I 



ELISHA MITCHELL SCIENTIFIC SOCIETY. 143 



5. The tube is so constructed as to serve as a snare, and not 

 merely as the silken linin<>; of a "burrow." Being mainly sup- 

 ported by its upper extremity, its walls, 'for the most part free, 

 readily transmit any mechanical disturbance up or down. See 

 Plate, Figs. laud III. 



6. The tube is closed above by the convergence of its walls in 

 part, and in part by the simple approximation and adhesion of 

 its outer wall to the supporting object. 



7. The threads composing the walls are seen under the micro- 

 scope to have a general longitudinal direction, — a fact which 

 explains the ease with which longitudinal rents may be made. 



8. Usually the greater part of the tube is above ground.* 



9. The work appears to be done only at night. 



The feeding habits of our spider may now be noticed. Rev. 

 O. P. Cambridgef refers to E. Simon as authority for the state- 

 ment that spiders sometimes prey upon earth-worms. I have 

 not access to the journals in which the observations were 

 reported, nor can I say that the reference is to the particular 

 spider that concerns us here. Whether Mr. Banks' remark, 

 quoted early in these notes, rests upon the same authority, or 

 upon what authority it rests, I am not aware. But my own 

 observations of Atypus niger Hentz (?) lead me to doubt that its 

 "food is earth-worms." It is not improbable that so succulent a 

 morsel would be appropriated in case it should come in the way; 

 but it is improbable that the spider goes about to capture earth- 

 worms, or that they are its chief reliance for subsistence. No- 

 where in the nest have I noticed any feature of adaptation to 

 this end. On the contrary, it is evidently constructed for the 

 capture of insect food. The evidence is not merely inferential; 

 it is positive as well, for I have looked on at the spider's feast 

 and have seen the fragments that remained. 



The behavior of my first captive specimen when a fly was 

 crawling on her tube suggested to me that the disturbance made 



*Measurements of a certain tube: Total length, 38 cm. Length of aerial portion, 21 

 cm. Length of subterranean portion, 17 cm. Diameter at surface, 2 cm. 



fEncycl. Brit., Vol. 2, p. 298 (9th Ed.). 



