356 MR. E. B. POULTON ON THE COLOUR-RELATION BETWEEN EXPOSED 
proof of the susceptibility of these larvae to the dark colour of the bodies of their 
neighbours. This experiment is also a very good test of the influence of white surfaces 
upon the larvae, and compares in an interesting way with the other divisions of this 
series in which gilt surroundings were used. 
B. The experiments upon shorn larvae had shown that the sensitive surface for 
which I was searching did not exist on the bristles, and the blinding experiments had 
also led to negative results. Yet the theory that the ocelli do represent the desired 
terminal organ affected by the coloured surroundings seemed so probable that I was 
desirous of obtaining other support for the experiments, which seemed at first sight to 
overthrow the theory so completely. These larvae are very active, and when blinded 
or irritated in other ways fling their heads about in the most violent manner, and 
I feared that the varnish might in this way have been accidentally knocked off one 
or more of the six ocelli on one or both sides. In the most cai’eful experiments 
the application of varnish had been repeated, so that three coats had been given in 
many instances, but the larvae had been violent whenever the process was repeated. 
Again, it seemed possible that if the varnish had not been removed it might have 
been made thin and partially transparent over the ocelli, most of which project 
considerably. In spite of these arguments I did not see how my repeated and careful 
application of varnish could have failed in all cases ; I did not expect to find the 
blinded larvae as dark as those exposed to black surfaces, but I did expect that they 
would not be equally brilliant with normal larvae when both were exposed to white 
or gilt surfaces, that is, if the ocelli represented the desired terminal organs. But the 
above considerations made it very desirable to test the ocelli in some other way. I 
had long thought of another experiment, in which the head and the body of a larva 
were to be exposed respectively to two different colours, producing the most opposite 
effects, but the difficulty was in the mode of application. In fact, this difficulty 
seemed insuperable for Stages I. and II. of the preparatory period, but perhaps might 
be overcome with a little contrivance for Stage III., in which the larva rests suspended. 
It was, however, first necessary to decide whether the larva is sensitive to colour 
during this stage. Previous experiments (V., B.; VI., C.; VII., B.) had not thrown 
much light on the question, but the numbers employed had been insufficient. I 
therefore determined to devote the majority of the larvae of this series to the settle¬ 
ment of this question. It seemed that a decisive answer could be best obtained by 
the transference of larvae in sufficient numbers at the beginning of Stage III. from a 
colour to which they had been up to that time exposed into one which tended to 
produce the most opposite effect. Quite apart from the chief object to be served by 
the experiment, i.e., the possible introduction of further experiments, the results 
would be most interesting’ on their own account in the further light which would be 
thrown upon the preparatory period. At 2.30 P.M., August 31, it was seen that 
the larvae were nearly all mature, and were beginning to collect on the sides and roof 
