866 Charles Paul Alexander 



unites with Ml +2 for a short distance, obliterating the medial cross-vein. 

 In some genera — Bittacomorpha (Plate XXX, 3), Bittacomorphella 

 (Plate XXX, 4), Hexatoma (Plate XXXVII, 112), Diotrepha, and many 

 species of Trentepohlia — but one branch of the media reaches the wing 

 margin, and in these cases the posterior branch has cither fused with 

 the cubitus (as in Hexatoma) and reaches the margin by this fusion, 

 or has been lost by atrophy. Needham (1908:227-229) beheves the 

 posterior branch of the media is lost only by atrophy, and undoubtedly 

 this is true in most instances; the series of Polymera, however, a tropical, 

 American genus studied by the author (Alexander, 1913 b: 526-535), 

 showed an interesting condition indicating that the veins may be vmited 

 by fusion, and similar conditions may exist in the genus Rhaphidolabis 

 and in the South African species Gonomyia hrevifurca. The entire end of 

 M-i is lost by atrophy in four known species of Dicranomyia, one of these 

 being D. whartoni (Plate XXXI, 15). The cell 1st M^ (discal) is in many 

 cases opened by the atrophy of part of Ms, leaving the tip of Mz attached 

 to the medial cross-vein (as in Ormosia, Plate XXXIV, 59-64, and 

 in Gonomyia, Plate XXXVI, 92 and 93) ; in other cases it is the medial 

 cross-vein (w) that is atrophied, opening the cell (as in Dicranomyia, 

 Plate XXXI, 14 and 16, in Cryptolabis, Plate XXXVII, 101, and in 

 many genera of the Pediciini, Plate XLI, 172-174). 



The cubitus (Cu, fig. 128, a), lying between the media and the anal veins, 

 is the most constant and, after the radius, the most powerful vein of 

 the wing. There are always two branches, which are never lost. At the 

 fork, the anterior branch, Cui, is directed strongly forward, so that in 

 all but the most generalized forms it 'simulates a cross- vein and from 

 its conspicuous size it has long been termed the great cross-vein; this 

 deflection is the basal deflection of Cui of the Comstock-Needham system, 

 and the pars ascendens of Bergroth. In the more generalized groups, 

 such as Tanyderidae (Plate XXX, 1), Ptychopteridae (Plate XXX, 2^), 

 a very few Limnobiinae — as some species of Tricyphona (Plate XLII, 

 184 and 185) — and many of the TipuUnae (Plate XLIII, 195-197), the 

 medial-cubital cross-vein (rn-cu) is persistent, but in the great majority of 

 cases it is lost by the fusion of Cui and Ms + i. As already stated, this 

 fusion may be very short — merely a point of contact (punctiform), as 

 in most species of Tipula (Plate XLVI, 222) — ^ or it may be subequal 

 in length to the cell 1st M2, the deflection of Cui entering the media at 



