80 BRITISH BUTTERFLIES. 



and Anderson and Spry, in their little book "Victorian Butterflies," 

 note that Holochila absimilis, Hypochrysops delicia, Ogyris abrota, 

 lalmenus evagoras, I. ictinus, etc., are all accompanied by ants, and these 

 authors remark that the ants that are affiliated to the last-named 

 species are particularly large and fierce. [See also preceding vol., 

 pp. 35-36.] One wonders whether the highly-developed gregarious 

 habit, exhibited by the larvae of lalmenus evagoras (and Indian 

 Thaduka multicandata), has been evolved for the purpose of mutual 

 protection, due to excessive sluggishness, or because of the ease with 

 which the swarms of little black ants that live with them can overlook 

 them. Further observation is needed on this subject. 



In Europe, as already noted, the connection between Lycaenid 

 larvae and ants has been known for at least 130 years, but the presence 

 of the honey-gland on the 7th abdominal segment was not demon- 

 strated till 1867. Even then Guenee did not connect his anatomical 

 discovery with the recorded observations of the symbiotic habit 

 existing between the ants and larvae. Edwards' discovery, however, in 

 1877, connected the two, and observations on the habits of Plebeius 

 aegon, P. argyrognomon, Agriades corydon, A. bellargus, Polyommatus 

 icarus, P. clamon, P. hylas, Aricia astrarche, Nomiades tolas, Lycaena 

 avion, Celasirina argiolus, and other species, have been detailed at 

 length. (References to these are given, postea pp. 323-4.) 



These glandular structures appear to be present in all species that 

 we unhesitatingly regard as Lycaenine, and equally wanting in all 

 typical Theclids and Chrysophanids, so that it is hardly going too far 

 to postulate that any larvae possessing these structures are Lycae- 

 nines, those without them, though with less certainty, belong to some 

 other group. Thus we may presume Curetis to be a Lycaenine 

 derivative, and the association with ants, though of so different a 

 character, may even suggest a place here for Liphyra rather than 

 with the Theclids. 



Chapman notes that, in connection with the apathy shown by the 

 larvae when the ants are running over them, i.e., apathy, not as 

 regards paying the ants for their services, but as concerns any show 

 of inconvenience due to the sharp claw^s of the ants on their delicate 

 cuticle, it is not at all improbable that the remarkably complex 

 structure of the hairs with their stellate bases, and their abundance, 

 both most notable in the later instars when their size requires that 

 the ants should crawl over them, are an adaptation of a protective 

 character. The claw T s of the ants would obtain a good purchase 

 amongst these hairs, and would be held away from the actual skin, so 

 that their sharp tips would rarely, if at all, touch it. 



How far the suggestion made by various observers, to the effect 

 that ants accompany certain Lycaenid larvae for the purpose of keeping 

 clean their burrows by devouring their frass, is to be accepted, we do 

 not know. Graves claims that, in Egypt, ants accompany the larvae of 

 Hypolycaena lima for this purpose (Ent. Bee, xvi., p. 203) as also 

 those of Tarucus theophrastiis (Ent. Rec, xvi., p. 19), whilst Taylor 

 writes (de Nicer. Butts, of India, iii., p. 479) : "The larvae of Virachola 

 isocrates are attended by the ant, Formica nigra, w T hich clear away 

 their droppings and act as sweepers, as well as guard their pupae." 

 This matter certainly wants further elucidation. Further informa- 

 tion on the connection between Lycaenid larvae and ants may be found 

 in our preceding volume, pp. 30 et seq. 



