10 THK KNTOMOLOGIST. 



which yellow is rapidly and satisfactorily affected (dissolved as it 

 happens) are but a small minority of the whole. It is incumbent 

 on me, therefore, to justify my assumption, and to endeavour an 

 explanation of the seeming discrepancy : and it appears to me 

 that one all-important clue may be furnished by the distinction, 

 which I emphasised in section E, between the two kinds of pigment 

 colour, viz., the soluble and the alterable. Yellow I regard as 

 universally a " solution " pigment, although — paradoxical as the 

 expression may appear — it cannot in the majority of cases be 

 dissolved. The evidence furnished by the behaviour of yellow 

 will be found — as it seems to me — to afford additional and 

 valuable confirmation to my views on this subject, as already 

 explained.* 



But here we are confronted by an illustration of the fact, so 

 often alluded to already, that it is impossible to go very far in an 

 experimental investigation without adopting some provisional 

 hypothesis. We cannot even lucidly or connectedly present the 

 actual facts obtained without stringing them on the thread of a 

 working hypothesis ; any more than we can orderly arrange a 

 row of beads without a thread to string them on. I will there- 

 fore preface our examination into the behaviour of individual 

 yellow species by a statement of the theory that I have been led 

 to form on the subject; merely warning my readers that this 

 theory is at present extremely provisional and hypothetical, and 

 — even so — cannot be rigidly or universally applied. 



I assume, then, that the yellow pigment, when first evolved, 

 is exceedingly sensitive and very susceptible of solution by various 

 reagents. We have examples of this stage in several of the Pieridse, 

 and, still more interestingly, in Melanargia galatea, where we can, 

 so to say, actually behold the pigment evolved — and dissolved. 



compared with grossulariata on the other, it will be noticed that whilst in the 

 former the under side of the wing is a " faded " facsimile of the colour of the upper 

 side, there is in the latter case no sign of yellow below. Seeing, then, that in 

 prossulariata the yellow is confined to one side, and on dissolving off leaves the 

 ground colour of pure white, it at first appeared to me that the difference might be 

 attributable to the fact that in one instance the colour was (if 1 may be allowed to 

 use a somewhat loose and figurative expression) simply "laid on " the upper side , 

 while in the others it — so to speak — " went through " the wing. Such an explana- 

 tion might, perhaps, be correlated with the statements of Hargen and Wallace 

 ('Tropical Nature,' p. 184) as to hypodervial and non-hypodermal colours. Sub- 

 sequent investigations having, however, given a different turn to my thoughts, I did 

 not follow up this idea (as perhaps I ought to have done), although it suggests 

 considerations that I shall be bound to take into account at a further stage in my 

 work. It still appears to me not impossible that at least a plausible case might be 

 made out for this hypothesis, although in the instance of Colias edusa one would find 

 a troublesome objection ; and since it is not quite clear how far such an hypothesis 

 could be harmonised with, or how far it would be antagonist to, the " development" 

 theory provisionally adopted in the text in explanation of these phenomena, it is 

 certainly incumbent on me to put both views before my readers, that they may 

 be enabled to form their own estimate as to the tenability or otherwise of my 

 explanations. 



* See p. 372, above. 



