418 MR. E. B. BOULTON ON THE COLOUR-RELATION BETWEEN EXPOSED 
These results negative the view that the ocelli represent the sensitive organs sought 
for, for the two sets of pupae were, on the whole, as equal as possible. The latter set 
of three pupae included one lighter than any of the other set, and two somewhat 
darker; but the differences were slight in all cases. The larval heads remained in 
proximity to their respective pupae, so that the success of the blinding could be tested 
afterwards, and the varnish, when examined with a lens, appeared to completely cover 
the ocelli and the surrounding area. 
The light pupae produced by the white surfaces are well seen in the 12 pupae of 
these experiments. 
It will now be interesting to give a tabular analysis of these results, showing the 
effects of white surfaces with different degrees of illumination in the case of P. rapoe. 
Degrees of colour. 
(1) 
(2) 
Dark 
(3) 
Light 
(3) 
Dark 
<4, 
(4) 
Light 
(4) 
Pale 
(5) 
Deep 
(5) 
II., A. White surface in almost 
complete darkness 
1 
1 
— 2 
„ C. VWhite surface in strongj 
. . 
, t 
. , 
2 
1 
2 
. , 
. # 
5 
„ D.j illumination [ 
1 
. , 
1 
3 
I 
1 
= 7 
,, B. White surface in very 
strong illumination 
. • 
• * 
* * 
' * 
2 
1 
1 
12 
3 
* * 
- 19 
Total .... 
i 
33 
This analysis shows well the increasing effect produced by increasing illumination, 
for the description of the apparatus used in C. and D., on the one hand, and in B. on 
the other, indicates that the differences between their degrees of illumination must 
have been very great. Furthermore, the background of B. was entirely composed of 
an intensely white and strongly reflecting surface, while in C. and D. the roof (and to 
some extent the floor) alone possessed this property to a considerable extent, the 
sides of the cylinder being composed of white tissue-paper. On the other hand, there 
was a large amount of food-plant which remained until after pupation had taken 
place in B., while only a small quantity was placed in the cylinders, and this was 
removed when the larvae had ceased to feed. 
III. Red. 
A. 
On September 13, 15 mature larvae of P. brassicce, and the next day many larvae 
of P. rap>ce, were placed in a cylinder 8 '2 centimetres in internal diameter and 17'8 
centimetres in height, lined internally with deep-red opaque paper for about three- 
quarters of its circumference, and with a roof and floor of the same substance. The 
experiment was conducted as follows. This and the following colours made use of in 
these experiments are shown on Plate 2 G, figs. 1 G-21 inclusive. 
