1888.] NATURAL SCIENCES OF PHILADELPHIA. 213 



habit of getting the sand out at a point nearer the ground than the 

 top of the tube. On the contrary the particles dumped from the 

 top or through slits in the side, and which al&o readily entangle 

 within the silk strands as they fall, are permitted to remain inas- 

 much as they are not inconvenient. The idea of a protective pur- 

 pose cannot, however, be wholly excluded ; for it is found that in 

 repairing the rents made in the tube in order to draw in the strick- 

 en prey, the new material spun over the rent is quite invariably 

 sanded. This indicates a deliberate intention. 



On the whole, in view of the above facts, and reasoning from 

 them by analogy it appears that (1) much of the sand and bark- 

 dust which covers the outside of the nests of Atypus is an incidental 

 result of the act of excavation ; (2) that, however, the spider does 

 at times deliberately add to this coating; (3) that the purpose of 

 this act is probably i)rotective at least in the way of strengthening 

 the tube; (4) that there is no positive proof that protective mimic- 

 ry has any part in the habit ; yet (5) as a matter of fact this exterior 

 coating does better adapt the tube as a snare both to decoy insects 

 to a light and enable them to travel upon it. 



V. MATERNITY HABITS. 



Much remains to be determined of the life-history of the Purseweb 

 spider, but we may venture the prediction that in many points it 

 will be found to difler little from the habits of its British cono-ener 

 as described by various observers. We know from Abbot's note 

 above cited that the young, like the offspring of Lycosids, domicile 

 upon the back of the mother after they are hatched. The cocoon 

 containing the eggs is of course retained within the purseweb, and 

 probably in that portion which is beneath the surface of the ground. 

 Atypus piceus suspends her egg-cocoon in a pretty hammock of silk 

 an inch long, attached to the top and bottom of the pouch.^ The 

 number of eggs within the cocoon of Piceus is from one hundred to 

 one hundred and fifty. They are deposited in midsummer, July or 

 early August, and the young issue from the cocoon about the latter 

 part of September. They remain with their mother in the maternal 

 nest during the winter, and Mr. Enoch found the female and her 

 young together March 31st, and again as late as April 5th, About 



1 See Mr. Knock's paper, p. 392. See also a good figure representing the same 

 habit in Mr. Simon's paper, Annals Entomological Society of France, 5th Series 

 torn. 3, 1874, plate 4; also "Spiders of Dorset", Rev. O. P. Cambridge, page 

 xxxiii, Introduction. 



