146 THE THEORY OF EVOLUTION 



istics. Such islands as Bermuda and Madeira are 

 annually visited by birds from the adjoining con- 

 tinents and they have hardly any peculiar forms, 

 while St. Helena and the Hawaiian group have 

 almost all their birds and reptiles peculiar to them- 

 selves and so far changed, that it is difficult to make 

 out their geographical relationships. If species are 

 immutable, why should this factor of inaccessibility 

 have any influence upon the peculiarity of the species 

 found in a given island? Another curious series of 

 facts is displayed by the family of birds known as 

 the Rails, which are distributed all over the world 

 except the polar regions. The family includes fifty- 

 five genera, twenty-five of which inhabit islands 

 and of these ten have lost the power of flight. A 

 genus, with only one or two species, will be confined 

 to a single island, while the genera of the continents 

 have more numerous species and all of them can fly. 

 It is an obvious inference that the loss of flight and 

 the separation as distinct genus and species must 

 have taken place after the original progenitors had 

 reached a given oceanic island, for the birds could 

 not have crossed wide seas by swimming, nor could 

 the flightless ones have arrived in any other way. 



"If we briefly resume these surveys of island 

 faunas, the result for our problem is as follows: 

 The islands must have received their animals in 

 some fashion from the mainland, the continental 

 islands when they were a part of it, the oceanic not 

 till after their appearance above the surface of the 



