CHEMISTRY OF INSECT COLOURS. 13 



become partly transformed into red, while yet in the primary 



unstable stage.* It would be difficult to exaggerate the interest 

 and importance which these two species possess for me, since 

 their behaviour both furnished the last and important link in the 

 chain of evidence as to the genetic relations of red, yellow, and 

 white ; and also showed that the difference between soluble and 

 insoluble yellows was one of degree only, and not of kind. 



But after emphasising so strongly the interest attaching to 

 these species, I must not quit the subject without a word of 

 description of them and their behaviour. They are white species 

 of Pieridae ; on the under side of the hind wings is a broad patch of 

 yellow, and adjoining this a red blotch. So that here we have co- 

 existing the three stages in coloric evolution. Great was the inte- 

 rest with which I experimented. On adding the usual reagents 

 the red was instantly changed to yellow, — that of course ; but the 

 question was, what would happen next ? I had no need to wait 

 long for an answer ; in a very short time the whole of the yellow 

 was dissolved, leaving a pure white iving. So at last I had 

 succeeded in finding an instance in which red was changed to 

 yellow, and then that yellow dissolved. 



This instance of Delias shows, therefore, that although red is 

 usually preceded by a stable yellow, this is not indispensable ; 

 red may be evolved direct from an unstable yellow. It would be 

 exceedingly interesting to follow up this subject into its biological 

 bearings did not the urgent limits of space forbid ; I must, there- 

 fore, defer further speculations on this matter to a possible future 

 period. Three points still remain to be indicated ; firstly, that, 

 throughout, the flexible biological explanation, demanding nothing 

 more than the known fact of a yellow capable of progressive 

 change, and the action of Natural Selection, which may develop 

 one yellow into a brilliant red and retain another and precisely 

 similar yellow in its initial condition, harmonises far better with 

 the complicated facts than does a rigid and more purely chemicalt 

 explanation involving "necessary" developments and '"inherent 

 tendencies": secondly, that regarding the distinction between 

 " soluble " and non-soluble but " alterable " pigments as marking 

 a real difference in constitution, we can understand how, if 

 yellow belong (as it does) to the soluble class, it should be in- 

 capable of alteration by reagents even when by a slight modi- 

 fication in constitution it has become insoluble ; this lessens the 

 difficulty seemingly presented by the unaffected yellow species : 

 thirdly, we have here, in the existence of one or more red species 

 in an}' genus, the criterion above referred to of the phlygenetic 

 age of all the yellow species of that genus ; but the criterion, as 



* These are the exceptional instances to which I alluded in the footnote on 

 p. 31' I. 



t This may seem somewhat a misnomer, but I wish to distinguish between the 

 explanation that involves a reference to the totality of the conditions of life, and 

 that which involves the chemical processes of life and nothing more. 



