March, 1910.] BrUES : PARASITIC HyMENOPTERA. 9 



at least four genera belonging both to the Aulacinc-e and Evaniin^e. 

 This proportion is much greater than would occur in a general col- 

 lection of recent insects, so we may safely infer that the Evaniidae 

 were more abundant in Tertiary times than at the present day. From 

 a morphological standpoint this family also appears to be more closely 

 allied to the phytophagous hymenoptera in several respects. In the 

 first place the costal cell of the front wing is preserved while it is 

 absent in the Ichneumonidse and Braconidse, although present also 

 in many other families. In the Aulacinae at least the wings have a 

 more complete and primitive venation than in the Ichneumonidse, 

 and than in most of the Braconidse as well. The absence of the 

 costal cell in the Ichneumonidje would thus appear to exclude them 

 at once from the line of descent of aculeate forms. 



Of the Braconidae and Ichneumonidae, the former are much more 

 nearly related to the Evaniidae through forms like Cocnoceliiis (Anla- 

 codes) which has been variously placed by different writers in both 

 families although it has been even considered by Cresson ('65) as 

 more closely related to the Ichneumonidae. On the other hand the 

 small and peculiar family Stephanidae bridges the gap between cer- 

 tain other Evaniidae (Fceninae) (Bradley, '08) and one of the groups 

 of true Braconidae (Spathiinae). The Stephanidae are further pecu- 

 liar in having a horned structure of the head, recalling that of certain 

 Oryssidae, as has already been pointed out by Ashmead ('00), a char- 

 acter which gives additional evidence of their primitive character. 

 Only a single Stephanid of dubious relationship, Protostephamis, has 

 been found fossil, at Florissant (Cockerell, '06). This group of 

 three families is therefore very evidently to be regarded as a natural 

 association. The Ichneumonidae however present more difficulties and 

 I cannot subscribe to Handlirsch's opinion that they are more primi- 

 tive than any of the families heretofore mentioned. Their wings, 

 which are very constant in venation, always lack the costal vein 

 present in the Evaniidae and Stephanidae, while the basal section of the 

 cubital vein is invariably absent though normally present in other fam- 

 ilies. On the other hand they cannot be derived from the Braconidae 

 as known from any Tertiary genus on account of the presence of the 

 second recurrent nervure, which is invariably absent in the Braconidae. 

 In other respects the two families are very similar and both must, I 

 think, be derived from as yet unknown forms possessing common 



