It would seem that some hallucinations are best to be explained in this way. We have in fact a regular series of facts which can all be formulated under the single law that _the substantive strength of a state of consciousness bears an inverse proportion to its suggestiveness_. It is the halting-places of our thought which are occupied with distinct imagery. Most of the words we utter have no time to awaken images at all; they simply awaken the following words. But when the sentence _stops_, an image dwells for awhile before the mental eye (see Vol. I. p. 243). Again, whenever the associative processes are reduced and impeded by the approach of unconsciousness, as in falling asleep, or growing faint, or becoming narcotized, we find a concomitant increase in the intensity of whatever partial consciousness may survive. In some people what M. Maury has called 'hypnagogic' hallucinations[126] are the regular concomitant of the process of falling asleep. Trains of faces, landscapes, etc., pass before the mental eye, first as fancies, then as pseudo-hallucinations, finally as full-fledged hallucinations forming dreams. If we regard association-paths as paths of drainage, then the shutting off of one after another of them as the encroaching cerebral paralysis advances ought to act like the plugging of the hole in the bottom of the pail, and make the activity more intense in those systems of cells that retain any activity at all. The level rises because the currents are not drained away, until at last the full sensational explosion may occur. The usual explanation of hypnagogic hallucinations is that they are ideas deprived of their ordinary _reductives_. In somnolescence, sensations being extinct, the mind, it is said, then having no stronger things to compare its ideas with, ascribes to these the fulness of reality. At ordinary times the objects of our imagination are reduced to the _status_ of subjective facts by the ever-present contrast of our sensations with them. Eliminate the sensations, however, this view supposes, and the 'images' are forthwith 'projected' into the outer world and appear as realities. Thus is the illusion of dreams also explained. This, indeed, after a fashion gives an account of the facts.[127] And yet it certainly fails to explain the extraordinary vivacity and completeness of so many of our dream-fantasms. The process of 'imagining' must (in these cases at least[128]) be not merely relatively, but absolutely and in itself more intense than at other times. The fact is, it is not a process of imagining, but a genuine sensational process; and the theory in question is therefore false as far as that point is concerned.