Studies of North American Bees 1 1 



joint 3 of the antennae is longer than 4 (e. g., adducta), but all 

 such cases are in autumnal species obviously allied to certain 

 species of Holonomada \vhich also fly in the fall. As with 

 Heminomada, however, the bulk of the species fall distinctly in 

 one or the other group, while the aberrant forms are not numer- 

 ous or decided enough to throw the recognition of Holonomada 

 open to serious question. 



Xomada and Heminomada come very close together, and their 

 separation is based largely upon color characters. Heminomada 

 differs from Nomada in the possession of yellow ornaments on 

 the head and thorax of the female, and the male has the yellow 

 bands on abdominal tergites 1-6 entire and continuous except 

 sometimes on i, while Nomada has the head and thorax without 

 yellow ornaments and the bands on the tergites interrupted, or 

 if entire with separated spots on the extreme sides of tergite 5. 

 But of course these characters do not hold absolutely, for some 

 species of typical Nomada (e. g., taraxacella, mediana, nigro- 

 fasciata, accepta, etc.) have yellow spots on the corners of the 

 face, at least in some specimens, while some species of Hemi- 

 nomada (e. g., collinsiana) entirely lack the yellow ornaments 

 of their close relatives. Even in the type species of the genus 

 (ruficornis) the male has entire and continuous yellow bands on 

 tergites 2-6 without separated lateral spots cut off on 5, at least 

 in the specimen before the writer, and thus would fall in the 

 genus Xanthidinm as defined by Robertson, or, as we are calling 

 the group, the subgenus Heminomada. But there are very few 

 species that cannot at once be satisfactorily placed in one or the 

 other of these two groups by the characters ascribed, and those 

 falling in each group seem to be for the most part closely allied 

 inter sc, so that on the grounds of expediency it would seem 

 advisable to continue to recognize, for the present at least, the 

 distinctness of Nomada and Heminomada as subgenera, realizing 

 at the same time that color characters chiefly applicable to one 

 sex only and not invariably in that are a very poor basis for 

 even a subgenus to stand upon, 



Gnathias is certainly a good subgenus, if it is not worthy of 

 generic rank. The dentation of the mandibles is not a character 



1 1 



