488 Mr. W. A. Lamborn on the 
26. Cupido (Catochrysops) malathana, Boisd. 
The associated ant was Camponotus akwapimensis v. 
poultont. Also, in the house, Pheidole rotundata, var. 
No. 581. ¢. The following note refers to the larva of 
the specimen :— 
*‘ Jan. 7-8, 1912. On Jan. 5 I discovered a small 
Lycaenid larva which had eaten an oval hole in the side of 
a seed-pod and was half inside and half out of it. The 
seed-pod belonged to a Leguminous climbing plant with a 
flower very like that of our French bean. 
“ Five § min. and 1 § maj. C. akwapimensis poultoni were 
running to and fro on the pod, the centre of interest being 
the larva. On the dorsal aspect of the visible half of its 
body was a small dull-red area over which an ant was 
standing head down and holding his ground all the time 
against the attempts of others to come there. I secured 
all these ants and they are now sent. On arriving home I 
discovered a drop of fluid rather larger than a pin’s head 
on this patch. I shut the larva up closely and left it for 
the night. On the same day I had found two Oboronia 
pupae in a head of the plant Costus afer, and had left them 
on my verandah table. I found next morning that the 
small black ‘ house-ants,’ P. rotundata, var., had made an 
attempt to cover them with bits of wool, excreta of cater- 
pillars and tiny fragments of wood, so I took away the 
pupae, and the ants scattered over the table. I then 
opened the box containing the Catochrysops larva which 
had left the seed-pod, and the ants very soon swarmed 
over it. I removed all the ants save three so that I was 
then able to watch the proceedings without difficulty. 
“ The ants undoubtedly obtained food-material from the 
red patch, but behind and to its outer side I observed, on 
either side, a little white tubercle which was protruded and 
drawn in again repeatedly, and with such great rapidity 
that one could hardly count its movements. I then took 
away two more ants. The remaining one fed at the red 
patch, and now that the other ants were no longer running 
over the larva the two tubercles remained protruded. 
Presently the ant made a rush at one of the tubercles which 
was promptly withdrawn. It then ran over to the second 
tubercle. This was withdrawn in turn, but the first 
tubercle was thrust out again and incited the ant to rush 
back a fraction of a second too late to secure anything, for 
