PARASITES OF GIPSY-MOTH CATERPILLARS. 



201 



Two lots of each generation were collected after the healthy parasites had issued, 

 and the results follow: 





First generation. 



Second generation. 





Cocoons. 



Per cent. 



Cocoons. 



Per cent. 





543 

 162 

 22 



89 



66.5 

 20 

 2.5 

 11 



636 



1,207 



102 



M69 



26 





50 





4 





120 







Total 



816 



100 



2,414 



100 



i Oct, 20. 



Other collections of cocoons from colonies where the Apanteles was liberated too 

 late in the season to permit of two generations showed a high rate of hyperparasitism 

 in the single generation, actually the first but corresponding to the second. Com- 

 parison in this instance is valueless, as local conditions enter in which can not be 



gauged. 



This rather lengthy summary of a study in hyperparasitism has 

 been prepared and is here presented with the object of illustrating 

 the somewhat modified stand which it has been necessary to take 

 concerning the subject in its relation to the project of parasite intro- 

 duction. Were it within the bounds of possibility to introduce into 

 America the parasites of the gipsy moth (Apanteles fulvipes, for 

 example) without introducing the secondary parasites which preyed 

 upon them abroad, it would unquestionably be possible to secure a 

 greater meed of efficiency in America than that which the same para- 

 sites were capable of attaining in their native countries. This is on 

 the supposition that the parasites themselves are no more likely to be 

 attacked by the American hyperparasites than their hosts are likely 

 to be attacked by the American primary parasites. 



That the assumptions are fallacious, to a certain extent, is well 

 proved by the results following the temporary establishment here of 

 Apanteles fulvipes, as recounted above, and that the same results 

 as those which followed the exposure of this parasite to American 

 hyperparasites will result in the instance of others among the im- 

 ported parasites is more than likely. In the case of Compsilura 

 concinnata and Apanteles lacteicolor Vier. it is proved. 



The truth of the matter is that the secondary parasites are very 

 far from being as closely restricted to one or two species of hosts as 

 are the primary parasites. This is in part due to the fact that they 

 represent for the most part a much more degraded form of para- 

 sitism. Species like Dibracliys houcheanus, which is perhaps as 

 generally abundant and omnivorous as any of the parasitic Hymen- 

 optera, will attack anything which is dipterous or hymenopterous, 

 provided it is physically suitable as food for its larvae. Apanteles 

 fulvipes and caterpillar parasites generally are governed in their host 



