terminal segments of Aculeate Hymenoptera. 257 



ture short ; the apices of the stipites on their inner 

 margins each with a wide membranous lateral pro- 

 duction, the left-hand membrane folding over the right. 

 In this form of armature this genus seems to stand 

 alone. 



_ CerceridcB. — Eighth dorsal segment simple, 8th ventral 

 bidentate ; armature very variable, but with the stipites 

 not flattened and dilated towards the apex (see PL XIII., 

 fig. 7) ; 8th segment simple. 



Crabronidce. — In the genus Crahro the stipites are 

 produced into a long membranous wing at the apex, and 

 often may be seen projecting beyond the apex of the 

 abdomen ; and the cardo in many species is very long 

 (see PI. XIII. , fig. 9). In Entomognathus and Lindenius 

 the stipites are shorter ; and in Oxyhelus, although the 

 cardo is long, the stipites are not wing-shaped, but 

 simple and elongate. This last genus, however, I 

 scarcely think belongs naturally to the Crahronidce. 



Trypoxylonidce. — Terminal segments simple ; arma- 

 ture with the cardo short ; the stipites not wing- shaped. 



The characters above noticed would draw together the 

 Heterogyna, MutilUdce, Pomjnlidce, and Astatidce, on ac- 

 count of the appendages of the 8th dorsal segment ; the 

 Scoliidce, Tij)hiidce, Bemhecidce, and Nyssonidce, on ac- 

 count of the form of the armature, and the processes of 

 the 8th ventral segment ; the PemphredonidcE and 

 MimesidcB would come close to the Nyssonidce, on 

 account of the apical process of the 8th ventral seg- 

 ment ; the Crabronid(B, on account of their dilated 

 wing-like stipites, also show a distinct relation to these 

 families. The Sphegidie and Larridcs seem to come 

 together naturally, both as regards habits and structure. 

 Still I wish to avoid offering any new arrangement until 

 I have had the opportunity of examining more of the 

 exotic genera. 



DiPLOPTERA. — In this section the 8th segment is 

 simple, and the armature is very similar in nearly all 

 the genera, being very convex and highly polished ; the 

 stipites each bearing a long spine, which may be often 

 seen in life projecting beyond the apex of the abdomen 

 (see PI. XIII., fig. 20). In Vespa, however, this spinels 

 obsolete, or nearly so ; the sagittaB are generally united. 

 This section, to my mind, bears strong proof of the 



