68 BRITISH BUTTERFLIES. 



supposed to be — and very probably are — scent-organs to attract the ant 

 to the gland, it was noticed during these experiments that they were 

 most active when the larva appeared to resent the attempts of the ant 

 to obtain fluid from the gland, as was sometimes the case. At this 

 time the gland was contracted and withdrawn below the surrounding 

 surface of the segment, and the rapid erection and withdrawal of the 

 tubercles generally resulted in' momentarily distracting the ant's 

 attention, causing it to leave the gland, to which, however, it usually 

 quickly returned. He further observes (in litt.) that, in its later stages, 

 the larva of Ayriades coridon prefers to rest in the daytime. on a leaf- 

 stalk at about an inch above the surface of the ground, and that, when 

 thus discovered, the larva has generally been accompanied by ants. 

 He also states that, on the afternoon of May 11th, 1907, he found four 

 little larvae thus resting on a small isolated plant of Hippocrepis comosa 

 growing on the top of a small nest of Lasius jiavus, and so many ants 

 were busy running about and over them, that his attention was at 

 once attracted. On the other hand, Prideaux states (in litt.) that he 

 found larvae common on Eeigate Hill in June, 1895, but quite un- 

 attended by ants. Wood notes [in litt.) that, in his experience, wild 

 larvae are only attended by Formica rufa, although Lasius flavits is also 

 to be found in the spots they frequent. Sometimes as many as five or 

 six F. rufa are to be found in attendance on one larva, whilst, in one 

 instance last summer, no fewer than twelve were battling on the body 

 of one ; they were so close together that the colour of the larva could 

 not be seen, and the whole looked simply a slug-shaped mass of jostling 

 ants. Donisthorpe writes (Fnt. Record, xix., p. 256) that, having two 

 larvae of A. coridon, he placed them into plaster-nests with glass tops, 

 and introduced separately six or eight examples of each of the follow- 

 ing ants to them, viz., Formica rufa, F. sanyuinea, F. exsecta, F. fusca, 

 and Lasius fuliginosus. One Formica sanguinea seized a larva and 

 dragged it along a little way, the caterpillar remaining quite rigid, but 

 soon dropped it, and all the ants mostly sat on the larva at different 

 times, tapping the posterior part of the body with their antennae. He 

 adds that he " left the caterpillars with a species of ant for a week at 

 a time, and put one into a h\ rufa nest itself, where it crawled 

 about among hundreds of ants without coming to any harm." Krodel 

 states (All;/. Zeits. fur Fnt., ix., pp. 104 et seq.) that, at Wiirzburg, in 

 the Main valley, he always found the larvae by day exclusively under 

 stones in the neighbourhood of //. comosa, and in all cases in the 

 company of ants. Stones which, on being turned over, showed no 

 ants, afforded no larvae of A. coridon cither. As a rule it was tin* 

 common Formica nigra, more rarely F. rufa, which sought the A. 

 coridon lame ; with the small yellow F. jlara (Lasius Jiavus), he never 

 found any. 



Larva. — First instar: When newly-hatched the larva is of a 



* An ovum opened on January 4th, 1908, disclosed an almost fully developed 

 larva about lmm. long and yellowish-white in colour. The head rather dingy 

 ochreous, and the parts withdrawn within the prothoracio segment showing 

 through the dorsal surface of the latter Leaden -grey i The thoracic legs very pale 

 shining ochreous, and bearing a number of short, pointed, white hairs ; the claws 



fully Formed, ochreous, with a shade of red-brown. The abdominal legs difficult 

 In define, and probably not fully developed. The spiracles Large and conspicuous, 

 the margins imperfect and appearing as a cirole of black hooks curving inwards 

 towards the centre. The body surface covered with akin-points or granulations, 



