'Sleep from the psychological point of view,' says Foucault, 'is a state of profound distraction or total inattention.' And Mourre shows by dreams of his own that any exercise of will in dreaming leads to awakening, and that the deeper the sleep the more absent is volition from dreams. Hence the involuntary wavering and perpetually mere meaningless change of dream imagery. Such concentration as is possible during sleep usually reveals a shifting, oscillating, uncertain movement of the vision before us. We are, as it were, reading a sign-post in the dusk, or making guesses at the names of the stations as our express train flashes by the painted letters. It is this factor in dreams which causes them so often to baffle our analysis. There is thus a failure of sleeping attention to fix definitely the final result--a failure which itself may evidently serve to carry on the dream process by suggesting new images and combinations. It can scarcely be said, however, that the question of attention in dreams is thus settled. It would be inconceivable that the terrible occurrences that may overtake us in dreams and the emotional turmoil aroused should be accompanied by 'total inattention and distraction.' Nor can it be said that that supposition agrees with the vivid memory which our dreams sometimes leave. We can probably account for the phenomena much more satisfactorily by adopting Ribot's useful distinction between voluntary attention and spontaneous attention.[16] Voluntary or artificial attention is a product of education and training. It is directed by extrinsic force, is the result of deliberation, and is accompanied by some feeling of effort. It always acts on the muscles and by the muscles; without muscular tension there can be no voluntary attention. Spontaneous or natural attention, on the other hand, is that more fundamental kind of attention which exists anteriorly to any education or training, and is the only kind of attention which animals and young children are capable of. It may be weak or strong, but always and everywhere it is based on emotional states; every creature moved by pleasure and pain is capable of spontaneous attention under the influence of those stimuli. These two kinds of attention are at the opposite poles from each other, and are incompatible with each other. There can be no doubt that, as Ribot himself pointed out, it is voluntary attention that is defective (though it may not always be entirely absent) in dreams;[17] the muscular weakness and inco-ordination of sleep involve this lack of attention which is indeed an essential condition of the restoration and repose of sleep. But all the characters of spontaneous attention are present. The attention we exercise in dreams is mainly of this fundamental, automatic, involuntary character, conditioned by the emotions we experience, and for the most part escaping all the efforts of our voluntary attention. Further, it has been ably