FOOn PLANTS OF PLATYSAMIA CKCROPIA. 21 ^^ 



leaves of tliis peacli I had often jdaced young cecropia larvfe, invari- 

 ably they refused to eat and starved to death. Twice I put larv}« 

 about half grown, with the same result. 



When the larva- on the plum were about half giown, a gale of wind 

 blew the branches of the trees togetlier. A few- days afterwards I 

 noticed the uppei- shoots of the i)each denuded of leaves. On close 

 inspection I found four cecropia larvte — having evidently worked over 

 from the plum — feeding greedily and thriving well. These all 

 matured and spun upon the peach. 



The horse-chestnut is entered from finding eight cocoons so situated 

 that the larva- could not have fed on any other tree. 



The Symphoricaryus vnlgarifi is entered from one convincing 

 exajuple — the slirub was isolated a long distance fiom any other 

 shrub or tree — the cocoon was found soon after completion, and 

 several shoots denuded of leaves gave evidence of the ])resence of the 

 larvse. 



Some doubts have been exj)i-essed as to tlie elm being a food plant. 

 The rej>eated finding of cocoons, often 30 ft. or more from the ground, 

 seems quite conclusive, but for several seasons I have reared a large 

 number of larvte on an elm tree in my garden and the only noticeable 

 difference was the generally small size of the cocoons. Cocoons are 

 sometimes found on the common jnivet, but. although it is most pro- 

 bably a food plant, in absence of sufficient evidence it is omitted from 

 this list. 



1 liave no doubt it will lie found that the larva- U^ei^x^ indiscriminately 

 on all our species of willow and poplar. 



In all our northern lumbering sections, a few years after the sweep 

 of a bush fire, you find a dense growth of sambucus birch and willow. 

 Thus, with abundance of food and absence of parasites, this would be 

 the Cecropia's paradise, were it not for the woodpeckers which per- 

 forate the cocoons and feed on the pupje during the winter. 



I am indebted to Mr. R. Morey and Mr. W. Scpiires for valuable 

 assistance in the collection of material for this list. 



