ISOLATION 285 



terflies belonging to two different families present 

 a greenish-blue colouration which has not been ob- 

 served in any other butterflies, and yet, neither of 

 those groups possesses any means of self-protection. 



The same can be said of the differences in colour- 

 ation between island and mainland butterflies. Is- 

 land species are generally less pigmented and some- 

 times larger sized. In the Andaman archipelago, 

 lighter hues prevail not only in butterflies, but in birds 

 as well. Certain colours seem to be localised in cer- 

 tain regions. For instance, there are no red parrots 

 anywhere except in the Molucca Islands and in New 

 Guinea. Such specific differences are incontestably 

 due to geographical distribution. Wallace remarks, 

 however, that they are not due merely to the bare fact 

 of isolation but to new conditions which modify the 

 mode of action of natural selection. 



This question was given much attention by two 

 scientists who reached their conclusions simultane- 

 ously though independently — Romanes, 2 a famous 

 Darwinian, one of Darwin's direct disciples, and 

 Gulick, 3 a missionary who for fifteen years observed 

 the land and fresh water molluscs of the Sandwich 

 Islands. While Gulick devoted himself exclusively 



2 J. T. Romanes. Physiological Selections (1885) and Darwin and 

 After Darwin (London, 1892-1897). 



3 J. T. Gulick. Divergent Evolution through Cumulative Segregation 

 (J. Linn, Soc, XX, 189-274), and Intensive Segregation. (Ibid., XXIII, 

 pp. 312-380.) 



