126 THE ENTOMOLOGIST. 



being unduly biassed by previous conceptions. The phenomena 

 of colour variation are essentially of three distinct kinds : — 



1. Due to a change in the nature or arrest in the metabolism 

 of a pigment (chemical). 



2. Due to a change of structure (physical). 



3. Due to a change in the proportion of the normal colours. 

 The third of these is generally easy to distinguish ; the first 



and second are easily confused, especially when descrij^tions only 

 are accessible of the varieties under consideration. It will be 

 useful to review these three classes : — 



1. We have seen that red may be replaced by yellow, and more 

 rarely yellow by red. A white pigment may also be changed to 

 yellow, and, again, yellow is sometimes replaced by white (as in 

 Runiia, see Entom. xxi. 15 ; and in Heliodes arbuti albescens, 

 which has white in place of yellow on the secondaries, see 

 Newman, ' Brit. Moths'). A caustic alkali will change the white 

 of Lyccena to yellow, and a damp cyanide bottle will dye the 

 yellow of Colias and of some Hymenoptera a deep red : these are 

 chemical changes, presumably due to a change in the nature of 

 the pigment. But alkali will not turn a white Pleris yellow ; and, 

 in fact, according to Dr. Dimmock, there is very little pigment at 

 all in the white of a Pieris, — it is a sort of optical illusion, com- 

 parable to the white of finely-ground glass. So here at once is 

 introduced a difficulty, which seems to throw such variations as 

 Pieris rapce novanglice, P. napi Jiava (upper side entirely canary- 

 yellow, C. G. Barrett, E. M. M., 1888, 81), and Bryophila perla 

 Jlavescens into the third class instead of the first, and prove them 

 to be caused by an enormous preponderance of that yellow 

 pigment which is generally more or less visible about the typical 

 forms. Abraxas grossulariata lutea I had already recognised (p. 2) 

 as being of this nature. In the face of this evidence that changes 

 from yellow to white and white to yellow, though apparently 

 identical in their nature, really represent two radically different 

 phenomena, one feels afraid of theorising about the other varia- 

 tions. Dimorphism of pigment, or dichroism, most certainly does 

 exist, and is even frequent ; but it is perfectly plain that to decide 

 which varieties really belong to this class requires something 

 more than a superficial examination. It is by the action of 

 chemical reagents that we must test them. Pale forms, if due to 

 a non-metabolism of pigment, will come properly under the first 

 division ; but it is possible that the green of Venilia macularia 

 viridimaculata is really a deceiving combination of black and 

 yellow, like the green of Anthocharis. Will anyone who has an 

 example of this variety please examine it with a microscope, and 

 report ? Some suffused varieties, which one would naturally 

 class under the third division, must belong to the first, because a 

 damp cyanide bottle gives a suffused appearance to Danals 



