The formation of a sexual ideal, while it furnishes a predisposition to be attracted in a certain direction, and undoubtedly has a certain weight in sexual choice, is not by any means the whole of sexual selection. It is not even the whole of the psychic element in sexual selection. Let us take, for instance, the question of stature. There would seem to be a general tendency for both men and women, apart from and before experience, to desire sexually large persons of the opposite sex. It may even be that this is part of a wider zoölogical tendency. In the human species it shows itself also on the spiritual plane, in the desire for the infinite, in the deep and unreasoning feeling that it is impossible to have too much of a good thing. But it not infrequently happens that a man in whose youthful dreams of love the heroine has always been large, has not been able to calculate what are the special nervous and other characteristics most likely to be met in large women, nor how far these correlated characteristics would suit his own instinctive demands. He may, and sometimes does, find that in these other demands, which prove to be more important and insistent than the desire for stature, the tall women he meets are less likely to suit him than the medium or short women.[176] It may thus happen that a man whose ideal of woman has always been as tall may yet throughout life never be in intimate relationship with a tall woman because he finds that practically he has more marked affinities in the case of shorter women. His abstract ideals are modified or negatived by more imperative sympathies or antipathies. In one field such sympathies have long been recognized, especially by alienists, as leading to sexual unions of parity, notwithstanding the belief in the generally superior attraction of disparity. It has often been pointed out that the neuropathic, the insane and criminal, "degenerates" of all kinds, show a notable tendency to marry each other. This tendency has not, however, been investigated with any precision.[177] The first attempt on a statistical basis to ascertain what degree of parity or disparity is actually attained by sexual selection was made by Alphonse de Candolle.[178] Obtaining his facts from Switzerland, North Germany, and Belgium, he came to the conclusion that marriages are most commonly contracted between persons with different eye-colors, except in the case of brown-eyed women, who (as Schopenhauer stated, and as is seen in the English data of the sexual ideal I have brought forward) are found more attractive than others.