on some Carnivorous Insects. 323 



down, followed it, also in sections. By this time the ants 

 had entirely abandoned the four Mylahris and the three 

 Epilachna. One of the former was on its back motionless, 

 another was remaining quite still — head held down — in 

 the centre of the stream of ants ; many of them felt it with 

 their antennae as they passed, but no attack was made 

 on it, and the other two as well as the three Epilachnas 

 were calmly marching off. They drew attacks in doing 

 so from the ants they passed, and the two Mylahris in 

 particular were occasionally overwhelmed for a moment 

 and perhaps dragged back a little. But they were always 

 abandoned again after a time and they made good their 

 escape. I took them up and found that both Mylahris 

 were still able to expel juice fairly freely, the Coccinellidae 

 (with one exception, not, however, so freshly caught 

 when given), apparently not. The Zonocerus continued 

 to be the centre of attraction for large numbers of ants, 

 but they were making very little headway in the process 

 of its dismemberment. 



I now inserted four larvae, one of Papilio demodocus 

 at about the end of the brown-and-white stage, one of 

 Amauris alhimaculata (about ^ grown) and two of Acraea 

 acara {\ and \ grown respectively). All were at once sur- 

 rounded, but the attack on the Amauris larva was not of 

 a very formidable nature : the ants used their antennae 

 rather than their jaws and evidently disliked it intensely, 

 for they very speedily abandoned it entirely. In making 

 off, which it did apparently quite undamaged, it drew a 

 few attacks, but these proved to be no more formidable 

 than the first, and the larva was each time quickly aban- 

 doned. The A. acara larvae were more seriously attacked, 

 but the ants had the greatest difficulty in getting " in " 

 at them, the much-branching spines beaded at all the 

 extremities with the yellow protective juice everywhere 

 barring their way. A number of these spines were in 

 each case shorn by the ants (some close oft' to the skin) 

 during the attack, but they finally abandoned both larvae, 

 and, though I frequently replaced both these larvae and 

 that of the Amauris in the centre of the ant-column 

 during the remainder of the experiment, the subsequent 

 attacks, when any were made at all, were less serious 

 than the first, and the larvae were each time allowed to 

 escape. Very different was it with that of the Pajiilio. 

 It at once extended its red filaments on being attacked, 



