244 NATURALISTS CABINET. 



Experiments by M. Poupart. 



ready to eal when food is offered to it. It al- 

 ways appears starved and small when kept thus; 

 and if a fly be given to it in that hungry state, 

 it will suck out all its juices so perfectly that the 

 remaining shell may be rubbed to powder be- 

 tween the fingers, whilst the body of the crea- 

 ture that has sucked it appears remarkably 

 swelled and distended. 



M. Poupart put an ant-lion, for the sake of 

 experiment, into a wooden box with some sand, 

 and covered it \vith a glass, so as tb exclude 

 every other insect. Here it formed its cone, and 

 \vatched as usual for prey, though in vain. Thus 

 he kept it for several months, while in an adjoin- 

 ing box he kept another of the same species, 

 which he supplied with food by giving it ants 

 and flies pretty regularly. He could perceive 

 no difference between the movements or actions 

 of the two; but, when he took them from their 

 holes, he found the abdomen of that which had 

 received no food was shrunk to a very diminutive 

 size, whilst the other retained its proper shape. 



If the ant-lion form its pit in a bed of pure 

 sand, it is made and repaired with great ease ; 

 but, where it meets with other substances among 

 the sand, that labour becomes much more em- 

 barrassing. If, for instance, when the creature 

 has half formed it, it comes to a stone of some 

 moderate size, it does not desert the work on 

 this account, but goes on, intending to remove 

 that impediment the Jast. When the pit is fi- 



