ANCIENT GEOGRAPHY OF PATAGONIA 135 



most Striking feature of the group of forms we found is their 

 entire isolation in relationships, from any of the Eocene or 

 later forms on any continent. It looks as though, from the 

 beginning of the Eocene at least, South America was an 

 island continent, as it remained until in the Pliocene when 

 the Isthmus of Panama connected it with North America. 



However, during the Cretaceous and probably during 

 the upper half of this period South America must have 

 received its original mammals from somewhere. The most 

 primitive members of this fauna seem to resemble the 

 primitive stock from which the mammals of North America 

 arose; and, as it now seems, these two continents must have 

 received their original stock from the same source, possibly 

 then from Africa, more probably from the north over 

 Greenland, etc., from northern Europe, whence it spread 

 into South America, and then the two continents were 

 separated and the animals of each continent developed 

 each in its own way, adapting themselves to similar circum- 

 stances often in more or less parallel manners. 



Our collections contain bones representing over 300 in- 

 dividuals, some represented only by a fragment of a jaw or 

 by a limb bone, but most by at least a jaw, a skull, or in 

 four cases by more or less complete skeletons. The most 

 striking specimen is the complete skull of Pyrotherium, 

 a form previously known only by the teeth. The skull, 

 thirty-eight inches in length, is greatly elongated, with the 

 nostril openings nearly half way back from the front of 

 the snout, indicating a flexible proboscis, though not a 

 pendant one. In the front of the upper jaw are four heavy 

 tusks, which is double the number that was expected. 

 Each of them projects some ten inches to the front, and 

 they are matched with two similar ones in the lower jaws. 

 The back teeth are very large, each with two high ridges 

 across it, and the two rows so broadened that almost the 

 whole palate is covered with the dental armor. It looks 

 as if the food required very efi^ective crushing and grind- 



