212 JS^FXT ARCHITECTURE. 



one leg in this operation — the one, namely, which is 

 nearest the centre of the circle. Were it to employ 

 the others in digging av/ay the sand, it would en- 

 croach upon the regularity of its plan. Working 

 with great industry and adroitness in the manner 

 we have just described, it quickly makes the round of 

 its circle, and as its works backwards it soon arrives 

 at the point where it had commenced. Instead, 

 however, of proceeding from this point in the same 

 direction as before, it wheels about and works a 

 round in the contrary du'ection, and in this way it 

 avoids throwing all the fatigue of the labour on one 

 leg, alternating them every round of the circle. 



Ant-Lior.h Pitfalls, in an expcrbncnting-hov. 



Were there nothing to scoop out but sand or loose 

 earth, the little engineer would have only to repeat 

 the operations we have described, till it had completed 

 the whole. But it frequently happens in the course 

 of its labours, sometimes even v.'hcn they are near a 

 close, that it will meet with a stone of some size which 

 would, if suffered to remain, injure materially the 

 perfection of its trap. But such obstacles as this do 

 not prevent the insect from proceeding: on the con- 

 trary, it redoubles its assiduity to remove the ob- 

 struction, asM. Bonnet repeatedly witnessed. If the 

 stone be smal, it can manage to jerk it out in the 

 same manner as the sand ; but when it is two or 



