October, '16] SMITH: DEFINING HOST RELATIONSHIPS 481 



cidoid parasite Perilampus hyalinus,^ and is, so far as I know, the first 

 case on record of an indirect parasite. Perilampus hyalinus has an 

 hiatus in its known life-history, since we know nothing at the present 

 time of its oviposition habits. But we do know that it attacks the 

 larvae of Hyphantria cunea, not for the purpose of breeding upon Hy- 

 phantria, as it is unable to do this, but for the sake of the primary 

 parasite which it harbors. Strangely enough, in the case of this par- 

 ticular parasite, it does not matter much what the primary parasite is, 

 just so it is an internal parasite of H. cunea. It will be seen then that 

 in the type of host-relationship known as indirect parasitism there are 

 alwaj^s thi'ee insects necessarily concerned simultaneously if the in- 

 direct parasite is to succeed in reproducing: first, the host of the prim- 

 ary parasite; second, the primar}"- parasite, and third, the indirect 

 parasite. No other host conditions will suffice. It will readily be 

 seen that this type of host-relationship represents a very different kind 

 of parasitism from that occurring where a parasite oviposits directly 

 into the primary and yet both have always been known as secondary 

 parasitism. Types of host-relationship so widely different should be 

 distinguished by different terms. I would restrict the term indirect 

 parasitism to the type of symbiosis similar in a general way to that 

 occurring in Perilampus hyalinus. As a definition of indirect parasi- 

 tism I would suggest the following : Indirect parasitism is that type of 

 symbiosis in which the one parasite attacks a host insect upon which it 

 itself is incapable of breeding, for the sake of the primary parasite 

 which it may harbor. Since the biology of so few parasitic insects is 

 known it is impossible to say just to what extent indirect parasitism 

 occurs in nature. Besides Perilampus hyalinus one or two other species 

 of this genus are known to have this habit, although several species are 

 known to be true primary parasites. In the Ichneumonoidea this 

 type of parasitism is known to occur in Mesochorus pallipes, a parasite 

 of the Braconid, Apanteles fulvipes, which is in turn a parasite of the 

 gypsy moth in Europe. Since several other species of this genus are 

 parasitic on Apanteles species, it is probable that many of them have the 

 same habit. To this class belong also a number of hyperparasites of 

 scale insects. Among them are species of the genus Eusemion which 

 are parasites of Microterj^s and Aphycus, in their turn parasites of the 

 soft brown scale, species of Cheiloneurus which breed on various 

 parasites of mealy-bugs and scales, and Cerchysius, a parasite in one 

 case of Microterys on soft brown scale, and in another on Scutellista 

 cyanea and Tomocera calif ornica, parasites of the black scale.^ Species 



1 The Chalcidoid Genus Perilampus, and its Relations to the Problem of Parasite 

 Introduction. Bui. 19, Tech. Ser., pt. IV, Bur. Ent., U.S.D.A. 



2 Vide Timberlake, "Parasites of Coccus hesperidum, " Jour. Econ. Ent., vol. 6, 

 pp. 293-303. 



