LEPIDOPTERQUS PUP2E AND THEIR SURROUNDING SURFACES. 
437 
seem to be much deeper in significance and far more difficult to understand. The black 
pigment patches and minute black dots are cuticular and superficial, while the ground¬ 
colours are sub-cuticular and deep-seated ; and in the most brightly coloured pupae they 
are mixed colours due to the existence of different pigmentary (and probably chloro- 
phylloid) bodies present in the different elements and at different depths of the 
sub-cuticular tissues of the same pupa. In other pupae no trace of such colours can be 
seen. Hence we see in these most complex and varied effects of the stimulus 
provided by the reflected light, which deepen into their permanent pupal condition 
very many hours after the stimulus has ceased to act, the strongest evidence for the 
existence of a chain of physiological processes almost unparalleled in intricacy and 
difficulty, while a theory of comparatively simple and direct photo-chemical changes 
induced by the stimulus itself without the intervention of such a physiological circle 
seems entirely inadequate as an explanation of the facts. 
Observations upon the Colours of the Pupce in the Genus Ephyra. 
After the consideration of the variable pupae of many species of Rhopalocera it is 
interesting to compare the results obtained after an examination of the equally exposed 
and variable pupse of a single genus of the Heterocera—the genus Ephyra. In 1883 
I had the opportunity of studying the life-histories of three species of this genus 
(E. pendulana, E. omicronaria, and E. orbieularia), and an account of the investiga¬ 
tion is published in the ‘ Transactions of the Entomological Society of London/ Pt. I., 
1834, pp. 50-56. A short summary of the results obtained is given below. The 
larvae of E pendularia are dimorphic in the last stage, appearing in the two most 
usual colours, green and brown; those of E. omicronaria are similarly dimorphic, but 
the brown forms are relatively rare; while the larvae of E. orbieularia are variable. 
The dimorphism of the two former species extends into the pupal stage, the brown 
larvae always becoming brown pupae, and the green larvae green pupae. (The two 
forms of E. omicronarm are shown on Plate 26, figs. 22 and 23, natural size.) Hence 
the colour of the pupa can only be affected through the influences which determine 
the larval colour, and it liars not yet been shown that the colours of these larvae can be 
controlled, although, from many experiments on other larvae, I think that the proof 
of such a relation to surrounding colours is likely to be afforded by experiment. 
The pupal and larval dimorphism has no relation to sex or to any observable 
irnaginal character. Statistics appeared to prove that the brown forms of E. pendu¬ 
laria (alone observed in sufficient numbers) are relatively abundant in the winter 
(larvae and) pupse, and green in the (larvae and) pupae of the summer broods. It was 
also shown that the relative preponderance of either form could be greatly increased 
by breeding from parents which possessed the same colour in the earlier stages. 
Observations upon the situations selected for pupation failed to establish any colour- 
relation ; but the results were not convincing against the existence of such a relation, 
