458 ANNUAL EEPORT SMITHSONIAN INSTITUTION, 1925 



had no difficultj^ in retaining it by means of its second and third 

 pairs of legs, the first pair was used to take up loose fragments 

 and guide them to the mouth. He saw a loose leg thus taken up and 

 eaten, and in this way every particle of the unfortunate little " hop- 

 per " was secured. 



After the meal the legs of the first pair were drawn one after the 

 other between the mandibles and then polished on the outer side 

 of the face, after the maimer of a cat. 



He continued to watch the larva closely. It remained without 

 movement for about 20 minutes and then approached another jassid. 

 This one was evidently not satisfied as to the honest}'^ of its purpose, 

 for immediately the larva commenced to tickle it the jassid ran 

 away up the stem to a distance of about half an inch. However, 

 the larva followed on and overtook it, and in due course it shared 

 the fate of its predecessor, the series of actions by the larva being 

 precisely the same as described in the former case. 



On the following day he brought in more of the food insects, 

 and the larva first found made up for its long fast by catching 9 

 out of 11 between 8 a. m. and 3 p. m. 



The procedure was so interesting to him that he actually wit- 

 nessed the caressing, capture, and eating of most of these, and he 

 has seen it many times since. The caterpillars eat both nymphs 

 and imagos of the jassids, but seem to secure more of the former, 

 for, though they are able to jump and run with surprising activity, 

 quite as fast as an ordinary ant, the imagos sometimes escape bj' 

 flying. When imagos are eaten the hard anterior wings are usually 

 rejected. 



Mr. Lamborn remarks that in nature the slow-moving lycaenid 

 larvae must depend for their very existence on the fact that these 

 insects are gregarious and if disturbed frequently reassemble at the 

 same spot. 



If the caterpillar is on a broad surface it raises itself anteriorly 

 when grasping a victim, but when on a narrow surface it drags 

 the insect off its support by simply bending to one side or the other. 



The first two larvae found by Mr. Lamborn were feeding on the 

 jassid Nehela ornata^ an active jumping insect which they lull 

 to a false sense of security by simulating the attentions of ants. 

 The third larva found would not eat the jassid which was eaten by 

 the other two. It was found in an ant shelter on Triuinfetta cordi- 

 folia which covered a large number of little green hopping crea- 

 tures which proved to be immature membracids. Another larva was 

 found in a shelter containing membracid larvae {Gargara varie- 

 gata) . He saw this larva feeding on a large larva of the membracid 

 type, though he did not actually witness the attack. The cater- 

 pillar ate the body and then part of the head but was unable to 



