sexual differentiation is not always so marked as it is in adults; and it may happen that the sexes may exchange their roles. Cases observed by Seitz have been published by Groos and also by myself.[46] I have myself watched a young cow which repeatedly attempted to mount another young cow; I have also on several occasions seen young bitches attempt to cover dogs. To this part of our subject belongs the observation of Exner, that when dogs are playing wildly with one another one hardly ever sees a bitch among them. But if an exception should occur, the bitch is usually a young one. In animals, sexual differentiation is not complete until sexual maturity is attained, and the same is true of the human species, although, as I have shown above, children already manifest sexual differentiation in their games, their inclinations, and their general conduct. I have thought it desirable to refer to the play of animals in this place, as well as to treat of the subject in its direct relationship to the sexual impulse. What is true of play is true also of the other interests and inclinations of the child, many of which are also associated with the sexual life; these have been described earlier, so that here I need merely allude to the matter in passing. Hitherto I have described the sexual life of the child in so far as it is the subject of direct observation or can be recalled to memory. But it was explained at the outset that there is still another way of gaining clear knowledge of the subject, namely, by experiment; and it was shown that castration may be regarded as such an experiment. Although the reproductive capacity of the male is not developed prior to the formation of spermatozoa in the testicles, nevertheless we learn from the effects of castration that the testicles exhibit important functional activity much earlier in life. This fact was long overlooked, and its importance is even to-day largely underestimated, because we have been accustomed to regard the provision of an external secretion as the only function of the testicle. But it is now firmly established that these glands exercise influence in other ways. We know that bodily and mental development are affected by the removal of the testicles; and that the influence is greater the earlier in life the castration takes place. A number of secondary sexual characters remain undeveloped. The beard does not grow; in many instances a thick _panniculus adiposus_ is formed; there are changes in the growth of the bones; the voice remains a soprano; and the other reproductive organs are imperfectly developed, the penis and the prostate remaining comparatively small An early castration does not, of course, result in the obliteration of all differences between the male and the female; we must rather say that a part only of the typical differential characters of sex remain undeveloped. The earlier assumption, that the secretion of semen