MIGRATIONS OF INSECTS. 281 



consisted of cavities pierced in the earth, containing 

 sufficiently spacious apartments, generally covered 

 with fragments of straw, and resembling small ant- 

 hills. We might there observe some sentinels doing 

 daily duty, that is to say, opening and closing [the 

 gates of the ant-hill morning and evening. Sometimes 

 these asylums become little colonies, which maintain 

 a close connexion with the principal ant-hill; ^they 

 are different habitations, common to the same ants, 

 serving them for places of refuge on any derange- 

 ment of what we might term their capital.'* 



The only analogous instance of a number of es- 

 tablishments formed in the vicinity of the parent 

 nest, occurs among spiders, who do not, Hke the 

 ants, live in communities, but every individual 

 forages for itself Redi and some 'recent naturalists 

 mention the experiment of confining young spiders 

 for a long time without food, and talk of their even 

 devouring the bodies of their brethren; but however 

 that may have been after their separation, we are 

 certain that it could not have happened before: at 

 least we have very often confined them together, in 

 the same box without food for weeks together, and 

 never saw anything to countenance the supposition. 

 But it is chiefly their proceedings after leaving the 

 nest which here claim our notice. From fifty to a 

 hundred or more are usually produced by one 

 mother, and as soon as they are strong enough to 

 make their way, they quit the maternal nest, to com- 

 mence war upon every insect that flies. When the 

 nests of several of the geometric spiders are placed, 

 as they often are, near the iron railings of our squares, 

 every interval may be seen filled with the nets of the 

 little emigrants, as if in their journeyings from home 

 each had appropriated to its exclusive use the half of 



* Huber on Ants, p. 166. 

 VOL. XII. 24* 



