HORIZONTAL SNARES AND DOMED ORBS. 161 



of navigation by means of filament sails. The first experiment was re- 

 peated a number of times with various details upon several spiders, and al- 

 ways with the same result. The gossamer thread was undoubt- 



Sails of ^j US ed as a sail, and the action of the wind bore the little 

 'Silk Fila- 



te navigator to the shore. I frequently blew the tiny craft out to 



sea, either with my breath or by fanning with my hat. My stick 

 could at any time arrest the thread by placing it a foot or more above the 

 spider, and having entangled it, I could draw her thereby in any direction. 

 The filament was plainly seen floating above the spider, waving to and 

 fro, generally bending above her back. 



The legs during motion were raised upward and bowed, thus holding 

 the body well up from the surface, exactly in the attitude of a spider about 

 to take aeronautic flight. They were kept quite rigid and motionless. The 

 feet were spread out, describing the outline of an irregular octagon. I fancied 

 that they were united by threads and that thus the spider sat upon a delicate 

 raft of silk. Before spreading her sails the first act of the spider was to 

 drop the abdomen to the 

 surface, at which moment, 

 I inferred, an attachment to 

 the legs was made. I can 

 hardly bring myself to be- 

 lieve that the threads were 

 attached to the water, al- 

 though it may so have been. 



While the spider was un- Tm ' 15S ' Epeira tising ber cocoon M a float - 



der sail the feet made a very noticeable ripple of wavelets as they were 

 hurried along. She could accomplish short distances from the shore by 

 running, without spreading sail, but when put well out she always re- 

 sorted to the latter mode. However, any floating object which she met 

 during the voyage was pretty sure to be taken advantage of as con- 

 venient harborage. A downy seed which fell upon the water and drifted 

 within reach of one of my aranead sailors was immediately seized by the 

 creature's fore feet, the spinnerets were set in motion, and the seed was 

 overspun with a delicate floss, which converted it into a sort of float. 



In this connection I record an incident which may throw some light upon 

 the development among Orbweavers of this interesting water habit. A large 



female Epeira sclopetaria was collected, along with an abandoned 



cocoon, behind which it had spent the winter, at Atlantic City. 

 Float ^ ne s pider was accidentally dropped into the Inlet, together with 



the flossy ball of the cocoon. She immediately threw out threads 

 around the ball (Fig. 153), to which she remained attached as to a buoy, 

 and thus the two drifted along safely under the floor of a boat house 

 and so out of sight. One associates such an action with the water walking 

 and navigating of Tetragnatha, and wonders whether Epeira could ever, 



