I trust you are taken by surprise by the establishment of this fact, and that you will ask me why I did not refer to this proof before in the case of the slips, where we credited the man who made a mistake in speech with intentions he knew nothing about and which he denied. "If a person believes he knows nothing concerning experiences, the memory of which, however, he retains," you might say, "it is no longer so improbable that there are also other psychic experiences within him of whose existence he is ignorant. This argument would have impressed us and advanced us in the understanding of errors." To be sure, I might then have referred to this but I reserved it for another place, where it was more necessary. Errors have in a measure explained themselves, have, in part, furnished us with the warning that we must assume the existence of psychic processes of which we know nothing, for the sake of the connection of the phenomena. In dreams we are compelled to look to other sources for explanations; and besides, I count on the fact that you will permit the inference I draw from hypnotism more readily in this instance. The condition in which we make mistakes most seem to you to be the normal one. It has no similarity to the hypnotic. On the other hand, there is a clear relationship between the hypnotic state and sleep, which is the essential condition of dreams. Hypnotism is known as artificial sleep; we say to the person whom we hypnotize, "Sleep," and the suggestions which we throw out are comparable to the dreams of natural sleep. The psychical conditions are in both cases really analogous. In natural sleep we withdraw our attention from the entire outside world; in the hypnotic, on the other hand, from the whole world with the exception of the one person who has hypnotized us, with whom we remain in touch. Furthermore, the so-called nurse's sleep in which the nurse remains in touch with the child, and can be waked only by him, is a normal counterpart of hypnotism. The transference of one of the conditions of hypnotism to natural sleep does not appear to be such a daring proceeding. The inferential assumption that there is also present in the case of the dreamer a knowledge of his dream, a knowledge which is so inaccessible that he does not believe it himself, does not seem to be made out of whole cloth. Let us note that at this point there appears a third approach to the study of the dream; from the sleep-disturbing stimuli, from the day-dreams, and now in addition, from the suggested dreams of the hypnotic state.