tNTRODUCTORY REMARKS. 81 



mee k sa base dans une 

 gaine coriace qui s'em- 

 boite sur les cotes, dans 

 les machoires. 



This character agrees very well with the insects of 

 this class, with the exception of ailes vehu'es, 

 which, as I just now observed, are not common to 

 the whole : but still 1 prefer that of Linneus, on 

 account of its including the aculeus^ its most re- 

 markable and striking distinction. 



Instead of subdividing his genera, this authdf 

 divides the class into sections and families, the 

 characters of which are drawn from the antenna, 

 OS, lahium inferius, lingua, maxiihe. Fab. tubus, 

 and palpi. He gives forty-three Hymenopterous 

 genera, without including those of Fabricius's sup- 

 plement : the characters of these are taken from 

 the antenme, lahium superius, manditulce, lingua, 

 maxilUe, Fab. tubus, and palpi. There cannot be 

 a stronger proof that differences in the antennce 

 and instrumenta ciharia :\vc not the best foundation 

 for characteristics of genus in this class, than the 

 great and unnecessary multiplication of genera by 

 those who build their system upon them. Cha- 

 racters of families are mistaken for distinctive 

 marks of a genus ; and thus natural genera are 

 taken to pieces, and, if this practice continues, we 

 shall have no such thing as a subdivision in any. 

 The different families of the Linnean genus Tt.i- 

 tliredo, are all characterized from variations in 



G thp. 



