482 Mr. "P oulion' 8 ftirth 67' eocperimenis upon 



Here the direct transmitted light was the same, but the 

 amount reflected was different; for white paper returns 

 the whole, while the coloured background only reflects 

 a dim green band out of the transmitted light. 



The above ('X})lanation c(n-r(>sponds to the fact which 

 is apparent in the last table, that coloured screens, when 

 they differ at all from backgrounds in strong light, only 

 do so by producing green pupao in place of dark ones, and 

 never dark instead of green. 



At first sight the tempting converse explanation of the 

 phenomena is suggested, viz., that the terminal parts of 

 the spectrum, and especially the actinic blue, are instru- 

 mental in producing the animal pigments, while the 

 orange and yellow rays merely fail to produce them. 



I do not think that such a view can be sustained for a 

 moment in face of the facts already adduced. The 

 common appearance of dark i)igraent(!d larvae and pupfe 

 in complete darkness (and on the blackest backgrounds 

 in light), shows that the pigment is a normal product of 

 the animal organism, entirely independent of the agency 

 of light. Furthermore, the region of the spectrum, by 

 which such formation is normally checked, corresponds 

 to the region of greatest intensity of light, and so sup- 

 ports the view that it acts as a definite stimulus, and not 

 merely i)assively. All the facts hitherto brought forward 

 (except perhaps the golden forms of V. urticce), su})port 

 the opposite view, that the pigmented form is the primi- 

 tive one, and is still produced, as a rule, in the absence 

 of any definite stimulus ; but that certain conditions in 

 the life of certain species have encouraged, by natural 

 selection, a special susceptibility to certain stimuli which 

 check the formation of pigment, and so produce an ap- 

 pearance which harmonizes with that from which the 

 stimuli arise. 



I am here alluding only to the power of becoming 

 green by the non-appearance of true superficial pigment, 

 and not to the i)ower of altering the colour of the latter 

 as in A. Jxjtidana, &c. This indicates another complex 

 adaptation which has been already briefly considered 

 (see pp. 353, 356, 369). 



3. Experiments in 1892 upon the pup^ of Vanessa 



ATALANTA AND CyNTHIA CARDUI. 



A few experiments upon V. atalanta are described in 

 my previous paper (/. c. pp. 398, 399), and the lighter 



