There are certain additional points about which it is unnecessary to write here, for the reason that these have all been considered in some appropriate connexion earlier in this book. For example, I have insisted upon the importance of anyone who possesses children's confidence taking steps for the removal of corrupted children from the environment of uncorrupted ones. Where we have reason to believe, in the case of a particular child, that a perverse mode of sexual sensibility is developing, we shall occasionally find it preferable rather to attempt to hinder the growth of the perversion, than to try to check the general manifestations of the sexual impulse. Thus, in the case of a boy of fourteen, who is continually affected with homosexual imaginings, we shall find it far more difficult to repress sexual manifestations altogether, than to divert the homosexual sensibility into heterosexual channels. If a boy affected in this way be thrown much into the society of girls, or conversely, a girl into the society of boys (at dances, games of lawn-tennis, &c.), the subsequent effect is likely to be good, because the sexual pervert, even if his perverse tendency be congenital, can nevertheless be educated out of his perversion. It should hardly be necessary to state expressly, that when I speak of finding for the homosexual associates of the opposite sex, I am not thinking of suggesting intimate sexual intercourse. Apart from moral considerations, we could not, in the cases under consideration, expect any benefit to accrue on medical grounds; my reference was to a purely platonic association. No one need suggest that all these recommendations are superfluous, for the reason that, according to my own previous account of the matter, the undifferentiated condition of the sexual impulse is spontaneously replaced by the normal heterosexual impulse. For, first of all, the signs that give rise to anxiety may not be manifestations of the undifferentiated sexual impulse, but may be the first manifestations of a developing congenital perversion; and, secondly, it is by no means improbable that, even in the entire absence of any congenital tendency to sexual perversion, unfavourable external conditions may lead to the further development of the perverse manifestations of the undifferentiated period. I may refer in this connexion to what was said upon p. 312 _et seq._