rTYCnOPTERA. 311 



the appearance of a double forceps ; but the upper pair seems to 

 represent the last dorsal segment of the abdomen, only very much 

 excised, and with the sides developed into elongated, forceps-like 

 appendages. Interpreted in such a manner, the structure of the 

 forceps in the American and in the European species is perfectly 

 homologous. The abdomen of the female is also narrowed at 

 the basis, broader in the middle ; the upper valves of the ovi- 

 positor of P. rufocinda are broad, convex above. Feet rather 

 strong, especially the hind tibite and tarsi ; coxte moderately de- 

 veloped ; tibias with strong, divaricate spurs at the tip ; the 

 tarsal joints in the male are attenuated at the extreme basis, 

 which is not the case in the females ; the foui'th tarsal joint of the 

 male has the basis incrassated, and with a tuft of hair ; the ungues 

 are very small, the empodia rather large and not linear, as usual 

 among the Tip. brevipalpi, but short and transverse. Wings 

 of moderate breadth, in some species comparatively broad ; the 

 surface is clothed with a microscopic pubescence, visible under a 

 lens of moderate power, and especially dense in the apical portion. 

 The peculiarities of the venation are numerous ; the principal 

 ones are : the absence of the subcostal cross-vein ; the length of 

 the first submarginal cell, which is usually twice the length of the 

 second ; in other words, it is not the second longitudinal vein 

 which is forked, as in most Tipulidse brevijjalpi, but the third ; the 

 central cross-veins are nearly in the middle of the wing ; there is 

 no discal cell ; the anterior branch of the fourth vein alone is 

 forked, and this fork is very short ; the latter portion of the fifth 

 vein is bisinuated ; a striking fold, almost like a spurious vein, 

 runs along the last longitudinal vein, crosses the anal cell, and 

 ends at the tip of the fifth vein ; beyond the fifth longitudinal 

 vein, there is only a single vein, and not two, as in all the Tipu- 

 lidae brevipalpi ; this vein apparently represents the seventh 

 longitudinal vein, the sixth being obsolete ; it is strongly arcuated 

 at the tip (compare the wing of P. rufocinda, Tab. II, fig. 19). 

 In the first posterior cell there is a longitudinal fold, which has 

 not attracted any attention before, but deserves to be noticed, as 

 it seems to foreshadow the supernumerary longitudinal vein, 

 which, in Protoplasa and its congeners, divides the first posterior 

 cell in two parts. This fold is especially distinct in the European 

 r. albimana, where it assumes the appearance of a spurious vein, 

 abruptly terminating somewhat beyond the inner end of the 



