CHEMISTRY OF INSECT COLOURS. 191 



of these phenomena to those that I have produced bj'^ means of 

 chemical reagents ; and since, in the latter case, I have shown 

 that the pigment is dissolved out, leaving a pure white wing, it 

 seems a fully justifiable conclusion that these natural albino or 

 "sun-bleached" varieties are likewise due to the absence — i.e., 

 to the non-development — of pigment. In a word, I should 

 regard them, as I have already stated, as unhealthy or patho- 

 logical cases. (As showing the relationship of yellow and chest- 

 nut, I may refer to the yellowish varieties of A . prunaria in the 

 Doubleday collection.) 



B. White (colourless) varieties of yellow species. Theoreti- 

 cally these should be as common as the corresponding varieties of 

 chestnut, but I have not been able to find any record of such 

 varieties. In the Doubleday collection is a specimen of A. plan- 

 taginis with the hind wings white instead of yellow, a specimen of 

 R.cratcegata "half washed out," of S. atomaria partly whitish, 

 and in the 'Entomologist' (xxiii. 382) is chronicled a variety of 

 V. io having yellow replaced by white. 



C. Yellow or white varieties of green species. I am 

 acquainted with no instances of such, although they doubtless 

 exist, unless we reckon the fading of some green Geometrse to 

 white as a case in point. 



D. Chestnut varieties of red species. These must neces- 

 sarily be scarce, since so few red species are descended from 

 chestnut. I have no examples to hand, but the want is partly 

 compensated by the colours of subspecies or allied species. In 

 Vanessa indica, a subspecies of atalanta, the scarlet is replaced by 

 orange or chestnut ; and in Anartia amalthea the same bands are 

 scarlet in one specimen, and chestnut in another;* while my 

 reagents convert the former into a chestnut identical with that 

 normally occurring. 



E. Yellow varieties of red species. This seems to be by far 

 the most common form of all coloric variation, a fact at which we 

 need feel no surprise, seeing how closely related are red and 

 yellow, and how very little constitutional disturbance is probably 

 requisite in order to upset the normal processes, and arrest the 

 pigmental development at the yellow stage ; in such cases also it 

 is clear that I must regard the variation as a pathological 

 symptom. It is almost superfluous to quote examples of this 

 type of variation, but the following are of interest : — Yellow 

 varieties of Z . filipendulce (Doubleday collection), A. caia with 

 red replaced by orangef (Doubleday); also bred varieties of caia 

 varying from buff to brick-red t (Entom. xxiii. 134), orange 



* I do not know whether or not this is a sexual difference. Entom. xxiv. p. 91. 



t In this specimen also the hind wings were more than normally, and the fore 

 wings nearly wholly, black. This may be considered an instance of general physio- 

 logical disturbance. 



\ Vide supra, p. 186. 



