1 82 Leland O. Howa^-d 



been left for this unfortunate group. Its definition lacks that 

 trenchant clearness characteristic of Mayr's other generic 

 characterizations, and what are really diverse types to-day 

 bear this generic name. 



In glancing through the host insects which species of 

 Encyrtus affect we find in Europe a wasp larva, a beetle larva, 

 the eggs of noctuid and bombycid moths, the larva of a 

 micro-lepidopter, the larva of a syrphus fly, dipterous galls, 

 plant lice and bark lice. We have thus 8 types of host insects. 

 In America we have also a wasp larva as well as two other 

 hymenopterous hosts, viz. : an Ichneumon and a cynipid gall. 

 We have also the eggs of a bombycid moth and of a beetle, 

 the larva of a micro lepidopter, larva of a syrphus fly, dipter- 

 ous gall-makers, plant lice and bark lice, and two new ele- 

 ments in addition to the beetle eggs and the hymenopterous 

 insects mentioned above, viz. , heteropterous eggs and Psyllidae. 

 We have then 12 quite distinct types of hosts, all told, 8 of 

 them occurring in Europe and all in America.* With this 

 view of the biology of the genus it at once becomes important 

 to make a closer study of the morphological aspect of the in- 

 dividual forms than has yet been done. One would naturall}- 

 expect to find an assemblage of characters grouping together 

 those species which prey upon a common type of host, and, 

 such characters being found, shall we not be justified in giv- 

 ing them greater classificatory weight than parallel separating 

 characters which are not correlated with important, not to say 

 vital, biologic facts ? 



No attempt has hitherto been made in this direction. To 

 test provisionally the aptness of the idea, a brief survey of 

 the synoptic table of European species shows that while no 

 attempt has been professedly made to form natural groups, 

 yet the characters hit upon to conveniently analyse the species 

 have brought into immediate juxtaposition the species para- 

 sitic upon lepidopterous eggs ; those parasitic upon dipterous 

 larvae are brought into close connection ; the bark-louse para- 



*Should the species of Encyrtus described by Girard as coming from 

 a Psyllid gall prove to belong to this genus, Europe will have represen- 

 tatives of 9 of the types. 



