BY FREDERICK A. A. SKUSE. 777 



Rostrum the length of the thorax only in G. lutulenta and 

 annulata; as long as the thorax and head taken together in G. iiicta 

 and/?<sca. Antennae rather short, joints elliptical, sessile; sub- 

 cylindrical in G. fusca. The basal joint of the palpi is always 

 long and slender, about twice the length of the second joint 

 (PI. XXIV., figs. 46-48). 



Venation similar to that of Dicranomyia. Auxiliary vein in 

 G. ^J^c^« reaching costa nearly opposite but somewhat beyond 

 origin of prsefurca, in the other species reaching considerably 

 beyond. Sub-costal cross-vein always close to the tip of the 

 auxiliary vein. The second longitudinal vein is rather angularly 

 bent near its origin in the four species known to me, in G. picta 

 and annulata Avith even a short stump of a vein at the angles. 

 The sub-marginal cell is much longer than the first posterior cell. 

 The discal cell is open in one specimen of G. lutulenta, and 

 coalesces with the third posterior cell. In G. picta the great 

 cross-vein is situated considerably before the inner end of discal 

 cell. 



In the male forceps the rostriform appendage of the fleshy lobes 

 bears two short stifi" bristles ; in G. fusca this is situated much 

 lower down the lobe than usual. The falciform appendages are 

 long and curved. The anal style is large in G. picta (PI. xxiv., 

 fig. 49), but small and hammer-shaped m fusca (PI. xxiv., fig. 50). 

 These difi'erences however may be only of specific importance. 



Until further species have been studied it is impossible to fully 

 define the three following sub-genera ; other characters may be 

 ultimately found to be constantly associated with the differences 

 in the palpi. 



1. Sub-genus Geranomyia. Palpi two-jointed. Proposed for 

 the single species G. picta (PI. xxiv., fig. 46). 



2. Sub-genus Triphana. Palpi three-jointed. Proposed for 

 the reception of t>vo species, G. lutulenta and annulata (PI. xxiv. 

 fig. 47). 



