NATURAL SCIENCES OF PHILADELPHIA. 



207 



1888.J 



fly through the opening, and after several tugs the insect was pulled 

 Avithin the tube, and the spider backed downward holding its prey- 

 fast in its falces, leaving a rent a quarter inch long by three-sixteenth 

 inch wide. After an interval of three minutes the spider cautiously 

 reascended the tube to the opening, and taking hold of the ragged 

 ■edges of the rent, drew them towards each other until they almost 

 touched. She then backed a little and turned her abdomen so that 

 the spinners approximated the united edges. Then by a number of 

 zigzag movements with the spinners across the juncture, she com- 

 pletely closed the rent, and when it was neatly repaired returned 

 apparently to feed upon her prey. The next morning the rent was 

 covered with sand so carefully that Mr. Enock could scarcely detect 

 where it had been. When the spider was satisfied with food, it 

 would draw in the tube in a determined manner, and would retain 

 her hold in this position sometimes for several hours. 



1. A time-measure of the spinning-work. — Being foiled by the 

 persistent secretiveness of this spider in natural sites, I captured sev- 

 eral specimens and placed them within glass jars in order to observe 

 their behavior under these artificial conditions. Some important 

 facts resulted, particularly as to the mode of constructing the tubu- 

 lar nests. The bottom of each jar was filled with sand, and a stick 



inserted within, in order to give a 

 natural position for the establish- 

 ment of a nest if the spider should 

 be inclined to weave one. One indi- 

 vidual, after long continued explo- 

 ration of its quarters, at last estab- 

 lished itself at the foot of the stand- 

 ing stick and began to burrow a 

 little hole. I was compelled to 

 leave at this point, and did not re- 

 turn to my room until evening, after 

 twelve hours absence. During this 

 time a vertical tube of white silk 

 one and one-fourth inch long and 

 about the thickness of the spider 

 had been spun along the side of the 

 stick. The outside of the tube was 

 sparsely covered over with particles 

 of sand which of course had been 

 brought from below. 



A unit for measuring the time re- 

 quired to construct a tube was also obtained. The inch and a 



Fig. 5. First section of Purseweb spider's 

 tubular nest. 



