1886.J Larva of Smerintkus ocellatus and its Food-plants. 161 



advanced stage. When the larva is smaller and eats nvnch less it 

 remains on the same leaf for many days. Bnt directly the larva rests 

 on the stem the tints of its immediate environment alter, for they 

 are then due to the colours of both sides of the leaves and of the stem 

 itself. The relative predominance of the colours of the two sides of the 

 leaves depends upon the position of the larva and the arrangement of 

 the leaves. But the position of the larva is uniform (except when it 

 is wandering to a fresh twig), the head being always directed towards 

 the apex of the stem. Hence in the case of food-plants whose large 

 leaves regularly droop over from the vertical twigs, or are curved in 

 the usual way with the concavity downwards, the tints of the under 

 side still predominate after the larva has retired to the stem, and they 

 will still form almost the only effective colour as was the case in 

 earlier periods. When, however, the leaves hang irregularly or 

 spread out horizontally from horizontal twigs, the colours of the two 

 sides may be equally important, or may depend (in the latter case) 

 upon the side of twig on which the larva rests. 



This explanation of course will apply but little to leaves of which 

 the upper and under surfaces are approximately similar in colour, 

 and accordingly there is very little conflicting testimony from such 

 food-plants (Salix rubra, S. babylonica, S. triandra), and such as there 

 is, is mainly explicable by variations in the other two factors which 

 go to influence larval colour. But even in these plants there is some 

 difference between the colours of the two sides which would have an 

 effect upon the larva?, as was proved by the experiment in which the 

 bloom was rubbed off in the case of S. triandra. But in the case of 

 leaves with strongly white under sides such an explanation accounts for 

 all the conflicting evidence met with in the field (due to this factor and 

 recognised by the uniformity cf results from trees with leaves of a 

 particular size and arrangement). Thus when I expressed the opinion 

 that Salix viminalis produces white larvae I was thinking of our 

 common long-leaved variety. The leaves of this variety often grow 

 6 inches long near Oxford, and quite three-quarters of an inch 

 wide. Such leaves would retain a larva until the end of the fourth 

 stage, and often far into the last stage. Furthermore the long 

 leaves droop over very regularly from the higher vertical twigs upon 

 which the larvae are generally found, and so present their under 

 sides to the stem and to a larva resting upon it. On the other hand, 

 the leaves of the other variety are much smaller (from memory I 

 should say that they are often about an inch and a-halflong and three- 

 eighths of an inch wide), and often hang in irregular wisps from the 

 more vertical twigs, or droop vertically from the more horizontal 

 branches. Thus such a variety of leaf would retain the larva for a 

 comparatively short time, and after its retirement to the stem the 

 colour of the immediate environment would be as largely due to the 



