876 - On Aquatic Carnivorous Celcoptera or Dytiscide 
as in Metronectes. ‘The unique species is found in Corsica, and must rank as a low 
form ; on the whole its nearest systematic ally is the Agabus cephalotes, which also 
is peculiar to Corsica. 
I. 46.—Genus AGABUS. (Vide p. 493.) 
This is a very extensive aggregate, comprising no less than ninety-five species: 
the size is moderate, varying from 6 to 12 m.m. of length, the colour is usually 
sombre, but in a few species is variegate, and a brassy tint of the upper surface is 
not uncommon, the sculpture of the wing-cases nearly always consist of a fine reticula- 
tion ; the meshes formed being variable in size and shape according to the species; 
the terminal joints of the palpi are not, or scarcely, broader than the preceding ones; 
hind coxee never extremely large; wings of metasternum externally generally of 
moderate length, but in a few species, very short, and parallel sided. Epipleuree of 
elytra narrow behind the middle; coxal lines always present, never greatly 
approximated, (in A. serricornis, however, much more approximated than normally) 
divergent in front, coxal lobes well formed, with distinct border. Swimming legs 
often slender, never very highly developed, their femora, with setigerous patch (often 
ill developed) at the postero-externo angle, their tarsi with the lower parts of the 
hind margin of the basal joints not lobed, 7.e., not produced so as to considerably 
overlap the following joint), their claws simple, equal, and more or less curved. 
The above characters are drawn up to define a large number of species showing 
great variety in parts of their structure, but all possessing the characters alluded to, 
which is not the case with any of the allied aggregates. 
I have sought in vain to arrange the ninety-five species in separate aggregates, 
but have failed in doing so, and have been obliged to content myself with arranging 
them in twenty-three groups, to each of which I have assigned, as will be seen, 
characters drawn from all such parts of the structure as I have been able to observe. 
There are other characters to be found which I have not used, owing chiefly to my 
not being able to examine them in the numerous cases of species where I have only 
a single individual for study. Such are the structure of the inner face of the wing- 
cases, and that of the various abdominal stigmata. The amount of variation of 
structure shown by the aggregate will be gathered by an examination of the 
characters I have mentioned as defining the various groups. C.J. Thomson (in his 
Skandinaviens Coleoptera, Vol. I, and Opusc. Ent. VI), has established several 
genera which I have not been able to adopt; in doing this he has pointed out a 
number of structural characters of great value, but he has unfortunately relied 
largely on one which is very unsuitable for the purpose of arriving at a natural 
classification ; I am alluding to the size and form of the wings of the metasternum; 
so variable is this character that scarcely any two species can be found exactly 
agreeing as to it, and on the other hand, species which are nearly or quite conform- 
