34 
There is a very striking corroboration to this view of the question 
which cannot fail to be of interest, and it is this. If the furrows are 
the result of the female requiring special concealment, then they have 
no doubt been produced entirely by natural selection, and the question 
of sexual selection has had nothing to do with the matter at all, or at 
least, we may argue so, since it is not considered scientific to give two 
causes for a given result when one is sufficient to cover the facts, and 
it has already been shown how difficult it is to account for the 
structure by sexual selection. But if it becomes an imperative necessity 
at any time during the history of a genus or species to have a change 
of structure brought about by natural selection to suit its environments 
or habits, then the sexual characters must if necessary alter, to suit 
the amount of change demanded by natural selection. I wish to make 
this point perfectly clear, as it leads up to one of those interesting cases 
where a sexual difference in a certain structure would never have been 
noticed had not the theory demonstrated that it might possibly occur. 
My point is this, natural selection is a far more important factor in 
the life-history of a species or genus than sexual selection, and 
consequently when an alteration by natural selection becomes 
necessary, sexual selection even if opposed to it would at most only 
retard the alteration demanded by natural selection, and it conse¬ 
quently follows that if natural selection be all powerful in the matter, 
then sexual selection must accomodate itself where necessary to the 
structural alterations brought about by natural selection. Here then 
we will take up Dytiscus once more. We will suppose that natural 
selection has brought about the furrows on the elytra of the female 
to give it better concealment, and by the ordinary laws of nature if 
this rendered copulation more difficult, then the male must develop 
some new structure or improve some existing one in order to get over 
the difficulty. So far the theory, and it remains to be seen if the facts 
are in accordance with it and whether we can find the structure which 
the theory suggests as probable. With this in view, I took all my 
specimens of Dytiscus out of the cabinet, both males and females, and 
I found that at the base of the anterior and intermediate femora in 
both sexes, was a small stiff pencil of bristles which, in the normal 
position of the legs while swimming, would point backwards but when 
the legs were grasping anything beneath the insect, the femora being 
twined round with the coxa towards the front of the insect, these 
bristles would point downwards and would therefore assist the insect 
in getting a better hold, if there were any structure in the object clung 
to, into which these bristles would fit. 
It had been suggested to me that the suckers on the anterior tarsi 
of the male were placed on the elytra of the female during copulation, 
which I had very much doubted, as the suckers being adapted for a smooth 
surface I had no doubt were placed somewhere on the thorax, and 
arguing on the analogy in the case of Hydrophilus pice us in which the 
onychium is dilated in such a way that the claw must catch over the 
lateral edge of the thorax, and also in the allied species Hydrous 
caraboides where the tarsal claw in the male is exactly the shape to fit 
the edge of the thorax, it seemed pretty clear that the tarsal claw in 
Dytiscus (and other members of the family) caught the edge of thorax, 
and the disc with the suckers was pressed down on the upper surface. I 
mentioned the matter to Mr. Donisthorpc, who very kindly obtained 
