416 MR. E. B. POULTON ON THE COLOUR-RELATION BETWEEN EXPOSED 
13 pupae were thickly crowded along the angles 
made by the roof and adjacent parts of the 
two sides with the back, and of these . . 2 were (3), very light for this degree, 1 pinkish and the 
other yellowish. 
10 „ (4), very little pigment on all but 1, and this 
1 but little more than normal; 4 pinkish, 
4 yellowish-pink, and 2 yellowish-white. 
1 was (5), very little pigment; pale, but distinct, 
yellowish-green ground-colour, with pink 
tips and ridges, 
3 pupae were isolated on the roof, sides, &c.; 
of these. .2 were (4), both pinkish: 1 typical, 1 with very little 
pigment. 
1 was (5), little more pigment than usual: very pale 
yellowish-green ground-colour. 
The strong effect of the white back-ground in producing light-coloured pupse is well 
seen in the above descriptions. It is noteworthy that the degree of colour represented 
by (4) harmonises far better with this background than that represented by (5). 
Hence the (5) were of a very pale-greenish colour, and not conspicuous as well-marked 
green varieties would have been. Furthermore, the (4) were in nearly all cases very 
deficient in pigment, so that in this respect they were quite equal to normal (5). 
C. Strongly illuminated, but not equal to B. 
A few larvte of P. rapes were made use of in some experiments to test, by blinding, 
whether the influence of surrounding colours acts through the ocelli. Two glass 
cylinders of equal size (8'2 centimetres in internal diameter and 1*81 decimetre in 
height) were covered externally with a single layer of white tissue-paper, and with a 
roof of white glazed paj:>er, and a floor of ordinary white paper. The cylinders were 
placed in a fairly strong light, several feet from a north-east window. The experiment 
was conducted as follows :— 
Dates. 
a. Blinded. 
/ 3 . Normal. 
Sept. 17, 10 P.M. . 
Experiment began. 
Experiment began. 
„ 18, 8.40 p.m. 
1 larva, which had been found in 
2 larvse had both gone up the side and 
Stage I. or II. on a plain wood fence at 
Avere sitting motionless ; food removed; 
Bembridge, was girdled; another, still 
they had been found on mignonette at 
feeding, having been found on migno- 
Seaview. Another, added at this time, 
nette at Seaview 
found in Stage II. under the cement 
coping of a gate pillar at Spring Vale. 
„ 19, 9.5 p.m. . 
No further change. 
No further notes were taken, but all five larvse produced living pupa). 
