COCOON LIFE AND BAr.YIIOOD. 209 



about half way vemovefl from the abdomen. The feet pushed downward, 

 and at the same time tlie altdomen appeared to be ))u!1(h1 upwai'd, until 

 the white ponc-h was gradual!}- worked off. (Fig. 241.) Tiie motion was 

 not unlike that of a child stripping ofp its niglit dress by ]iushing it down 

 the body and stepping out from the drajjer}'. 



Life within the cocoon is not wholly destitute of "moving accidents" 

 and " hairbreadth 'scapes," if we may believe Professor Wilder, who ar- 

 gues that the young of Argiope cophinaria eat one another wliile 

 Cocoon yg(^ within cocoon limits. His reason is that a comparison of 

 anni- ■ ^j^^ contents of cocoons opened early in the season with those 

 opened later showed that the spiderlings were fewer in number 

 but larger in size.^ He infers the same thing from the fact that after 

 egress the young do prey upon one another, Ijut without sufficient ground, 

 as the one fact by no means implies the other. My own observation has 

 been, of all species, that the young live together peacefully while within 

 the cocoon. However, I have chiefly studied the cocoonery of our more 

 northern latitudes. 



In southern latitudes, where the hatcliing probably occurs earlier with 

 some species, and the period of confinement after hatching is thus mucli 

 prolonged, or tlie appetite of the young quickened by the climate, hmiger 

 may assert its supremacy. Yet, even in the case of some southern spiders, 

 as examples of Zilla from southern California, reared during winter in my 

 study under conditions of temperature not very different from their native 

 latitude, tliere never api)eared a trace of cannibalism until after the j'oung 

 araneads had woven their fir.st independent snares. In the case of most, 

 probably of all, species in our more northern climate, during the greater part 

 of the four montlis intervening between hatching and egress, the young 

 are probably more or less torpid by reason of the cold, and thus with 

 natural appetite still in abeyance. Even in our Southern States the influ- 

 ence of season is seen by a general suspension of activity in the insect 

 and aranead world ;' and, independent of climatic influence. Nature doubt- 

 less gives a semidormant tone to the sjiider young. 



Whate^-er may be the truth as to Argiope and Nephila, I am certain 

 that many species do not have this cannibal habit within the cocoon, nor, 

 indeed, for .some time after egress therefrom. Mr. Pollock's observations'- 

 of Argiope aurelia, of Madeira, quite correspond witli this statement, for 

 the broods were always friendly within the cocoon, and indeed for a fort- 

 niglit after leaving it. 



III. 



The spiderlings themselves procure exit from the cocoon in most spe- 

 cies. This is frequently accomplished by cutting a small opening through 



' Proceed. Am. Asso., VII., 1873, page 2(50. ^ Ann. and Mag. Nat. Hist., 1865, page 4G0. 



