That even when the thoughts of the causal experience emerge in consciousness along with the vision a portion of the functioning complex—e. g., the perceptual elements—may still remain submerged is shown by the following example: The vision, one of several of the same kind, portrayed in pantomime an elaborate nocturnal somnambulistic act. It represented the subject walking in her sleep with eyes closed; then sitting before the fire in profound and depressing thought; then joyously dancing; then writing letters, etc., and finally ascending the stairs, _unconsciously dropping one of the letters from her hand on the way_,[96] and returning to bed. During the visualization the thoughts and feelings of the vision-self, even the contents of the letters, arose in the mind of the visualizer whose features and tone of voice betrayed the feelings. The point to be noted in this observation is that the _vision reproduced as a detail of the somnambulistic act the accidental dropping of a letter from the hand of the somnambulist who was unaware of the fact_; it reproduced what was not in conscious experience. How came it that an act for which there had been no awareness could appear in the vision? The only explanation is that originally in the somnambulistic state, as is so commonly observed in hypnotic somnambulism, there was a subconscious tactual perception (with secondary visual images?) of dropping the letter and now the memory of this antecedent perception, functioning subconsciously, induced this detail of the vision. The general conclusion then would seem to be justified that this hallucination was determined by a fairly large complex of antecedent somnambulistic experiences of which a part emerged as the hallucination and the thoughts of the somnambulist into consciousness, and a part—the tactual and other perceptions—remained submerged as the subconscious process. How much more may have been contained in this process the facts do not enable us to determine. An examination, then, of even the more simple artificial hallucinations discloses that underlying them there is a residual process which is quite an extensive subconscious memory of antecedent thoughts, perceptions and affective experiences. Whether this memory is only an unconscious functioning neurogram or whether it is also a coconscious memory, or partly both, cannot be determined from the data.[97] The bearing of these results upon the interpretation of _insane hallucinations_ is obvious.