GEOGRAPHICAL DISTRIBUTION 113 



is a remnant of the arctic flora which was spread over New 

 England during the later glacial period, and that, as the ice 

 melted and the arctic flora retreated northward, some species 

 persisted in more southern latitudes by ascending the moun- 

 tains, the cold of whose higher altitudes resembles the arctic 

 climate to which these species are adapted. 



Certain cases of distribution which at first glance seem to 



O 



be anomalous are found on careful scrutiny to support our 

 hypothesis. For example, the opossums of North and South 

 America are very different from all the other mammals of the 

 same region, so different as to be properly placed in a distinct 

 subclass, the Marsupialia. In no other region are similar 

 animals found except in Australia and its adjacent islands. In 

 Australasia, however, there are, with two exceptions, no indig- 

 enous mammals except those belonging to the same subclass 

 as the opossum. It seems at first sight absurd to postulate any 

 communication between Australasia and America by which 



^ 



one may have become peopled from the other. It looks as if 

 the opossum type must have arisen independently in the two 

 areas, a thing which would be contrary to our knowledge of 

 the ways of evolution. Paleontology here comes to our aid. 

 The fossil fauna of America is rich in species of the opossum 

 type, the opossums being the only living representatives of an 

 at one time very extensive marsupial fauna. The marsupial 

 type is more primitive than that of the other Mammalia. 

 There is evidence that at one time, before the higher I\Iamma- 

 lia came into existence, the marsupials were spread over the 

 whole eastern and western hemispheres, and that as the higher 

 mammals arose they exterminated the mammals of the more 

 primitive marsupial type, except that in Australia the earlier 

 forms persisted and in America the opossums remained. 



