FOOD OF INSECTS. 427 



I have described the whole process by which it excavates 

 its trap, arid the ingenious contrivances to which^it has 

 recourse. 



Its first concern is to find a soil of loose dry sand, in 

 the neighbourhood of which, indeed, its provident mo- 

 ther has previously taken care to place it, and in a shel- 

 tered spot near an old wall, or at the foot of a tree. This 

 is necessary on two accounts : the prey most acceptable 

 to it abounds there, and no other soil would suit for the 

 construction of its snare. Its next step is to trace in the 

 sand a circle, which, like the furrow with which Romulus 

 marked out the limits of his new city, is to determine the 

 extent of its future abode. This being done, it proceeds 

 to excavate the cavity by throwing out the sand in a 

 mode not less singular than effective. Placing itself in 

 the inside of the circle which it has traced, it thrusts the 

 hind part of its body under the sand, and with one of its 

 fore-legs, serving as a shovel, it charges its flat and square 

 head with a load, which it immediately throws over the 

 outside of the circle with a jerk strong enough to carry it 

 to the distance of several inches. This little manoeuvre 

 is executed with surprising promptitude and address. A 

 gardener does not operate so quickly or so well with 

 his spade and his foot, as the ant-lion with its head and 

 leg. Walking backwards, and constantly repeating the 

 process, it soon arrives at the part of the circle from 

 which it set out. It then traces a new one, excavates 

 another furrow in a similar manner, and by a repetition 

 of these operations at length arrives at the centre of its 

 cavity. One circumstance deserves remark that it 

 never loads its head with the sand lying on the outside 

 of the circle, though it would be as easy to do this with 



