382 Insect Architecture. 



in the air by another line. When the current of air also is 

 strong, she will sometimes commit herself to it by swinging 

 from the end of the line. We have even remarked this when 

 there was scarcely a breath of air. 



We tried another experiment. We pressed pretty firmly 

 upon the base of the spinnerets, so as not to injure the 

 spider, blowing obliquely over them ; but no floating line 

 appeared. We then touched them with a pencil and drew 

 out several lines an inch or two in length, upon which we 

 blew in order to extend them; but in this also we were 

 unsuccessful, as they did not lengthen more than a quarter 

 of an inch. We next traced out the reservoirs of a garden- 

 spider (Epeira diadema), and immediately taking a drop of 

 the matter from one of them on the point of a fine needle, 

 we directed upon it a strong current of air, and succeeded 

 in blowing out a thick yellow line, as we might have done 

 with gum-water, of about an inch and a half long. 



When we observed our long-bodied spider eager to 

 throw a line by raising up its body, we brought within 

 three inches of its spinnerets an excited stick of sealing- 

 wax, of which it took no notice, nor did any thread 

 extend to it, not even when brought almost to touch the 

 spinnerets. We had the same want of success with an 

 excited glass rod ; and indeed we had not anticipated any 

 other result, as we have never observed that these either 

 attract or repel the floating threads, as Mr. Murray has 

 seen them do ; nor have we ever seen the end of a float- 

 ing thread separated into its component threadlets and 

 diverging like a brush, as he and Mr. Bowman describe. 

 It may be proper to mention that Mr. Murray, in con- 

 formity with his theory, explains the shooting of lines in 

 a current of air by the electric state produced by motion 

 in consequence of the mutual friction of the gaseous par- 

 ticles. But this view of the matter does not seem to afifect 

 our statements. 



