LEPIDOPTEROUS PUP^E AND THEIR SURROUNDING SURFACES. 
329 
not 
2 deformed and with larval heads attached, both 
apparently.light (2). 
1 normal.light (2). 
suspended to the glass side f 2 deformed as above, both apparently. . . . (5). 
of cylinder |_1 normal.light (3). 
suspended to the roof (about l - 9 centimetres apart).both (4), but with little gold 
for this stage. 
lying on the floor, 
attached to the food 
The deformity is clue to the fact that the larvae, when blinded, dash their heads 
about, and so tend to spread the varnish over a much larger area, and it is thus liable 
to harden on the top of the head, and to prevent the latter from splitting in the 
complete manner which is necessary for ecclysis. 
I have been obliged to quote an “ apparent ” degree of colour for the deformed 
pupae, because it was difficult to judge them by the same standard as the normal ones. 
In the result it is to be noted that the white roof produced a strong effect on the 
pupae suspended from it, and that those attached to the side of the low cylinder also 
came under its influence. The floor was white, but the relative proximity of the dark 
leaves and stems of the food-plant may have made all the difference. On the other 
hand, these pupae were not crowded together, as in the majority of the A. division. 
This last important difference in arrangement, which must certainly produce some 
considerable effect upon colour, together with the possible sources of error introduced 
by the food necessary for many larvae, induced me to alter the conditions of these 
comparison experiments in a manner which will be described below. 
Allowing for all uncertainties, there is no sufficient ground in the respective results 
of these two divisions (A. and B.) for the belief that the ocelli represent the larval 
sensory surface through which the colour influence makes itself felt. 
C. Nearly all the remaining larvae were placed (August 16, evening) in a large 
cylinder (2 - 39 decimetres in internal diameter and 1‘02 decimetres in height) covered 
with one layer of black tissue-paper, and with a similar roof and floor. Thus, 
although the surrounding surfaces were all black, a considerable amount of light could 
enter the cylinder. These larvae had previously been in the green-glass covered 
cylinder from which those of the B. division were taken. Compared with the others, 
and with Series II., the results were :— 
3 suspended from the roof 
1 suspended from the side 
2 on the floor. 
2 near together, both (3), one of them light (3). 
1 isolated and . . . (3), darkish. 
.(4). 
. . were both (2). 
6 
Further experiments will show that these results were largely due to the exposure 
previous to the evening of the 16th. I was unable to take notes of the hours at 
which suspension or pupation took place in this or the B, division, but, as the latter 
MDCCCLXXXVII.—B, 2 U 
