352 MR. E. B. POULTON ON THE COLOUR-RELATION BETWEEN EXPOSED 
From the above Table it is possible to gain a very correct estimate of the duration of 
the stages and of the whole period : Stage III. is seen to be fairly constant in length, 
but the others very irregular. The boundaries of the different stages are fixed with 
great exactness in nearly all cases, owing to the number of the observations and the 
fact that the experiment began about two-and-a-half hours after the larvae were taken 
from their food plant. I think that this Table gives the duration of the different 
stages more completely than they can be found in any of the other series. 
B. The four larvae found motionless, but not suspended (Stage II.), upon the porch 
of Binsey church were made use of in a transference experiment; the three found 
upon grey stone (light) being placed in black surroundings, and the one found upon 
the oak door (dark) being placed in white surroundings. Between them capture and 
subjection to these influences they were kept in ordinary chip-boxes of light wood, 
partially transparent. The experiment is described below :— 
Dates, &c. 
(a) The larvae found on light stone. 
(/3) The larvae found on dark oak. 
Aug. 26, 7 P.M. . 
Found on stone ; kept in chip-box till 
put in dark cylinder 
Found on oak; kept in chip-box till 
put in white cylinder. 
„ 10.45 p.m. . 
Placed in small cylinder covered with 
1 layer of black tissue-paper and black 
floor 
Kept in chip-box till put in white 
cylinder. 
„ 11.30 p.m. . 
No change. 
Placed in small cylinder covered with 
white paper. 
Aug. 27, 9.40 a.m. . 
All suspended from roof. 
Suspended from roof. 
„ 9.30 p.m. . 
55 55 59 . 
59 55 
„ 10.58 p.m. 
2 just pupated. 
5 5 5 5 
Aug. 28, 10 a.m. 
The last pupated very many hours . 
Pupated very many hours. 
Results, pupae com¬ 
pared Aug. 29. 
Of the 3 pupae suspended from roof, 
1 was light (3). 
2 were . (3). 
3 
The 1 pupa on roof was (3). 
There are not sufficient data from which to make estimates of the length, of the 
stages. 
There was extremely little difference between the pupae, and the results agree well 
with those of other transference experiments, for it is probable that the pupal colour 
was in both (a) and (/3) largely influenced by the fact that Stage II. had been passed 
(at any rate in part) upon surroundings which, in each case, tended in a different 
direction from the environment to which the pupae were exposed in Stage III. 
Thus 15 pupae -were obtained in this series. 
