have already so fully seen, a question of but secondary importance; since, whether or not the physiological form of homogamy ever acts alone, even Mr. Wallace now allows, or rather argues, that it acts in combination--and this so habitually, as well as with so much effect, that it constitutes a usual condition to the origination of species. Nevertheless, although the only relevancy of his numerical computation of chances--whereby he thinks that he overturns my theory _in toto_--is such relevancy as it bears to this question of secondary importance, I have thought it desirable to refer the question, together with Mr. Wallace's views upon it, to the consideration of a trained mathematician. [62] His sentence, "all fertility not correlated with some _useful_ variation has a constant tendency to effect its own elimination," still further restricts the possible action of physiological selection to cases where at least one of the other forms of homogamy with which it is associated is natural selection. Or, in other words, it is represented that physiological selection must always be associated with natural selection, even if it be likewise associated with any other form of exclusive breeding. But as this further limitation appears to me self-evidently unjustifiable (seeing that utility is not the only possible means of securing effective isolation) I here neglect it, and take the wider ground marked out above. It is needless to say that this is giving Mr. Wallace every possible advantage, by not holding him to his still narrower ground.