148 BRITISH BEES. 



comprising a family (our genera Ccelioxys and Stelis) ; 

 and the second is divided into the four families, a, ft, 7, 

 8, (a being the modern Megachile ; /3, Anthidium; 7, 

 Ckelostoma and Heriades conjunctively, and 8 is our 

 Osmia) . The subsection d has two subdivisions, 1 and 

 2, the first being a family (our Eucera) ; and the se- 

 cond is divided into the two families a and /3 (a com- 

 prising our Saropoda, Anthophora, and Ceratina), and 

 the family /3, consisting of the genus Xylocopa, then 

 supposed to be indigenous, but whose native occurrence 

 has not been substantiated. 



The fifth subsection, e, is split into two divisions, 1 

 and 2, each containing a family (1 is our Apis, and 2, 

 our Bombus) . 



In this last of his families Mr. Kirby had already 

 noticed, with the same sagacity with which he had pre- 

 viously conjectured the cuckoo-like habits of some of 

 the solitary bees, the distinctive structure of some of the 

 species, which incapacitated them from providing the 

 sustenance of their own young, and which thus reduced 

 them to the same category ; but he left the idea in its 

 supposititious condition, being too modest to use it as a 

 mark of separation, but which Newman, on our side of 

 the Channel, and St. Fargeau on the other side, subse- 

 quently, and both nearly about the same time, but with 

 the advantage in favour of Newman, distinguished, and 

 separated generically, respectively by the names of Apa- 

 thus and Psithyrus ; the former, having the priority, is 

 adopted, according to the rights of precedence in nomen- 

 clature. 



The above description of Mr. Ivirby's system Avill 

 perhaps be difficult to understand, unless T append the 

 naked scheme itself, which is as follows : 



