outline of the Life History of Lycaena arion. 303 



went directly towards the opening to the next (more central) 

 compartment, though it is, of course, possible the route 

 followed by the ants had left some impression of the correct 

 road. It finally got into the next compartment and ad- 

 vanced some 20 mm. therein. This journey did not take 

 very long, and in the course of it various ants walked past 

 it and over it, paying it no attention. At length one ant 

 seemed interested, examined it, went round it and examined 

 it with its antennae with care, the process occupying 

 several minutes ; at length it addressed itself to the honey- 

 gland region in the orthodox way, standing behind and 

 tapping the sides of the larva with its antennae, then 

 passing its mouth over the last segments of the larva nearly 

 everywhere, as if expecting something not very clear to it, 

 and finally seemed to steady at the honey-gland, which it 

 had previously passed over unavailingly, but which now 

 obviously afforded something. On a further dealing with 

 the larva, the latter bunched itself up in an attitude I had 

 not previously seen ; the ant then let it alone, but returning, 

 the ant antennaed the larva variously, and the latter again 

 bunched itself up — the head much retracted underneath, 

 the thoracic segments swollen up, and the segments behind 

 very attenuated, giving the larva a decided approach to 

 the well-known outline of a Buprestid larva. I had never 

 seen any approach to this form before, either in this or any 

 other Lycaenid larva. In a few seconds, whilst I was 

 marvelling over the matter, the ant passed its jaws over it 

 in various directions, and seemed quickly to find the right 

 place, picked it up by somewhere about the second ab- 

 dominal segment, directly over the dorsum, and, the larva 

 remaining in its curious attitude, carried it, as it would 

 an ant larva or pupa, right away to the inhabited portion 

 of the nest. The other larva wandered about a little, like 

 the other unnoticed by the ants ; but whilst I have been 

 writing this note it has disappeared, and as the distance to 

 cover was more than it could cover in the interval, it must 

 have been carried off like the first one. 



Later I placed a larva in a plaster nest of M. scabrinodis ; 

 after some minutes of neglect, the larva was in the outer 

 chamber, an ant became much interested, and milked the 

 larva over and over again. The process was curious : the 

 larva would be walking along and the ant examining it, 

 then the ant specially attended to the honey region, and, 

 stationed \isually behind the larva, it tapped it towards 



