24 R. W. CROSSKEY 



segregate ; but it should be emphasised that the diagnosis may require amendment 

 when the other stages of herero or turner i become known, or when further species 

 are discovered (as seems likely when the habitats favoured by Ethiopian Pro- 

 simulium s.l. species are more thoroughly investigated). Furthermore, I think it 

 almost certain that when the characters of Australian Prosimulmm-like forms, and 

 possibly also those of Mexico and Guatemala, are studied in detail it will be found 

 necessary to bring these species into the subgenus Paracnephia, when a widening of 

 the definition of the concept may be needed (as, for example, the larvae of some 

 Central American forms possess ventral papillae, and the pupa of at least one 

 Australian species has large lateral pleural plates on the abdomen). 



Some species of Paracnephia have fully toothed mandibles and maxillae in the 

 female and are presumably biting forms that may be anautogenous ; these have 

 from 10-15 outer teeth and 24-35 inner teeth on the mandible, and from 10-15 

 outer teeth and from 16-18 inner teeth on the maxilla, and include turneri, brincki 

 and thornei. Other species have the apices of the female mandibles and maxillae 

 atrophied, without definite teeth but with a jagged margin on the mandibles and 

 with haired apex on the maxillae : these forms presumably cannot bite and are 

 autogenous ; they include the type-species, muspratti, and also harrisoni and 

 barnardi. The condition of the female mouthparts in herero is unknown (it is 

 undesirable to break up the head of the unique holotype until this should prove 

 essential) . The condition of the female mouthparts does not correlate in any obvious 

 way with other characters, so that there is evidently no natural species-grouping 

 solely on this character, but the species of Paracnephia are here treated in three 

 groups using other criteria (which may not necessarily hold up as satisfactory 

 characters for delimiting groups when more becomes known of this poorly collected 

 and represented subgenus). The three groups are : — 



brincki- group. Vein i?i with hair-like setae only. Calcipala small, similar to muspratti- 

 group. Pupal gill with 6 divergent filaments (Text-fig. 41) . Fourth abdominal segment of pupa 

 ventrally with only one hook on each side. Terminal hooks of pupa relatively short and spike- 

 like (Text-fig. 44), not sinuous. 



Included taxon. Prosimulium (Paracnephia) brincki (de Meillon) comb. n. 



muspratti- group. Vein i?i with hair-like setae only. Calcipala small or of medium size 

 (Text-figs. 23, 24). Pupal gill with 10-22 filaments (of type shown in Text-fig. 40). Fourth 

 abdominal segment of pupa with two hooks each side ventrally. Terminal hooks of pupa not at 

 all reduced, long and slightly sinuous (Text-figs. 42, 43). 



Included taxa. Prosimulium (Paracnephia) barnardi (Gibbins) comb. n. ; 

 P. [P.) harrisoni (Freeman & de Meillon) comb. n. ; P. (P.) herero (Enderlein) 

 comb. n. ; P.{P.) muspratti (Freeman & de Meillon) comb. n. ; P. (P.) thornei 

 (de Meillon) comb. n. 



turner i- group. Vein R± with spiniform as well as hair-like setae (similar to costa). Cal- 

 cipala large and of the Simulium type (Text-fig. 25). [Pupal characters unknown.] 



Included taxon. Prosimulium (Paracnephia) turneri (Gibbins) comb. n. 



