258 PROCEEDINGS OF THE NATIONAL MUSEUM vol. io9 



and laterally. The basal arch of the aedeagus is connected to the 

 posterior excavation of the ninth sternum by a hyaline membrane 

 which in a few species is partially covered with spinules or spicules. 

 The parameres are usually a pair of sclerotized internal rodlike sclerites 

 with knobbed bases and ventrally directed distal points, but they 

 are subject to great modifications in the shape and direction of the 

 basal knob, the middle stem, or the distal point. In some groups of 

 species the parameres may fuse mesad and become platelike. 



Systematic Discussion 



The modern classification of the genus Culicoides had its beginnings 

 as recently as 1937, when Root and Hoffman grouped the North 

 American species according to the characters of the male genitalia 

 supported by a number of external characters. Edwards followed 

 and enlarged upon this system in 1939 with a grouping of the Palae- 

 arctic species. In 1948 Fox erected the subgenus Hof mania for a 

 number of Neotropical species which he differentiated from Culicoides, 

 s. str. Ortiz (1950) accepted and strengthened Fox's concept of the 

 subgenus Hof mania. Wirth (1952a) attempted only to follow Root 

 and Hoffman's outline in assigning the California species to groups 

 without respect to subgenera. None of the recent regional revisions — 

 such as those by Causey (1938, Siam), Fiedler (1951, South Africa), 

 Gutsevitch (1952, Russia), Lee and Reye (1953, Australia), and 

 Foote and Pratt (1954, eastern United States) — attempt any group- 

 ing of species. 



The first important attempt to bring the genus Culicoides into a 

 phylogenetic system was by Khalaf (1954), who placed the species 

 of the world in four subgenera — Culicoides, s. str., Oecacta Poey, 

 Selfia Khalaf, and Monoculicoides Khalaf — assigning all the species 

 whose genitalic characters had been figured in the literature to five 

 complexes and 21 groups within these subgenera. Although Khalaf 's 

 concept of important phylogenetic characters was basically sound, 

 he was hindered by the limited information available in the literature. 

 In a genus whose taxonomy is expanding so rapidly we may expect 

 numerous reshufflings of groups as new species are described and our 

 knowledge of the older species improves. Vargas' (1953a) proposal of 

 the subgenus Beltranmyia for species of the crepuscularis group and 

 Fox's (1955a) erection of the subgenera Avaritia for the obsoletus 

 group and Macjiella for the phlebotomus group furnish badly needed 

 subgeneric names, but numerous species will probably pass through 

 various taxa before being assigned to their correct phylogenetic niches. 



We believe that in a genus as large and compact as Culicoides 

 the number of subgenera will eventually have to be increased fourfold 



