361 



CLASSIFICATION OF THE CARABID.'E 



his definitions, and from many happy suggestions regarding the affinity of difficult groups. 

 The classifications adopted hy English naturalists are mainly based upon Latreille's sys- 

 tem, very slightly modified. That offered by Haliday (Entomologist, 185,) is a remarka- 

 ble exception, the structure of the prosternum being for the first time used as a character 

 for natural division in this tribe. 



However applicable and satisfactory these arrangements may be when applied to a 

 local fauna, they have all failed when applied to a more extensive series of genera, so 

 that after all these systems, there are yet many genera, which, when placed in the groups 

 with which they seem properly to belong, violate the rules of the system. Thus if Galerita 

 enters Erichson's group Brachinini, it would be impossible for the nearly allied Helluo to 

 do so; although these are certainly much more closely related than Brachinus and Lebia: 

 while in other cases (Anchomenini, Pterostichini, Trechini, Chlamiini) the characters are 

 almost entirely sexual. 



The great difficulty, in all investigations of this nature, is the determination of the true 

 value of the differences of organization, which are co-extensive with the groups: in pro- 

 portion as these are properly subordinated our classification becomes natural. Having 

 once established the limits of our smaller groups upon distinct characters, we may then 

 use those which are dependent on sex, or other condition, for the purpose of investi- 

 gating the relations existing among these groups. In proportion to the extent of a series, 

 and the number of types existing in it, will be the difficulty of giving a clear limit to the 

 variation of the series, and of furnishing a definition which will distinctly indicate such 

 series. 



The characters founded upon modifications indicating particular modes of life, are in 

 general, I think, unsafe; because animals must be looked upon as machines, which work 

 always in the most effective manner; and thus the habits of life would follow from the 

 modifications of structure, which form part of the system of variation, and but rarely from 

 those great and fundamental differences which distinguish the types in nature. Thus, for 

 example, the division of birds into land birds and water birds, is eminently unnatural; for 

 although it may be asserted, with some plausibility, that the webbed feet of the latter 

 were given them/or the purpose of enabling them to swim, no earthly reason can be given 

 why the embryo of a land bird at an early stage of development should also have webbed 

 feet, unless we stupidly shut our eyes to all provision or design in nature, and say: the 

 animal lives in a liquid, and therefore it must have swimming organs, entirely irrespective 

 of what may be its habits when fully developed. In mammalia this system of classifica- 

 tion has long been discarded, and we find the aquatic species partitioned among the va- 

 rious groups, to which, by their assemblage of characters, they seem most closely allied, 

 leaving only a small group, Ccte, which are typically aquatic. 



Applying similar considerations to the large series of pentamerous predaceous Coleop- 

 tera, I have been led to select other characters than those usually adopted for the division 

 into Adcphaga and Hydradephaga, which depend on the form of the feet. On closely in- 

 vestigating the nature of the parts supporting the feet, which assume this peculiar nata- 

 tory form, I was delighted to find that the coxa showed a modification in form which has 



