xvm. 9 ISLAND BIRDS 529 



between subspecies. The species, of course, do not cross in nature, and 



such as have been tested in captivity mate only with their own species. 



It is still not possible to give an entirely clear idea of the factors 



QiUngdor, \ MBEU 





Scale in land miles 



\ _ ) 10 20 30 40 50 

 \ Bind/oeQ J ff ' ' ' ' "* 



\ 

 \ 

 \ 



\ / James 



V 

 \ 



\PSITTACULA , 

 o^ \(SENS STRICT) 



\> \ 7 Indefatigable 



DunoanK 5 \ 



^ — ^ Barrinqton 



/ / / 



1 1 PSITTACULA 

 'SrLh* ( SENS - STRICT) 

 (O) AND 



CHARLES ^-^ PAUPER C5 



.pauper 



Caff in is 



C. psittacuia 



C. habeli 



Fig. 315. Development of distinct forms from a single one in the Galapagos 



finch Camarliynchus. Charles Island has two forms derived independently from 



the central form C. affinis. (From Lack, partly after Svvarth.) 



responsible for the development of varied animals, even in this much 

 simplified case. There is no clear line between subspecific differences, 

 say of the beak that are not obviously adaptive and those with clearly 

 adaptive character. Since favourable mutations will spread through a 

 population the easiest assumption is that all these differences are in 

 fact adaptive, or linked with unseen adaptive characters. The easiest 

 assumption is not necessarily correct, but the demonstrably adaptive 



