CHEMISTEY OF INSECT COLOURS. 117 



In column 2 we have an exceedingly interesting and equally 

 puzzling group — one, too, that has caused me not a little mental 

 exercitation. Originally, I had split this into two groups, the 

 second of which possessed one solitary representative, T. rubi ; 

 but further reflection and experiment have led me to class 

 T. rubi with the other species, as in the present arrangement. 

 The difficulty with this group is to obtain some decided evidence 

 that shall justify us in stating certainly that these greens are or 

 are not "physical" colours. The facts that I have found no 

 solution of pigment (indeed, the change from green to brown is 

 so instantaneous as to well nigh preclude any supposition of 

 solution taking place), and have no evidence of a " reversible " 

 effect comparable with that in the case of red, both tell against 

 the existence of a green pigment in these instances. But this 

 evidence is not conclusive. Considerably more conclusive is the 

 fact that in Limenitis procris 1 have been able to effect the 

 change from green to brown (or rather, in this case, a sort of 

 dun-grey) by simple water ; and this change was permanent. In 

 T, rubi, too, it is possible to abolish the green colour by water ; but 

 the colour rapidly returns. In Argynnis I have been less successful 

 in inducing this change by water only ; but in both these two last 

 species I have effected it by alcohol. In the case of rubi the green 

 returned very quickly, as soon as the wing was dried in fact. In 

 Argynnis the return was less rapid. I must point out that the 

 argument from the action of alcohol must be applied with con- 

 siderable reservation. 



Summing up the case of group 2, it seems to me that we may 

 lay down the following conclusions : — That the green of Parthenos 

 gainbrisius is physical; this is strongly supported by the un- 

 doubtedly physical character of the green on the upper surface of 

 that species. The similar green of Limenitis procris* is also 

 physical ; and the green of Ino statices and globularia is, in all 

 probability, of the same nature. As to T. rubi and the Argynnis 

 species, the evidence is strongly in favour of the former being of 

 a physical green ; while as to the latter, the evidence, so far as it 

 goes, favours the same supposition, but is less strong perhaps.t 



Turning now to group 3, — the most interesting of all, since 

 here we have clearly pigment colours in evidence, — we shall find 

 a brief notice, in addition to the facts published already in the 

 tables, I sufficient. Taking, first, the English species, which in all 

 cases, except the ambiguous one of Cidaria miata and psittacata, 

 dissolve, leaving a white wing, the case is evidently exactly coni- 



* This derives additional support from the fact that in L. sibylla we have this 

 colour represented by a physical pale blue. See next subsection. 



t I shall be greatly obliged to any correspondents who will inform me, either by 

 letter or through the images of the ' Entomologist,' of any species marked with 

 apparently the same green as rubi or ai-fji/iinis ; and also of any varieties (if any 

 such be known) of these species — varieties, I mean, so far as concerns the green. 



I Entom. xxiii. 2-52. 



