SHARP : HYDRADEPHAGA OF LANCASHIRE AND CHESHIRE. 193 
it is to be noted that they taper to a point instead of being uniform 
inthickness, as the Geodephagous antennz is throughout. In 
Gyrinus, however, the antennz are altogether shortened almost to 
a knob, and the joints are reduced to the form of concentric rings. 
and hence they have been modified away in the Hydradephaga. 
(B). Secondly, let us consider the legs, and here we have two, 
if not three, objects to be attained. Firstly, and most important, 
propulsion ; secondly, reduced retardation ; thirdly, a purpose which 
is generally assumed to be facility in pairing, and in this the anterior 
pair of the male only are concerned. 
' Now, propulsion is principally effected by the second and third 
pairs of legs, and if we compare those pairs in Had/ip~/us with the 
Geodephagous type, we can detect little, if any, difference. The 
tibize in neither are very much flattened, and in both forms the tarsi 
are set on angularly to the tibize, and have angular tarsal joints ; nor 
is the pubescence of either tibize or tarsi particularly long or close. 
But in the genera from Hydroporus to Acilius we observe a great 
advance in the direction of the oar; we see a compressed tibiz and 
a tarsus set in a curved line with the tibize; we see a distinctly 
horizontal instead of vertical play of the joints ; and we see the tarsal 
joints mia cylindrical, and tapered, and fringed together with 
the tibiz on the lower edge with long stiff hairs, which project 
a good way oan the tarsal claw, and spread out like a feathering 
paddle at the backward stroke of the leg, which impels the insect 
forward through the water. 
ut now if we consider the posterior leg of the genus Gyrinus to 
which we alluded above, what a divergence do we find! This limb 
is neither lengthened nor furnished with swimming hairs, but instead 
the whole leg, femur, tibia, and tarsus is flattened into broad blades, 
and the tarsal joints, instead of being imposed one on another, are 
fixed like paddles round a common centre, and can be shut up like 
the leaves of a fan. The adaptation is perhaps not much more 
perfect than in the Dytiscus type of leg, but it is in a direction far 
more remote from the original. 
In considering leg modification, I mentioned as the second object 
to be attained, diminished lia sa:<P5g This is more important as 
one might say ‘forward’ than ‘aft,’ and as the antennz have been 
with the same purpose Sein 3 so the anterior legs, which in the 
Geodephagous type are nearly as long as the others, in Ha/ip/us are 
slightly shorter, and in the Dytiscidee very much reduced in length. 
July 1892 N 
