The Crane-Flies of New York — Part I 865 



the base of the sector and finally being lost by atrophy or by fusion 

 with Ri] thr exotic genus CeratocKeilus shows this intermediate condi- 

 tion very remarkably, and indicates clearly the manner in which this 

 extreme reduction of the sector in Toxorhina was brought about. 



The media {M, fig. 128, a), or medial vein, like the radial sector, in the 

 hypothetical type of an insect wing is twice dichotomously forked, the 

 closed cell, 1st M2, lying in the first fork. There are no known crane- 

 flies that show this condition except the^^doubtful fossil genus Rhabdino- 

 brochus, which is apparently based on an abnormal and imperfectly pre- 

 served specimen, and occasionally freak specimens of Tipula which indicate 

 this condition by spurs of varying length. These specimens show that 

 the single posterior branch of the media which persists is Mi, the spur 

 always lying on the cephalic side and representing the atrophied M3. 

 Corastock (1918:349, fig. 360) has interpreted the venation of Protoplasa 

 fitchii as showing all four branches of media, Mi being fused with Cui dis- 

 tally. That this is not the true interpretation is indicated by a study of 

 the other species of Protoplasa. The vein in cell M3 which Comstock 

 interprets as being the downward deflection of Mi is a supernumerary 

 cross-vein. In this remarkable family of flies, such cross-veins are very 

 often found in different cells of the wings. That the presence of a vein 

 in cell Ms is a specific character only is shown by the fact that it is lacking 

 in the related Protoplasa vipio O. S. Mi and Mi, comprising the anterior 

 fork of the vein, are either separate or fused at the wing margin; such 

 genera as Limnophila (Plates XXXVIII-XL) show a perfect succession, 

 from deep forks as in the exotic Limnophila epiphragmoides (Alexander, 

 1913 b: 543), thru less deep forks as in L. montana (Plate XL, 148), to 

 L. brevifurca (Plate XXXVIII, 125), which has a very shallow fork that 

 is sometimes fused clear to the wing margin, and further to the numer- 

 ous species of the genus (Plate XL, 150-157) in which there is a 

 permanent and constant fusion between these veins extending entirely 

 to the wing margin and obhterating the cell Mi. In all except the 

 most generahzed species, including nearly all of the Limnobiinae, the 

 medial-cubital cross-vein (ni-cu) is obliterated by the fusion of M3 + 4 with 

 the upward deflection of Cui; this fusion may be short or long, and is 

 discussed in connection with the cubitus. After breaking away from 

 the cubitus, Ms generally runs free to the wing margin, but in some 

 eases (as in Styringomyia and Phaiacrocera, Plate XXX, 8 and 9) it 



