426 MR. E. B. POULTON ON THE COLOUR-RELATION BETWEEN EXPOSED 
Three living pupae of P. rapes were similarly compared, and— 
Of these, 1 was isolated on the side of 
the box, and it was a . . . . (4), pink ground-colonr, and very little pigment. 
2 were isolated on the glass 
front, and of these ... 2 were (3), one typical, with a yellowish ground-colonr; the 
other very light for this degree, with a pinkish 
ground-colonr. 
The influences in this division seem to have been much weaker as far as the 
prevention of pigment formation is concerned than in the two preceding experiments, 
and this is probably on account of the colour employed. The green, although bright, 
was of a very bluish delicate tint, and was easily removed or rendered paler. On the 
other hand, there was a special peculiarity about the ground-colour of the majority of 
the pupae of P. brassicce (one of which is represented in Plate 26, fig. 27, X 2) which 
harmonised well with these surroundings, and appeared to indicate a power of special 
colour adaptation to minute differences or peculiarities in the surroundings which 
was not seen to an equal extent in any of the other experiments. 
D. 
A mature larva of P. rapes was found upon cabbage in a garden at Oxford about 
the beginning of September, and was placed in a small cylinder covered with one 
layer of green tissue-paper, and with a roof and floor of the same substance. 
Mignonette was introduced as the food-plant, but the larva, without feeding, pupated 
upon one of the leaves. The pupa was a deep green (5), hut with an unusual develop¬ 
ment of the black markings for this stage. The markings were present to the same 
extent as in a typical (3), although the grey dusting was very deficient. The distinct 
and bright ground-colour left no doubt of the real degree of the pupa. This pupa is 
represented in Plate 26, fig. 31, natural size. 
E. 
October 5, a large number of pupae were found on a very deep-green gate in a 
shaded lane at Seaview, but the vast majority contained the larvae of parasitic 
Hymenoptera, and had dried up when the comparison was made. 
It is also to be noticed that the pupae were nearly always concealed in the shadows 
of the mouldings and under the overhanging parts of the framework, &c. 
Only 6 pupae remained alive when the comparison took place. 
Of these, 1 was (1), typical, yellowish-pink as far as the ground-colour could be seen through the 
abundant pigment. 
2 were (3), 1 typical and apparently faintly yellowish-pink, the other light for this degree and 
distinctly pinkish. 
3 ,, (4), all about normal, and yellowish-pink. 
There must have been quite 20 more which had died, and there certainly was not a 
(5) among them, for I remember being astonished at the time at the small effect 
produced by the green background. One P. brassicce was found at the same time, 
which also died from the same cause, but it was a distinct (l), and probably the 
common (1), 13. 
