CHEMISTRY OF INSECT COLOURS. 11 



After a time (I know not whether years or centuries or yet 

 longer periods) the yellow pigment becomes slightly altei'ed in 

 constitution* (probably accompanied by a greater or less change 

 in tint), and altered in the direction of greater stability ; or rather, 

 confining ourselves to what we know, altered to the extent of 

 becoming far less easily attacked by solvents. Of this stage we 

 have examples in rhamni and various other species, as tabulated 

 in column 2. Concerning this secondary change, however, I 

 make no assumption as to whether it is an inevitable sequel of 

 the chemical changes which first led to the development of the 

 yellow pigment, or whether it is an additional and independent 

 development, brought about — as we may suppose — by the agency 

 of Natural Selection ; since, whatever biological ends an attrac- 

 tivef colour, such as yellow, may serve, we must assume that a 

 distinct advantage would accrue to the species by such colour 

 being rendered more stable and — possibly its usual accompani- 

 ment — more brilliant.! A question of this sort cannot be settled 

 a priori, but only by distinct entomological evidence of facts : of 

 facts that is as to the comparative ages of various yellow species ; 

 and in saying this I am not oblivious that such facts can only be 

 of the nature of inferences from comparative entomology and the 

 study of phylogeny. (But, as I shall show immediately, we have 

 already one criterion — though a fallible one — of comparative age ; 

 a criterion, however, that must be used with delicacy and discrimi- 

 nation.) If we found that the most recently developed yellow 

 species are uniformly possessed of a readily soluble pigment, 

 whilst the oldest (I speak always in a phylogenetic sense, of 

 course, and not in an autogenetic) are very difficultly aff"ected, we 

 should then have good reason to assume that a yellow pigment is 

 in its very nature always destined to undergo such further modi- 

 fications in course of time (or at the very least that — being 

 susceptible of such change — it is certain to be so affected through 

 the agency of Natural Selection). In this light one might regard 

 the behaviour of E. cardamines, since the pigment here is readily 

 soluble, and may perhaps be regarded as of (comparatively 

 speaking) very recent development.§ But, on the other hand, 

 there would, I apprehend, be much difficulty in the case of the 

 Colias species, which I fancy to be very old established. The 

 subject, however, is a very wide one, and it is out of the question 

 to attempt dealing with it here; but, in quitting it, I will add that 

 I shall be greatly obliged by any direct information bearing on 

 this point, which appears to me both of interest and importance. 



* This change might be, for instance, in the direction of higher oxidation. 



t It must not be overlooked, however, that yellow pigments may in many cases 

 be protective by their resemblance to flowers. Any further evolution towards 

 stability, if entailing also modification of tint, might therefore, in a given species, 

 be fatal, and so checked by Natural Selection. 



J But E. cardamines must not be forgotten, where the colour, though very 

 soluble, is very brilliant. 



\ Scfj Darwin, ' Descent of Man,' chap. xi. 



