the colours of certain Lepidoptera. 465 



and pupae. And this is what has been found to occur. 

 I attach great importance to the colour of copper arsenite 

 in this respect. I hope to try it again on a larger scale, 

 covering it with varnish so as, if possible, to prevent any 

 poisonous effect. 



The larvae and pup^e are probably sensitive to diffused 

 rather than regularly reflected light, the strong effect of 

 "gilt" being explained by its absorption of blue rays, 

 and the consequent greater prominence of yellow, as 

 well as by its power of returning so high a proportion of 

 the light which falls on it. Silver and tin with this same 

 power, but without that selective absorption which gives 

 prominence to yellow, exert a far inferior influence upon 

 these insects. ' The effect of white paper and opal glass 

 is easily explained on the principles laid down above. 



I have hitherto only considered the production of 

 green or bright pupae and larvae. But the table of back- 

 grounds at once proves that the case is far more com- 

 plex in certain species, nnd notably in AuipJtidasis 

 hetidaria. These larvae behave like the pupae as regards 

 green, black, and orange backgrounds, but entirely diffe- 

 rently as regards brown, white, and, to some extent, deep 

 blue. These do not make the larvae green, but produce a 

 special form of true pigment, in two cases corresponding 

 to the coloured surface which emitted the rays forming 

 the stimulus. It cannot be doubted that these effects 

 also follow from the constitution of the diffused light 

 reflected from the background (see also pp. 35B, 356, 

 359). 



The same contention is true of R. cratmjata, the true 

 pigments of which can certainly be modified, as well as 

 dismissed (see p. 3*26), and probably of all sensitive larvae ; 

 for it is unlikely that the great difference between the 

 dark and light browns is only a question of quantity of 

 pigment. 



We are justified in concluding that a larva of a species 

 which possesses this power of adjustment (as regards 

 pupa or larva) is effected, during the sensitive period, by 

 certain constituents of the diffused light reflected from 

 surfaces in its immediate neighbourhood, diluted as it is 

 by other constituents, and far more by the direct white 

 Ught which falls on every part of its surface. It is 

 sensitive to this very small proportion of effective rays, 

 and can, as a response to the stimulus, produce true 



