168 BRITISH BEES. 



Having thus cursorily skimmed the surface of the 

 method I suggest, I have next to give my reasons for 

 proposing it in lieu of adopting any yet extant. 



My exhibition of Kirby's grouping, in the preceding 

 section, where I treat of the scientific cultivation of 

 British bees, will fully explain why I could not adopt 

 that arrangement. 



Why I cannot follow Latreille's, is, that in his last 

 elaboration, in his ' Families Naturelles/ published in 

 1825, which must be considered as his final view, he 

 does not satisfactorily divide the Andrenidae, of the ge- 

 nera of which he has made a complete jumble. With the 

 Apida in his group of Dasygasters, he intermixes Cera- 

 t'ma, separating it from the group of Scopulipedes, where 

 it truly belongs by every characteristic, and he mingles 

 also with them the two cuckoo genera Stelis and Cceli- 

 oxys, which are merely parasites upon these Dasygasters, 

 and can only be associated by the structural conformity 

 of the two submarginal cells to the superior wings, and 

 the length of the labrum, the latter being a character of 

 very secondary importance ; and further, he dissevers the 

 Scopulipedes in placing Panurgus at the commencement 

 of the Apidce, and the rest proximate to the social bees. 



Westwood, in his modification of Latreille's system, 

 certainly divides the Andrenidae better than his master 

 had done, but he does not go far enough. Besides, he 

 interposes Halictus and Lasioglossum, (the latter ad- 

 mitted as a genus merely out of courtesy to Curtis, who 

 had elevated it to that rank in his ( British Entomology/ 

 although it is nothing more than a male Halictus] , 

 between Sphecodes and Andrena with Cilissa, these 

 having lanceolate tongues with lacerate paraglossse, 

 whereas Halictus has a very acute tongue, and its para- 



