LEPIDOPTEROUS PUPH3 AND THEIR SURROUNDING SURFACES. 
373 
upper and more trustworthy analysis, although depending on insufficient data, further 
supports the conclusion that the influence acts upon the skin by showing preponde¬ 
rating effects from the colour to which the larger area of skin has been exposed ; and 
the differences are small, both between the effects wrought in the pupae and between 
the two areas of skin affected by opposite stimuli. 
E. Another rather more elaborate plan of conducting the conflicting colour 
experiments occurred to me, in which each larva was to be kept separate from the 
others during Stage III. I made a number of short cardboard tubes with two 
perforated discs in each of them, one disc at one end, and the other near the middle, 
b. Black, c. Cardboard, g. Gilt. gl. Glass, s. Boss of silk. (The black line indicates 
the black coating, the white margin the gilt coating, while the lines between represent 
the cardboard substance of the tube.) 
dividing the tube into two compartments. The size and shape are shown in fig. 3, 
which is drawn twice the actual size in all dimensions. In each tube the colour 
lining one compartment was gold, while black lined the other. The method of 
application is also shown in fig. 3, B, the upper aperture being slipped over the head 
of the larva, which was then assisted through the lower aperture by the use of 
forceps; a little glue had been placed on the upper surface of the upper disc, and this 
was pressed tightly with a slight screwing motion on to the glass, from which the 
larva was hanging, and in all cases tightly adhered. The larva nearly always 
remained quiet in the tube, and did not retract its head into the upper compartment. 
Fearing lest the larva might stretch its head beyond the lower rim of the tube, the 
externa] surface, as far as it could be seen from such a point of view, was always of 
the same colour as the lower compartment, but I do not think that the larvae ever 
stretched so far; in fact, the tubes were of such a length that it would have been very 
difficult for them to do so. All the dimensions were adopted after careful measurement 
of larvae in Stage III. The upper compartment was illuminated through the upper 
aperture, and the lower by the open end of the tube, and also partially through the 
