324 MR. E. B. POULTON OR THE COLOUR-RELATION BETWEEN EXPOSED 
G. Another lot of larvae were placed in a small cylinder surrounded by two layers 
of black tissue-paper, and with a roof of two layers. Only a single larva reached the 
pupal state, but this was an exceedingly dark and well-marked (1). 
This result also contributed greatly towards the frequent use of black surroundings 
in subsequent experiments. 
H. Another lot were placed in a cylinder covered, as before, with one layer of 
green paper, yellowish on the sides, but recently renewed and greener on the roof. 
Of the seven pupse, four were on the roof and three on the food-plant. The results 
were :— 
Of the 4 pupse on the roof ... 3 were (3). 
1 was (5), with the usual extreme development of gold. 
„ 3 ,, „ food-plant . 1 „ (1). 
2 were (3), 1 with rather more gold than usual. 
7 
There is nothing to add to what has been said before concerning the colour of pupte 
in the green cylinders. On August 13 three larvse were removed from this cylinder 
and treated in a manner which will be described. 
1. Although all the above-mentioned pupse in Series II. were compared together on 
August 19 (because by that date nearly all had reached the pupal stage), I had been 
watching the results seen in the various cylinders for many days, and had started 
further experiments, which w T ere suggested by those already described, so that both 
series of results could be compared together on the above-mentioned date. The effect 
in the black cylinder (G.), and that in E., having convinced me that black and white 
would be good colours with which to experiment, it seemed clear that, if successful, 
there would be a more decided colour-relation between the dark pupse and the black 
surroundings than between the brilliantly metallic pupse and their white environment; 
and if the former relation was, as it seemed to be, most real and protective, it appeared 
advisable to offer the pupa a surface which would harmonise with its gilded form as 
completely as the black surface did with the darker variety. Having already seen a 
few instances of the gilding developed upon No. (5), it was quite clear that nothing 
could harmonise so well with it as the brilliant colour of metallic gold-leaf. On entering 
upon such an experiment, I was not prepared for the extraordinary success with which 
it was attended. But the conclusions from the other experiments led up to this as 
a crucial test, and it certainly seemed well worth the trial. I procured some paper 
covered with gold-leaf, and with it lined the inside of a box which was provided with 
a glass front, and was placed facing a strong east light. The side of the box, which 
formed the floor in this position, was covered with brown paper. A vertical partition, 
also gilded, separated the box into two equal compartments, which possessed every 
