case closely corresponds to those somnambulic dream-states which have recently been subjected to a penetrating study in England and France. These lapses of memory, which at first seem without content, gain a content by means of accidental auto-suggestion, and this content builds itself up automatically to a certain extent. It achieves no further development, probably on account of the improvement now beginning, and finally it disappears altogether as recovery sets in. Binet and Féré have made numerous experiments on the implanting of suggestions in states of partial sleep. They have shown, for example, that when a pencil is put in the anæsthetic hand of a hysteric, letters of great length are written automatically whose contents are unknown to the patient's consciousness. Cutaneous stimuli in anæsthetic regions are sometimes perceived as visual images, or at least as vivid associated visual presentations. These independent transmutations of simple stimuli must be regarded as primary phenomena in the formation of somnambulic dream-pictures. Analogous manifestations occur in exceptional cases within the sphere of waking consciousness. Goethe,[14] for instance, states that when he sat down, lowered his head and vividly conjured up the image of a flower, he saw it undergoing changes of its own accord, as if entering into new combinations. In half-waking states these manifestations are relatively frequent in the so-called hypnagogic hallucinations. The automatisms which the Goethe example illustrates are differentiated from the truly somnambulic, inasmuch as the primary presentation is a conscious one in this case; the further development of the automatism is maintained within the definite limits of the original presentation, that is, within the purely motor or visual region.