All these are fragmentary indications,--useless for practical purposes if you will,--of sleeping faculty exercised on the same order of things as waking faculty, and with comparable or even superior power. But we were bound to push our inquiry further still--we were bound to ask whether the self of sleep showed any faculty of a quite different order from that by which waking consciousness maintains the activity of man. We found that this was so indeed; that there was evidence that the sleeping spirit was susceptible of relations unfettered by spatial bonds; of telæsthetic perception of distant scenes; of telepathic communication with distant persons, or even with spirits of whom we can predicate neither distance nor nearness, since they are released from the prison of the flesh. The inference which all this evidence suggests is entirely in accordance with the hypothesis on which my whole work is based. I have assumed that man is an organism informed or possessed by a soul. This view obviously involves the hypothesis that we are living a life in two worlds at once; a planetary life in this material world, to which the organism is intended to react; and also a cosmic life in that spiritual or metetherial world, which is the native environment of the soul. From that unseen world the energy of the organism needs to be perpetually replenished. That replenishment we cannot understand: we may figure it to ourselves as a protoplasmic process;--as some relation between protoplasm, ether, and whatever is beyond ether, on which it is at present useless to speculate. Admitting, for the sake of argument, these vast assumptions, it will be easy to draw the further inference that it may be needful that the soul's attention should be frequently withdrawn from the business of earthly life, so as to pursue with greater intensity what we may call its protoplasmic task,--the maintenance of the fundamental, pervading connection between the organism and the spiritual world. Nay, this profounder condition, as responding to more primitive, more fundamental needs, will itself be more primitive than the waking state. And this is so: sleep is the infant's dominant phase: the pre-natal state resembles sleep rather than waking; and so does the whole life-condition of our lowly ancestors. And as the sleeping state is the more _primitive_, so also is it the more _generalised_, and the more _plastic_. Out of this dreamy abeyance between two worlds, the needs of the material world are constantly developing some form of alert activity, some faculty which was _potential_ only until search for food and the defence against enemies compelled a closer heed to "the life of relation," lest the relation should become only that of victim to devourer.