THE GREGARIOUS HABIT IN BUTTERFLY LARVAE. 19 



with in the preceding volume, but exactly similar habits to those 

 of A. urticae larvae in Europe are noticed in the larvae of Aglais 

 milberti in America. Of the latter, Scudder notes (Butts. New 

 England, i., p. 426): "The young larvae, on escaping from a 

 cluster of eggs, do not stop to devour the eggshells in the least, but, 

 after eating a portion, or the whole, of the leaf on which they are born, 

 climb to the summit of the plant by weaving a silken path; within a 

 day they smear the whole of the summit with a web, and may be seen 

 swarming (for they are highly gregarious in early life) upon the dried, 

 curving projections of the leaves, upon which they soon fasten them- 

 selves for a moult. They feed crowded side by side, and, on the least 

 disturbance, raise their heads and front part of the body at right angles 

 to the rest, and wag them slowly in concert, producing a ludicrous 

 effect .... They are generally found on the upper surface of 

 the leaf, and, until half-grown, make no attempt whatever to conceal 

 themselves. After the third moult, when they have attained half their 

 size, they quit these webs and scatter over the neighbouring plants, 

 living singly, or in small companies of three or four, ' leaving their 

 deserted habitations mere leafless stalks, covered with the dense and 

 cloth-like web, and with the excrement and sloughed skins of the 

 caterpillars ' (Gosse). At this time, they occasionally collect together 

 in larger or smaller numbers in incompletely closed leaves of 

 nettle, but they leave these nests to feed in the most exposed manner. 

 These nests are thoroughly closed next the base of the leaf, the edges 

 having been drawn closely together with silk along the basal half of the 

 leaf, to effect which, an irregular, triangular notch is eaten close to the 

 base, cutting through one or both of the principal lateral ribs which 

 spring from the very base of the leaf, leaving two considerable flaps, 

 which are flattened beneath the stem by their tips, thus bending the 

 leaf at a strong angle between the deepest parts of the notches ; the 

 edges of the notch are then united, closing completely the pocketed 

 base of the leaf ; the opposite extremity, however, flares completely 

 open, but, by the bend in the leaf, is hidden from view above," etc. 

 Comparison of this account of the larval habits of Aglais milberti with 

 those of A. urticae makes one feel doubtful whether the insect can be 

 really specifically different from A. urticae (see Nat. Hist. Brit. Lep., 

 viii., p. 58). 



It will be observed that, more or less, the gregarious habit of the 

 larvae of Aglais milberti and A. urticae (and, similarly, that of those of 

 Vanessa io) fails after the third instar, but those of Kugonia polychloros 

 and Euvanessa antiopa continue through the whole of larval life. Chap- 

 man observes that the young larvae of both these species cover their eggs, 

 and the neighbourhood where they were deposited, with a silken web, 

 not spun, as it were, of set purpose, but the result of journeying to the 

 nearest leaves to feed and returning again to the central position for 

 resting. They appear often to feed in turns, one lot going out to feed 

 whilst others have just returned to rest. As they get larger they move 

 their headquarters, again apparently, according to such exigencies as 

 may occur from the form of the branch they are on, to make another 

 position more central to the available food, than to any instinct that 

 makes them move at any particular stage or instar. Different broods 

 seem to vary a good deal as to how far they remain gregarious in the 

 last instar or become quite solitary. If food remains at hand, few 



