In man, however, freedom from hereditary taint cannot so easily be secured. Individuals cannot be selected scientifically for breeding purposes. Furthermore, the human body is more delicately constructed than that of the lower animals, and the nervous system is more highly developed and specialized, so that it is reasonable to suppose that in man degeneration would set in earlier in the process of inbreeding, and that it would be impossible to breed as closely as with the lower animals. Instances are well known, however, where incestuous unions have been productive of healthy offspring, and successive generations of offspring of incestuous connection are not unknown; but, although statistics are lacking, it seems to be very often true that children of such unions are degenerate. It may be that the reason for this is that with the laws and social sentiments now prevailing in all civilized communities, only degenerates ever contract incestuous alliances. Desirable as it may be from a social point of view that this strong sentiment against incest should continue, it is not yet _proven_ that even the closest blood relationship between the parents is directly injurious to the offspring. The "instinctive horror of incest" is a myth, for although a horror of incest does very properly exist in civilized, and in some tribal societies, it is purely a matter of custom and education, and not at all a universal law. Double heredity may account for all the observed ill effects of consanguineous marriage, including the high youthful death-rate, the higher percentage of idiocy, deafness and blindness, and probably also the scrofulous and other degenerate tendencies; nevertheless, there may be in some instances a lowering of vitality which this hypothesis does not fully explain. The tendency of inbreeding in animals, it is well known, is to fix the type, the tendency of crossing, to variation. Inbreeding then, tends to become simple repetition with no natural variations in any direction, a stagnation which in itself would indicate a comparatively low vitality. Variation and consequent selection is necessary to progress. "Sex," according to Ward[96] "is a device for keeping up a difference of potential," and its object is not primarily reproduction, but variation.[97] [Footnote 96: _Pure Sociology_, p. 232.] [Footnote 97: Pearson (_Grammar of Science_, p. 373) points out that variation does occur in asexual reproduction. But that sex is at least a powerful stimulus to variation can hardly be questioned.]