When we observe a person sleeping, we notice that he moves, that he reacts to sensory stimuli, that he covers himself up again if one takes the bed-clothes away, that he not infrequently speaks, groans, or leaves off snoring when told to do so, and even that he answers when spoken to, and may get up and do things occasionally. Certain people sleep very lightly and very quietly, and awaken at the faintest sound. These people show more connection with the outer world.

We only know our sleep subjectively - that is, know the linking of the reflection of our waking consciousness - by the remembrance of our dreams. We feel that our dream consciousness is different from our waking consciousness, but that it approaches the latter more nearly the lighter our sleep is. The sleep consciousness can be differentiated, above all, from the waking consciousness by the following facts, as far as our dream consciousness permits the latter to gain an insight into it:

(1) Sleep consciousness does not show a sharp division between inner conceptions and perceptions. All conceptions are more or less hallucinated - i.e., they have the subjective characters of perceptions, and simulate real occurrences.

(2) The sharpness and precision of the waking perceptions, which are produced by outer processes, are usually absent during these sleep or dream hallucinations. The latter appear, however, with very intense accentuations of feelings, and may exercise powerful reactions on the central nervous system. A dream can produce sweating, convulsive muscle contractions, extreme terror, etc. Erotic dreams produce pollutions without mechanical stimulation of the penis, while erotic perceptions during waking rarely can do this.

(3) The dream hallucinations are very faultily associated, in contradistinction to thinking and perceiving when awake. As a rule, only loosely connected outer associations link one with the other. The organized unconscious logic of thinking during waking, which becomes instinctive, and which is gradually automatized by the psychical dynamisms during the course of life, is not applicable to thinking in sleep. The brain ob-viously is in a condition of relative inactivity or inhibition during sleep. The most abject, consummate nonsense is therefore dreamed, is associated and perceived quite falsely as far as time and place is concerned in dreams, and is even believed in. As a rule, it is only during light sleep, and rarely during deep sleep, that a higher or lower degree of logical control is produced. At times this logical control exists side by side with the dreamed nonsense. It is as if two consciousnesses were present simultaneously - the one that of the dream chain, which believes in the nonsense; and the other that of the waking logical associations, which says: "No, this is all dream nonsense; I am lying half asleep in bed."

The three typical characteristics of the dream existence are, at the same time, the criteria of hypnotic consciousness. They are: hallucinations of perception, exaggerated feeling and reflex actions of the same, and dissociation of the organic logical associations of the engram complexes. They are the best foundations for a marked suggestibility.

Awakening, the reverse of going to sleep, shows the same suggestive phenomena as the going to sleep. One usually awakes at a certain accustomed time by means of associations. A light sleep frequently forms a gradual transition from the sleep to the awakening, and the remembrances of dreams are left behind. Dreams not infrequently awaken the subject. The capability possessed by many people to awaken at a definite chosen time is curious. Here time is exactly measured during sleep. We meet with the same thing in hypnosis.

Liebeault distinguishes in normal sleep, as in hypnosis, the light sleep, with recollections of dreams, from the deep sleep, which usually is not accompanied by such recollections. The characteristic of the latter is the total amnesia on awakening. But we find, nevertheless, that people who sleep deeply are just those who exhibit the phenomena of somnambulism and "sleep drunkenness," during both of which they can walk, do things - sometimes even ordered and complicated things - speak, and even exert violence. These are phenomena which have been recognized in jurisprudence as a ground for irresponsibility. This shows that the amnesia after deep sleep is only amnesia, and proves that the consciousness is by no means blotted out during deep sleep, but is only cut off from the waking consciousness. Still, the lethargic sleep evidences itself in a different way than in somnambulism, with its narrowed consciousness; but one is not justified in deducing the existence of complete immovability of the cortex from the immovability of the motor area. Friedrich Heerwagen published under Kraepelin's direction in Wundt's "Philosophical Studies" his "Statistical Investigations of Dreams and Sleep," which is based on the personal statements of many people. The statement of those persons that they dream a lot, dream little, or do not dream at all, is, according to Heerwagen, to be accepted, and this forms the foundation of bis statistics. But since the study of hypnotism and many experiences of normal sleep prove that one must not rely on these subjective recollections of dreams, or on the non-recollection of them, I cannot ascribe any value to these statistics, but believe all the more that everybody dreams continuously during sleep. Many people forget all their dreams, and the majority forget the greater part of their dreams (autosuggestion of amnesia). I cannot be awakened so unexpectedly at any time of the night that I do not catch on at all events the last portion of a dream chain; but I forget this immediately unless I write it down at once, or energetically reperceive it during the waking condition. That which remains, then, in my memory is the picture of the perception renewed in the condition of waking, and not the direct recollection of the dream, for the latter is almost always obliterated very soon after awakening.