This being so, I cannot see how we can phrase our definition more simply than by saying that any act or condition must be regarded as conscious if it is _potentially memorable_;--if it can be recollected, under any circumstances, by the subject concerned. It does not seem needful that the circumstances under which such recollection may occur should arise while the subject is still incarnated on this planet. We shall never on this planet remember the great majority of our dreams; but those dreams were presumably no less conscious than the dreams which a sudden awakening allowed us to keep in memory. Certain hypnotic subjects, indeed, who can be made to remember their dreams by suggestion, apparently remember dreams previously latent just as easily as dreams previously remembered. And we shall have various other examples of the unexpected recollection of experiences supposed to have been entirely devoid of consciousness. We are bound, I think, to draw at least this negative conclusion: that we must not take for granted that our apparently central consciousness is something wholly different in kind from the minor consciousnesses out of which it is in some sense elaborated. I do indeed believe it to be in an important sense different; but this difference must not be assumed on the basis of our subjective sensations alone. We must approach the whole subject of split or duplicated personalities with no prepossession against the possibility of any given arrangement or division of the total mass of consciousness which exists within us. Before we can picture to ourselves how that mass of consciousness may _disintegrate_, we ought, were it possible, to picture to ourselves how it is in the first instance _integrated_. That, however, is a difficulty which does not begin with the constitution of man. It begins when unicellular develop into multicellular organisms. It is, of course, a mystery how a single cell can hold together, and what kind of unity it can possess. But it is a fresh mystery when several cells cohere in a conjoint and independent life. In the collective unity of certain "colonial animals" we have a kind of sketch or parody of our own complex being. Higher intelligences may possibly see us as we see the hydrozoon--a creature split up into different "persons," a "hydriform person" who feeds, a "medusiform person" who propagates, and so on--elements of the animal differentiated for different ends--interconnected from one point of view as closely as our stomach and brain, yet from another point of view separable existences, capable of detachment and of independent regeneration in all kinds of different ways. Still more composite, though less conspicuously composite, is every animal that we meet as we rise through the scale; and in man we reach the summit both of colonial complexity and of centralised control.