970 On Aquatic Carnivorous Coleoptera or Dytiscidce. 



forms to which we give this collective name, but a vast number of more imperfect 

 forms which there is reason to suppose may have existed in remote geological 

 periods, it is quite possible — I think I may say probable — that among such forms 

 would be found the true ancestors of the present Dytiscida-. But the term 

 " Carabidse " has only real scientific value inasmuch as it expresses certain 

 creatures whose organization is known to us, and is expressed by a certain definition 

 drawn from, and generalizing this known structure, and to apply the term Carabidae 

 to include these distant and different creatures, is to alter its meaning and cannot 

 but lead to logical bewilderments and mazes of a quite perplexing nature. I con- 

 clude therefore, that in no correct manner can the Carabidse and Dytiscidie be 

 placed in a genetic relation to one another on any genealogical chart. 



On the other hand it seems quite probable that if we knew a great deal more 

 than we do know about the functional values of the structures we discuss, we should 

 be able to comprehend some of, if not all, the intricate affinities exhibited by the 

 two families as being the result of, or correlative with, approximations in the mode 

 in which function is or has been performed, and to be quite independent of genetic 

 community. Thus the elongation and amalgamation of the internal laminae of the 

 hind coxse along the mesial line, might be found to be connected with the fact that 

 the hind legs have had to use an increased degree of force, and take a more than 

 usually large share in the process of locomotion, and to act in a more directly 

 horizontal direction ; this would be the case with the Dytiscidse moving through 

 the water, and probably with the Pseudomorphini dwellers under the bark of trees. 



There is another question in connection with the classification of the Dytiscidse and 

 Carabidse, which I may briefly allude to, although only to show the complex and 

 difficult nature of the problems with which classification has to deal. It is this ; 

 is it — keeping in view the intimate relation existing between these two aggregates — 

 a correct course to keep them separate or should they not be in some way or other 

 united ? To answer this requires a valuation of structural characters of a most 

 difficult nature. I will give an instance. We have in the Dytiscidse two series, 

 Dytisci fragmentati and Dytisci complicati, of these the latter is distinguished by 

 all its members possessing a character which is completely wanting not only to the 

 other series, but to the Carabidse and all the other beetles : I allude to the fact that 

 four principal pieces of the body form the walls of the middle coxal cavities ; while 

 the other series, Dytisci fragmentati, have these cavities formed by only three pieces. 

 Now the Carabidse also consist of two series distinguished by the structure of these 

 cavities, the first series having, like the Dytisci fragmentati, three pieces forming the 

 coxal cavity, while the second series has only two pieces entering into the com- 

 position of these parts. Thus we have four series, viz., (1), Carabici second series, 

 with coxal cavities formed by two pieces ; (2) and (3) Carabici first series and Dytisci 

 fragmentati, with coxal cavities formed by three pieces ; and (4), Dytisci complicati 

 with coxal cavities formed by four pieces. Now if we bear in mind the fact that 



