When repose, the herald of sleep and dreams, begins, my thoughts wander in an irregular and somewhat confused manner.  As they are gradually subjected to the associations to which they successively give rise, they are transformed into more vivid images, a vividness which is always in inverse proportion to the attention.  This gradually produces the state which has been described by Maury and others as hypnagogic hallucination; that is, the images seem to be real, although the subject is still partly awake, and the voluntary exercise of thought is lost from time to time in this species of incipient chaos.  It is at this point that images are really most intense, and that every idea assumes a body and form, every image a reality:  finally, when the body and the brain have reached the physiological conditions of sleep, thoughts which had been changed into hypnagogic images in the intermediate stage between sleep and waking, are altogether transformed into the real images of dreams.

By an effort of will I have often been able to surprise myself in this intermediate stage, and the same thing has been done by others, and it always appears that this is the real moment of transition from wakefulness to dreaming, I have been able to verify the fact that the first dream is only the continuation of our last waking thoughts, which have now become dramatic and real I have also observed that this intermediate stage between waking and dreaming, during which the images are real and vivid, although we are still conscious of our real condition, goes on for a long while, sometimes for a whole night, with brief intervals of sleep.  This has occurred to me when I was kept awake, either when travelling at night, or when I had taken a large draught of water before lying down (other liquids or food does not produce the phenomenon) or if I have been looking during the day at objects illuminated by dazzling sunshine.  In all these circumstances the bright and vivid images appear reduced to an almost microscopic scale, although very distinct in form and colour; in ordinary cases, the images appear of the ordinary size, but not without a tendency to become smaller.

I believe that there is a physical cause for the reduction and attenuation of the images in the excessive excitement of the retina, or central encephalic organ in which images are formed in conscious concurrence with the cortical part of the hemispheres.  Owing to the excitement caused by wakefulness, by fatigue, by sunshine, or in some cases by the condition of the nerves of the stomach, the objective projection on psychical space, partly transmitted by heredity and gradually formed by associations and local signs,[34] is arrested by the innate force of the image on the organ, and it appears to be smaller and in proportion with the relative smallness of the image which is produced by minute vibrations and by the susceptibility of the cellule.  This intermediate and persistent stage of hypnagogic images serves in every way to explain the physical genesis of involuntary hallucinations.