120 CRABRO. 



eyes large, subtriangular, with the angles rounded, approximated 

 at the base of the antennae, distant on the vertex; steramata 

 generally placed in the centre of the vertex; antennas geniculated, 

 inserted at the base of the clypeus, the flagellum filiform in the 

 females, in the males sometimes subfusiform, with the basal 

 joints fornicate and fimbriated within, the apical joints serrated 

 on the exterior, in others emarginate beneath, sometimes appa- 

 rently consisting of twelve joints ; clypeus transverse, longitu- 

 dinally carinated in the centre, and covered with golden or silvery 

 pubescence ; mandibles bidentate. Thorax ovate, with the collar 

 transverse ; scutellum with the hinder margin rounded, the 

 metathorax short and obtuse, with an enclosed space at its base, 

 subquadrate or subcordiform, sometimes obsolete ; the anterior 

 wings with one marginal and one submarginal cell, the marginal 

 cell truncated at its apex, and slightly apj)endiculated ; the legs 

 short and robust, usually very spinose; tarsi longer than the tibiae; 

 the anterior pair, in the males of some species, have the basal 

 joint broadly dilated into a concavo-convex plate. Abdomen 

 subsessile, lanceolate or clavate ; in some species the basal seg- 

 ment prolonged into a petiole, w hich is subpyriform and nodose 

 at its extremity. 



In this genus the views of Shuckard are adopted. The known 

 species of the genus Crabro have been divided and subdivided 

 by different authors into about fifteen genera, most of them 

 based upon what are here considered as constituting specific cha- 

 racters only. We cannot agree in the propriety of establishing 

 a new genus for the reception of a single species, only distin- 

 guished from the rest by having a deep notch at the base of 

 the mandibles, as in Crabro brevis : nor do we deem it advisable 

 to separate generically those species which have the ocelli placed 

 in a curve, from those in which they are placed in a triangle ; we 

 find numerous instances in which they are placed in an inter- 

 mediate position. Many sections formed upon such specific 

 distinctions are extremely convenient w r hen examining an entire 

 family; but when observation is restricted to the fauna of a 

 district, and that a small one, these sections are too apt to be 

 regarded as possessing a degree of importance, to which in 

 reality they have not any legitimate claim : it is only through a 

 knowledge of entire groups that we can hope to arrive at a 

 correct generic distribution. The most satisfactory character 

 hitherto employed in forming the generic divisions of this tribe 



