53Q EVOLUTION OF BIRDS xvm. 9-10 



character of many of these differences certainly constitutes a case for 

 considering that the others are also of this nature. In very small 

 populations ( < 1 ,000) unfavourable characteristics may become estab- 

 lished by chance ('genetic drift', Wright, 1940), and this factor may 

 be responsible for some of the island races. 



It is especially interesting that in the Galapagos animals special 

 reasons have made it possible for variety to arise. If, as seems likely, 

 the earliest Geospizas were among the first birds to arrive in the 

 archipelago, they found there neither competitors nor enemies. Be- 

 cause of the distance from the mainland this condition has remained 

 with little change ever since. Even today the six passerines not be- 

 longing to the Geospizinae do not provide serious rivals, and the 

 local owls and hawks are apparently not serious predators on the 

 finches. 



The other great factor that has led to differentiation is the splitting 

 up of the area into a number of isolated units. This has allowed slightly 

 different races to emerge, and we may suppose that if these again 

 came into contact with each other they would find slightly different 

 optimal conditions, and therefore, with partial or complete inter- 

 sterility, would continue as distinct species. This has almost certainly 

 happened with the Canary chaffinch (p. 523), and in Galapagos there 

 is a similar case in that two species of the large insectivorous tree- 

 finch Camarhynchus occur together on Charles Island. Both are 

 derived from a single population, offshoots from which have inde- 

 pendently colonized the island (Fig. 315). 



10. Birds on other oceanic islands 



While the case of Darwin's finches is very striking it is important to 

 recognize that similar radiation from a few species is not found on all 

 oceanic islands. For instance, there are numerous birds on the Azores 

 and they differ little from those of Europe. Whereas migrants are 

 frequent in the Azores they are rare in the Galapagos; evidently there 

 are special factors producing the isolation of the latter. 



A radiation similar to that of the Galapagos has, however, occurred 

 in Hawaii, where there are only five passerine forms and one of these, 

 a finch-like bird, the sicklebill, has produced an even greater variety 

 than Darwin's finches (Fig. 316). Some of these sicklebills feed on 

 insects, others on nectar, fruit, or seeds, and the beaks have developed 

 accordingly. One species climbs like a woodpecker, digs for beetles 

 with a short lower mandible, and probes them out with a long curved 

 upper one. 



