64 R. W. CROSSKEY 



pupal abdomen are all characters conforming with their Division A groups. Con- 

 sidering the characters of both sexes and all stages together it seems certain that the 

 affinities of these Ethiopian species lie with those forms of Eusimulium from other 

 regions that have multiple parameral hooks, a large transverse male ventral plate 

 without median keel, simple male styles, four or six pupal gill filaments, and no 

 secondary annulation on the larval antenna, and they are here placed without 

 hesitation in the subgenus Eusimulium. It should be noted, however, that the 

 larval stage is known only of loutetense and the larval group characters cited in the 

 foregoing group definition may need modifying when larvae of rutherfoordi and 

 narcaeum are available. 



The Holarctic forms of Eusimulium in which there are many parameral hooks in 

 the parameral organ of the male hypopygium do not group so readily as those in 

 which there is one large main hook (latipes-gvonp, ruficorne-group, attreum-group) , 

 and a satisfactory group treatment for them has not yet been worked out : instead 

 of fitting the Ethiopian species to one of the named Holarctic groups, therefore, a 

 group is here proposed and defined for them with loutetense as lead species. One 

 North American species, S.(E.) furculatum (Shewell), however agrees so well with 

 loutetense and the definition of the loutetense-group that it is here assigned to the 

 group. 



S.(E.) furculatum differs from S.(E.) loutetense by having a simple non-necked 

 cocoon and an 8-filamented (instead of 6-filamented) gill, but in all other characters 

 shows an extraordinarily close agreement. The females of the two species have the 

 claw-tooth minute or virtually absent (an uncommon condition among Eusimulium 

 species), and the larvae of both have a very similar large mitre-shaped postgenal 

 cleft (Text-fig. 245) and unusually bold head-spot pigmentation. The larval 

 abdomen is unusual amongst species of Eusimulium but is similar in the two species : 

 the ventral papillae are small and bluntly rounded in profile, rather inconspicuous, 

 but are represented by a slightly swollen ridge in a more lateral position than normal 

 (i.e. not downwardly directed and subconical as in typical members of the sub- 

 genus), and the cuticle is sparsely covered with minute colourless hairs. The 

 agreement in unusual characters is so close that true phyletic affinity probably 

 exists. 



In the Oriental Region and Japan several species occur that, to judge from the 

 figures and descriptions, are closely similar to the species from Africa here forming 

 the loutetense-group. The evidence is not sufficient at present to assign these species 

 positively to the group, but attention is drawn to them here as the Ethiopian species 

 very likely have rather close affinity with at least some of them : the species referred 

 to are S.(E.) mie Ogata & Sasa, S.(E.) yamayaense Ogata & Sasa, and S.(E.) sasai 

 Rubzov from Japan, S.(E.) philippinense Delfinado from the Philippines and S.(E.) 

 rufithorax Brunetti from India. The male genitalia of Simulium feuerborni Edwards 

 and S. fuscinervis Edwards, together with other characters so far as known, resemble 

 those of the loutetense-group, but in these Indo-Malayan species the fore tarsus is 

 quite exceptionally slender and it is not certain that they are assignable to Eusimu- 

 lium at all. 



