Halban ("Die Entstehung der Geschlechtscharaktere,” Archiv fuer Gynaekologie, 1903, pp. 205-308) argues that the primary sex glands do not necessarily produce the secondary sex characters, nor inhibit the development of those characteristic of the opposite sex.  It is indeed the rule, but it is not the inevitable result.  Sexual differences exist from the first.  Nussbaum made experiments on frogs (Rana fusca), which go through a yearly cycle of secondary sexual changes at the period of heat.  These changes cease on castration, but, if the testes of other frogs are introduced beneath the skin of the castrated frogs, Nussbaum found that they acted as if the frog had not been castrated.  It is the secretion of the testes which produces the secondary sexual changes.  But Nussbaum found that the testicular secretion does not work if the nerves of the secondary sexual region are cut, and that the secretion has no direct action on the organism.  Pflueger, discussing these experiments (Archiv fuer die Gesammte Physiologie, 1907, vol. cxvi, parts 5 and 6), disputes this conclusion, and argues that the secretion is not dependent on the action of the nervous system, and that therefore the secondary sexual characters are independent of the nervous system.

Steinach has also in later experiments ("Geschlechtstrieb und echt Sekundaere Geschlechtsmerkmale als Folge der innerskretorischen Funktion der Keimdrusen,” Zentralblatt fuer Physiologie, Bd. xxiv, Nu. 13, 1910) argued against any local nervous influence.  He found in Rana fusca and esculenta that after castration in autumn the impulse to grasp the female persisted in some degrees and then disappeared, reappearing in a slight degree, however, every winter at the normal period of sexual activity.  But when the testicular substance of actively sexual frogs was injected into the castrated frogs it exerted an elective action on the sexual reflex, sometimes in a few hours, but the action is, Steinach concludes, first central.  The testicular secretion of frogs that were not sexually active had no stimulating action, but if the frogs were sexually active the injection of their central nervous substance was as effective as their testicular substance.  In either case, Steinach concludes, there is the removal of an inhibition which is in operation at sexually quiescent periods.

Speaking generally, Steinach considers that there is a process of “erotisation” (Erotisieurung) of the nervous center under the influence of the internal testicular secretions, and that this persists even when the primary physical stimulus has been removed.

The experience of veterinary surgeons also shows that the sexual impulse tends to persist in animals after castration.  Thus the ox and the gelding make frequent efforts to copulate with females in heat.  In some cases, at all events in the case of the horse, castrated animals remain potent, and are even abnormally ardent, although impregnation cannot, of course, result.[8]