334 MR. E. B. POULTON ON THE COLOUR-RELATION BETWEEN EXPOSED 
The most isolated pupa was . 
The pupa next in degree of 
isolation was. 
S'4 centimetres from the one 
nearest to it. and it was (4), very golden. 
5'4 centimetres from the last, and y It was light (3), unusually pink. 
3'4 „ „ another pupa 
The pupa next in degree of 
isolation was. 
'I 
3'4 centimetres from the last, and j 
4-0 „ „ the nearest of \ ^ w a s light (3). nn^naUy golden, 
the group. J 
The 3 remaining pupae were arranged in a group, at the angles of 
an isosceles triangle of which the sides were about l - 3 centimetre -= 
in length and the base about 2T centimetres. 
They were all (3), 2 of them dark, 
although with 
pink colour 
distinct. 
The effects produced by the white surroundings are, perhaps, rather stronger in 
this case, and are to be accounted for by the especial care with which the less mature 
larvae were selected, so that the influence was of longer duration. The delicate sus¬ 
ceptibility of the larvae to the dark bodies of others near to them is extremely well 
shown in the difference between the three isolated pupae and the three in a small 
group, and also in the difference between the most isolated pupa in the former 
number and the other two. 
(/3) At the same time six exactly similar larvae were selected from the same cylinder, 
blinded, and placed in an “ opal ” globe, under precisely similar circumstances of food, 
light, &c. The results were :— 
Of 3 pupae lying on the floor, but not attached to 
the food,.2 were deformed, and were apparently 1 dark (3), 
1 light (3). 
1 was perfect, and was.(3). 
„ 2 suspended more than halfway up the sides of the globe and T05 decimetre apart, both were (3). 
In the relative results of (a) and (/3) there is no sufficient reason for believing that 
the blinding produced any effect. Three of the blinded larvae pupated in close 
proximity to the dark food, and therefore were in this respect under darker 
influences than the unblinded larvae ; but, on the other hand, the latter were more 
crowded, and so were under darker influences from another cause. Such an uncertainty, 
introduced into an experiment conducted with considerable care, chiefly induced me to 
separate the larvae to a far greater extent, so as to diminish the errors due to both 
causes. 
Thus 91 pupae were obtained from Series IV., and very few of the larvae died. It 
is, therefore, likely that I had obtained nearly the whole of a smallish company, and 
that none of them had previously left the nettle-bed. The results are open to the 
criticism upon Series III., but to a rather less extent, because, the larvae being slightly 
less advanced, there was more opportunity for the influence of the coloured sur¬ 
roundings when the re-arrangement took place on the evening of August 16. The 
