486 . PROCEEDINGS OF THE \ATIU\AL MCtiEUM. vol. 44. 



6. Cornell University, Ithaca, New York. The rather extensive 

 collections made by Mr. Parish in the Province of Para, Brazil, in 

 January and February, 1912. 



7. Pomona College, Claremont, California. A collection of 5 

 specunens, 3 species, from Mexico; received through Prof. C. F. 

 Baker. 



8. Museu Rocha. Ceara, Brazil. Ten specimens of a Geranomyia; 

 received from Mr. P. Rocha. 



9. Staudinger and Bang-Haas, Dresden, Germany. A collection of 

 South American species; 22 specimens. 1-5 species: in author's 

 collection. 



10. Mr. E. B. Williamson, Bluffton, Indiana. Ten specimens, rep- 

 resenting 3 species, collected by Mr. Williamson on his recent trip 

 (see p. 483) ; in author's collection. 



In a very recent paper,' Dr. G. Enderlein has added much to our 

 knowledge of tropical crane-flies. The specific descriptions and the 

 figures are excellent, but the very free erection of genera must be 

 criticized. Of the 15 genera proposed in this paper, several are rank 

 synonyms of older genera, while the majority of the others are based 

 on trivial differences of venational or antcnnal charactei-s. In regard 

 to this subject it may be well to quote the words of Osten Sacken,* 

 as follows: 



To these successora I am free to give a piece of advice, as the result of more than 

 30 years' experience with the Tipulidse, and this is not to introduce new genera 

 prematurely. Large accessions of new forms, or of variation.s of already well-known 

 fonoa, must be expected from as yet unexplored, principally tropical, regions; but 

 tliese accessions although large will be slow in coming. Do not introduce new genera 

 for every slight deviation from a well-known tj-pe, because you would soon have no 

 end of new genera and a growing difficulty in discriminating between them. But do 

 not hesitate to establish a new genus for a form that can not be forced into any of the 

 existing genera and which shows distinctive characters in more than one organ of 

 its body. Such forms are not very common. 



The opinion that I have formed of Enderleins genera, after a 

 careful study of their characters, is given below. Mr. F. W. Edwards, 

 the well-known British authority on the Limnobinse, writes me that, 

 in his opinion also, most of the Enderlein genera rest on a very insuffi- 

 cient basis. 



t. Ctenacroscelis (p. 1) equals Eolorusia Loew. {E. rubiginosa 

 Loew has cross-vein r present in normal individuaLs, and connected 

 with R,+3 instead of Rj.) 



2. Icnomastax (p. 9) may be a valid genus, although the characters 

 on wliich it is based are rather trivial. 



1 Oiinther Enderlein. Studlen Qherdie Tipulidcii, LImnoljiidca C.vlindrotomiden,und I'tychopteriden, 

 ZoCilogische JatirbQcher. Abteilung fUr systematik, gcographie und Biologie der Tiere, 1912, vol. 32 

 pt. l.pp. 1-S5, fig. .11. 



' Osten Sackea, Studies ou Tlpulldii', pt. 2, B.Tliii. Km -/..it^.h- v.i :;i.p laa. 



