148 , COLOUR IN NATURE chap. 



thought to be of importance in colour-production. 

 The two membranes are connected by bridges of 

 chitin, and pigment granules may be deposited both 

 in these bridges and in the outer membrane. 



Having regard to the small size of the scales, it 

 will be readily understood that there is frequently 

 great difficulty in determining whether a particular 

 colour is due to pigment or structure ; the fact that 

 a scale showing structural colour frequently contains 

 pigment in addition further complicates the matter. 

 We thus find that there is much difference of opinion 

 among observers as to the cause of particular colours. 

 This occurs especially in the case of dichroic scales, 

 those displaying one colour by transmitted light and 

 the complementary colour by reflected light. Some 

 regard this effect as purely structural, others main- 

 tain that the colours are due to dichroic pigments. 

 Making due allowance for difficulties of observation, 

 it seems, however, certain that blue in butterflies is 

 always a structural colour, while green, black, and 

 white are at least usually due to structure. Other 

 of the structural colours are readily recognised by 

 their metallic brilliancy and changing glow, which 

 give to the butterflies possessing them an appear- 

 ance of surpassing beauty. Many will recall Mr. 

 "Wallace's description of how his heart beat fast 

 and his brain reeled when his perseverance was 

 rewarded, and he captured with his own hands 

 one of the finest of these living gems in the Malay 

 Archipelago. 



In connection with these colours Urech (1893) 

 notices one little point of some interest. He found 

 that in some cases the scales display under the 



