fhr colnnrs of certain Lepidoptera. 431 



similar backgrounds, and to the behaviour of the larva 

 of Amphidasls hetularia, as regards the production ot 

 green varieties (see pp. 355-357). Furthermore the 

 conditions imposed are in strict accordance with those 

 which obtain in nature. The wild larvffi and those which 

 pupate upon coloured backgrounds are freely exposed 

 to bright davHght. Mixed with this, but nnmensely 

 diluted°byit, are the rays reflected from adjacent sur- 

 faces and the yellow and orange constituents ot these 

 reflected rays determine the appearance of green pupie 

 bv checking the formation of true pigment. 



We are compelled to conclude from these results and 

 those upon the Pieridcs [l. c.) that the greens of nature 

 (due to chlorophyll) do not produce their efiects m making 

 arv^ and pup^ green, because o their brightest con- 

 stituent, the green rays, but m virtue ot the partially 

 absorbed, but still bright, yellow and orange rays con- 

 tained in their reflected light. And we must further 

 conclude that if these yellow and orange rays were 

 removed, the green rays, bright as they are, diluted by 

 other reflected rays, and, above all, by the immense pre- 

 ponderance of direct white light, would be unable to 

 check the formation of pigment and produce the green 

 pup^ and larv.^. Diluted in this way, only the orange 

 and yellow possess the power to effect such a change. 



When, however, we employ coloured screens the con- 

 ditions are entirely altered. The larva is not exposed 

 o dhect white light, but only to the light transmitted 

 through the screen, and the same alter reflection from a 

 iabt background. Hence the rays tall upon the arva 

 surface in an undiluted comparatively concentrated 

 form, and their efficiency is correspondingly increased 

 extending beyond the orange into the red and beyond 

 ?he yellow into the green. The effect begins to die away, 

 however, in the feebler blue rays, even when presen m 

 this concentrated state. When screens are emp oyed m 

 this way, it is still the reflected rays^ rather than the 

 direct tmnsmitted ones which are effective ; thus in 

 Experiment 21 upon V. urticce (see p. 376), a box 

 fLXXVII.) with green glass windows and lined witli 

 d^-k green paper {the paper similar to that employed 

 in Experiment 81 upon V. io), produced dark pupa3 ; 

 while the same box, lined mih white v^ipervvodnced 

 uniformly bright pup^ of T^ io (Experiment 16, p. 401). 



