On Aquatic Carnivorous Coleoptera or Dytiscidce. 233 



and hind coxa, while in Eretes where there exists considerable capacity for 

 thoracic movement, and very little adaptation of the elytra to the sides of the 

 body, the whole of the large epimeron is horny. 



The more important of the characteristics of the metasternum in the Dytiscidae 

 are, 1 , the peculiarly formed side wings ; 2, the intercoxal process grooved for the 

 reception of the prosternal process, and 3, the absence of a transverse suture on 

 its posterior part. None of these characters however are constant enough to serve 

 as marks of distinction between the Dytiscidte and Carabidas ; the first of them is 

 excessively variable in the Dytiscida ; it is chiefly dependent on the great develop- 

 ment of the hind coxae, and in Amphizoa and Pelobius, and even in Colpius and 

 Suphis, the form of the metasternum is quite Carabideous. The groove of the 

 intercoxal process too is far from being constantly present in the Dytiscidse, it is 

 absent as we have seen in all cases where the middle coxse are widely separated, 

 and all that can be said is that it never occurs in the Carabidte, but is present in the 

 great majority of the Dytiscidas. As regards the transverse suture in the 

 metasternum, it occurs in the Dytiscidse only in Pelobius and Amphizoa, while on 

 the other hand it becomes indistinct in some Carabidae, although I have not succeeded 

 in finding any in which it is effaced. 



The metasternum of the Haliplides has none of the peculiarities of the Dytiscidse ; 

 the intercoxal process is not grooved ; the outline is completely Carabideous, and 

 the transverse suture is present, though without dissection it is not easily 

 apjireciated owing to a series of very large punctures placed on it. 



The contiguity or amount of separation of the middle coxse in the carnivorous 

 Coleoptera appears to me to be dependent rather on the metasternum than on the 

 mesosternum. In the Carabidte these coxas are nearly ahvays very considerably 

 separated by the intercoxal process, so that the one group of that family — the 

 subfamily Ozsenidse — in which the coxse are unusually approximate is, according to 

 Leconte, well distinguished by this character from all the rest of the family. In 

 the Dytiscidse, greater variety in respect to the contiguity of the middle coxse is 

 found, but still the character appears to me to be one of great importance ; indeed 

 if I might exercise my imagination I would suppose that in the early history of the 

 Dytiscidse the fact whether the middle coxse were well separated, or were approxi- 

 mate, exercised a most important, or indeed predominant influence on the future 

 mode of evolution. When they were distant this fact appears to me to have 

 facilitated rapid co-adaptation of the contiguous parts for the purpose of protection 

 and keeping out water from entering the body by the great fissure between 

 the prothorax and the after-body by a process of mere placing together of accurately 

 fitting surfaces ; and this involved complete disuse of any mobility or extension 

 between these parts, and rendered sedentary habits of advantage, for in all 

 movements there was the danger of these parts becoming separated, and it was 

 necessary to keep the pieces of this great joint without any motion, and yet they 



