awaken; but, as a matter of fact, this picture is strong only because the dreamer is already very near awakening when it appears. “Un r�ve c'est un r�veil qui commence.”

It has already been emphasized by Dugas that Goblet was forced to repudiate many facts in order to generalize his theory. There are, moreover, dreams from which we do not awaken, e.g., some dreams in which we dream that we dream. From our knowledge of the dream-work, we can by no means admit that it extends only over the period of awakening. On the contrary, we must consider it probable that the first part of the dream-work begins during the day when we are still under the domination of the foreconscious.

The second phase of the dream-work, viz. the modification through the censor, the attraction by the