32 BRITISH BUTTERFLIES. 



ants to be confined to the summer broods of caterpillars of Cyaniris 

 pseudargiolus, and, even then, to those on Cimicifuga, and suspects that 

 the larvae feeding on Comus or Actinomeris cannot exude so sweet 

 a fluid, the flower of Cimicifuga being of exceeding sweetness, 

 whilst that of Actinomeris is bitter to the taste. 



Saunders notes [Can. Ent., x., p. 14) that the larvae of Rusticus 

 scudderii (closely allied to the European Plebeius argus) are accompanied 

 by ants, and that the discovery of the larvae was made comparatively 

 easy from the invariable presence of these active attendants. The 

 ants were observed actively running about the leaves on which these 

 caterpillars were found, and repeatedly over the caterpillars themselves, 

 which did not seem in the least disturbed by them. 



Edwards also gives a most interesting account of the connec- 

 tion between ants and the larva of another Lycaenid, Rusticus melissa 

 (Papilio, iv., pp. 92-3). He notes that, on June 9th, he introduced 

 a small ant to a larva of this species, which was confined in a 

 glass tube. The ant soon discovered the larva and ran about 

 it in great excitement, caressing it with its antennae. Immediately 

 the tubes, not hitherto seen, began to play, and one or the other, or 

 both together, were exposed for some minutes, and, indeed, so long as 

 the ant was near. Sometimes the tubes were fully protruded, with 

 the tentacles expanded, at other times they were partially withdrawn, in 

 that case coming together in a pencil just as has been observed in 

 C. pseudargiolus. The ant always ended its caresses by putting its mouth 

 to the orifice of the gland on the 7th abdominal segment ; and, by its 

 motions, evidently found the fluid it sought. Next day, two ants of a 

 larger species were turned in at the same time ; they ran about the 

 glass, alarmed at finding themselves in confinement, and accidentally 

 one soon touched the larva ; at once a drop of green fluid bubbled out 

 of the orifice before the tubes made any movement. The ant saw it, 

 rushed at it, and then the tubes began to play although they had been 

 quiet for fully five minutes before ; they now played intermittently for 

 two or three minutes, the tentacles fully expanding and then partly 

 retracting. The ants drank of the drops four times and then 

 desisted, running about the glass again ; then they were liberated, and 

 a small ant of the species experimented with the preceding day, was 

 introduced ; almost at once it found the larva, caressed it gently, and 

 was favoured with the coveted nectar, the tubes being all the time in 

 motion. On June 12th, the larva now T being mature, another ant was 

 introduced ; as usual, as soon as the manipulations began, the tubes 

 commenced to play, and, presently, a large drop issued ; in ten seconds 

 another followed, but for some time after there was no more, though 

 the ant begged urgently for it. The ant left the orifice, ran up and 

 down the body of the larva, caressing the anterior segments, and then 

 returned to the orifice and begged again ; this was repeated several 

 times, but the larva was obdurate, probably it was exhausted, being 

 near pupation. The solicitations are made by the antenna 1 alone, which 

 fly about drumming here there and everywhere, the ant manifesting 

 great excitement. Edwards notes that he was observing larvae of Cyaniris 

 pseudargiolus at the same time, and the behaviour of the two species 

 was identical. 



Niceville observes that many Lycaenid larvae in India are provided 

 with an oval opening on the dorsal line of the 7th abdominal 



