144 JOURNAL OF THE 



by an insect might be simulated by gently rubbing the tube with 

 a tender twig. My first experiment in the pines received a 

 speedy response. The spider came up by starts, her position in 

 the opaque tube being shown by the slight bulging of its walls 

 beneath her feet. This became my habitual method of ascer- 

 taining whether "my lady" was in. She seems unable to distin- 

 guish such disturbance from that caused by an insect. On one 

 occasion I rubbed a tube with a fascicle of dead pine leaves. 

 The spider came up and, the fascicle being kept at the same 

 point, she thrust her fangs through the wall to grasp it, but 

 failed. It was not long before she made another effort, that time 

 seizing it firmly with both fangs and sustaining its entire weight 

 (the three leaves were 15 cm. long) until I blew upon the tube, 

 when she instantly let go and hurried below. She held the 

 fascicle in the position indicated in the Plate, Fig. V. Her 

 eyes can furnish little aid for the reason that the tube is not 

 transparent, and the experiment just described suggests that the 

 sense of smell is of no service. I conclude her habit is to try 

 to catch whatever disturbs her snare in the hope that it will 

 prove to be acceptable food, having ascertained its position by 

 the sense of touch in her [)alpi and feet. 



On the morning of August 21st I put a cockroach on a tube. 

 The spider was soon up to the same level and, thrusting her 

 fangs through the wall, caught one of the insect's legs. She 

 seemed afterwards to hold it with her feet, for the fangs were 

 seen striking through here and there in the neighborhood of the 

 insect. They finally got a good grip on it. On my return after 

 a few minutes I found that the spider had made a longitudinal 

 slit 13 mm. long and was drawing the transversely placed cock- 

 roach into the tube. She had some difficulty in getting it in. 

 Still holding it with her fangs, she pressed the lips of the rent 

 out and over it with her feet. She dragged the insect below, 

 and for fifteen minutes I saw nothing of her. I went away and 

 returned after another fifteen minutes to find the rent repaired so 

 perfectly that it could not have been detected by one who did not 

 know its position. The same day a cockroach was similarly 



