CLASSIFICATION OF THE BEES. 53 



linous and parasitic Gynipids are offshoots from the plant- feeding 

 Cynipids, and that the parasitic Chacidids are offshoots from the 

 plant-feeding Chalcidids, Agaonidse, Eurytomidse, etc. 



This same kind of development or evolution is also found existing 

 among the wood, sand and digger wasps: Masaridse and Chrysididse 

 originated from the Vespidse and Eumenidse, while Ceropales, a 

 genus in the Pompilidse, living parasitieally in the nests of other 

 species of Pompilids, is an offshoot from Pompilus. 



Among the bees, Psithyrus (— Apathus) is clearly an offshoot 

 from Bombus, as Friese has shown. 



The Nomadidse evidently came directly from the Anthophoridse 

 and other pollen-gathering bees, while the Stelididse for the most 

 part had their origin among the Megachilidse — the subfamily Steli- 

 dinse coming from the Anthidiinse, the Coelioxinse from the Mega- 

 chilinse, etc. 



The Panurgi<Ue, or at least part of them, are of quite recent 

 origin, some genera being evidently only recently evolved from 

 Andrena and Halictus, while others seem to have come from the 

 Anthophoridse. 



In our classificatory work I believe this law of evolution or de- 

 velopment (for certainly evolution is no longer a theory, bat a demon- 

 strated law of the universe), should be borne in mind, and, as far as 

 possible, the origin and affinities of the complexes, such as families, 

 groups, genera, etc., should be interpreted and shown in our tables. 



With this object in view, in the present work, therefore, I have 

 not recognized the three sections of bees defined by Schmiedeknecht, 

 since, to do so, would separate widely closely allied families; but 

 instead I have interpolated, in my tables, the inquilinous and para- 

 sitic bees among the true honey-making bees, in order to demon- 

 strate, what I conceive to be, their true affinities and relationship to 

 each other. 



The families into which the Apoidea are now divided may be 

 separated by the use of the following table : 



Superfamily I. Apoidea. 

 Table of Families. 



Labium or tongue flattened, most frequently shorter than the mention, rarely 

 very much louger (some PanurgidseJ ; basal joints (if labial palpi cylin- 

 drical, the first joint sometimes very elongate or thickened, hut still 

 neither flattened nor uulike the following 6. 



TRANS. AM. ENT. SOC. XXVI. MAY, 1899. 



