114 Mr. F. "W. Edwards 07i the 



at, and the following points of resemblance between the two 

 genera become more apparent: — (1) the radial sector forks 

 basally to the R-M cross-vein, or, in other words, the small 

 cross-vein is situated on the third longitudinal vein, instead 

 o£ on the prsefurca ; (2) the lower branch of the cubitus is 

 distinctly sinuous, the cell Oui having a convexity on the 

 lower side towards the base. The resemblance between the 

 venation of Mycetohia (fig. 5 a) and Olhiogaster (fig. 5 b) is in 

 some respects still more marked, and it is worthy of particular 

 notice that in Olhiogaster africamis, Edw., and still more 

 conspicuously in 0. sackeni, Edw., the lowest of the three 

 veins arising from the discal cell (M3) is less strongly 

 cliitinized than the other two, suggesting that the venation of 

 Mycetohia lias arisen directly from that of Olhiogaster through 

 the obsolescence of M3. 



In almost all other MycETOPHILID^ the radial sector, when 

 it forks at all, does so nearer the wing-apex than the position 

 of the R-M cross-vein, which, besides, is usually sloping, and 

 not straight as it is in Mycetohia. The only exception to this 

 rule is the genus Pachyneura^ in which the radial sector forks 

 exactly at the R-M cross-vein. In all otlier MYCETOPHiLiDiE, 

 with the exception of jLfW?»?/ia {Glaphyroptera) and its allies, 

 the cell Cui is concave instead of convex on its lower margin. 

 Ditomyia and Symmerus, genera which, together with Myce- 

 tohia^ have been made to form the subfamily Mycetobiin^, 

 agree in both these respects with the other Mycetophiltd^, 

 and I therefore consider that they are not at all closely 

 related to Mycetohia. The genus Mesochria, recently described 

 from the Seychelles Islands, is, on the other hand, closely 

 related to Mycetohia. Its venation is very interesting, as the 

 media is evanescent ; it evidently represents a further stage 

 of evolution from Mycetohia, in which the lowest branch of the 

 originally three-branched media has already disappeared. 



A comparison with other Diptera as regards the position 

 of the radial fork reveals the fact that in the Tipulidse, 

 Culicidse, Psychodidae, Orthorrhapha Brachycera, and 

 Oyclorrhapha it always takes place anteriorly to the R-M 

 cross-vein, wliereas in the Chironomidse, Simulidre, and 

 Bibionidffi the radial sector is usually simple, but when it 

 forks does so beyond the R-M cross-vein. An apparent 

 exception to this rule, however, is the Bibionid genus Eupei- 

 tenus. It is also noteworthy that in the families of the 

 first group the media is primitively three-branched, while in 

 those of the second it is never more than two-branched. 



