MATTHEW, CLIMATE AND EVOLUTION 283 



As for Miolania, it occurs in the Notostylops Beds of Patagonia and in 

 the Pleistocene of Australia. The Notostylops Beds are Eocene, as here 

 advocated. The persistence of so highly specialized a genus for so long a 

 period appears surprising; if they are Lower Cretaceous, as Ameghino 

 asserts, it is quite unprecedented. My acquaintance with chelonian anat- 

 omy is not adequate to warrant my venturing an opinion as to how far 

 parallel evolution from less specialized Pleurodira might account for this 

 anomaly. But we certainly do not know to what extent this genus or a 

 less specialized pleurodiran ancestor may have been aquatic or even ma- 

 rine in its habits. And unless we suppose that it had some such semi- 

 marine adaptations which would enable it to cross a marine barrier im- 

 possible for terrestrial mammals, I do not see how to account for its reach- 

 ing Australia without any of the Notostylops mammalian fauna accom- 

 panying it. We cannot believe that a placental fauna ever reached Aus- 

 tralia, for if it had we should not see the development of a marsupial 

 fauna on analogous adaptive lines to take its place. Miolania, then, could 

 cross some barrier, presumably an ocean barrier, which land mammals 

 could not; and it becomes merely a question of how wide a barrier this 

 extinct chelonian of unknown habits could cross. The present lines of 

 the continents within the continental shelf would not present materially 

 greater difficulties in its reaching Australia via Antarctica than Testudo 

 has managed to surmount in reaching Mauritius and the Seychelles, and 

 I think we are justified in saying that the occurrence of Miolania has no 

 weight as evidence of former Antarctic connections of the Southern conti- 

 nents and, in fact, is opposed to any actual land connection. 



The following notes on the distribution of the land Chelonia are sum- 

 marized from Dr. Hay's monograph : 



Cryptodira are the dominant group of turtles and compare with the pla- 

 centals among mammals. All continents except Australia. 



Chelydridce. — Central America, eastern North America and New Guinea. 

 Apparently a relict-distribution, but the family is unknown fossil. 



Dcrmatemydidre. — Part of Central America. Found in abundance in North 

 America in the Upper Cretaceous and in reduced numbers during the Tertiary. 



Emydidiv. — Chiefly Holarctic and Oriental. A few have reached South 

 America, none in Ethiopia, Madagascar or Australia. First known in Holarctic 

 Lower Eocene. 



Testudinidce. — Very abundant in Tertiary Holarctica but now mostly re- 

 stricted to its southern margin. Alnindant now in Ethiopia and a few species 

 in Neotropical and Oriental regions: also in oceanic islands. Present in Su- 

 matra, absent in Java, present in Celebes but absent in Borneo. These and 

 other features are very suggestive of man's having had much to do with the 

 local extinction of Tortoises. For obvious reasons this family would be pecu- 

 liarly subject to his ravages. 



