984. This seems to show that a man's subliminal self may sometimes perceive his own approaching death, and may transmit this knowledge to the empirical self, sometimes by aid of a hallucination. Now we know that the subliminal self may sometimes communicate to other persons knowledge which it cannot or does not communicate to its own empirical self. This is familiar enough in hypnotic experiments, or in spontaneous automatic script, which script may be (for instance) written in a position turned away from the automatist, and may remain unknown to him, although its content must have come from, or passed through, his own deeper being. We know also that an agent has sometimes succeeded in transmitting a phantasmal image of himself to a percipient at a distance, without knowing whether he has, in fact, been successful or no.

It is natural, therefore, to ask whether there is anything to show that the subliminal self ever reveals the approach of death, not to its own empirical self, but to other persons; - showing, perhaps, by a phantasmal image, the source from which the information comes.

To this question we have some ground for returning an affirmative answer, for my readers will remember that there are various cases where the phantom of a person destined soon to die has been seen by a percipient at a distance; nor does it seem that such an apparition depends upon the decedent's own supraliminal effort. On the contrary, it often appears while he is asleep or in a comatose condition (see, e.g., cases in Proceedings S.P.R., vol. xi. pp. 440-454).

While there are thus many precognitions which are in various ways explicable without postulating any direct knowledge of the future, since they may be due only to some kind of inference from a knowledge of existing facts wider than our own, it is possible that other cases may be due to inference of some supernormal kind, - some perception of the future more direct than any which our ordinary minds enjoy. Such are some of the dream-predictions quoted in the Appendices to section 425 (see vol. i. pp. 402-413). It is conceivable that predictions of these and other types may be communicated by disembodied spirits, to whom may also be attributed the cases that suggest an unseen guidance or protection (see 824 and the cases given at the end of 663 A).

985. Finally, we must admit the possibility of a knowledge which comes to a man from no individuated, or at least from no human source; which is no longer inference but the reflection of Reality itself; of the World-Soul as the Future; of a pre-existent Cosmorama of infinite fates.

But before turning to this last line of reflection, I must say a few words as to the relation of our evidence to the problem of Free Will. Here I have a suggestion to make which even in this time-worn controversy is, I think, absolutely novel. It is that we have now a possibility of making the question between liberty and determination a matter of actual experiment.

Let us put that old question in this specific form: "Is there evidence that any power can show me a picture involving my own (so-called) voluntary actions in the future, which picture I cannot by any effort in the smallest degree hinder from becoming actual fact?"

For mere ordinary prevision this would of course be impossible. But we have here certain foreshadowings which depend on no ordinary prevision, and which are more wholly outside ourselves than any information of equally definite character which we can otherwise receive. The scenes or statements thus given in complete detail seem sometimes to be fulfilled with equal completeness. But must they, or must any of them, inevitably be thus fulfilled? Here it is that a possibility of experiment comes in. The experiment indeed cannot be conclusive either way. But suppose that - as in some folklore story - we were to make vigorous effort to avert some incident, and were yet to find that incident fulfil itself, perhaps by dint of that very effort, exactly after the dreaded fashion, - should we not then have some reason to infer that earth-life was not really modifiable by anything which we feel as free-will?

Assuming such a result of our experiment, analogy would at once suggest a further possibility. For our life on earth would then be seen to resemble the experience of the hypnotised subject, fulfilling unwittingly in waking hours the suggestions previously made to him in the trance. We should ask whether in our own history some epoch may have existed in which a self-suggestion may have been given which could similarly dominate our earthly career. Our complex organism, the result of a long previous history, is felt to restrict our so-called voluntary action within narrow limits; and if we possess also a soul independent of the body, it is surely likely that the soul's previous history also - for some previous history any entity so highly specialised as a man's soul must have had - may exercise a determining influence, even more profound than the organism's influence, upon the thoughts and actions of this incarnation. There may, in short, be a kind of alternating personality, expressing itself first in an incorporeal and then in a corporeal state, in such a way that the incorporeal state is the deeper and the more permanent, and that suggestions thence derived influence corporeal life, although the empirical consciousness which governs that life may never know it.

This idea, of course, is not new to religion or to philosophy, in East or West, and it has long since been suggested that our earthly existence may be the inevitable sequel of our past eternity; a predestined pilgrimage on which our true soul looks with calm content, since not one of earth's phantom sorrows can find her unwilling or strike her unaware. The soul foretaught, the body forewrought, - these will move onwards as they must and may; but meanwhile the problem of Liberty and Necessity will no longer be one for earthly experience to discuss; it will be lifted into a pre-natal region, among the secrets of the transcendental world.

All this must be conceived as possible; yet I do not think that our evidence thus far collected does in fact make for this view of pre-deter-mined earthly fates. Rather we have seen that in many cases monitions have averted incidents which would doubtless have occurred had the percipient received no warning. And where dangers have been foreshown and yet not averted, this seems often to have been because no adequate effort was made to avert them. The problem which our narratives more urgently suggest is how to reconcile so much foreknowledge with so much freedom. I have suggested elsewhere that this problem of free human wills amid the predictable operations of unchanging law may resemble the problem of molecular motion amid molar calm. Clear and stable is for us the diamond; the dewdrop is clear and still; yet within their tranquil clarity a myriad molecules jostle in narrow orbits, or speed on an uncom-puted way. So to "the spectator of all Time and of all Existence" may the Cosmos be "as one entire and perfect chrysolite"; and yet man's petty hopes and passions may make endless turmoil among its minutest elements and in its infinitesimal grains.

Those movements, too, must be ruled by unknown law; yet on a wide view they will average out, and will admit of predictions fulfilled immutably, and overriding the small Wills of men.