the Cretaceous-T ertiary Problem, 223 



disappear at the end of the Paleocene, others survive to 

 the end of the Eocene, and some still exist precariously; 

 the soft-shell turtles alone are flourishing. The dino- 

 saurs and plesiosaurs apparently become extinct before 

 the close of the Paleocene. 



The true Eocene, however, is sharply distinguished by 

 the sudden appearance of the modern orders of mammals, 

 and of the principal modern chelonians, the same new 

 fauna appearing both in Europe and America and 

 flourishing and evolving throughout the Cenozoic, while 

 the older groups of vertebrates drop out one after another. 

 This is a great and sudden faunal break, clearly due to 

 migration, and represents the incoming of the Tertiary 

 vertebrates into the known parts of Holarctica. It is a 

 faunal change comparable with that of the Pleistocene, 

 not more marked as to percentage of change in the 

 families than that at the end of the Eocene, but consid- 

 erably greater than the Oligocene-Miocene or Mio-Plio- 

 cene breaks. It is important, however, as introducing for 

 the first time the Cenozoic mammals and reptiles, save for 

 the few forerunners noted. It coincides quite nearly with 

 the classic line between Cretaceous and Tertiary in 

 Europe, but would include the Thanetian in Cretaceous, 

 as it would include the Tiffany in America. 



Diastrophism and the Migration of Land Faunas. ^^ 



The hypothesis that appears to fit best with the data 

 at present known is : 



(1.) That Asia was the primary center of evolution 

 and dispersal of the terrestrial vertebrates. 



(2.) That epochs of diastrophism, uplift and con- 

 tinental connection afforded opportunity and environ- 

 mental pressure that caused them to spread out thence 

 into the marginal parts of the Holarctic realm and ulti- 

 mately into the marginal regions (Africa, South America, 

 Australia). 



(3.) The long intervening epochs of rest, peneplana- 

 tion, subsidence and isolation induced evolution in loco, 

 parallelism and expansive evolution of lowland, marsh 

 and shallow water faunas. 



^^ See in this connection Matthew, 1915, ' ' Climate and Evolution, ' ' Ann. 

 N. Y. Acad. Sci., vol. 24, pp. 171-318, where the distribution of land verte- 

 brates is interpreted in accordance with this hypothesis. 



