Such considerations have now led most scientific investigators of dreaming to regard these problems of dream memory as settled. Woodworth's observations on the hypnagogic or half-waking state revealed no remarkable rapidity of mental processes. Clavière showed by experiments with an alarm clock which struck twice with an interval of twenty-two seconds that speech dreams at all events take place merely with normal rapidity, or are even slightly slower than under waking conditions. The imagery of sleep, Clavière concluded, is not more rapid than the imagery of waking life, though to the dreamer it may seem to last for hours or days. It is often slackened rather than accelerated, says Piéron, who refers to the corresponding illusion under the influence of drugs like hashish, though in some cases he finds that there is really a slight acceleration. The illusion is simply due, Foucault thinks, to the dreamer's belief that the events of his dream occupy the same time as real events. This illusion of time, concludes Dr. Justine Tobolowska, in her Paris thesis on this subject, is simply the necessary and constant result of the form assumed by psychic life during sleep.[194] If this peculiarity of memory in dreaming is not difficult to explain as a natural illusion, there are other and rarer characteristics of dream memory which are much more puzzling. In attempting to unravel these, it is probable that, as in explaining the illusion of rapidity, we must always bear in mind the tendency of memory-groups in dreams to fall apart from their waking links of association, so well as the complementary tendency to form associations which in waking life would only be attained by a strained effort. Apperception, with the power it involves of combining and bringing to a focus all the various groups of memories bearing on the point in hand, is defective. The focus of conscious attention is contracted, and there is the curious and significant phenomenon that sleeping consciousness is occasionally unconscious of psychic elements which yet are present just outside it and thrusting imagery into its focus. The imagery becomes conscious, but its relation to the existing focus of consciousness is not consciously perceived. Such a psychic mechanism, as Freud and his disciples have shown, quite commonly appears in hysteria and obsessional neuroses when healthy normal consciousness is degraded to a pathological level resembling that which is normal in dreams.[195] In such a case the surface of sleeping consciousness is, as it were, crumpled up, and the concealed portion appears only at the end of the dream or not at all. A simple example may make this clear. In a dream I ask a lady if she knows the work of the poet Bau; she replies that she does not; then I see before me a paper having on it the name Baudelaire, clearly the name which should have been contained in my query.[196] In such a dream the crumpling