LEP1D0PTER0US PUPH3 AND THEIR SURROUNDING SURFACES. 
337 
All four pupae were on the roof—three in a line along the right side, the anterior 
pupa 3'2 centimetres from the middle one, and the latter 2'2 centimetres from the 
posterior one. The fourth was on the other side of the roof, beside the vertical 
partition, and 9 - 5 centimetres away from the nearest of the others. Thus all 
were rather isolated, especially the last, but it was in a position which was on the 
whole less strongly illuminated. The colours were compared at the same time with 
the (a) sub-division :— 
Of the 4 pups3 the 1 by the vertical partition was light . (3). 
The 3 others by the side of the box were all (4), very golden, but not equal to the (4) 
pupae of the shorn larvae, except in 
— one case, which is near a (5). 
4 
The comparison between these two sets is, as far as it goes, completely destructive 
of the theory that the bristles contain the terminal organ which was sought for. The 
pupse of the shorn larvae are on the whole rather the more brilliant. Both sets show 
well the extremely powerful influence of the gilded surface upon the larvae. There 
are also certain data from which we can arrive at a fairly accurate estimate of the 
length of the preparatory period and of Stage III. 
B. On the morning of August 23 the two cylinders containing the stock of larvae 
were examined, and it was seen that one larva had left the food, and was resting on 
the roof. This larva was joined by five others about 11 a.m. They were all sitting 
motionless (Stage II.), but when removed and placed on fresh food they would some¬ 
times eat a little. At noon on the 23rd five larvae were removed from the two 
“ opal ’’-roofed cylinders, and were placed in separate cylinders of nearly the same 
height, and all of the same internal diameter, i.e., 6‘1 centimetres. Two of these 
larvae (numbered 1 and 2) were among the six larvae which had already entered 
Stage II. Each cylinder had a roof of white glazed paper, and two-thirds of the 
circumference was also surrounded (externally) by the same material, while the floor 
was of ordinary white paper. The cylinders were placed with the clear uncovered 
side facing a strong east light, about two feet from a large window. On the floor of 
each a single nettle-leaf was placed, which was renewed, if necessary, but always 
removed when the larva had ceased to feed. Thus the conditions appeared to be as 
nearly equal as it is possible to make them. The cylinders were numbered 1—6, 
and the larvse w 7 ere known by corresponding numbers. In comparing the results, 
it must be remembered that the larvse numbered 1 and 2 were taken from the roof of 
the large cylinders in which the stock w r as kept, while those numbered 3, 4, and 
5 were taken from the food-plant, Number 6 being added much later, and taken 
from the food-plant in one compartment of the gilt box previously described. The 
chief object of the experiment being to determine the lengths of the three stages of 
the period preparatory to pupation, the cylinders were frequently examined, the 
results being shown in a tabular form below :—- 
MDCCCLXXXVIT.-B. 
2 x 
