90 HYMENOPTERA. 



character as to be almost worthless except when supple- 

 mented by other characters ; and the former when used in 

 such a general way as " face below antennae yellow," is also 

 worthless, since it may fit any one of several species. Most 

 of the earlier papers, then, must be ignored unless supple- 

 mented by later ones. 



Subsequent writers, realizing the inadequacy of the early 

 papers, went considerably more into detail ; but they too in 

 many cases based their distinctions on utterly inconstant dif- 

 ferences. One of the most confusing of these is the color 

 and venation of the wings, which has been emphasized 

 strongly by some. The actual shade of the wing, whether 

 darkened or not, is of no value whatever, as almost any 

 large series of specimens will show ; and the minor differ- 

 ences in wing venation such as, " second submarginal cell 

 little narrowed toward marginal " or " first recurrent nervure 

 entering second submarginal cell", show as much variation 

 in a species as between species, and should not be used. 

 Another character causing confusion is the presence or abs- 

 ence of a spot on the tegula. This is probably the most 

 variable of the color markings, and yet it has been repeat- 

 edly used as a diagnostic character. Along with these a 

 number of color characters have given trouble. Certain of 

 them, such as the yellow on the collar or the shape of the 

 face marks, are valuable characters, if shozvn to be fairly con- 

 stant, and furnish very easy means of distinguishing species 

 which are also differentiated by other characters. In some 

 species, however, and when not supported by a large series, 

 they cannot be depended upon. This fact, as already men- 

 tioned, was formerly overlooked, and in consequence we 

 have numbers of specimens described as new because of 

 some such difference. In the species viodestus alone, no less 

 than eight or ten, and perhaps more, forms have been de- 

 scribed as new species for this reason. 



Up to 1911 fifty-eight species of Prosopis had been described 

 from the United States and Canada. The number is not 

 large in comparison with that of other allied genera, but in 

 the final analysis probably not more than one-third or one- 



