890 Charles Paul Alexander 



2. Vein Ri obliterated by atrophy (this is also the case to a lesser extent in Tipula sub- 



fasciala and T. penohscot), or else (as in Brachypremna) the second anal vein very 

 short, not more than one-third the length of the first anal vein; legs very slender, 



filiform Dolichopezini (p. 928) 



Vein Ri present for its entire length (except in Tipula suhfasciata, T. penohscot, and 

 other species); second anal vein longer, one-half the length of first anal vein; legs 

 stouter and usually shorter than in Dolichopezini 3 



3. Antennae without verticils (see Stygeropis, below); flagella of the male antennae 



pectinate Ctenophorini (p. 930) 



Antennae verticillate (except in Stygeropis and most species of Holorusia); flagella 

 of the male antennae not pectinate (in some species of Nephrotoma and Tipula the 

 ventral face of the segments is often deeply incised, producing a serrate appearance, 

 but the antennae in the northern forms are never pectinate) 2'ipulini (p. 932) 



4. Four branches of radius reaching the margin (see note on Gonomyia blanda, below) 5 



Two or three branches of radius reaching the margin 9 



5. Tibiae spurred at tip 6 



Tibiae without spurs at tip. {Gonomyia blanda, p. 905, has R2 in close proximity to 



Ri at the wing margin, so that but three branches of the radius appear to reach the 

 wing margin; the tropical antochine genus Paratropeza will also run to here, and has 

 been mistaken by some authors for a Gnophomyia; the investigator must always 

 be on the lookout for such aberrant genera and species, especially when dealing with 

 tropical material.) Eriopterini (p. 901) 



6. Antennae with from 6 to 10 segments Hexatomini (p. 920) 



Antennae with more tlian 10 segments 7 



7. Sc> beyond the origin of i^s Limnophilini (p. 913) (except genus Ula) 



Sc2 before the origin oi Rs 8 



8. Antennae 17-segmented; wings pubescent Genus C/Za, tribe Limnophilini (p. 913) 



Antennae 13- to 16-segmented; wings glabrous Pediciini (p. 923) 



9. Tibiae spurred; an apparent fusion of Ri-\-2-\-3 to the wing margin so that but two 



branches of the radius are present (except in Phalacrocera neoxena, in which three 

 branches are present). The European hexatomine genus Cladolipes runs to here but 

 has only eight antennal segments; the South American species Psaronius ahnormis 

 also comes here, but may be readily separated by the very elongate subcosta. 



Subfamily Cylindrotominae (p. 926) 

 Tibiae without spurs; no contiguity of Ri and K2+3 at their tips 10 



10. Antennae 12-, 15-, or 16-segmented; claws usually without teeth on their lower side.. . 11 

 Antennae 14-segmented; claws with teeth on their lower side Limnohiini (p. 890) 



11. Cross-vein r lacking; Sc ending before the origin of the short Rs; Rz+d upcurved at 



the end, Ri+b bent strongly toward the apex of the wing producing a trumpet-shaped 

 cell R3; cell 1st Mi, when present, pointed at the inner end. 



Subgenus Leiponeura, genus Gonomyia, tribe Eriopterini (p. 905) 



Cross- vein r present or lacking; if lacking, Sc ends far beyond the origin of Rs; /?2+3 not 



strongly upcurved at end, and 7^4+5 not bent strongly toward the apex of the wing; 



inner end of cell 1st M2 not pointed Antochini (p. 897) 



The nearly wingless snow fly, Chionea, belongs to the tribe Eriopterini 

 (page 902). 



SUBFAMILY LiiTinobiinae 



Tribe Limnobiini 



The genera of the tribe Limnobiini may be separated by the following key : 



1. Rostrum elongated, longer than head and thorax together Geranomyia Hal. (p. 891) 



^ Rostrum not elongated, shorter than the head 2 



