38 BRITISH BUTTERFLIES. 



first being cylindrical, and later slug-shaped, the head retractile. 

 The larvae feed on the aphides, pressing them against the plant with 

 head and forelegs, sometimes holding them in the forelegs quite away 

 from the plants ; a few bites dispose of an aphis, and the larva then 

 licks and cleans its legs, just as a mantis does ; some aphides must 

 have a better flavour than others, as the larvae pick and choose, moving 

 their heads up and down over the backs of the insects, evidently smelling 

 them ; as a rule, the creatures seem to make little attempt to escape 

 till they are actually bitten, when struggling is useless. When not 

 engaged in feeding, the larvae rest among the aphides, or crawl about 

 between, or over, them, and the aphides do likewise, the larvae sometimes 

 covered with them. The larvae have been observed to feed only 

 on two kinds of aphides, one slate-coloured with white efflorescence, 

 the other greenish, with four dark green patches, some of them being 

 fringed with white, probably moulted skin. The larval state lasts about 

 fifteen days, and it was reckoned on the average, from first to last, 

 that a larva ate some twenty aphides per day, but it would require 

 many larvae to make much impression on the crowds of aphides one 

 sees, for often a yard of bamboo stem, two or three inches in diameter, 

 is absolutely covered with these insects. One of the features of 

 this larva is the calm way it moves about among the aphides and 

 selects its prey, and the indifference with which the latter apparently 

 accepts its fate. Although Kershaw notes that the aphides are overrun 

 by a host of ants of the two species, Polyrrhachis dives and Dolichoderus 

 bituberculatus, he notes no signs of enmity between the ants and larvae, 

 possibly because of the abundance of the aphides. 



Holland records (Can. Ent., xix., pp. 61-62) the receipt of a large 

 $ of Liphyra brassolis, captured in Penang, upon which was, at the 

 time of capture, a quantity of fluff-like mildew, particularly thick on 

 the abdomen and underside of the wings. This fluff was proved to be 

 the mealy covering of certain "shield lice," a few specimens of a large 

 species of which were in the same consignment. Holland at once 

 concluded that — (1) The captor had caught the specimen of Liphyra 

 near a colony of scale insects, which were so large as to attract 

 attention, and lead him to put a few into papers. (2) This £ was 

 engaged in oviposition just before she was captured, and that (3) the 

 mealy-white deposit described as " fluff," and which was compared 

 with mould or mildew, was nothing less than the fragments of the 

 white covering of the scale-insects over and among which the 

 butterfly had been flying whilst engaged in the act of laying her eggs. 

 Dodd observes (Ent., xxxv., pp. 153 et seq.) that, in July, 1900, he 

 noted a $ of Liphyra brassolis depositing eggs upon a tree in complete 

 possession of the wonderfully interesting green tree-ant, (Ecophylla 

 smaragdina, which exists in vast numbers in Tasmania, on the coast 

 and mountain scrubs. Upon this tree were several nests of the ants, 

 and several eggs were deposited singly on the tree, on the underside of 

 branches, or protected side of the trunk. Searching the ants' nests for 

 larvae was unsuccessful, until a half-grown larva was found accidentally, 

 whilst searching an ants' nest for other insects. After this, other larvae 

 were found, the lozenge- shaped body peculiarly flattened, the head, 

 legs, and claspers, in a groove, the edges of the body closing down 

 tightly all round, except during progression, when the body is raised 

 a little, On one occasion a larva was seen to deliberately seize a half- 



