340 THE CANADIAN ENTOMOLOGIST. 



Halictoidifice, Macropis, Panurgiiue except Panurginus, Megac/iilince, 

 Eucera, Affimobates, Pasties, Biastes. 



In Perditinoe the long submarginal cell is the first, the next is 

 the third, and the dividing nervure is formed by the coalescence of the 

 first and second transverse cubilals. 



These three types of venation are no evidence of affinity, but must 

 have had an independent origin from wings with three submarginal cells. 



Characters connected with nectar-sucking, pollen-collecting, and the 

 inquiline habit ate, as compared with the venation, more physiological. 

 They no doubt usually indicate relationship, but they often obscure it. 

 I think they maybe relied upon when they are confirmed by the venation. 

 On the other hand, when the venation indicates relationship it is hard 

 to disprove it by evidence drawn from the tongue, pollen scopte, 

 and characters connected with the inquiline habit. If the venation is of 

 a low type, such as that o{ Andreiia or Halictus, I think the genus should 

 be assigned to a low position in spite of the tongue and scopje. Usually 

 specializations of venation, tongue and scopte go hand in hand, but often 

 they do not. A slightly specialized tongue may go with a highly special- 

 ized scopa, as in Macropis. Or a slightly specialized scopa may go with 

 a highly specialized tongue, as in Ceratina. In such cases I assign the 

 bee to a low or high position according to the character which seems to 

 be supported by the venation. 



To my mind the most egregious errors in the classification of bees 

 seem to be in the location of the inquilines. The analogy of the case of 

 Bembus and Psithyrus seems to me of prime importance in the solution ot 

 the question. No one doubts that they are more closely related to 

 each other than either is to any host bee or any inquiline. The 

 only differences Psitliyrus shows are such as are correlated with the 

 inquiline habit. Delia Torre's last subfamilies of bees are : 12 Bombinje ; 

 13 Psithyrinse, 14 Apinte. To give Psithyrus this rank, I think, involves 

 a great systematic error only equalled in the old physiological classifica- 

 tion of Shuckard. The latter author calls those bees which carry pollen on 

 their legs scopulipedes ; those with abdominal brushes, dasygasters. 

 Under Nudipedes he includes all British bees without scopre, except 

 Prosopis, Sphecodes, and Psithyrus. But Coelioxys and Stelis are 

 developed from nudigasters and are related to the Megachilin^e. Delia 



