Diptera of St. Vincent [IVetit Indies). 297 



were originally described as having twenty-eight joints, 

 and the true number, sixteen, was not known till speci- 

 mens were examined by Loew. The joints in this sex 

 are very closely united, and it is only by close exami- 

 nation that the number can be made out. From 

 P. fiisca, Wiedemann, and P. ohscura, Macquart, which 

 seems to be a distinct species, the present appears to be 

 so closely allied, that the short descriptions will hardly 

 distinguish them. However, Wiedemann figures the 

 female antennte as elongate, like those of the males, but 

 ■with simple, not constricted joints. In the present 

 species, the female antennas are short, in fact not longer 

 than the thorax, and are of the ordinary Tipulid struc- 

 ture ; that is, the joints are oval, slightly hairy, and are 

 easily distinguishable. The first joint is not cylindrical, 

 and is not provided with short, closely-set hairs, as is 

 described. The figure given by Macquart shows short 

 and abundant hairs on the swellings of the male 

 antennas, very different from the long, delicate hairs of 

 the present species. 



Specjes Ikcert^ Sedis. 



Two species of Tipulidw, represented by single, more 

 or less mutilated specimens, I cannot locate. I give 

 figures herewith, which will, I believe, render the identi- 

 fication not doubtful. One of them (PI. X., 6gs. 72 and 

 72a) possibly represents a new genus. The tip of the 

 antennas is broken ofi", otherwise the specimen is com- 

 plete. If the antennae are but 14-jointed, the species 

 would be located in RJiipidia, from which, however, the 

 structure of the male organs show distinct differences, 

 and the antennae are hardly sufficiently pectinated. If 

 there are sixteen joints, Schiner^s and Osten Sacken's 

 tables will carry the species to Antocha. This genus and 

 Thanmaslopteva are almost the only ones in this group 

 which I do not know. Still, the neuration is sufficientlv 

 distinct to render its location with ^?i^ocAa practically out 

 of the question. 



