February, 1914.] 25 



111 two cases an ant was seen actually carrying or trying to carry 

 a larva much larger and many times heavier than itself. In one case, 

 before the plant was in any way disturbed, a larva was seen advancing 

 rapidly up the stem of a leaf, much more quickly than the ordinary 

 pace of a Lyctena larva. A closer scrutiny showed that it was actually 

 being carried or dragged by an ant. In the other case, after some 

 interference with the plant, an ant was found to be trying to drag a 

 larva underground and to some extent succeeding. Much the same 

 process they employ with their own larvae and pupse when a nest is 

 disturbed. Many of the larvae down near the root- stock were well 

 covered by the loose material piled up by the ants. How far this con- 

 dition was due to the larvae burying themselves, how far to their being 

 guided by the ants, and how far to the ants piling the material over 

 them, could only be settled by much longer observation than was avail- 

 able, but I should fancy that each of these processes had its share in 

 the result. It is certain that the larvae could not have buried themselves 

 so deeply and so comfortably, if at all, on the unaffected plants. There 

 were usually several, and even half-a-dozen or more larvae to a plant, and I 

 could not help asking myself the question which had previously occurred 

 to me in regard to the larvae of corydon on Hippocrepis : did the butter- 

 flies lay eggs only on these affected plants ? Supposing they laid on 

 all the plants fairly impartially, were larvae absent from the unaffected 

 plants because they had perished for want of ant protection, or, as I 

 fancied to be more likely, because the ants collected the young larvae 

 and carried them to the conveniently placed plants ? 



Two pupae (certainly early ones, looking to the abundance of 

 larvae) were found : one attached to a bit of dead stem, nearly an inch 

 below the surface, the superincumbent material being the lightly and 

 loosely placed coarse sand mixed with vegetable debris piled up by the 

 ants ; the other was very similarly placed, but was less disturbed in 

 the discovery, and there was found to be round the pupa an actual 

 cavity, in which were also some ants and^two half -grown argyrognomon 

 larvae. 



On the Orwbrychis plants here there also occurred two larvae 

 of damon. Argyrognomon were also found on Helianthemum, on 

 Anthyllis, and also on Hvppocrepis along with those of corydon, but 

 only sparingly on each of these plants. 



Larvae placed on Hippocrepis and on Helianthemum ate readily in 

 both cases. They were placed on Antliyllis only along with sainfoin, and 

 would not touch it with this choice before them. Larvae fed on Hippo- 



