190 F. B. LOOMIS ^ORIGIN OF SOUTH AMERICAN FAUNAS 



marsupial families had become differentiated as early as the beginning of 

 the Eocene. We are already used to thinkmg of the opossum family as 

 distinct at the beginning of the Eocene. The Thyalocines must also have 

 been distinct, and haA'e been in North America in the early Eocene, and I 

 anticipate representative's of the family will turn up in the Fort Union. 

 The Diprotodont division of marsupials must thus early have been estab- 

 lished.^ Gidley, after describing the skull of Ptilodus, concludes that the 

 Allotheria and Multituberculata are marsupials related to, though not 

 ancestral to, the diprotodonts, the two groups having an ancestry still 

 further back.* Last winter, on making a further study of Pyrotherium, 

 I was convinced that its relationship was with Diprotodont, making it a 

 marsupial instead of an elephant, as I earlier thought. In the early 

 Eocene, then, we have didelphids, allotheres, and plagiaulicids common 

 to both continents. In South America there are sparassodonts, csenoles- 

 toids, and pyrotheres as yet not found in North America. On the other 

 hand. North America developed creodonts to take the place occupied by 

 sparassodonts in the south. 



Turning to Edentata (Xenarthra only), we find members of this order 

 represented by teeth and scutes in the Notostylops beds. The Torreon 

 and Puerco of North America yield several genera (Psittacotherium, 

 Onychodectes, Conoryctes, etcetera) which Wortman united as Gano- 

 donta, making clear that they belonged to the edentates. This would 

 indicate that the order originated in Cretaceous time and was in both 

 North and South America in the early Eocene. In North America it 

 slowly died out by the end of the Eocene represented by such genera as 

 Calamodon, Stylinodon, and Metacheiromys ; in South America, on the 

 other hand, they become a little more abundant in the Oligocene, and 

 then in the Miocene suddenly increase in numbers, variety, and size, 

 diverging into the armadillos, ground-sloths, and glyptodons until over 

 half of all finds in these and later beds belong to this group. When the 

 Americas united, in the Pliocene, they migrated to North America in 

 numbers sufficient to make a real element in the fauna of the Pliocene 

 and Pleistocene, but they were in company too fast for them and have all 

 died out on this continent. At the same time our types of carnivores 

 moved to South America, and, though more slowly there, the edentates 

 gave way; so now only diminutive remnants of that great fauna are left 

 in a few armadillos, sloths, and ant-eaters. 



The rodents offer one of the most puzzling problems as to their origin in 

 South America. No rodents are found in the Paleocene of North America. 



^ Gidley has recently described a pouched ant-eater from the Fort Union, assigning it 

 to the same genus. Mymecobius, as the form living today in Australia. Proc. U. S. Na- 

 tional Museum. 191.'i, vol. 48. p. 395. 



* Gidley: Proc. U. S. National Museum, 1909, vol. 36, p. 611. 



