February 28, 1896.] 



SCIENCE. 



309 



in later Oligocene beds, it does not ap- 

 pear in North America till the end of the 

 Miocene or beginning of the Pliocene, and is 

 very scantily represented here to-day. The 

 Ungulates are much more distantly related 

 to those of the north and can be connected 

 only by remote ancestors, for the divergence 

 is very striking in the oldest South Ameri- 

 can forms yet recovered. If the connection 

 with the north was not by means of North 

 America it can only have been through 

 Africa. Admitting such connection, it is 

 much more likely to have been due to the 

 junction of both continents with the Ant- 

 arctic land mass than to a Transatlantic 

 bridge. Such a mode of connection would 

 explain the very wide divergences in the 

 character of the mammalian faunas which 

 still exist between Africa and South Amer- 

 ica, for a circumpolar land would very 

 likely oppose climatic barriers to migration, 

 and confine that migration to comparatively 

 few groups. (2) The presence of numerous 

 marsupials of distinctively Australian type 

 in the Tertiary rocks of South America is 

 very strong evidence indeed that both of 

 those continents were connected with the 

 Antarctic land. The Australian'marsupials 

 have been much misunderstood and many 

 observers appear to think that Australia is 

 a sort of museum which has preserved 

 ■Jurassic types to this day. As a matter of 

 fact, these marsupials are an extremely di- 

 versified and modernized assemblage of 

 forms, which have pp,ralleled the placental 

 orders in a remarkable way. Their struc- 

 ture is, it is true, fundamentally primitive, 

 but their many and divergent adaptions are 

 modern. That these marsupials indicate a 

 land connection between South America and 

 Australia can hardly be denied, for none of 

 them have ever been found in any northern 

 continent. If it be asked why this sup- 

 posed migration was all in one direction, 

 and why South American mammals did not 

 reach Australia, several possible explana- 



tions suggest themselves, (a) The mar- 

 supials may have originated in South 

 America and, covering the South Polar 

 lands, have reached Australia, which was 

 then severed from Antarctica, before the 

 Placentals had made their appearance in 

 South America, (b) Placentals may have 

 reached Australia but not kept a foothold 

 there, finding conditions unfavorable to 

 them. These possibilities seem very un- 

 likely and much more probable is a third 

 explanation, (c) The Australian connec- 

 tion with Antarctica first existed and al- 

 lowed the marsupials to spread over the 

 polar lands. Before South America became 

 connected with the circumpolar area, the 

 latter was severed from Australia. Until 

 Tertiary mammals are recovered in Austra- 

 lia, explanation of these curious circum- 

 stances must remain conjectural. What is 

 known of Australian Pleistocene mammals 

 indicates that nothing had reached that 

 continent from South America. 



Another line of evidence which trends in 

 the same general direction as that which 

 we have already considered is given by the 

 Pleistocene birds of the southern hemi- 

 sphere to which attention has been directed 

 by Forbes, and more recently by Milne Ed- 

 wards and others. The weight which 

 should be given to evidence of this kind is 

 very difficult to determine, because of the 

 uncertainty which still obtains concerning 

 the real relationship of the birds in ques- 

 tion. The extinct types of wingless rails 

 which are found in New Zealand, the Chat- 

 ham Islands, the Mascarene Islands are be- 

 lieved by many to indicate land bridges, 

 while jEpyornis, of Madagascar, the Moas 

 of New Zealand, the Emeus of Australia, 

 and the gigantic Tertiary birds of the 

 Argentine Eepublic (Broutornis, Phoro- 

 rhacus, Opisthodactylus) , are supposed to 

 be branches of the same stock of Eatitse. 

 UntU, however, we learn a great deal more 

 than is known at present with regard to 



