60 AMERICAN SPIDERS AND THEIR SPINNINGWORK. 



is parasitic on Epeira diademata and also upon supposed spider cocoons ; 

 Pezomachus fasciatus is guest within the egg nests of Agroeca brunnea, and 

 on a species of Theridion ; Hemiteles sirailis parasitizes Epeira diademata 

 and an unknown spider cocoon. Hemiteles rufocinctus, H. fulvipes, and H. 

 forniosus are reported as body parasites, while other species of the genus 

 are parasitic upon sjiider eggs. This generalization appears to me remark- 

 able, and contrary to what one would naturally suppose from the ordinary 

 specialization of such instincts. If correct, it shows much elasticity of 

 habit among these parasitic Hymenoptera, for there would seem to be an 

 immense distance on the maternal side between tlie instincts which prompt 

 to the several acts ; and on the side of the offspring, between development 

 within a spider's egg and within a spider's body. 



Mr. Howard, who was kind enough to look over this part of my manu- 

 script, expresses the doubt here suggested in a more positive way. He thus 

 writes me : " None of the spider Pimplas are typical, but belong to a 

 restricted group, possibly a subgenus, all of whicli, I believe, are parasitic 

 in spider egg bags. All of the Polysphincta, in my ojiinion, are external 

 parasites of spiders ; and this genus, with the genus Acrodactylus, I believe 

 contains all of the external hymenopterous spider parasites." This opinion 

 from such an authority would justify the omission of the above inference ; 

 but as the facts as reported are not changed in Mr. Howard's published 

 lists, I permit the item to remain as qualified. The observations of early 

 date which here compose the list are probably erroneous; and, indeed, 

 they are confessedly defective. 



Fourth, the general spinning and other characteristics of spider species 



appear to make no marked difference in liability to parasitic attack. Such 



sedentary spiders as Epeira, Theridion, and Agroeca are not exempt 



Relations fj-om body parasites by their occupation of snares. Yet, in such 



. cases, one wonders how the ingenuity of the insects' maternal 



Habits. instinct could overcome, as we know it does, the local difficulties, 



and drojj an egg upon the spider's body while swinging upon or 



lurking within its web. The difficulty is not lessened, but rather enhanced, 



if we suppose that some of the parasitic larvse are traveling parasites, and 



seek their hosts after independent hatching. The condition of the wingless 



female parasite would appear to make her task more formidable, but does 



not limit her ovipositing to periods when the spiders are off their snares 



and reposing on adjacent objects, or within their dens and tents. 



Fifth, the special cocooning habits of sjyiders appear to have no relation 

 to their exposure to parasitic attack. Here, also, one meets apparently con- 

 tradictory facts. For example, the egg cocoon of Argiope 

 Of Co- cophinaria is one of the most skillfully con.structed to protect 

 H^'b^t'^^ the enclosed young. (See Vol. II., pages 76-80.) The eggs are 

 encased in a silken sheet, overlapped by a thick, compact oval 

 blanketing, which in turn is encased in a tough glazed sac. The whole is 



