account of his experiments to the Entomological Society of London on 

 December 2nd last, he said his results were absolutely nil. Mr. 

 Poulton, who had similarly experimented on larvae of Gnophos obscu- 

 rata, reported equally negative results. This fully bears out my 

 supposition when I stated in my work on ' Melanism and Melanochroism 

 in British Lepidoptera,' p. 50, " At present I am unable to see any 

 connection between cause and effect in this direction." 



With regard to the coloration of lepidoptera, we now come to a most 

 interesting branch of our study, and, at the outset, we may classify 

 the colours under two heads. The colours produced simply by inter- 

 ference &c., of light waves on the surface of the wings = non-pigment 

 colours ; those produced by the light entering the scales and being 

 partly absorbed and partly reflected, the resultant colour being given 

 by the reflected portion = pigment colours. I have seen the unhappy 

 term " physical/' occasionally applied to the former class, as if both 

 classes be not in reality physical colours. It is probable that every 

 shade of colour exists in some insect or other, both as a pigment and 

 non-pigment colour, not even excluding white and black, and that 

 although in almost all cases the natural evolution of colour has been 

 from white, the result is frequently obscured by the colour being 

 developed in or on normally black scales, or in or on a scale which is 

 normally some other colour than white, and where retrogression in 

 colour appears to travel not towards white, but undoubtedly towards 

 black, and hence the essentially different results frequently met with, 

 when our premises appear to warrant us with expecting at any rate a 

 somewhat similar result. It is difficult, too, at times to differentiate 

 clearly between a pigmentary and non-pigmentary colour, and transi- 

 tions are very frequent in a state of nature, i.e., of essentially 

 non-pigmentary coloured species presenting traces of pigmentary 

 development, and it is not always clear whether this should be looked 

 upon as a progressive or retrogressive development. But I think, in 

 almost every instance, it is a progressive development, when normally 

 non-pigmentary species (or rather scales) develop pigmentary matter. 

 Pigment is now generally considered to be essentially a waste 

 product. The excess, or otherwise, of the food may, in this way, have 

 some direct action on the general coloration, especially in the shade &c., 

 developed. The kind of food, so far as relates to its chemical com- 

 position, I believe to have but little, if any, action whatever, and this 

 I fully explained some time since in ' Melanism and Melanochroism in 

 British Lepidoptera,' pp. 58-60. It seems impossible that this should 

 be so, because the vital processes of assimilation taking place in the 

 larva, change the vegetable into animal tissues and thus decompose 

 the food into its various constituents, elaborating that required into its 

 own tissues, and excreting the remainder as waste. When we come 

 to consider the quality and quantity it is different, and an excess of 

 food with a consequent excess of waste in the tissues may, and probably 

 does in most instances, become converted into pigment-producing 

 matter. An illustration of this kind has come under my observation. 

 Some years ago, I bred and interbred Selenia illustraria, and I must 

 own that I was very careless as to food (both quantity and quality). 

 The result was that the brood " ran out," became very small, and what 

 should have been the richly coloured spring brood, differed but little 

 in coloration from the less highly coloured (summer brood. I know of 



