352 M. W. R. de V. GRAHAM 



the flagellar segments are quite strongly transverse ; and the pecten of the hind 

 tibia is at least slightly developed. 



From Stenophrus compressus Forster, the new species differs in the characters given 

 in the key ; also in its strongly flattened thorax, in having the ocelli disposed in an 

 obtuse-angled triangle, and in lacking a pecten on the hind tibia. The male of 

 bouceki also differs from compressus in having the eyes small and widely separated, 

 remote from the ocelli, their facets of uniform size ; the scape strongly swollen, with 

 a large distal excavation ; the proximal segments of the flagellum at most slightly 

 transverse ; and the clava shorter than the funicle. The female of bouceki also 

 differs from that of compressus in having shorter and less conspicuous hairs on the 

 eyes. 



At first I thought that bouceki might be referred to Stenophrus, but now consider 

 it better placed in Pirene. It must be admitted, however, that these two general 

 are pretty close to one another. 



Biology. Unknown. 



PTEROMALINAE 



This is regarded as the central subfamily of Pteromalidae ; it is the largest, and 

 includes a formidable number of genera. The other subfamilies are for the most part 

 distinguished from it by more positive characters which are given in my key to 

 subfamilies. Pteromalinae, however, show less positive characters, for which reason 

 not attempt is made here to give a formal definition of the group. The main difficulty 

 which the worker may encounter is that of distinguishing certain Miscogasterinae 

 from Pteromalinae, because in such cases the characters given in the key are rather 

 slight. The majority of Miscogasterinae have the notauli complete, though some- 

 times superficial posteriorly. Most Pteromalinae have the notauli incomplete ; of 

 the genera which have complete notauli, only four (Perniphora, Dorcatomophaga, 

 Coruna, Platecrizotes) have them sharply impressed throughout, whilst in the others 

 they are superficial posteriorly. Many Miscogasterinae have two distinct apical 

 spurs on each hind tibia, whilst the majority of Pteromalinae have only one spur. 

 In Miscogasterinae this character is correlated to some extent with absolute size, 

 larger species having as a rule two spurs, small species only one ; most Sphegi- 

 gasterini, however, have only one spur irrespective of size, whilst Micradelini, 

 Termolampini and Pirenini also have one spur. Sphegigasterini have a conspicuous, 

 distinctly sculptured gastral petiole, which is found in only a few genera of 

 Pteromalinae. Micradelini, Termolampini, Pirenini, and nearly all Ormocerini have 

 only 12 antennal segments, whilst nearly all Pteromalinae have 13 antennal segments. 

 The various exceptions to the above general rules are allowed for in the keys and it is 

 hoped that careful attention to the other characters used will ensure correct identi- 

 fication in most cases. 



I have not found it practicable to divide Pteromalinae into tribes ; it may be 

 possible to do so when the genera are surveyed on a wider basis. 



