104 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 5 1 



Fourth vein rounded at bend, second antennal joint short or only appre- 

 ciably elongate ; vibrissse inserted close on or only appreciably above 

 oral margin, about as far above as length of second antennal joint; 

 male claws usually not elongate, but longer than the short female 

 claws Latreillimyia, nom. nov. ( Latrcillia preocc.) 3 



3. Vibrissse inserted close on the oral margin 4 



Vibrissas inserted appreciably above the c ral margin, male claws 



strongly elongate. (Brazil, Pennsylvania.) L. (aberrant form) 



4. Second antennal joint appreciably elongate, female claws rather short. 



1 Minnesota.) L. ( intermediate form) 



Second antennal joint short, not at all elongate. ( Brazil. Mexico, New 



V rk.) Latreillimyia (typical forms) 



bifasciata J. C. Fabricius, leucopyga (van der Wulp) Williston. 



The character of the ciliate faeialia is more important than the 

 venational character and the same holds good of the vibrissal charac- 

 ter and the elongation of second antennal joint. As already pointed 

 out in a previous section of this paper, the relative length of the sec- 

 ond and third antennal joints will not hold for generic separation, 

 since the length and size of the third joint in these flies is largely a 

 sexual character. But the actual length of second joint taken inde- 

 pendently furnishes a good character. Only a few genera have the 

 second joint elongate. It may be compared in length with the first 

 joint. The first two joints do not vary sexually. 



Brauer and von Bergenstamm state that the claws of male are 

 elongate in Willistonia and short in Latreillimyia. It is doubtful 

 how far these characters can be relied upon, since they are also 

 sexual. The same authors also give as a character of Willistonia a 

 stump at the angular bend of fourth vein and the angle more ap- 

 proximated to hind margin of wing. These may hold good, espe- 

 cially the latter, but are not necessary for the separation of the forms 

 at present known to us. Further material will probably call for 

 their use. 



The writer pointed out in 1892 (Trans. Am. Ent. Soc, xix, p. 

 89) that bifasciata has the faeialia strongly ciliate and bicincta has 

 not; that five specimens of bicincta from New Mexico had the third 

 antennal joint scarcely longer than the second, which means that the 

 second joint was strongly elongate; that three specimens from New 

 York were easily referable to bifasciata, and one from Jamaica to 

 bicincta; and that, while the parafacials are bare in both species, the 

 whole anterior aspect of head is altogether more bristly in bifasciata, 

 which possesses also greater hairiness of cheeks. 



The elimination of Bclvosia, argued for by Brauer and von Ber- 

 genstamm, is not permissible under the rules of the International 

 Code. Its maintenance fortunately does not conflict with the genus 

 JJlllistonia, since bicincta differs generically from csitricns. 



