598 PROCEEDINGS OF THE NATIONAL MUSEUM vor,. 87 



central South American fauna is very distinct from that of southern 

 South America, the dividing line being from about the mouth of the 

 Rio de la Plata to north-central Chile. There is abundant faunal 

 and floral evidence of a former bar to the spreading of species of 

 animals and plants northward or southward across this line. It 

 seems most likely that this obstacle to migration was salt water, an 

 arm of the sea or an ocean channel, but satisfactory geologic evidence 

 of this has not been presented. The bar, of whatever sort, was 

 effective in many groups of animals and plants. South of this 

 obstruction the fauna and flora show resemblance to those of Australia 

 and New Zealand in very many items, more than they do to the 

 animals and plants of central South America. Von Ihering made 

 this point clearly, and it has since been confirmed by many others, 

 e. g., Eigenmann in his studies of teleosts. 



By what route did Hyla pass between tropical America and Aus- 

 tralia? Not by any Northern Hemisphere route, for hylids are not 

 now in Euro-Asia except one species, with its several subspecies, a 

 comparatively recent immigrant from North America; not by way of 

 Patagonia and Antarctica, for no hylids are today in Patagonia. 

 This restricts them to a trans-Pacific route across the southern 

 Pacific Ocean, by way of the extensive lauds present there during the 

 Cretaceous period (Haug, 1907-1911; Scharff, 1911; Arldt, 1907; 

 Berry, 1930; Joleaud, 1931; and many others). Study of the Pacific 

 islands and the ocean bottom shows many branching ridges (fig. 146), 

 interconnecting in many ways. Former rising and sinking of Pacific 

 lands is indicated by several things. The coral islands indicate much 

 depression ; the Hawaiian ridge is today rising at one end and sinking 

 at the other; volcanic action and earthquakes, frequently associated 

 with changes of elevation, are and have been present through the 

 Pacific area; it is generally recognized that repeated changes of eleva- 

 tion have occurred among the Malay islands, especially along their 

 southwest-northeast axis (Merrill, 1931). The hypothesis of connec- 

 tion between the South American tropics and Australasia, probably 

 northern Australasia, by means of land ridges, or perhaps land waves 

 such as are illustrated today in Hawaii, is not a far-fetched one. But, 

 if this was the migration route for the Hylidae from America to 

 Australasia, why are there not at least some relic forms among the 

 central and eastern Pacific islands? We fuid Hylas in the islands 

 only of the Papuasian region of the extreme western Pacific. Nothing 

 but subsidence of the southern Pacific lands seems adequate to ac- 

 count for such extermination of former Anura. The repeated forma- 

 tion, expansion, and shrmking of the Arctic (and Antarctic?) ice 

 sheets are estimated to have caused fluctuations of ocean level of over 

 60 feet, but that could be, of course, only a minor factor for islands 

 with mountains of considerable elevation. Evidence of a tropical 



