January 28, 1916] 



SCIENCE 



117 



America then had no hares, rabbits, pikas, 

 squirrels, marmots, beavers, rats or mice, 

 but in place of them had a host of cavies, 

 chinchillas, agoutis, and several other 

 families, all of which still abound in the 

 southern continent. The South American 

 monkeys present a problem of much dif- 

 ficulty; all our present knowledge justifies 

 us in saying that they could not have come 

 from North America, for that continent 

 never had any monkeys and even the 

 lemurs became extinct here after the 

 Eocene. 



The Miocene faunas of North America a: e 

 in the sharpest possible contrast to those 

 of South America. They contained several 

 families of true Carnivora, wolves of many 

 diverse kinds, saber-toothed tigers and true 

 cats, weasels, martens, otters, raccoons, and 

 the like ; a great abundance and variety of 

 hoofed animals belonging to the artiodae- 

 tyls and perissodactyls, horses, tapirs, rhi- 

 noceroses, ehalicotheres, peccaries, camels, 

 llamas, deer, antelopes and certain fam- 

 ilies, such as the highly characteristic 

 oreodonts, which are now extinct, while the 

 earliest of American mastodons had already 

 made their way in from the Old World. 

 The rodents, though including some very 

 bizarre forms, now extinct, all belonged to 

 the familiar northern families, rabbits, 

 squirrels, marmots, beavers, pocket-gophers, 

 jumping mice, kangaroo rats, vesper mice, 

 etc., etc. There were no marsupials, eden- 

 tates (with an exception to be noted subse- 

 quently) , monkeys, rodents of the southern 

 families, nor any of the remarkable hoofed 

 animals which then swarmed in such multi- 

 tudes in South America. Nothing could be 

 more obvious or more assured than the con- 

 clusion that the two Americas had long 

 been so completely separated that no migra- 

 tion of land animals from one to the other 

 was possible and, in this long interval, the 

 operation of divergent evolution had 



brought about this total disparity and un- 

 likeness of faunas. 



The junction of the two Americas, ,ty 

 way of the Isthmus of Panama, would seem 

 to have been effected early in the Miocene 

 epoch, or possibly even at the close of the 

 Oligocene, and then began the slow process 

 of the intermigration of land animals in 

 both directions. The earliest indication 

 of the animal of Notogsean forms in North 

 America is the claw of a ground-sloth, dis- 

 covered by Sinclair in the middle Miocene 

 of Oregon but it was not till the Pliocene 

 and, more strikingly, in the Pleistocene, 

 that the southern immigrants arrived in 

 large numbers, and in South America no 

 northern types have been found in beds 

 older than the Parana formation, which I 

 believe to be Pliocene, but which may prove 

 to belong to the later Miocene. 



As thus recorded, the process of mam- 

 malian diffusion might seem to be in- 

 credibly slow, but there are several con- 

 siderations which help to explain the ex- 

 treme tardiness with which the exchange 

 of animals between the two continents was 

 carried on. (1) The Miocene mammalian 

 faunas which have as yet been recovered 

 are in the far north and the far south and 

 we know nothing of the intermediate 

 regions, Central America, northern and 

 middle South America. (2) As previously 

 pointed out, the so-called migration of 

 mammals is merely a gradual spread, a wider 

 and wider range, as increasing numbers de- 

 mand fresh sources of food. (3) For a con- 

 siderable period after the upheaval of a sea- 

 bottom into land, it must remain impassable 

 to most mammals, because devoid of vege- 

 tation, and until plants have taken posses- 

 sion of it, it can serve as only an imperfect 

 and difficult means of communication. (4) 

 Another and very important obstacle to 

 migration between the Americas was that 

 it took place along the lines of longitude 



