SPIDERS. 351 



of it with one of its feet, by way of security, it placed 

 its body in a vertical position, with its spinnerets ex- 

 tended outwards ; and immediately we had the pleasure 

 of seeing a thread streaming out from them several 

 feet in length, on which the little aeronaut sprung up 

 into the air. We were convinced, from what we thus 

 observed, that it was the double or bend of the thread 

 which was blown into the air ; and we assigned as a 

 reason for her previously attaching and drawing out 

 a thread from the glass, the wish to give the wind 

 a point d'appui — something upon which it might 

 have a purchase, as a mechanic would say of a lever. 

 The bend of the thread, then, on this view of the 

 matter, would be carried out by the wind,— would 

 form the point of impulsion, — and, of course, the 

 escape bridge would be an ordinary line doubled. 



Such was our conclusion, which was strongly 

 corroborated by what we subsequently found said by 

 M. Latreille— than whom no higher authority could 

 be given. " When the animal," says he, " desires to 

 cross a brook, she fixes to a tree or some other object 

 one of the ends of her first threads, in order that the 

 wind or a current of air may carry the other end 

 beyond the obstacle;"* and as one end is always 

 attached to the spinnerets, he must mean that the 

 double of the thread flies off. In his previous publi- 

 cations, however, Latreille had contented himself 

 with copying the statement of Dr. Lister. 



In order to ascertain the fact, and put. an end to all 

 doubts, we watched, with great care and minuteness, 

 the proceedings of the long-bodied spider above men- 

 tioned, by producing a stream of air in the same 

 manner, as it perambulated the brim of the glass. 

 It immediately, as the other had done, attached a 



* « L'un des bouts de ces premiers His, afin <rae le vent 



ou un courant d'air pousse l'autre extremite de l'un d eux au de 

 U de 1' obstacle."— Diet. Classique d'Hist. Nat., vol. 1. p. 510. 



x 2 



