homes in the sub-Himala^-an region 

 eastward through Burma, China, and Japan, and perhaps as far as 



* These Indian forms agree probably with the American mastodon in having 

 but one pair of enamelless tuslis. They may represent the Mammut stage but 

 in an entirely different phylum, hence should not bear the same generic name. 



