39 



autumn are these periods. In summer we have found 

 it sometimes impracticable to determine their ascent : 

 they have detached themselves, after several vibrations, 

 and fallen to the ground. On one day (May, 1823,) this 

 remarkable fact was proved in the case of numbers. 

 The insect seems to be sufficiently aware when the 

 threads are buoyant, and perhaps the temporary sus- 

 pension in the horizontal plane may communicate this 

 information: aeronautic spiders make their appearance 

 early in the season. 



Several circumstances concur to show the pheno- 

 menon of ascent to be electric : the propelled threads 

 do not interfere with each other ; they are divellent, 

 and this divergence seems to proceed from their being 

 imbued with similar electricity; and the character of 

 that electricity appeared to us to be an interesting 

 subject for subsequent research. When a conductor 

 is brought near the thread by which it suspends 

 itself, but, above all, to the flocculi or balls, they are 

 considerably deflected from the perpendicular, and 

 the horizontal fibre is attracted by the point: when 

 a stick of excited sealing-wax was brought near the 

 thread of suspension, it seemed to be repelled; con- 

 sequently the electricity of the thread is negative. 

 The descent of the thread is instantly determined by 

 bringing over it the excited sealing-wax. On the 3d of 

 July, 1322, at 4> P. M., thermometer 66 F., when two 

 aeronautic spiders, on separate threads, were brought 

 near to each other, a mutual repulsion supervened. 



In one experiment made, the ascent of the insect 

 was so slow and tranquil, from the humidity of the 

 lower atmosphere and wetness of the terrestrial sur- 

 face, that I could easily catch it by following its pro- 

 gress : it moved in a plane parallel to the point of 

 departure. On the 4-th August, 1822, at 3 P. M., ther- 

 mometer 66, the ascent was slow and beautiful, the 



