2 PROCEEDINGS OF THE NATIONAL MUSEUM vol. 68 



as a member of the family Myrmosidae, though at one time the 

 species now placed in the genus were considered as belonging to 

 MutillidiTR and were linked with species of Cyphotcs under that generic 

 designation. It is m^^ opinion that Brachyclstis })elongs to the 

 Scoliidre. of which it should form a subfamily, or if one considers the 

 major group as a super-family then it should be considered as a 

 family in the complex. This family or subfamily status is largely 

 a matter of opinion and until we have advanced farther in our knowl- 

 edge of the Hymenoptera as a whole thi'oughout the entire range of 

 their occurrence a dogmatic opinion would be ill-advised. 



Realizing that this work is merely a small contribution to the 

 elucidation of the classification of the group upon which it is based 

 and tliat with future accessions botli of species and of workers it must 

 be carried much farther I have deposited practically all my dissected 

 specimens and types as well as all uniques and specimens not abso- 

 lutely necessary for carrying on the work of the bureau of which I 

 am a member in the collection of the United States National Museum 

 where they will be accessible to any interested student. 



Family SCOLIIDAE 



In this family the pronotum is carried backward at its upper pos- 

 terior angles to a point above the level of the tegulae where there is 

 a rather sliarp angle (fig. 1 , la); the mesosternum is carried backward 

 over or between the bases of the mid coxae in the form of a cleft 

 plate; the apical ventral segment of the abdomen in the males is 

 produced in the form of a single curved spine (figs. 2, 3) or three such 

 spines; the hind wing has the anal lobe present; the basal tergite of 

 abdomen has a linear sulcus below spiracle extending from base to 

 apex, beyond which the margin is flexed ventrad, closely adhering to 

 surface of the sternite. 



I present a key for the separation of the four subfamilies of Scoliidae 

 that occur in North America. The females of Brachycistiinae are 

 unknown, unless my surmise is correct that MutilJa pecuHaris 

 C^resson belongs here, an opinion that is expressed later on in this 

 paper. 



Subfamily Brachycistiinae 



This subfamily is distinguished from the M3'rmosidae by the vena- 

 tion of the fore wings (figs. 4, 5, 6), in whicli they more closely re- 

 semble Photopsinae; the forward opening antenna! sockets; ])resence 

 of a scalelike extension of the mesosternum between the bases of the 

 mid coxae; the spinelike structure of the eighth abdominal sternite, 

 in which character they are similar to Ti})hiinae; the structure of the 

 male hypopygium; and the lack of a platelike extension on hind coxae. 



