ANT-LIONS 



was thrown up precisely as when the trap was made, 

 more or less violently, and so vigorously at times that 

 it appeared to boil. This agitated the surrounding 

 particles, which began to move towards the centre; 

 and these setting in motion those above and yet above, 

 at last withdrew those beneath the feet of the ant, which 

 slid along with the tiny sand avalanche into the apex. 

 There it was seized, unless, as sometimes occurred, it 

 was fortunate enough to escape. 



The ants often showed a strange fascination for the 

 pit even after they had escaped therefrom. A large 

 black carpenter ant was trapped and seized, but man- 

 aged to get loose and rushed out of the pit. Then it 

 paced excitedly around it, ran into it, and so in and 

 out and around again and again, as though dazed. 

 Meanwhile, Myrmeleon, no less agitated than the big 

 carpenter, kept up a furious bombardment that fairly 

 filled its pit with flying sand pellets. These were tossed 

 up with force enough to cast them out of the bowl over 

 the brim to the distance of several inches on the table — 

 that is, from thirty to forty times the grub's own length 

 and many times its height. Pellets as large as grains 

 of rice were thus ejected. But they flew in all direc- 

 tions, on the side opposite to the ant or upon it or 

 around it equally and indiscriminately. 



My conclusion was, from many observations, that 

 there was no direct aim, simply the usual and general 

 instinctive movement, which, of course, was apt to 

 reach anything within its compass. In short, the volley- 

 ing was at haphazard, but as it was in all directions it 

 was likely sometimes to hit the object of attack. But 

 this seemed to be accidental, not of set purpose. In 

 other words, the larva's weapon of offence was a vortex, 



135 



