106 Mr. E. B. Poulton. On a special Colour-Relation [Feb. 10, 



(j3.) Wliite. — In this case also the effects were stronger (as shown 

 in the prevention of pigment formation) as the surface was more 

 highly illuminated (P. rapce only). 



(7.) Colours of the Spectrum. — All the colours were used except 

 violet, and the effects upon pigment formation in the two species were 

 so graduated in the successive colours that it was possible to approxi- 

 mately represent the results by a graphic method, making the abscissae 

 of the scale of wave-lengths of the visible spectrum, and each ordinate 

 of a length which corresponded to the average amounts of pigment 

 obtained from all the pupaa subjected to any one colour, each ordinate 

 being made to diverge at its base, and to include the degrees on the 

 scale of wave-lengths which were shown by the spectroscope to corre- 

 spond to the rays reflected (or transmitted) by the colour in question. 

 Joining the summits of all the ordinates, the lines obtained were 

 strikingly similar in the two species. 



The effects may be summarised as follows : — 



Colours. 



P. brassicce. 



P. rapce. 



Black {for com- 

 parison). 

 Dark red. 



Deep orange. 



Pale yellow. 

 Green. 



Pale bluish-green. 

 Dark blue. 



Largest amount 



of pigment. 

 Almost the same. 



1 



Smallest amount Marke ?. f ect * 



of i ment on 



" ° ' groundcolour, 



especially in 



the case of 



Rather more. orange. 



Rather more. 



Much more, almost equal to 

 black and red. 



Largest amount 

 of pigment. 



f Marked 

 J effects 



Smallest amount j on tint 

 of pigment. | of 



| ground 

 ^ colour. 



Rather more. 

 Intermediate between the 



two last. 

 More than in yellow. 



Still greater amount, hut 

 not nearly equal to black. 



The colours which most retard the formation of pigment were 

 shown by the spectroscope to contain certain rays in common, i.e., 

 those from W.L. O00057— W.L. 0'00059, or 0*00060. The whole of 

 the experiments on these species seemed to show that, of the light 

 incident on the larval surface, the direct white light produces no 

 effect at all (until after it has been reflected). Further experiments 

 mast decide whether direct light can be equally efficient with re- 

 flected light, when it contains the same spectroscopic components. 

 The green tissue-paper was quite insufficient to prove this, for it 

 must have been largely coloured by absorption from reflected as well 

 as transmitted light. 



