18 G. E. J. NIXON 



3 Notaulices deeply impressed as intricately rugose furrows. 



Ocelli in a high, virtually equilateral triangle. W. Indies 



aciculatus- group (p. 181) 



- Notaulices never impressed, though their course is very frequently indicated by a 



band of duller, denser, sometimes rugose, sculpture ..... 4 



4 Propodeum with an areola and almost always with costulae that define on each side 



a postero-lateral area ; if the costulae are absent and the areola is weakly defined, 

 then the propodeum tends to be dull and sculptured all over and/or the vannal lobe 

 beyond its widest part is concave and without trace of a hair-fringe here . . 5 



- Propodeum rarely with trace of an areola and never with costulae ; if there is a faint 



indication of an areola, then the edge of the vannal lobe beyond its widest part is not 

 concave, though it may be virtually straight here and shows the merest trace of a 

 fringe or the fore wings are dappled . . . . . . . . 12 



5 Edge of vannal lobe beyond its widest part usually deeply concave, the concave edge 



being completely free from projecting hairs ; if, rarely, this edge is virtually 

 straight and shows an indication of a hair-fringe, then either the median field of 

 tergite 1 is more or less coextensive with the whole surface of tergite 2 and almost 

 as long as tergite 3 or, tergite 1 is very long, strongly narrowed posteriorly and 

 wedge-shaped or, the costulae of the propodeum are situated anterior to middle so 

 that the postero-lateral areas do not appear as transverse ..... 8 



- Vannal lobe beyond its widest part evenly convex and with a distinct hair-fringe 



throughout. 



Tergite 1 never wedge-shaped ; postero-lateral areas of the propodeum always 

 distinctly transverse ; mesoscutum with a prevailing sculpture of sharp puncta- 

 tion at the posterior end of the imaginary course of the notaulices ... 6 



6 Tergites 2 and 3 subequal in length. 



Tergite 1 broadly triangular, not longer than apically wide .... 7 



- Tergites 2 and 3 not equal, 2 being shorter than 3 or, if not, then tergite 1 is distinctly 



longer than apically wide. 



Ovipositor sheath always projecting freely beyond the apex of the gaster. Old 

 and New World ........ ulfor-group (p. 126) 



7 Tergite 3, as well as 2, densely rugose all over ; hind spurs short, subequal, the inner 



one not reaching middle of hind basitarsus. N. America bucculat ricis-group (p. 148) 



- Tergite 3 polished ; tergite 2 coarsely rugose-striate ; hind spurs not subequal, the 



inner one much longer than the outer one and reaching the middle of the hind 

 basitarsus. 



Ovipositor sheath hardly exserted ; 1st abscissa of the discoideus much shorter 

 than the 2nd. Costa Rica ...... paradoxus- group (p. 127) 



8 Hypopygium short, evenly sclerotised throughout, not folded along the middle line 



and without trace of lateral, longitudinal creasing ; cheeks with a whitish, semi- 

 transparent patch ; tergite 1 on its posterior, horizontal surface with a narrow, 

 deep furrow that has strongly raised margins ; these margins unite or almost unite 

 anteriorly to form a short, conspicuous keel. N. and S. America nigriceps- group (p. 126) 



- Hypopygium, even if short and not readily visible in the dead insect, then at least 



folded along a more weakly sclerotised middle line and nearly always with dis- 

 tinctly lateral creasing ; if tergite 1 shows a longitudinal furrow (frequent in 

 ater-group) then this is broader, less sharply defined and lacks raised, lateral 

 margins ; cheeks without a whitish patch ....... 9 



9 Tergite 2 without a delimited median field, the whole tergite densely rugose-striate, 



and fully as long as tergite 3. 



Hind coxa and ventral surface of gaster reddish yellow and together with the 

 almost entirely yellow hind femur in sharp contrast with the blackish hind tibia ; 



