of certain lepidopterous larva and pupa. 35 



subdorsal has been retained very late in the j^hylogeny 

 of the genus Smerinthus. 



5. The prohahle phytophagic character of the ground 

 colour in S. ocellatus. — This conchision is suggested by 

 Mr. Meldola in the notes to his translation of Weismann's 

 book. I have no doubt that experiment will settle the 

 question in the affirmative. In the meantime I give a 

 list of the trees upon which I have found S. ocellatus 

 during the past summer, with the colour of the leaves 

 (under sides) and the larvae. The invariable resemblance 

 of the latter to the former in many species of plants is 

 much in favour of the cause being phytophagic, unless 

 we imagine that the larvae are truly dimorphic, and that 

 the moths of each variety lay their eggs on the appro- 

 priate trees only. I have found the bright yellowish 

 green variety on Salix cinerea and S. rubra, and formerly, 

 I believe, on a crab-apple tree at Beading. The under 

 side of the leaves of the crab are whitish, but brighter 

 in appearance than those of other apples, which have a 

 peculiarly "dead" colour very characteristic of the 

 whitish green larva. I have found the latter upon Salix 

 viminalis and commonly upon apple (see Mr. Meldola's 

 note on the subject). 



6. Protection sometimes gained hy the changes in colour 

 before pupation. — The adult larva of S. tilice turns a 

 dark purplish brown before pupation, and has usually 

 assumed this tint by the time that it has come down 

 from the tree upon which it was feeding. The same is 

 true of Sphinx ligustri, which becomes quite brown on 

 the back, and in this condition may often be found 

 hurrying along a road or garden path in search of a 

 suitable place to bury itself. I have found a larva of 

 S. ligiistri, still upon its food-plant, with distinct indi- 

 cations of the brown colour. 



In these two species the dark coloration before the 

 final change is very marked in amount and very early 

 in appearance. Conversely the larvae of Smerinthus 

 ocellatus and S. populi show but a slight tinge of brown 

 when they have left the trees and are wandering about 

 before burying themselves. In all these cases I am 

 speaking of the change of colour in the full-fed larva 

 when it ceases to feed and makes preparation for the 

 final change : I do not refer to the colours (already 

 described in S. ocellatus) assumed in the quiescent con- 

 dition immediately preceding pupation. 



