October, '16] SMITH: DEFINING HOST RELATIONSHIPS 483 



parasitic species.^ In the Ichneumonoidea secondary parasitism oc- 

 casionally occurs, especially in the Cryptinae. 



As a definition I would suggest the following: Secondary parasitism 

 is that type of symbiosis where a parasite destroys a primary parasite 

 by direct attack, and not through the medium of the host of the prim- 

 ary parasite. 



Since this type of insect has the same relation to primary parasites 

 as the latter have to insect pests, it naturally follows that they are an 

 extremely important consideration in the control of injurious insects. 

 They are in our native fauna responsible in many cases for the in- 

 effective work of primary parasites which would otherwise be of great 

 practical value. In the introduction of new beneficial insects it is 

 obviously of greatest importance to guard against the introduction of 

 these secondaries. Many primary parasites of little importance in 

 their native habitat might, by introducing them into new localities, 

 become of great practical value through the elimination of their sec- 

 ondaries, especially if these secondaries have no counterpart in the new 

 locality. Occasionally, however, the newly introduced parasite is 

 immediately attacked by secondaries native to the new locality and 

 which had as their original host species of the same genus as the parasite 

 introduced. A noteworthy instance of this kind occurred at the 

 Gypsy Moth Parasite Laboratory of the U. 8. Department of Agri- 

 culture.^ Apatiteles fiilvipes, a common parasite of the gypsy moth in 

 Europe and Japan, was introduced as a most promising species. In 

 Europe it was found to be attacked by something like twenty-five 

 species of secondaries and indirect parasites, and other hyperparasites. 

 In Japan at least thirty species of hyperparasites occurred. During 

 the first generation on American soil seventeen species of hyperpar- 

 asites, for the most part different species but the same genera as those 

 occurring in Europe and Japan, attacked Apanteles fukipes. 



In this case the elimination of the secondaries probably did not have 

 a very important bearing on the success of the introduction, since the 

 introduction of hyperparasites having the same habits as native par- 

 asites would merely serve to eliminate to a large degree the native 

 hyperparasites, leaving the total percentage of mortality about the 

 same as before. The introduction of secondaries which have no coun- 

 terpart in the new fauna, however, would have an entirely different 

 effect and the greatest care should be exercised to eliminate any hy- 

 perparasites which would form a new element in the local fauna. 

 Obviously the only safe way of doing this is to eliminate them all. 



1 Since writing the above I have come across a record of true secondary parasitism 

 among the Proctotrypoidea by Mr. Swezey of the Sugar Planters' Station of Honolulu. 

 A species of Ceraphron was found to parasitize Haplogonalop^ts. a Dryinid. 



2 Howard and Fiske: Bui. 91, Bur. Ent., U.S.D.A. 



