40 INTRODUCTORY REMARKS. 



those parts distinguish the class, which in others 

 are peculiar to certain genera. Thus, in the Cole- 

 optera, the variations, with respect to number of 

 joints, in the tarsus, will serve as an excellent cha- 

 racteristic of many genera, which, as they now 

 stand, are not sufficiently discriminated (z) ; while, 

 in the Hymenoptera, the same number obtains in 

 every genus. Again, in the same class, characters, 

 which in some genera are constant and proper 

 generic diagnostics, in others vary in the several 

 families, and even in subdivisions of the same family. 

 Thus, in most Hymenopterous genera, the number 

 of the articulations of the palpi is constant, while 

 in Apis it varies continually. Fabricius, had he 

 been aware of this circumstance, would have divided 

 this into more genera than he has done. This 

 observation shews the fallacy of his assumption, 

 that the Tnstrumenta cibaria afford constant cha- 

 racters. The labium or lingua is the only one of 

 his diagnostics, that, as far as I have examined it, 

 does not vary in the same genus ; and therefore, 

 where it may be readily inspected, it will furnish an 

 excellent one. Linneus has excluded this part and 



(?) The present character of Ckrysomela, for instance, is at 

 variance witli a large number of species that are arranged un- 

 der it ; but if the number of articulations of the tarsi, which 

 is constant in the genuine Chrysomelce, was added to its cha- 

 racter, it would furnish a clue to disting"uish the true from the 

 false, and be a great improvement. The same observation will 

 ppply to Tenebrio, 



its 



