HOST RELATIONS OF COMPSILUBA CONCINNATA 13 



Later a collection of 171 larvae was obtained, but no tachinid parasite 

 other than Compsilura was reared. It was not until a small collec- 

 tion of five last-stage larvae was received on September 20 that any of 

 the native tachinids were recovered, two specimens of Phorocera 

 claripennis being bred. 



The results obtained from the collections secured at Brooklyn, 

 N. Y., and Philadelphia, Pa. (1921-22), are interesting. The tachi- 

 nids recovered from this material were the same species as those bred 

 at the laboratory in 1910 and 1911. Furthermore, with hardly an 

 exception, the relative importance of each species was similar. The 

 similarity of records is probably due to the conditions existing ; that 

 is, a heavy infestation and the absence of Compsilura. 



It is obvious from the records that in a territory where Comp- 

 silura is established a light infestation of Hemerocampa leucostigma 

 is comparatively free from native tachinid parasites. Such a condi- 

 tion would hardly exist, however, were there not some interference 

 from the exotic parasite. To what extent this interference reaches is 

 problematical, but two things seem assured : (1) There is no excessive 

 parasitism by the native species in localities where the introduced 

 parasite is absent, and (2) the host species must be materially 

 affected by the presence of this additional enemy. 



LASIOCAMPIDAE 



The lasiocampids Malacosoma americana Fab. (the tent caterpil- 

 lar) and Malacosoma disstria Hiibn. (the forest tent caterpillar), 

 although close relatives of the liparids and having some points in 

 common, are among the most unfavorable of hosts. Their appear- 

 ance in the field during May and early June, when there is not an 

 overabundance of insect larvae of sufficient growth for the early 

 issuing Compsilura, would lead one to believe them most desirable. 

 Such is not the case, however, for hundreds of larvae of both species 

 have given forth but few parasites. 



It is difficult to explain why Compsilura so steadfastly ignores 

 these species. In the case of Malaco soma- americana, where the 

 species is of the tent-making kind, some interference might be ex- 

 pected on account of this habit ; but when the same species is isolated 

 and used in laboratory experiments the results are likewise negative. 

 Reproductive experiments, using the tent-making arctiid Hyphantria 

 cunea Dru. (the fall webworm) as a host, have proved that far better 

 results can be obtained under laboratory conditions, where the larvae 

 are not allowed to web up, than can be obtained from field collec- 

 tions. The field collections nearly always result in failure as far as 

 Compsilura is concerned. It seems clear, in the latter case at least, 

 that the web must offer considerable protection to the larvae within 

 it. M. disstria, which does not make a tent, has a much better host 

 value than M. americana. 



Another species of this family, E picnaptera americana Harr., 

 appears in the field at a later period and the larvae are frequently 

 found during the last of August. Unlike the species of Malaco- 

 soma, this insect is solitary and is never found in abundance. Its 

 value as a host species is doubtful. 



