136 tHE ENTOMOLOGist. 



The most iinportant of these concerns the remarkable behaviour 

 of certain yellow species when submitted to the action of wet 

 potassium cyanide. An account of experiments on this subject — 

 originally written in November, for a footnote to subsection F, 

 but subsequently held back from considerations of space — will be 

 found as an appendix to this section. Again, on looking back, I 

 find that I had adduced* the reactions of T. ruhi and of the blue 

 Lycsenidse in support of mj^ contention that all the changes 

 eifected by my reagents were retrogressive. After the conclusions 

 set forth in the preceding subsections (H and I), however, it is 

 clear that neither of these reactions can properly be quoted as 

 evidence for retrogressive change in pigment colours. For the 

 defence of the position taken up on this subject of retrogressive 

 changes, I can very safely rely on the argument from red species, 

 as stated in subsection E. 



Furthermore, I see that in a footnote t to subsection F, as 

 well as in the text, I had referred to the behaviour of pink 

 species, and adverted to the possibility that these were changed 

 not to yellow but to white. This remark now seems to me very 

 ill advised and misleading, especially in view of the explanation 

 that I have been at some trouble to set forth in (F) previous 

 pages concerning the relations of red and yellow. In subsequent 

 examination of other pink species, I have found the yellow 

 so pale and feeble (most unlike the bold orange-yellow of, e. g., 

 Zygcena and Arctia) that it might be easily overlooked, and 

 carelessly mistaken for white. It seems, too, only reasonable 

 that a pale delicate pink should yield a far paler yellow than does 

 a prominent crimson or scarlet.l Since, moreover, these pinks 

 agree entirely with the other reds in showing the " reversible " 

 effect, it is certain that their constitution is the same. My 

 hypothetical supposition of a subclass for pinks is, therefore, — 

 as I now think on mature consideration, — entirely erroneous. 

 Indeed, I feel such thorough confidence in the argument from 

 the behaviour of red in general, that if I found a pink species 

 which to a positive certainty changed to white, but showed also 

 the " reversible " effect, I should feel bound to hold that there 

 was yet a yellow pigment there, though so slight and pale as to 

 be imperceptible.§ And, furthermore, if in any case an indubitable 

 white were produced, and no reversible effect were obtainable, we 

 should then have simply an exactly analogous case to that of 

 Delias, in which the red first changes to yellow, and then this is 

 dissolved, leaving a white wing. I have, perhaps, spoken at 



* Entom. xxiii. p. 371. t Ante, p. 15. 



X The amount of pigment present in some of these pink species seems to me 

 to be exceedingly small, as might be anticipated from the pale and delicate colour. 

 There is also special difficulty in observing the exact tint produced, since in many 

 cases the background (e. g,, in Sphinges) is very troublesome. 



§ It must be remembered that " yellow " is used in a broad sense. The essential 

 difference is between the pigment yellow (however ^ale) and the non-pigment white* 



