CONVERGENCE IN EVOLUTION 257 



by themselves apart from other mammals. Now, 

 it seems at first sight very remarkable that the 

 marsupials should, in the course of time, have 

 come to present what may be called superficial 

 .duplicates of several orders of the higher mammals. 

 The Tasmanian wolf converges towards a true wolf, 

 the banded ant-eater to a true ant-eater, the flying 

 phalanger to a flying squirrel, the swimming yapock 

 to an otter, the bandicoots to rats, the marsupial 

 mole to a true mole, and so on. The parallelism is 

 very interesting, for marsupials are not on the same 

 line as placental mammals; yet one is perhaps 

 inclined to make too much of it. It must be remem- 

 bered that the different kinds of habitat and the 

 different ways of getting a livelihood that are open 

 to mammals are not very numerous, and that Nature 

 was therefore almost bound to repeat herself. In 

 the same way it is not surprising that there should 

 have been terrestrial, aquatic, and aerial reptiles 

 in Mesozoic ages, just as there are terrestrial, 

 aquatic, and aerial mammals to-day. Some good 

 cases of convergence occur between rodents and 

 insectivores, e.g., between mouse and shrew, porcu- 

 pine and hedgehog, squirrel and tree-shrew; but 

 we have given illustrations enough. The climax 

 is to be found in the " mimicry "of unrelated types, 

 but this problem is better kept apart, since the 

 superficial resemblance in itself is here of survival- 

 value and may be the direct result of natural 

 selection. 



In his Creative Evolution Professor Bergson dealt 



