446 BRITISH BUTTERFLIES. 



the ivy, especially at the tips of new shoots, where they were milking 

 black aphides (Raynor, Ent. Rec, xviii., p. 299). Edwards had, however, 

 long before worked out in detail the connection between no fewer than 

 four species of ants and the larva of the American form of this species. 

 He states (Butts. Nth. America, ii., Lye, pp. 10-12) that he observed, 

 in 1877, the flower- spikes of Cimicifaga racemosa, on which larvae of 

 0. pseudargiolus were feeding, were much frequented by ants. It was 

 soon evident that the ants were attracted by the larvae, for they 

 caressed them with their antennae, running up and down their backs, 

 the larvae in no way resenting this familiarity, not even withdrawing 

 their heads from the buds they were excavating ; the ants seemed 

 especially to linger about the last segments, particularly the 11th (7th 

 abdominal). Lintner examined the larvae and found the processes on the 

 8th segment, whilst Mack found in addition the honey-gland on the 

 dorsum of the 7th abdominal. Between 1878 and 1883, observation was 

 almost continuously kept on the larvae, and it was noted that when 

 ants were discovered on a stem, they were almost invariably on or 



near the larvae At first only two species of ants were 



observed, of medium size, but, later, Edwards noted that at least four, 

 of which one is a very small species, not more than one-eighth of an 

 inch long, accompanied these larvae, operating in the same manner, and, 

 in one case, six of these were busy over one larva, but the movements 



of all the species were similar They run over the body, 



caressing incessantly w T ith the antennae, and undoubtedly with the 



object of persuading the larva to emit the fluid Much of 



the caressing is done about the anterior segments, and while the ants 

 are .... absent from the last segments, the tubes .... are 

 almost constantly exposed to their full extent, and so remain without 

 retracting, until the ants come tumbling along in great excitement, 

 and put either foot or antenna directly upon, or close by, the tubes, 

 when these are instantly withdrawn. The ants pay no heed to the 

 tubes so far as touching them with intention, but at once turn to the 

 median honey-gland, caress the back of that segment, put their mouths 

 to the orifice, and show every sign of eager expectancy. With a lens, 

 a movement becomes speedily apparent, and there protrudes a dark 

 green mammilloid membrane, from the top of which exudes a tiny 

 drop of clear green fluid. This the ants drink greedily, two or three 

 of them perhaps standing guard over it. The demonstrations of the 

 ants are of the most gentle nature, caressing, entreating, and, as the 

 little creatures drink in the fluid, lifting their heads, as if to prolong 

 the swallowing, there is a manifest satisfaction and delectation that 

 is amusing to see. They lick away the last trace and stroke the back 

 of the segment, and wait to see if their coaxing avails anything. If 

 not, they run about . . . . but presently return, and the 

 caressings go on as before. The intervals between the appearance of 

 the globule vary with the condition of the larva. If exhausted by 

 yielding to the frequent solicitations, some minutes may elapse, and 

 the tubes meanwhile will remain concealed ; but a fresh larva requires 

 little urging ; and the mere intimation of the presence of an ant in 

 the vicinity is enough to cause the tubes to play rapidly, and one 

 globule to follow another, sometimes without a retracting of the 

 membrane, and before the near approach of the ants. I have counted 

 six emissions in 75 seconds. The tubes are usually expanded when 



