LEPIDOPTEROUS PUP2E AND THEIR SURROUNDING SURFACES. 
339 
From the data the approximate duration of the respective stages was estimated by 
taking’ the mean of the times between which the beginnings and ends of the stages 
O O O O 
occurred, the following results being obtained 
Stages I. and II. 
Stage III. 
The whole preparatory period. 
Larva 1 
Incomplete 
About 23 hrs. 
Incomplete 
„ 2 
Incomplete 
Uncertain. 
Incomplete 
„ 3 
About 20f hrs. 
About 20f hrs. 
About 40-| hrs. fvery nearly correct) 
„ 4 
„ 14f „ 
9QJL 
')') U -‘2 >7 
About 36|- hrs. 
„ 5 
„ 16 
>> 20f ,, 
About 36f hrs. (very nearly correct) 
„ 6 
In complete 
>> 19 ,, 
Incomplete 
Averages 
Average of the 3 instances 
Average of the 5 instances 
Average of the 3 instances 
16| hrs. 
D 
21 hrs. 
37-f- hrs. 
Comparing these averages with the estimates in Division A. of this series, the length 
of Stage III. is here rather longer, and that of the whole period slightly shorter, 
but without much difference. It is to be noted that in three cases (Nos. 1, 4, and 5) 
Stage III. was passed in the dark, and that the duration of the stage was rather longer 
than in the other larvae (except in one case, in which the length was the same as that 
of the shortest of the stages passed in the dark). If the stages were somewhat 
protracted by this treatment, of course the whole preparatory period would be 
correspondingly lengthened. Evidence in favour of such protraction will be found in 
some of the later experiments upon this species, and also in the case of the Pieridae.* 
The exceptionally short duration assigned to Stages I. and II. in Larva 4 may be 
partially explained by supposing that the stage really began very soon after 1 p.m. 
on August 23 (instead of at 3.30 p.m., and thus halfway between 1 p.m. and 6 p.m.), 
and that it was thus about two hours longer. Such a supposition is rendered 
probable by a comparison of the times of pupation of Larvae 3 and 4. 
In such small cylinders the larvae wandered very little before fixing on the position 
in which to suspend themselves, and therefore Stage I. was reduced to a minimum. 
As to the colours of the pupae, there was very great uniformity, Pupa 6 being the 
lightest, then Pupae 1 and 2, and then, after an equal interval, Nos. 3, 4, and 5 ; but 
the whole of the difference being comprised in the slight interval between normal (3) 
and very light (3). Thus the white-paper back-ground produced much less effect 
than the gilt back-ground in Division A., and the lightest of these pupae, 6, had 
already passed Stage I. and much of Stage II. in the gilt box. The removal of three of 
the larvae into darkness during Stage III. produced no apparent effect, but the numbers 
* This probable effect of darkness appeared to be so important that I experimented upon 44 larvae 
during the past summer (1887) with the object of testing the above-mentioned conclusion. The larvae, 
placed in a strong light, were surrounded in some cases by gilt and in others by tin surfaces ; those in 
the dark being surrounded by black paper. Without giving the details of the experiment, I may say 
that its results conclusively proved that darkness does considerably protract the preparatory period. 
There did not appear, however, to be any evidence for the supposition that the gilded pupae pass through 
a shorter preparatory period than those which are less brilliant, when both are equally exposed to light.— 
September 10, 1887. 
2x2 
