I said advisedly, a moment ago, “if the hypnotic personality remembered the whole of the conscious calculation,” for, as a matter of fact, we find, when we examine several different hypnotic states in the same subject, that their memories for coconscious ideas are not coextensive, one (or more) being fuller than another. Indeed in certain states there may not be any such memories at all. It is necessary, therefore, to obtain by hypnosis a degree of dissociation which will allow the complete memories of this kind to be evoked. In the subjects I made use of this procedure was followed. Theoretically it might be held that, no matter how complete the memories evoked in the various states, some other state might possibly be obtained in which still more complete memory would be manifested. Theoretically this is true and all conclusions are subject to this criticism. Practically, however, I found, when making these investigations, that I seemed to have come to the limit of such possibilities, for, obtain as I would new dissociated arrangements of personality, after a certain point no additional memories could be evoked. There is still another possibility that there may be coconscious processes for which no memories can be evoked by any method or in any state. II. _Artificially induced visual hallucinations_ with which we have already become familiar can, as we have seen, only be interpreted as the product of subconscious processes. If only because of the important part that hallucinations play in insanity and other pathological states and of the frequency with which they occur in normal people (mystics and others), the characteristics of the subconscious process are well worth closer study. What is found to be true of the experimental type is probably true of the spontaneous variety whether occurring in pathological or normal conditions. Indeed, as we shall see, spontaneous hallucinations have the same characteristics. We have considered them thus far only from two points of view, viz. (1) as evidence of conservation of forgotten experiences, and (2) as evidence for specific residua of such experiences functioning as subconscious processes. Now, artificial visual hallucinations, like the spontaneous ones, may be limited—relatively speaking—to what is apparently little more than an exact reproduction of an antecedent visual perception, e. g., a person or object. But, generally speaking, it is more than this and when analyzed will be found almost always to be the expression of a complicated process. For instance, take the relatively simple crystal vision, of the subject smoking a cigarette in a particular situation during hypnosis, which I have previously cited. (Lecture III.) As a matter of fact, the subject had no primary visual perceptions at the time of the original episode at all. She was in hypnosis, her eyes were closed, and she did not and could not see herself (particularly her own face) or the cigarette or her surroundings.