Vi INTRODUCTION. 



other instances of the same kind. Mr. Adkin, at a recent meeting of 

 the Sth. Lond. Ent. Soc., commenting on this said, that by careful 

 breeding and by giving a plentiful supply of food, he kept up the 

 seasonal dimorphic condition in this species, and obtained large and 

 richly coloured spring broods from his interbred specimens. Here the 

 varying results arrived at are clearly referable to the difference in 

 food supply, and, if we look on the better fed specimens as having an 

 excess of the waste product necessary for pigment formation, it is 

 certain that the ill-fed specimens had no such excess and the colour 

 suffered accordingly. Applying this to the normal condition of this 

 species, what do we find ? The imagines that emerge as the spring 

 brood are feeding from July to October (about three months), whilst 

 those that emerge as the summer brood are only feeding from May to 

 June (less than two months), and hence, there is a smaller surplus in 

 this brood, exhibited both by its almost uniformly diminished size as 

 well as paler colour. Of course, I do not forget that " natural 

 selection " &c., have also something to do with this. In the allied 

 S. illunaria under parallel conditions, similar variation in size is ac- 

 companied by similar retrogression in colour, the brown ground colour 

 and purple lines, being changed to ochreous and red respectively. I 

 am here dealing with the primary elements on which " natural selec- 

 tion," " meteorological influence " &c., have to work in order to bring 

 about the final result as exhibited in the seasonal dimorphism of these 

 species. 



Having now noticed what pigment really is, we may attempt to 

 trace the pigmentary colours in a genetic relationship, and our studies 

 in variation lead us through the following changes : (1) white, 

 yellow, orange, red, brown, black. (2) white, yellow, green, red (or 

 brown), purple (or blue), black. Probably green is sometimes modified 

 almost directly into blue. 



Of the non-pigmental colours, the simplest are white and black. 

 The former is apparently caused by the utter absence of anything but 

 air in the cells, and is wholly and essentially due to the surface re- 

 flection of the incident rays of light. This is the case in normal 

 Hepialus humuli $ , Euchloe (Anthocharis), Pieris &c., and also in male 

 Aporia cratcegi, but there is another form, or condition, even lower, or 

 more retrograde, than this. I refer to the purely transparent scale which 

 neither absorbs nor reflects light, but simply transmits it. It is com- 

 mon in some genera, and among our British species occurs in the 

 female of Aporia cratcegi. The male of this species has (as I have 

 before remarked) essentially typical Pieris scales, non-pigmental, but 

 clear white ; in the female this is not the case, the scales (in the centre 

 of the wing especially;, becoming practically transparent, the membra- 

 nous structure only preventing the wing becoming absolutely so. 

 Here then we appear to get the simplest form of scale, on which the 

 typical white of Pieris &c., is a decided advance. So much for white 

 as a non-pigmental colour. We may now consider white as a pigment 

 colour. 



Perhaps the best example of white as a pigment colour is seen in 

 Mclanargia galatea. The pigment here is very unstable, so much so, 

 that the mere presence of the fumes of strong liquid ammonia is suffi- 

 cient at once to break up its original condition and produce a brilliant 

 yellow, although, if exposed to air a short time, the pigment reverts to 



