332 BRITISH BUTTERFLIES. 



surprising in their readily eating green peas or the soft beans of the 

 scarlet runner, when we remember that many species feed on the young 

 seeds of Collutea arborescem and other pod-bearing plants throughout 

 their larval life. It must also be remembered that of the larvae fed by 

 Mr. Frohawk in this manner, not one survived, which fact, though far 

 from being a proof, is an indication that there was something unnatural 

 about the time or manner of their feeding, or about the actual diet. 

 It may reasonably be objected that some of the Lycaenid larva? 

 which are known to be parasitic on ants and at the same time carni- 

 vorous, such as Euliphyra mirifica, are efficiently protected by a 

 hard chitinous skin, the flange of which projects so far over at the sides 

 as entirely to guard the ventral portions of the larvae, and that they 

 thus escape destruction ; and it may be doubted whether the ants would 

 tolerate the presence of an unprotected parasitical carnivorous insect 

 in their nests. To this it may be replied: first, that ants do tolerate 

 in their nests other insects which prey upon their larva? and pupa? ; 

 secondly, that the larva of L. avion has a well-developed honey-gland, 

 w T hich, though not strictly speaking protective, is very likely to make 

 its presence at least tolerated, and probably even welcome, in the nest ; 

 thirdly, that even if the larva is carnivorous, it does not necessarily 

 follow that the larva? or pupa? of the ants form its pabulum, as it may 

 feed upon one or more of the other insects that inhabit ants' nests, 

 though this is perhaps less likefy, and there is certainly so far no proof, 

 or even indication, of it ; and fourthly, that it is not certain that the 

 ants never do, after hybernation, resent the intrusion of the stranger, 

 since the conduct of the ants towards the larva? has only been tested 

 before that period, at a time when, if my hypothesis be correct, the 

 carnivorous instinct is still latent. Indeed, either of the two last- 

 mentioned proclivities would help to account for the usually very 

 sporadic appearance of the insect throughout its range, and if the larva? 

 feed on any other myrmecophilous insect, the survival of the species 

 in any locality would be dependent on the existance of that particular 

 insect in the identical nests into which the avion larva? found their 

 way. Mr. Frohawk has, with great fairness, pointed out in his last 

 communication to the Entomolof/ist (lot: cit., p. 322) that the small 

 size of the larva? when they leave off feeding, and their close colour- 

 resemblance to their surroundings may account for some of them 

 having been overlooked in ants' nests before hybernation, and the 

 fact that all the pupa? found by Mr. Percy Richards {Knt., xli., p. 

 183) were not only under stones in the ant's nest, but were also 

 enclosed in earthen cells of the size of the pupa?, may readily account 

 for the previous lack of success in finding the latter. Mr. Richards 

 suggests that these cells may have been made by the ants ; ib is 

 possible, and there is a way of accounting for it which may not be too 

 far-fetched ; possibly the larvse* hitherto tolerated, or welcomed, on 

 account of their honey-gland, may instinctively attempt to get out of 

 danger from the ants before making the pupal change, either by 

 making their way close to the surface, as in the case of those found 

 by Mr. Frohawk and Mr. Rayward, or by biding under stones, as in 

 the case of those found by Mr. Richards. In the latter event, they 

 may escape detection during the short time the pupal shell requires 

 for hardening, and the ants, regarding the now useless pupa as an 

 encumbrance may enclose it in the earthen cell, finding it too hard to 





