CHEMISTRY OF INSECT COLOURS. 189 



day collection) having the ground colour entirely yellowish instead 

 of white ; in this species the pigment is usually present only as a 

 narrow yellow hand : varieties of E. cardamines with more than 

 the normal extent of orange (Entom. xxiii. 228) : a variety of 

 A. caia having white replaced by pink. It must be especially 

 observed that these latter variations are all of them progressive in 

 character — advance guards, so to say, of the march towards higher 

 coloric development, and of what we may expect the normal 

 specific type to become in the course of time, if the struggle for 

 existence be sufficiently keen to make such variation advan- 

 tageous. 



The third type of variation in physical colour is when one 

 physical colour is replaced by another — that is to say, when the 

 microscopic structure varies from the normal. If the blue of 

 Lycsenidse be physical, then we have examples of this type of 

 variation in the slate-coloured or French grey varieties of P. 

 alexis, that have from time to time been chronicled in the ' Ento- 

 mologist ' by Mr. Sabine and others. I would draw especial 

 notice to the fact that exactly similar* artificial varieties have 

 been obtained in my experiments, from which we may conclude 

 that in both cases the effect is due to injury of the fine structure 

 of the scales, f If we consider that the females of the Lycwnce 

 preserve the original coloric type, from which the males have 

 advanced to their blue adornment, then we must consider such 

 slate-grey varieties as retrogressive. If, after all, the blue of 

 Lycsenidse should prove to be pigmental, then these varieties 

 would acquire far greater interest, and would be ranked as retro- 

 gressive varieties of pigment colours, just as those varieties which 

 must now be considered. 



As to other varieties of physical colours, I am unfortunately 

 entirely in the dark ; one would expect to find varieties of 

 Thecla ruhi without the green, in fact, of all the physically green 

 species with this green replaced by bronze-brown, &c. Whether 

 or not such varieties have been found I do not know, but I shall 

 be very glad indeed of any information on the subject. 



The naturally occurring varieties of pigmental colours are of 

 peculiar interest to me, since — as I think — they form an in- 

 dependent and corroborative chain of evidence in support of the 

 inferences drawn from my experiments ; in fact, we have observa- 

 tion and experiment mutually confirmatory. I need scarcely 

 remind my readers of the broad conclusions as to coloric develop- 

 ment summarised last month, viz., that red is evolved from yellow, 

 or occasionally from chestnut ; that green is probably evolved 

 •from yellow ; and that yellow and chestnut are closely related and 

 both developed not from any white pigment, but in a formerly 

 unpigmented and usually white wing. Now, if these inferences 



* I have not seen these varieties, but judge merely from the descriptions. 

 t Cf. statements in Entom. xxiv. p. 116. 



