24 G. E. J. NIXON 



- Tergite (2 + 3) always with some sort of delimited median field which, if it tends to 



be coextensive with the whole of tergite 2, is normally much shorter than 3, except 

 in a few species of the circumscriptus-group with strongly sculptured 1st and 2nd 

 tergite ; if such a median field is hardly delimited, then this is because the whole 

 of tergite (2 -f- 3) is smooth and polished ; tergite 3 never with obvious sculpture, 

 its surface having at most a dull sheen ........ 37 



37 Propodeum with a strong, medial keel ........ 38 



- Propodeum without such a keel . . . . . . . . .41 



38 Polished, lateral field of the scutellum large enough to cut off between itself and the 



disc a long, parallel-sided, foveate groove ; ocelli in a low triangle, with the trans- 

 verse, posterior tangent to the median ocellus cutting deeply into the posterior 

 ocelli. 



Inner spur of the hind tibia very long, fully three fifths as long as the hind 

 basitarsus. Indo-malayan region ..... sundanus- group (p. 170) 



- Polished, lateral field of the scutellum much smaller, short from back to front and 



hence cutting off between itself and the disc a rugose tongue that gradually narrows 

 posteriorly ; ocelli in a higher triangle, the posterior, transverse tangent to the 

 median ocellus at most touching the posterior ocelli ..... 39 



39 Cubitellan cell of hind wing extremely short on radiella and cubitella, hence much 



higher than long (so much higher than long, and thus narrow, in the annulicovnis- 

 group, as to be easily overlooked) ........ 40 



- Cubitellan cell less short and longer than high. America. Oriental region 



nen'on-group (p. 142) 



40 Cubitellan cell at most about one and a half times times higher than long on the 



cubitella ; tergite 1 only slightly narrowed posteriorly, its turned over, posterior 

 part rugose all over. 



Transverse cubitus rather long ; vannal lobe beyond its widest part straight or 

 slightly concave and here with at most an occasional projecting hair ; median field 

 of tergite (2 + 3) rugose, rectangular, strongly transverse ; inner spur of the hind 

 tibia much longer than the outer one and reaching slightly beyond the middle of 

 the hind basitarsus. Brazil. N. America .... sesiae- group (p. 144) 



- Cubitellan cell at least twice as high as long (Text-fig. 130) ; tergite 1 very strongly 



narrowed posteriorly, almost wedge-shaped. 



At least the mesoscutum and mesopleurum yellow ; propodeum, apart from the 

 median keel, highly polished ; inner spur of the hind tibia considerably longer 

 than the outer one and almost reaching the middle of the hind basitarsus ; anterior 

 declivous part of tergite 1 deeply channelled. W. Indies. Brit. Guiana 



annulicornis- group (p. 140) 



41 Tergite 1 usually markedly narrowed to apex, long, more or less wedge-shaped. 



Vannal lobe beyond its widest part more or less straight and virtually without 

 projecting hairs ; ovipositor sheath always longer than the hind tibia. Where 

 tergite 1 is almost parallel-sided as in inunctus, only the shape of the vannal lobe 

 separates the group from the laevigatus-group. Indo-oriental region 



tnycetophilus-Qroup (p. 171) 



- If tergite 1 is strongly narrowed to apex (as in nearly all species of the metacarpalis- 



and cir cunts criptus-gr onps) , then the vannal lobe has a more or less evenly convex 



edge which is fringed throughout ; or else the metacarp is unusually short . . 42 



42 Tergite 1 strongly constricted at apex, never so long, nor so clearly wedge-shaped, as 



in the mycetophilus-group. 



Hypopygium short and rarely with longitudinal creases on each side of middle 

 line (but cf. murinanae) ; median field of tergite (2 + 3) triangular, shorter than 

 the hind tibia (but cf. murinanae and dioryctriae) ; stigma often pale but never 

 with contrasting, pale basal spot ......... 43 



