Habits of the larva of Lycaena arion. 317 



to hibernation in the fourth instar. As to losing the fourth 

 moult, this would present few difficulties, the most parallel 

 case I know of being that of Acronycta alni, which has lost a 

 moult as compared with its nearest relatives. The loss is 

 associated with the remarkable change from the bird-dirt- 

 like young larva to the handsome and conspicuous last-stage 

 one. Now and then an individual resumes the lost instar 

 and presents a stage intermediate between the two forms — ■ 

 obviously a form that is neither flesh nor fowl and there- 

 fore has been lost, but not so completely as not to reappear 

 sporadically. Moma orion presents in each brood a larger 

 and a smaller race, the former with one more moult.* 0. 

 antiqua and many other species present examples showing 

 that the losing or gaining of a moult is not an infrequent 

 occurrence. Orgyia antiqua has, in fact, in each brood two 

 races — one with a moult less than the other, and each with 

 a moult less in the male than in the female ; this is probably 

 correlative, along with the long period over which the eggs 

 of one brood go on hatching, with the fact that this single- 

 brooded (in England) species is on the wing from early 

 summer till late autumn, fresh specimens emerging during 

 several months.f To return to L. arion. 



We may admit that whilst still plant-feeders they were 

 less cannibalistic than now ; but, as it is a characteristic not 

 infrequent in other blues, they would take not unreadily 

 to eating ant larvae, and we may assume that a certain 

 percentage survived the difficulties of a purely animal diet 

 (Calymnia trapezina, e. g., can be reared on oak (or other) 

 leaves, or entirely on other larvae), and the dangers from 

 the ants, which would be greater just after moults. If so 

 much be granted, and it seems all very simple and straight- 

 forward, the advantages of protection during the winter 

 and an assured dietary would largely balance the dis- 

 advantages and dangers. These would be eliminated as 

 certain larvae dropped a moult and were accordingly at an 

 advantage, and assuming as possible they were first carried 

 in in the third instar, a diminution in size at moults would 

 lead to a change to fourth instar for the translation. This 

 diminution in size would be favoured, if the carnivorous 

 habit was advantageous from a nutritive aspect ; since the 

 larvae that grew least on thyme and so got earlier into the 



* Ent. Record, vol. iii, p. 220. 

 t E. M. M., vol. xxiii, p. 224. 



