36G CLASSIFICATION OF THE CARABID^E 



separating the epimera and episterna, is advanced so that the posterior portion becomes 

 triangular, and the suture is diagonal. For convenience, in the descriptions which follow, 

 I have termed these quadrangular pieces "epimera;" and although it is evident they are 

 composed of the epimera and episterna together, it was thought that this term, though not 

 strictly accurate, was obvious in its meaning, and preferable to the introduction of an 

 unusual expression. The diagnosis of the first division will, therefore, be "epimera meso- 

 thoracis non diagonaliter divisa;" and, of the second, "epimera mesothoracis diagonaliter 

 divisa." 



In the isolation of the different types composing the first of these divisions, it was 

 found that Erichson's definition of his group Brachinini, would not apply to many of our 

 genera, which were evidently closely related with genera already contained in his group: 

 it thus became necessary to seek for new characters. In comparing together the diffe- 

 rent genera with truncate elytra, it was seen that Brachinus, distinguished by its singular 

 power of producing explosions, differed not only from all other genera with truncate ely- 

 tra, but from all other Carabidse, by having seven ventral segments in the abdomen ; thus 

 violating the diagnosis of the entire family, as given by Erichson. Yet it was evident 

 that the Brachini did not constitute a separate family, of the same rank as the Cicindelidse, 

 but were rather to be compared with the two divisions already established on the form of 

 the epimera of the mesothorax. In this respect the Brachini appeared to come closer to 

 the second series than to the first. 



We now, therefore, have three sub-families, easily distinguished from each other, and 

 which may be called Brachinini, Harpalini, and Scaritini. 



In taking up the second of these, containing the great mass of the Carabida?, and re- 

 curring again to those with truncate elytra, it was found that a certain set, typified by 

 Lehia and Cymindis, passed by insensible gradations into those having the elytra not trun- 

 cated, as Platynus, and thus to Pterostichus, all distinguished by having the anterior tibiae 

 slender, or but slightly thickened, while Helluo had the same organs broad and strongly 

 compressed, so that both the truncate elytra and the structure of the anterior tibiae failed 

 to define the group. While comparing together Helluo with Galerita, no difference 

 was perceived excepting those already pointed out; but, in comparison with Cymindis, the 

 ligula was found to be destitute of the paraglossee, which although connate with the ligula 

 in the greater number of genera having truncate elytra, are still distinct. On further 

 comparison this absence of paraglospoe was found to be a character of all the genera 

 which were suspected to have any affinity with Helluo, such as Galerita, Anthia, Morio, 

 however different might be the form of the tibiae and elytra. This, then, was obviously a 

 point of fundamental importance, since it brought together genera, the affinities of which 

 had been recognised, although by the ordinary classifications they had been widely sepa- 

 rated. On farther consideration I was induced to consider l'anagaeus as related with 

 these genera, rather than with Chlcenius or Badister. 



Another peculiarity which distinguishes this set of genera is the tendency to abnormal 

 pubescence on the antennae, the difference between the glabrous and pubescent points, so 

 obvious in must of the Carabidae, being here by no means plain: some genera, as Gale- 



