CHEMISTRY OF INSECT COLOURS. 87 



remarks in part of the last subsection were intended especially 

 to enforce, — j'et in certain instances, as in cardamines, a remark- 

 able brilliancy of colour has been attained without any decrease 

 in solubility.* But it should be noted, as of especial interest, 

 that this cardamines colour is, as it were, developed in a collateral 

 line, and not in the direct line of ascent to the stable oranges and 

 yellows.f I merely note this now, since it will be necessary to 

 recur to the subject. 



Furthermore, it is necessary for me to retract my remarks on 

 brown, as being distinct from chestnut. It will be remembered 

 that I designated as chestnut all those brownish species that proved 

 amenable to my reagents, and separated off as a distinct class, 

 under the name of brown, all those that were unaffected. But I 

 have since concluded that this is hardly justifiable ; at any rate, it 

 is both illogical and inconsistent, so long as we recognise among 

 the analogous yellows three stages, of which one is perfectly 

 indifferent to all my reagents. Whatever mode of treatment we 

 adopt in dealing with yellow, we must adopt the same with 

 regard to chestnut; and, accordingly, I now propose to restore 

 the eliminated brown species to the class of chestnuts, recognising 

 in chestnut several different stages exactly as in yellow. And, in 

 fact, so closely similar is the behaviour of chestnut to that of 

 yellow, that a very large portion of the remarks made on yellow + 

 are almost equally applicable, mutatis mutandis, to this colour. I 

 shall, therefore, to avoid needless repetition, refer the reader 

 back to subsection F for the bearing of the phenomena of 

 chestnut on natural selection and coloric evolution. 



The main headings, to which I wish to direct attention in this 

 present subsection, are the following : — 



1. — That in chestnut, as in yellow, there is reason to suppose 

 that the pigment has been developed on an — usually white — 

 wing ; that this pigment was at first of a very soluble nature, but 

 in process of time becomes increasingly stable. That we may 



* An exactly parallel case is found in the Delias species, already referred to. 

 See also following footnote. 



t Since the preceding subsection (F) — on yellow and red — was written, I have 

 been a good deal exercised in my mind about these species of Euchloe. At one time 

 it appeared to me that I had been wrong in assuming that cardamines orange had 

 been developed via pale yellow at all, and that rather it was a distinct and 

 independent development altogether. This view would, of course, entirely evade 

 the necessity of supposing that a yellow had progressed into so brilliant an orange, 

 while still remaining so soluble ; and I was inclined further to support this by 

 arguing from the chestnut species, such as L. phlceas, &c. But further reflection 

 and experiment have resulted in bringing me back to my original views. For, on 

 examining the species E. eupheno (with its orange tips — identical with those of 

 cardamines — on a brimstone-yellow background), I feel almost convinced that the 

 orange here is a development from the yellow. Unfortunately, I have been unsuc- 

 cessful in my attempts to obtain specimens for experiment ; but, in the very 

 analogous Gonopteryx cleopatra, I find that the orange is rapidly converted into the 

 ground colour of yellow, and then the wlwle yellow dissolves, leaving a lohitish wing. 

 This is exactly analogous to Delias. 



I Antea, pp. 11, 12, i:^. 



