THE SUPPOSED TERTIARY ANTARCTIC CONTINENT. 239 



of animals in existing northern continental rather than in a hypothetical 

 southern continental land, it can hardly be refused to extend it to plants as well. 



Unfortunately the fossil remains of flowering plants can rarely be interpreted 

 with any certainty. The appeal to paleontological evidence which is in cases 

 such as these which have been quoted sufficiently conclusive, is in respect to 

 plants almost wholly negative. Heer and Ettingshausen thought they had 

 obtained conclusive proof of the existence of Proteacece in eocene and miocene 

 formations in Europe. But Bentham, after a prolonged study of the existing 

 representatives of the order and a careful consideration of the supposed fossil 

 evidence, was forced to the conclusion that it had broken down. 



If, however, we go back to an earlier period, the Jurassic, and turn to Coniferce, 

 the structures of which lend themselves better to recognition in the fossil state, 

 we find in Araucaria, & genus now represented by a few species in both divisions 

 of the southern hemisphere, abundant evidence that it was once widely dispersed 

 in the northern. Going back earlier still, the strong case which has been made 

 out for Gondwana land is perceptibly weakened by the discovery of remains of 

 the Glossopteris flora in Russia. An economy of hypothesis is better served by 

 assuming a northern origin and a dispersal southward than by calling into exist- 

 ence a vast territory from the Indian Ocean. 



A more serious difficulty is to account for the presence in the widely dissevered 

 tropics of the Old and New World of identical groups of a higher order although 

 differing in their present representatives, and still more for such an extreme case 

 as the two species of Ravenala. This it must be confessed would be inexplicable 

 on Dana's view were it not for the apparently inevitable evidence for the exist- 

 ence in miocene and probably earlier times of a subtropical climate and vege- 

 tation in the Arctic regions. Purely tropical types would have under such 

 conditions a more northern range and their interchange between the two Worlds 

 would not be impracticable. 



The classical researches of Hooker first recognised the "continuous current 

 of vegetation" southward at the present time. He indicated three types each 

 with its own character: the Scandinavian which reaches Tasmania; the Persian, 

 Australia; and the Himalayo-Malayan, Polynesia. 



Half a century has nearly elapsed since Dana laid down his memorable 

 principle. Time has strengthened and in no way diminished its force. But 

 though adopted by Darwin and by Wallace it is still ignored by those who prefer 

 facile speculation to the sober contemplation of established facts. 



