Our own studies lead us to believe that it also applies in the case of feeble-mindedness, but this will be taken up in a later work to which we have already referred. We do not know that feeble-mindedness is a “unit character.” Indeed, there are many reasons for thinking that it cannot be. But assuming for the sake of simplifying our illustration that it is a “unit character,” then we have something like the following conditions. If two feeble-minded people marry, then we have the same unit character in both, and all of the offspring will be feeble-minded; and if these offspring select feeble-minded mates, then the same thing will continue. But what will happen if a feeble-minded person takes a normal mate? If feeble-mindedness is recessive (due to the absence of something that would make for normality), we would expect in the first generation from such a union all normal children, and if these children marry persons like themselves, _i.e._ the offspring of one normal and one defective parent, then the offspring would be normal and defective in the ratio of three to one. Of the normal children, one third would breed true and we would have a normal line of descent. Without following the illustration further, we see already that it is questionable whether we ought to say that the original feeble-minded individual should have been sterilized because he was feeble-minded. We see that in the first generation all of his children were normal and in the next generation one fourth of them were normal and bred true. We should not forget, however, that one fourth of his grandchildren would be feeble-minded and that two other fourths had the power of begetting feeble-minded children. We must not forget, either, that these are averages, and that for the full carrying out of these figures there must be a large enough number of offspring to give the law of averages room to have full play. In other words, any marriage which, according to the Mendelian principle, would give normals and defectives in the ratio of three to one might result in only one child. That child might happen to be one of the feeble-minded ones, and so there is propagated nothing but the feeble-minded type. It is equally true that it might be the normal child, with a consequent normal line of descendants; or still again, it might be one of the intermediate ones that are capable of reproducing again the ratio of three normal to one defective, so that the chance is only one in four of such offspring starting a normal line.