NATURAL SELECTION 281 



was better adapted than its predecessor, and the theory de- 

 mands a far more detailed correspondence between pattern 

 and normal habitat than we can usually perceive. On our 

 present knowledge we assume less if we suppose that the 

 greater part of specific divergence in colour has been due to 

 other processes, while in some cases selection has merely 

 checked the development of bright colours and maintained a 

 general brown, grey, or mottled ground colour. 



We will now consider some of the examples in which the 

 correspondence between colour and environment is more de- 

 tailed than in many of the examples described on pp. 236-42. 

 In nearly every instance the variations corresponding with a 

 differently coloured background are intraspecific (see p. 233). 

 We may first mention the power of colour-change in many 

 lepidopterous larvae and pupae (Poulton, 1892 ; Bateson, 1892 ; 

 cf. also Chapter II, p. 37), which enables them to har- 

 monise with their general background. This harmony is 

 acquired gradually during the life-history, and appears to 

 be due to a direct effect on the nervous system of the insect 

 through the eyes. A possibly similar state of affairs is seen 

 in the beetle Cleonus sulcirostris (Merryfield and Poulton, 1899) 

 and the adult moth Gnophos obscurata (Poulton, 1892), both of 

 which have marked local colour-variation corresponding to 

 changes in the nature of the soil. Such cases could be 

 multiplied, and Poulton (1926) has recently dealt rather fully 

 with the phenomenon in grasshoppers. All collectors of these 

 insects are aware of the general agreement between the colour 

 of a species and the background, so that we have green forms 

 on grass, green and brown forms on heather, sandy forms, black 

 forms, etc. Poulton deals with a large series of black or black 

 and pale streaked species occurring on areas of burnt grass 

 in Africa. 



It is not unlikely that the permanent colour-harmony 

 established in many inhabitants of deserts may be of the same 

 nature. What was once a power of response to the various 

 backgrounds on which the species had to live has now 

 become fixed, giving an unvarying and close correspondence 

 with the colour of what has become the permanent habitat. 



On the selectionist hypothesis it is supposed that it is the 

 power of responding to the colour of the background that 

 has been built up by selection, since the actual changes in 



