356 Mr. 'Bonlton' H fiirtJicr experiments upon 



purplish larva is represented on a spill of the dark blue 

 colour employed. The effects of lighter blue (XXXII.) 

 were far less uniform. It may be that the blue rays 

 must come from a surface of a certain depth of colour in 

 order to produce the effect seen in XXXI. 



Orange surroundings, on the other hand (XXXIII. and 

 XXXIV.), tend to produce typical green larvfe, although 

 the effect of orange paper was much stronger than that 

 of orange enamel. The larvae were, however, exposed to 

 the latter for a shorter time, and probably found it a less 

 congenial surface to rest upon. 



Here, too, when an artificial colour entirely different 

 from anything in the normal surroundings of the species 

 produces exactly the same effect as a totally different 

 appearance in the natural environment, the most 

 probable view is that there is some common qualit}^ in 

 the reflected light, and that this is the effective stimulus. 

 If there were any evidence for pathological change or 

 abnormal develo[)ment of any kind, the argument would 

 not hold ; but the larvae reared among orange spills and 

 sticks appeared to be as healthy, and in every way as 

 normal, as those reared among the green leaves and 

 shoots, which produce the same result. Nor is it at all 

 probable that the results are merely due to the quantity 

 of reflected light rather than its quality. Thus the 

 greater amount of light reflected from white paper does 

 not make the larvae greener than orange paper, but 

 utterly different in appearance. 



These considerations will be brought side by side with 

 those derived from the experiments on pupae in the 

 Conclusions at the end of the paper, where the spectro- 

 scopic composition of the light reflected from the back- 

 grounds will be described, together with its effect upon 

 the various species employed. 



Comparison with the earlier experiments on the pupae 

 of Pieris rupee and P. hrassicue strengthens this con- 

 clusion; for I have shown (Phil. Trans lioy. Soc, vol. 

 178 (1887), B, pp. 429 — 432) that in both these species 

 the same orange paper employed in XXXllI. and the 

 same blue employed in XXXI., produced green and dark 

 pupie respectively. On p. 43 J the extent in the spectrum 

 of the rays which were chiefly reflected from these colours 

 is represented in a diagram, and it is shown to be pro- 

 bable that the rays which check the formation of true 



