ARGUMENTS FOR EVOLUTION 141 



complex) throughout the ages. On the contrary, after 

 an exuberant development in certain lines, it has often 

 happened that some relatively primitive and insig- 

 nificant type has given rise to the group destined to be 

 dominant long after. This is so true, that we are accus- 

 tomed to think of highly specialized types as ends of the 

 branches of the tree of life, giving rise to nothing be- 

 yond. For example, though amphibians were derived 

 from fishes, it was not from the highest fishes, such as 

 the perch or sole, which have gone far beyond the point 

 where it would be possible for them to develop any 

 amphibian features. The paleontologist, convinced of 

 the truth of evolution, is greedy for every fragment of 

 evidence he can glean from the past. It is as though 

 some great book had been broken up, and the leaves 

 scattered far and wide. He knows that many of the 

 leaves must have been destroyed, others are lost and will 

 never be found ; but every page, every line, which he 

 can recover conveys part of the message of the book. 



7. The study of geographical distribution is also very Distribution 

 suggestive. If evolution has taken place, members of a jJjf g c 4 e n ~ 

 group having a relatively recent common ancestor might idea of 



. evolution 



be expected to occupy the same continent or hemisphere. 

 This is what we find in a number of cases ; for example, 

 the humming birds, with hundreds of species, are all 

 American. There seems to be no climatic or other 

 reason why humming birds should not flourish in the 

 Old World tropics, but they have never been able to get 

 there. A series of islands forming an archipelago will 

 often have a series of birds, mice, reptiles, or snails, 

 each island with its particular sorts. As Darwin noted 

 in the case of the Galapagos Islands, the nearness of the 

 islands, and the shallowness of the sea between them, 

 correspond in a marked degree with the degree of re- 



