386 MR. E. B. POULTON ON THE COLOUR-RELATION BETWEEN EXPOSED 
In the above Table it is seen that the difference in the position of the gilt chamber 
corresponds with a larger difference in pupal colour when the larvae were covered for 
part of Stage III. than when they were covered for the whole of it. And the difference 
is made by the increased darkness of the pupse of the former set in tubes with the 
gilt chamber above (especially in XIII.), while the pupae in the other tubes of the 
same set are as nearly as possible identical with those in similar tubes, but which 
had been covered for the whole stage. This is probably due to the fact that the gilt 
chamber when below was strongly illuminated, and could produce effects even when 
working for something short of the whole stage—although it is very likely that 
large numbers of experiments would show that such effects are not so great as when 
the influence worked for the whole stage, while the less illuminated upper chamber 
failed to have any effect except when working for the whole stage. 
If the results obtained were merely those of the last three lines of the complete 
total, it would make the conclusion as certain as the conditions of this experiment 
and the numbers employed could make it—that the influence does make itself felt 
through some anteriorly placed sense-organ situated in the lower chamber, and affected 
by its colour alone. Accepting the free larvae as on the whole normal, it is seen that 
when the head was in black surroundings the pupae are in no instances as light as the 
two lightest degrees attained among the normal ones ; but when, on the other hand, 
the head was in gilt surroundings the pupae never reach the two darkest degrees of 
the normal pupae. But such conclusions are quite upset by the further comparison 
with the far more trustworthy results of the pupae which were covered for the whole 
of Stage III. Here also the pupae in tubes with the gilt chamber below are rather 
lighter, but the others, ad though not equally light, are lighter than the free pupae. It 
has been suggested above that the free pupae are about normal, and I think that 
this suggestion is confirmed by a comparison with the pupae found on the food-plant 
and in the cylinders in which the stock was kept (see below, D.) The influence of 
a colour—black or gold—felt by a larva must be immensely different according as the 
latter is inside a tube or outside it, and it may well be that the effects in the last case 
are so slight as to be often inappreciable when the stimulus has been applied during 
Stage ILL only. Certainly in the above-described experiments the facts of one set of 
free pupae seeming to indicate a slight influence are compensated by those of the next, 
in which the results are highly irregular. I believe, however, that the effects were 
read in a few instances, although very slight, but there is insufficient evidence for the 
belief in these experiments alone. 
Assuming, then, that the free pupae are not far from normal, we see that the pupae 
in tubes for the whole of Stage III. are lighter, vffiatever be the position of the gilt 
chamber; the gold has a more powerful influence than the black with either system of 
relative position. But such results are quite inconsistent with the theory that the 
larval ocelli are influenced by the colour, for it is seen that the gold produces effects 
when it is shut off from the anterior part of the body together with the head. The 
