On Aquatic Carnivorous Coleoptera or Dytiscide. 965 
now in Colymbetes, at a distance of about 1 or 2 m.m. from the cavity, this edge is 
completely overwrapped and concealed by the mesothoracic epimeron; a little 
nearer to the cavity the epimeron however again diverges and exposes a part of 
the episternal edge, but close to the cavity the epimeron gives off a process which 
appears to unite with the side wing of the metasternum by a minute point, and thus 
to exclude the episternum from the cavity ; this is not however really the case for 
if the sutures be forced open so that the minute relations of the parts can be clearly 
seen, it will be perceived that, by the growth of the process of the epimeron just 
alluded to, the edge of the episternum has been twisted to one side and forced into 
the cavity in such a way as to leave only a thin lamina of the episternal edge 
visibly interposed between it and the metasternum ; this apparent exception there- 
fore shows actually how remarkably pertinacious is the point of structure I have 
alluded to. The varied aggregates composing the !)ytisci complicati are connected 
together in an intricate manner by numerous other points of structure,no one of which 
however is constantly present ; except that the posterior coxze possess a characteristic 
but variable form, their greatest anterior extension being near to the epipleura and 
far from the mesial line of the body (longitudinally). The remarkable Amphizoa 
forms however quite an exception to this integration of the Dytisci complicati ; it 
possesses certain peculiarities found in no other member of the aggregate, and it 
lacks the peculiar form of the hind coxze, and some other less persistent points of 
the other Dytisci complicati ; it is in fact quite isolated, and should perhaps not be 
united with the aggregate in the same synthesis as that in which the other component 
members are associated, but a special more distant (viz. 5th.) synthesis should be 
used for this purpose. 
Putting Amphizoa on one side, the aggregate is a very natural one, and although 
at first sight its tetramerous section appears very different from the pentamerous one, 
yet there exist some intermediate forms of the highest interest ; such an one is 
Celina of the Hydroporides, and still more the genus Methles, which cannot be 
associated in any of the larger syntheses, but remains an isolated form and a 
remarkable synthetic type. 
Although this is the case, yet careful examination fails to confirm any suspicion 
of genetic relationship between any of the components of the aggregate. For 
instance the Hydaticides are distinguished from the Colymbetides by the presence 
of a peculiar ciliation of the hind margins of the swimming feet ; this is a speciali- 
zation of high interest, and it might tor several different reasons be suggested that 
the Colymbetides stood in the relation of an ancestral group to the Hydaticides, 
that in fact Hydaticides were merely Colymbetides that had gained the said ciliation. 
There exists however a Hydaticid in which this ciliation appears in a rudimentary 
form. I allude to the genus Eretes. Here then we should expect the connecting 
link, at any rate, or evidences of a former connection between the Hydaticides and 
Colymbetides to be found; yet so far from this being the case Eretes is a most 
6 H 2 
