36 
pointed in D. circumcinctus, which we have just considered our oldest 
British form, this would appear not so. The posterior coxal processes act 
the very important function of keeping the swimming legs in a line 
with the centre of the body, so as to get the full advantage of each 
stroke, and the shape is no doubt dependent upon the build of the 
insect, and that the blunt ones are the more powerful, and where more 
power became necessary natural selection effected the necessary 
change. 
Having made clear the true meaning of the dull upper surface of 
the females, from the genus Dytiscus we can readily understand the 
other genera of the family, and instead of predicting that the dull 
form will disappear, we can now confidently predict that the shiny 
form of the female is the disappearing one ; and the fact that in some 
species the females are always dull, in others the majority are dull, 
in some only a few dull, in others again none are dull, becomes 
perfectly clear, and the whole process instead of being a confused mass 
of facts without a beginning or end is now seen to be the whole process 
of natural selection, being worked out in a perfectly clear and intelligible 
manner. I must not forget that in one or two instances, and that 
also where the male is less adapted for rapid movement (I refer more 
particularly to H. 12 -pustulatas and H. depressus ), the male is also dull 
like the female. Again in Hydroporus dorsalis, in some of the females the 
elytra have a red spot at the shoulder, which must also be looked upon 
as a higher development, seeing that it only occurs in the dull females. 
There is one other point in connection with the roughness of the 
surface in the females of our water-beetles, and it is that, although 
the same result is obtained, i.e., that of making the insect duller, the 
method employed is different (as in Dytiscus it is by sulci, in Hydro¬ 
porus by alutaceous punctuation, &c.) which is strongly in favour of its 
being a developing character, because were it a vestigial one, derived 
from a common ancestor, it would almost certainly have been built up 
the same way in both genera. 
This point opens up another exceedingly interesting side of the 
question, which can be readily studied in the secondary sexual 
characters of coleoptera, and that is the various means by which 
nature, tied down by the little-known laws of heredity, struggles to 
obtain by various means the same end. If one only knew the part 
played by heredity in governing and checking the adaptations brought 
about by natural selection, what a flood of light it would throw on 
these obscure structures which are so puzzling. Take almost any 
genus of beetles you like, and you find the secondary sexual characters 
run along certain definite lines, and in a way that cannot fail to be 
noticed ; in fact, one might almost tabulate some of the genera by 
these secondary sexual characters. If one takes the genus Pterostickus 
as a case in point, we find that all those displaying secondary sexual 
characters have them on the last ventral abdominal segment; in one 
species it is a depression, in two or three it is a raised tubercle, or 
ridge, but in all cases there is a close connection between them. In 
the genus Amara, again, a good number display secondary sexual 
characters, and here again wo see the characters exhibited are the 
clothing of the inner margin of the intermediate or posterior tibiie. In 
the genus llybius the characters in the males of different species are very 
closely allied indeed, and we find it a gonoral rule that the closer the 
