This most important question, as to what is actually happening within the limits of marriage, may legitimately be regarded as substantially equivalent to the question of the extent and nature of selection, for good or for evil, as it occurs in society to-day. If we remember that an overwhelming proportion of children are born in wedlock, that the death-rate of illegitimate children is gigantic, whilst the illegitimate birth-rate is generally falling, we shall be fully entitled to assume that the answer to the one question is the answer to the other; in a word, if under the present conditions of selection for marriage we find a eugenic tendency or an anti-eugenic tendency or a mere neutrality, the answer will be, _on the whole_, the approximate answer to the larger question as to the present state of selection for parenthood and therefore of our racial prospects, marriage or no marriage. The conclusion which we shall maintain is that _both forms of selection occur in society to-day_--the selection of the desirable and the selection of the undesirable. We shall go ludicrously wrong if we agree, with one party, that society in general to-day exhibits reversed selection; or, with the second party, that everything is going on admirably on the whole; or, with the third party, which jumbles the whole mass of facts and tendencies, and declares that there is no process of selection of any kind occurring in society to-day--an opinion which, in the face of disease, the enormous premature death-rate, and the fact that whilst vast numbers of women are unmarried, the choice of women for marriage does not occur by lot, beggars comment; is a girl with a birth-mark covering half her face, or a nose destroyed by transmissible disease, as likely to marry as a “beauty”? If not, surely we actually select to-day for beauty and therefore for whatever beauty depends upon--for instance, health. But really it cannot be necessary to deal seriously with the proposition that no selection occurs in society to-day. Let us attempt to state clearly the point at issue. There is granted, in the first place, that by far the greater part of all parenthood, in civilised and uncivilised communities alike, occurs within the limits of marriage; to which may be added that, owing to the excessive death-rate of illegitimate children, the proportion of effective parenthood, so to say, that occurs within the limits of marriage is even larger; and this intervention of marriage, and any selection that may be involved in it, steadily recur from generation to generation. Thus even those born outside wedlock will nevertheless be selected for parenthood, on their own part, mainly by the selective factors in marriage.