1903.] NATURAL SCIENCES OF PHILADELPHIA. 139 



for frequently both sexes may live in harmony for a considerable 

 length of time. The reason that in many species the females appear 

 more abundant than the males (particularly in Epeiridce), is not to 

 be ascribed to any initial disparity in the number of the sexes, nor yet 

 to the males being killed by the females, but to the females being of 

 stronger constitution. Thus, in the case of the individuals watched 

 by me, the males died sooner from hunger and thirst than did the 

 females, and each act of copulation appeared to diminish their vitality. 

 When one collects in the early part of the mating season the males 

 are generally found fully as abundant as the females, while later in the 

 year they may become, in some species, very rare. 



Parthenogenesis. — Two cases of apparent true parthenogenesis, with 

 the production of fertile eggs, have been described by Campbell (1S83) 

 and Damin (1893). A number of females observed by Blackwall 

 (1845), where he excluded males and kept the females through a num- 

 ber of moults, made cocoons and laid eggs, but the latter proved infer- 

 tile. Eggs frequently prove infertile even after copulation, according 

 to my own observations. Parthenogenesis, with the production of 

 fertile eggs, is then very rare among spiders, and it is probable that 

 most species do not show it at all. Blackwall's observations (1845a) 

 are quite decisive upon this point. 



A very curious case is that described by Doumerc (1840) of a The- 

 ridium triangulifer which made successive cocoons, out of some of 

 which only male spiderlings developed, out of others only females. 

 This is a phenomenon well worth critical study. 



Cocooning, Oviposition. — I shall not attempt to give here a com- 

 parative study of cocoon-architecture; that subject has been well 

 treated by Wagner, and much less accurately by McCook; but shall 

 simply call attention to the similarity that exists among all observed 

 species in the general modes of cocoon-making and oviposition. The 

 first stages of this process have not been frequently described, and the 

 following list comprises all the observations on the subject : Leeuwen- 

 hoek, Menge {Lycosa, Tegenaria, Gnaphosa, Micaria, Philodromus) , 

 Emerton {Epeira), McCook {Lycosa), Henking {Lycosa), Pokrowsky 

 {Pholcus), Warburton {Agalena), Wagner's excellent studies on a num- 

 ber of species, and my own {Lycosa stonei, L. lepida, L. punctulata, L. 

 ocreata pulchra, Pardosa nigripalpis, Pirata liber, Tegenaria derhami, 

 Theridium tepidariorum, Teutana.). In all of these the spider spins 

 first a base, oviposits upon it, then spins a cover over it; all ara- 

 neid cocoons are made in this way. McCook (1890) distinguishes 

 three kinds of cocoons, and states of the third kind (supposed to be 



