( clvi ) 



that has been bred, certainly the first from larvae taken in 

 the autumn, and made the following observations : — 



The life-history is interesting as parallel to, but differing 

 from, that of Lycaena arion. The young larva feeds in the 

 autumn in the flowers and other portions of Gentlana pneumo- 

 nanthe, and probably of other Gentians. So far it is exactly 

 parallel in growth to other Blues, such as many of our Plebeiids 

 that pass the winter in the third instar; when it reaches 

 the third instar it leaves the plant, wanders off, and, hitherto, 

 efforts to carry it further have failed . At this point it agrees 

 with L. arion in habits, but it is not like L. arion, which is 

 in a remarkably modified and concentrated (as regards skin 

 armature) fourth instar, but is in quite an ordinary third 

 instar. In its plant life it has differed also in that several, 

 often five or six, larvae live amicably in one flower, whereas 

 L. arion is solitary, and if by any accident two larvae meet, 

 as by a second egg having been laid on the same flower-head, 

 or especially when incautiously associated in captivity, they 

 are inveterate cannibals. 



The remaining history is that both arion and alcon live in 

 the nests of ants — I kept both species in those of Myrmica 

 scabrinodis — and pupate in the nest (re L. arion, teste Capt. 

 E. B. Purefoy). The differences are that L. arion eats the 

 ant brood, whereas L. alcon certainly sucks their juices 

 without eating them, and I could not prove that it ever 

 actually ate them, though I thought it did so when past the 

 winter, when its food would more often be ant pupae. 



L. arion is in its fourth instar, and provided with a skin 

 armature not unsuitable to it when full grown. L. alcon has 

 only a third-instar armature, and when full grown might 

 almost be described as without one, for, like L. arion, it docs 

 not moult after entering the ants' nest, but attains its lull 

 growth still in the third instar. The skin is then so attenuated 

 that the fat-masses are very obvious, and its general aspect 

 is like that of an internal feeder rather than that of a butterfly. 

 It is to be noted as exceedingly remarkable that a butterfly 

 larva should attain its full growth after only two moults. 



Monsieur Oberthur, who is familial' with localities where 

 L. alcon is common, and considered it highly probable that it 



