LEPIDOPTEROUS PUP^E AND THEIR SURROUNDING SURFACES. 
383 
perforated disc from each tube, believing that I had in the other experiments allowed 
far too much for the greater illumination of the upper chamber. The results of the 
experiment are shown below : — 
Dates, &c. 
1 
2 
3 
4 
5 
7 larvae free in glass box. 
Sept. 5, 12.5 p.m. . . 
„ 2.15 p.m. . . 
„ 4.30 p.m. . . 
,, 11 P.M. 
Cr 
Or 
Cr 
Cr 
Cr 
These were free to suspend 
themselves among the gilt 
and black tubes surrounding 
the other larvae. 
Time of pupation unknown. 
Results, Sept. 7. 
Time of 
pupation 
unknown 
4 on glass roof of cylinder : —- 
Gold 
below 
Black 
below 
Dead 
Dead 
Dead 
2 isolated, both (4), very 
light and pink, but little 
gold. 
(4) 
Very 
light 
and 
pink, 
but little 
gold 
(5) 
Rather 
dull for 
this 
stage 
1 close to a tube, gilt outside, 
(4) as above. 
1 close to 2 tubes, black out¬ 
side, very light (3) ; re¬ 
sembling the others, only 
dotted over with dark pig¬ 
ment. 
1 on side, isolated, (4) as 
above. 
2 on floor ; both (4) as above. 
7 
Most unfortunately only two larvae underwent the final change, and at this time 
the stock of larvae was exhausted, and no more could be obtained until another 
season. Had I been able to find them, I much wished to make a large number of 
experiments with tubes in which the upper disc was removed. However, as far as 
the experiment goes, it strongly supports my impression that the predominance of 
colour-effect produced by tubes with the gilt chamber below is entirely due to the 
much greater illumination from the wider lower opening. This will be alluded to 
below in discussing the analysis of all results obtained after the use of compart- 
mented tubes. In this experiment the pupa in the tube with the gilt chamber above 
was a somewhat dull (5), but nevertheless the only (5) obtained in any of the experi¬ 
ments with the tubes. Among the free larvae there is some colour-correspondence 
with the neighbouring tubes, which certainly suggests some influence on the part of 
the latter. It is interesting to note the extremely light results of experiment (y) as a 
whole. 
Before concluding the experiments of Series XIII., it is best—now that all the 
results obtained by the use of tubes have been described—to analyse the colours 
obtained in all subdivisions in which this method has been employed in this and the 
preceding series. The analysis is given below :— 
