This phenomenon has long been known, although its significance has not been perceived. Brierre de Boismont pointed out that certain vivid dreams are not recognised as dreams, but are mistaken for reality after waking, though he scarcely recognised the normal limitation of this mistake to the hypnagogic state. Moll compared such dreams, thus continued into waking life, to continuative post-hypnotic suggestions. Sully mentioned awaking from dreams which 'still wear the aspect of old acquaintances, so that for the moment I think they are waking realities.'[212] Colegrove, in his study of memory, recorded many cases in which young people mistook their dreams for actual events.[213] This persistence of the memory illusion of sleep into the subsequent hypnagogic state is obviously related to the allied persistence, more occasionally found, of the visual, auditory, and other sensory hallucinations of sleep into the hypnagogic state.[214] Visions thus seen persisting from dreams for a few moments into waking life are often very baffling and disturbing, as has already been pointed out, to ignorant and untrained people. Such visions may occur in the hypnagogic state, even when there has been no conscious precedent dream, and it is indeed probable, as Parish has argued, that it is precisely in the hypnagogic state, the narthex of the church of dreams, as I may term it, that hallucinations are most liable to occur. That illusions may momentarily occur in this state is obvious; thus falling asleep for a few minutes when seated before a black hollow smouldering fire, with red ashes at the bottom, I awake with the illusion that I see a curtain on fire, and have already leaned forward to snatch it away before I realise my mistake. Under normal conditions, the liability of a dream memory to be mistaken for an actual event seems to be greater when an interval has elapsed before the dream is remembered, such an interval making it difficult to distinguish one class of memories from the other, provided the dream has been of a plausible character. Thus Professor Näcke has recorded that his wife dreamed that an acquaintance, an old lady, had called at the house; this dream was apparently forgotten until forty or fifty hours afterwards when, on passing the old lady's house, it was recalled, and the dreamer was only with much difficulty convinced that the dream was not an actual occurrence. When we are concerned with memories of childhood, it not infrequently happens that we cannot distinguish with absolute certainty between real occurrences and what may possibly have been dreams.