On Aquatic Carnivorous Coleoptera or Dytiscda. 957 
showsat the front edge in the middle a distinct vertical face, placed at right angles to 
the plane of the prosternum; a superficial approximation to the definition of the pros- 
ternal structure of the Hydroporides is thus suggested ; but a very slight examination 
is sufficient to show that so far from this being the case, Coptotomus is of all the 
Colymbetides the farthest removed by its prosternal structure from the Hydro- 
porides. In Hydroporides it is the inter-coxal band of the prosternum that is placed 
on a different plane to the prosternal process, the very small transverse band forming 
the front of the prosternum being again nearly horizontal in plane like the prosternal 
process ; in Coptotomus the horizontal plane is continuous from the apex of the 
prosternal process to the front margin of the prosternum, and then there occurs the 
vertical free face of the incrassate portion. Dytiscus fuscipennis (Agabus, group 
21) makes however a genuine approach to Hydroporides in the prothoracie structure, 
the band between the front coxee being placed, in its relation to the prosternal process, 
somewhat asin that group: this interesting similarity is not however accompanied 
by any other approach in D. fuscipennis to the Hydroporides, so that we may say 
that those influences that have caused this insect to approach Hydroporides have 
been confined in their operation to its prosternal structure. 
The front and middle tarsi are always conspicuously five-jointed, the fourth joint 
being as far as its length goes, as much developed as the basal ones, and thus a 
conspicuous contrast with what is usual (but not absolute) in Hydroporides is 
exhibited. In the male sex of Colymbetides there is always a greater or less 
incrassation of the three basal joints of the tarsi, and we find in certain members 
of the aggregate that this mcerassation is displayed to a less extent by the fourth 
joint (vide Colymbetides), a most unusual character in the Dytiscidee. Still how- 
ever it is interesting to remark that in both Colymbetides and Hydroporides in 
certain exceptional cases, the sexual development exhibited by the three basal 
joints of the tarsi is accompanied by an unusual development of the fourth joint. 
Except in the case of sexual clothing of the males, the front and middle 
tarsi are bare beneath, or armed only with rigid ciliz : their structure being thus 
that characteristic of locomotion on rough surfaces, while in the Hydroporides 
we have seen the structure of the tarsi indicates an adaptation for locomotion on 
plants. 
The male tarsi do not show a great range of variation in the Colymbetides, 
although very different grades of development of their structure are included in 
theaggregate. As I have already remarked, the three basal joints of the male front 
feet are more or less incrassate ; in one or two cases (Metronectes, Agabinus), only two 
basal joints are thus affected, the third remaining undeveloped. This increase is 
in its rudimentary forms but slight, and is an incrassation rather than a dilatation, 
the increase in the size of the sole for bearing the sexual clothing being obtained 
by enlarging the diameter of the cylinder in both the horizontal and vertical 
directions; in the higher forms it becomes greater, and the incrassation is accom- 
6G2 
