CflEMISTRY OF iNSECT COLOtJES. 1^*? 



indirect influences of the surrounding conditions, which may 

 unfavourably affect the constitution and metabolic processes of 

 the insects. It must be remembered that — as I pointed out 

 some time since — all such products as pigments are almost 

 certainly produced by the decomposition of the protoplasm ; the 

 various food constituents are first built up into protoplasm, and 

 then this decomposes, with the result of producing inter alia 

 pigments.* We cannot safely assume any direct synthesis of 

 such pigments direct from the food, and it does therefore appear 

 to me a somewhat crude, and, so to say, almost too simply 

 mechanical a suggestion, that the presence of some unusual 

 chemical constituentf in the food may directly cause a different 

 pigment to be produced. However, there is certainly a wide field 

 for experiment in this direction open to any one who can obtain 

 abundant material for breeding. 



With regard to the local prevalence of any one colour, in- 

 cluding physical colours such as the blue-green referred to by 

 Wallace, I should rather ascribe them to a general process of 

 selection (for protective purposes probably), which rendered such 

 colours most advantageous : by the well-understood process of the 

 survival of any in the least favourable variations, the whole insect 

 fauna might in time become modified in any coloric direction. 

 There is no need at all to assume in such cases any chemical 

 action of soil or climate. 



We will now pass to consider the relation of the various 

 varieties naturally occurring to my experimental varieties, my 

 object being briefly to emphasise the point that the varieties 

 which we know to occur may equally well be produced by means of 

 chemical reagents, whether they be varieties in soluble pigments 

 or in " reversible " pigments ; and that, therefore, the existence 

 of such varieties is a strong confirmation of the theories (as to the 

 natural relations of the various colours) that I have deduced from 

 my experiments. There is one qualification necessary, however, 

 viz., that I can produce only retrogressive varieties and not 

 progressive, except in the case of the cyanide reactions of certain 

 yellow species. It will, however, probably be found, I believe, 

 that most of the varieties found naturally occurring are retro- 

 gressive varieties, and that progressive varieties are far scarcer ; 

 at least, that is my impression after collecting records of varieties 

 from several years' ' Entomologists,' and other sources. In other 

 words, then, most varieties will be found to be phenomena of 

 atavism, or degeneracy to a former less developed state.! 



First of all, though, it may be as well to say a word as to 

 variation among physical colours. There are three conceivable 



* See the whole question fully discussed in Vines' ' Physiology of Plants.' 

 t The jiresence of which, by the way, is wlioUy hypothetical after all. 

 X It would be deeply interesting to know whether or not specimens showing 

 coloric variation also display any variation in structure. Will any possessor of, 

 (,'. (J., yellow varieties carefully examine the limbs, &c., of his specimens. 



