breeding by securing several hybrids from quite distinct stocks to start with, and by having two or more sets of experiments carried on at once, so that crosses between the hybrids produced may be occasionally made. Till this is done no experiments, such as those hitherto tried, can be held to prove that hybrids are in all cases infertile inter se.
It has, however, been denied by Mr. A.H. Huth, in his interesting work on The Marriage of Near Kin, that any amount of breeding in-and-in is in itself hurtful; and he quotes the evidence of numerous breeders whose choicest stocks have always been so bred, as well as cases like the Porto Santo rabbits, the goats of Juan Fernandez, and other cases in which animals allowed to run wild have increased prodigiously and continued in perfect health and vigour, although all derived from a single pair. But in all these cases there has been rigid selection by which the weak or the infertile have been eliminated, and with such selection there is no doubt that the ill effects of close interbreeding can be prevented for a long time; but this by no means proves that no ill effects are produced. Mr. Huth himself quotes M. Allié, M. Aubé, Stephens, Giblett, Sir John Sebright, Youatt, Druce, Lord Weston, and other eminent breeders, as finding from experience that close interbreeding does produce bad effects; and it cannot be supposed that there would be such a consensus of opinion on this point if the evil were altogether imaginary. Mr. Huth argues, that the evil results which do occur do not depend on the close interbreeding itself, but on the tendency it has to perpetuate any constitutional weakness or other hereditary taints; and he attempts to prove this by the argument that "if crosses act by virtue of being a cross, and not by virtue of removing an hereditary taint, then the greater the difference between the two animals crossed the more beneficial will that act be." He then shows that, the wider the difference the less is the benefit, and concludes that a cross, as such, has no beneficial effect. A parallel argument would be, that change of air, as from inland to the sea-coast, or from a low to an elevated site, is not beneficial in itself, because, if so, a change to the tropics or to the polar regions should be more beneficial. In both these cases it may well be that no benefit would accrue to a person in perfect health; but then there is no
