98 BRITISH BUTTERFLIES. 



of retrogressive metamorphosis," or'change to original type, in pigments, 

 but, although there is nothing improbable in the conclusion, the data 

 offered do not at all afford sufficient proof that the underside of the wing 

 was originally brown, or that the change has any genetic value whatever. 

 To us, the probability appears to be in favour of considering the change 

 a chemical one, caused by the direct action of the water on the pig- 

 ment, although, as already noted, it may not be due at all to the 

 presence of a "pigment," or rather "pigment-factor," but to the 

 scales holding the vapour externally, and thus altering the reflective 

 power, and so influencing the ordinary diffraction. On the other hand, if 

 it be a pigment colour, the rapid change brought about by the 

 absorption of water-vapour is very remarkable. In 1892, Freer, 

 Riding, and others, discussed the matter at considerable length (Ent. 

 Rec, vi., pp. 35-40, 83-86, 107-111, 138-140, 204-208, 255-256), and 

 Riding notes (p. 86) : " The scales on the underside of C. rubi are 

 not arranged in superficial and deep layers, but overlap like the tiles 

 of a roof, and the lower portion of each scale contains less pigment 

 than the upper, whilst the green colour is confined to a little more than 

 the upper half of each scale, roughly speaking. An examination 

 of the scales shows the green to be much more brilliant than one 

 would anticipate. By reflected light, each green scale looks as though 

 dusted over with minute, brilliant, green scales, for a little more than 

 its upper half, and somewhat further down on the sides, whilst below 

 it is pale brownish. When wetted, the green becomes a complementary 

 brilliant red by reflected light. When the same scale is examined by 

 transmitted light, the green vanishes, and there is a pale brownish 

 scale, redder and yellower, where the green had been. Under a|-inch 

 power, the scale is seen to be finely striated vertically, and, behind the 

 striations, the coloured portion is divided into masses by irregular 

 lines, which, when wetted, seem to swell and become much more 

 distinct, giving a reticulated appearance, and the colour is somewhat 

 paler. None of these reticulations are observed in the brown scales, 

 so that there seems to be some connection between these and the green 

 colour, especially as they are actually seen to change when wetted, and, at 

 the same time, the green changes to its complementary red. We have, 

 therefore — (1) Brown scales, pigmented, slightly paler when wetted. 

 (2) Striated green scales, with a redder and yellower pigment (on the 

 green parts), apparently collected in masses, with intervening spaces, 

 through which water can pass, giving a brilliant green reflection when 

 dry, a brilliant red when wetted." So far, therefore, it seems connected 

 with pigment and structure. The only explanation that offers itself to us 

 is that the green colour of C. rubi appears to be a compound one, pro- 

 duced, in part, by diffraction, the result of the striations noted, and partly 

 by the presence of a pigmentary material in the scale. These striations 

 and facets, when wetted, may conceivably lose largely their diffractive 

 and reflective powers, and the result is that the brown or red, or brown 

 and red, pigments then give the colour to the scale, until the wetted 

 surface becomes dry and capable of its normal action again. This 

 view is supported by Prideaux (op. cit., p. 140), who states that, "by 

 delicate focussing, the curious tortoiseshell-like patches observable in 

 the scales of C. rubi, are within the substance of the scale and beneath 

 the ribs ; the patches certainly seeming to correspond with the broken 

 up coruscations of green colour which one sees by reflected light." 

 Burrows observes (ojj. cit., p. 139) that a compound of yttrium was 



