NO. 2219. A REVISION OP THE CREMASTINI—CUSHAtAN. 611 



the species should be run. In the first dichotomy, for example, 

 agreement with two of the three characters allies a species with a 

 group indicated thereby. These groups of character, rather than 

 single characters, are used in the table because, bj^ their use, the 

 species aiTange themselves in a manner more in accord with their 

 apparent natural affinities. 



The extent of secondary sexual modification is so variable in the 

 males that it seems inadvisable to attempt, in most cases, to associate 

 the sexes in the material at hand. This can, with assurance of 

 correctness, be done only by biological observations, either in regard 

 to mating or by rearing the sexes at the same time and from the 

 same host. On this account and because a number of species have 

 been described only in the male, the two sexes are tabulated 

 separately. 



KEY TO NORTH AMERICAN SPEaES (FEMALES). 



Petiole with ventral margins of tergite approximating below and nearly enclosing 

 sternite, never parallel; propodeum reaching to or beyond middle of dorsal 

 surface of hind coxae; stigma narrow with radius originating at or near the 

 middle (subgenus Cremastus Gravenhorst) 1. 



Petiole mth the ventral margins of tergite widely separated, parallel or nearly, 

 not enclosing sternite; propodeum not reaching to middle of dorsal surface 

 of hind coxae; stigma broad and with radius originating beyond middle 

 (subgenus Zaleptopygus Viereck) 19- 



1. Head subrostriform; malar space much longer than basal width of mandible. 



longigenalis, new species. 



Head not subrostriform; malar space rarely as long as basal width of mandible, 



most frequently much shorter 2. 



2. Thorax testaceous or yellow, sometimes more or less black dorsally or ven- 



trally 3. 



Thorax black, sometimes with yellow markings dorsally or laterally 14. 



3. Eyes distinctly, though slightly, divergent below 4. 



Eyes parallel within 8. 



4. Second tergite nearly six times as long as wide, at base; malar space as long as 



basal width of mandible; ovipositor sheath scarcely longer than first tergite. 



longiventris, new species. 



Second tergite distinctly less than six times as long as basal width; malar space 



usually less than basal width of mandible, occasionally subequal; ovipositor 



sheath much longer than first tergite ' 6. 



6, Scutellum with strong lateral carinae extending nearly o* quite to apex; head 



in front view subtriangular G. 



Scutellum with carinae developed only at base; head in front view suboval. . 7. 



6. Olypeus little more than half as long as interfoveal line (shortest distance between 



clypeal foveae) and subacutely rounded at apex; malar space subequal 



to basal ^-idth of mandible; spiracles of second tergite distinctly before 



middle; areola and petiolar area separated by a strong carina. 



ruficeps, new species. 

 Clypeus about three-fourths as long as interfoveal line, evenly strongly rounded 

 at apex; malar space distinctly less than basal width of mandible; spiracles 

 of second tergite in middle; carina separating areola and petiolar area sub- 

 obsolete gracilis, new species. 



