REVISION OF STEEPSIPTEEA PIERCE. 11 



The month structure, consisting of mandibles and maxilla>, but not 

 constructed for feeding on solid food, also draws the Strepsiptera 

 away from the Coleoptera, toward the Hymenoptera. 



The nervous system is reduced to a supracesophageal, a thoracic, 

 and a ventral ganglion from which nerves branch out to all of the 

 segments. Such reduction is frequently found in the Diptera. 



The larviparous reproduction is found in the Diptera, but not in 

 the Coleoptera. 



COLLECTION. 



The building up of a large collection of stylopids is necessarily 

 attended with great painstaking, for only occasional results can pos- 

 sibly attend indiscriminate collecting, but a systematic biological 

 study must yield astonishing results. Furthermore, the work must 

 henceforth be conducted in four orders — Orthoptera, Homoptera, 

 Heteroptera, and Hymenoptera. 



In Orthoptera (in the American sense) it is probable that the 

 Blattoidea and Grylloidea will be the only groups parasitized. These 

 insects inhabit damp places and frequently construct burrows. Para- 

 sites must be sought in the haunts of the hosts. 



The only heteropteron known to be parasitized is a large scutel- 

 larid from the Orient. Very possibly more extensive examination 

 in tropical regions will bring out other hosts. 



At the present writing the most important order parasitized is the 

 Homoptera. Perkins (1905) has recorded parasites in the Asir- 

 acida?, Issida?, and Tetigoniidse, and says that they are sometimes 

 quite abundant. This, then, opens a new field for the economic ento- 

 mologists. Although it is possible to collect parasites b}- beating in 

 the grass as Templeton (1838) did when he took Elenchus tenuicornis, 

 the best method must be to follow in the steps of Koebele and Perkins 

 (Perkins, 1905) and breed the parasites in breeding cages. The same 

 method will hold with Heteroptera. Mr. Koebele for several years 

 has been endeavoring to introduce American, Vitian (Fijian), and 

 Australian Elenchidse into Hawaii as parasites of the injurious cane 

 leaf-hoppers. The method used was to collect parasitized Homoptera 

 in large quantities and ship them in cages with a mesh fine enough 

 to prevent the leaf-hoppers from escaping. This mesh would give 

 plenty of room for the triungulinids to get out and attack other hosts. 

 Up to date no success has been recorded in this attempt ; in fact, there 

 seems to be no chance of success. Probabry the reasons are that the 

 Hawaiian leaf -hoppers are different species or genera and the parasites 

 will not change their host relations. 



The collecting and study of the hymenopterous parasites is by no 

 means as difficult or as exacting a work as that with homopterous 

 parasites. The hosts in this order belong to the families Formicidse 



