NOTES ON THE LIFE-IIISTORY OF LYC/RNA ARION. 323 



made, it is quite conceivable that some might have been over- 

 looked. If further, as seems probable on the analogy of Cupido 

 minima, which it much resembles in the larval state, the larva of 

 arion should feed up in the autumn and hibernate as a fully fed 

 larva, it would possibly be absent from the nest in the spring 

 when careful searches were also ^made. While the negative 

 result of these searches is thus explicable on Dr. Chapman's hypo- 

 thesis, it must necessarily, as was first pointed out by the Hon. 

 N. Charles Eothschild,* militate against it, and it must also be 

 noted that this hypothesis fails to account in any way for the 

 avidity of the larva for green peas. 



For the purpose of testing the rival hypothesis the author 

 was able to secure, through the kindness of Mr. H. St. J. 

 Donisthorpe, a number of living colonies of the yellow ant 

 (L. flavus) in observation nests. These observation nests are 

 ingenious contrivances by which one is able to study in close 

 detail colonies of ants living under circumstances closely re- 

 sembling their natural condition. Incidentally, it may be 

 mentioned here that the individuals composing a colony of 

 L. flavus hibernate during the winter. They do not feed at all, 

 nor is any food stored in the nest at this season. When a larva 

 which had completed its third moult was placed in an observation 

 nest containing L. flavus, it wandered from partition to partition 

 as if in search of something. It was occasionally milked by the 

 ants, but otherwise they took no notice of it, and it ate nothing 

 except a little honey which had been placed in the nest for the 

 ants. On the other hand, if a fresh green pea or a scarlet runner 

 bean was ojBfered to the larva it at once bored into it and com- 

 menced to feed. This experiment, which was repeated a number 

 of times without any material variation in the results, seems to 

 the author to prove conclusively that the larva of arion does not 

 feed on the ova, larvae, pupaB or imagines of the ants, nor on 

 their ejecta or excreta. It also proves that the ants do not feed 

 it, though they might of course procure food for it. 



In another experiment, a portion of an ants' nest was placed 

 in a glass cylinder with a little vegetation on the top. The 

 colony inside was small, but appeared to be otherwise in a 

 normally flourishing condition. A larva of arion which had just 

 completed its third moult was then placed on the surface of the 

 nest and carefully observed. The larva gave no indication of a 

 desire to burrow into the nest or to approach the ants. It 

 roamed about for a while, then rested beneath a small fragment 

 of a grass root and finally succumbed. This experiment also 

 appears to furnish convincing evidence that the larva does not 

 normally live within ants' nests. 



It occurred to the author as an alternative explanation that 



■'■ Ent. Rec. xxiii. p. 40. 



2 c 2 



