827. And here it must be strongly asserted that, however important it may be to work to the full that preliminary inquiry, it is still more important to collect the richest possible harvest of those more advanced cases. To such collection Mr. Moses' writings acted as a powerful stimulant; and ever since my first sight of his MSS. I have made it a principal object to get hold of automatic script from trustworthy sources.

During those twenty-seven years I have personally observed at least fifty cases where there was every reason to suppose that the writing was genuinely automatic; albeit in most of the cases it was uninteresting and non-evidential.

This number is, at any rate, sufficient to enable me to generalise as to the effects of this practice on healthy persons rather less inadequately than writers who generalise from mere hearsay, or from observation of hospital patients.

In two cases I think that the habit of automatic writing (carried on in spite of my warning, by persons over whom I had no influence), may have done some little harm, owing to the obstinate belief of the writers that the obvious trash which they wrote was necessarily true and authoritative. In the remaining cases no apparent harm was done; nor, so far as I know, was there any ill-health or disturbance in connection with the practice. Several of the writers were persons both physically and mentally above the average level.

My own conclusion is that when the writing is presumptuous or nonsensical, or evades test questions, it should be stopped; since in that case it is presumably the mere externalisation of a kind of dream-state of the automatist's; but that when the writing is coherent and straightforward, and especially when some facts unknown to the writer are given as tests of good faith, the practice of automatic writing is harmless, and may lead at any moment to important truth. The persons, in short, who should avoid this experiment are the self-centred and conceited. It is dangerous only to those who are secretly ready - and many are secretly ready - to regard themselves as superior to the rest of mankind.

828. What has now been said may suffice as regards the varieties of mechanism - the different forms of motor automatism - which the messages employ. I shall pass on to consider the contents of the messages, and shall endeavour to classify them according to their apparent sources.

A. In the first place, the message may come from the percipient's own mind; its contents being supplied from the resources of his ordinary memory, or of his more extensive subliminal memory; while the dramatisation of the message - its assumption of some other mind as its source - will resemble the dramatisations of dream or of hypnotic trance.

Of course the absence of facts unknown to the writer is not in itself a proof that the message does not come from some other mind. We cannot be sure that other minds, if they can communicate, will always be at the pains to fill their messages with evidential facts. But, equally of course, a message devoid of such facts must not, on the strength of its mere assertions, be claimed as the product of any but the writer's own mind.

B. Next above the motor messages whose content the automatist's own mental resources might supply, we may place the messages whose content seems to be derived telepathically from the mind of some other person still living on earth; that person being either conscious or unconscious of transmitting the suggestion.

C. Next comes the possibility that the message may emanate from some unembodied intelligence of unknown type - other, at any rate, than the intelligence of the alleged agent. Under this heading come the views which ascribe the messages on the one hand to "elementaries," or even devils, and on the other hand to "guides"or "guardians" of superhuman goodness and wisdom.

D. Finally we have the possibility that the message may be derived, in a more or less direct manner, from the mind of the agent - the departed friend - from whom the communication does actually claim to come.

My main effort has naturally been thus far directed to the proof that there are messages which do not fall into the lowest class, A - in which class most psychologists would still place them all. And I myself - while reserving a certain small portion of the messages for my other classes - do not only admit but assert that the great majority of such communications represent the subliminal workings of the automatist's mind alone. It does not, however, follow that such messages have for us no interest or novelty. On the contrary, they form an instructive, an indispensable transition from psychological introspection of the old-fashioned kind to the bolder methods on whose validity I am anxious to insist. The mind's subliminal action, as thus revealed, differs from the supraliminal in ways which no one anticipated, and which no one can explain. There seem to be subliminal tendencies setting steadily in certain obscure directions, and bearing as little relation to the individual characteristics of the person to the deeps of whose being we have somehow penetrated as profound ocean-currents bear to waves and winds on the surface of the sea.1

1 See Professor James's Psychology, vol. i. p. 394: "One curious thing about trance utterances is their generic similarity in different individuals. ... It seems exactly as if one author composed more than half of the trance messages, no matter by whom they are uttered. Whether all sub-conscious selves are peculiarly susceptible to a certain stratum of the Zeitgeist, and get their inspiration from it, I know not." See the account of automatic and impressional script, by Mr. Sidney Dean, which Professor James goes on to quote, and which is closely parallel to (for instance) Miss A.'s case, to be given below, although the one series of messages comes from the hand of a late member of Congress, "all his life a robust and active journalist, author, and man of affairs," and the other from a young lady with so different a history and entourage.

Is this indeed the drift of the Zeitgeist - as Professor James suggests - steady beneath the tossings and tumblings of individual man? Or is it something independent of age or season? Is there some pattern in the very fabric of our nature which begins to show whenever we scratch the glaze off the stuff?

All this may be better considered hereafter, apart from the evidential discussions with which this chapter must be mainly concerned.

Another point also, of fundamental importance, connected with the powers of the subliminal self, will be better deferred until a later chapter. I have said that a message containing only facts normally known to the automatist must not, on the strength of its mere assertions, be regarded as proceeding from any mind but his own. This seems evident; but the converse proposition is not equally indisputable. We must not take for granted that a message which does contain facts not normally known to the automatist must therefore come from some mind other than his own. If the subliminal self can acquire supernormal knowledge at all, it may obtain such knowledge by means other than telepathic impressions from other minds. It may assimilate its supernormal nutriment also by a directer process - it may devour it not only cooked but raw. Parallel with the possibilities of reception of such knowledge from the influence of other embodied or disembodied minds lies the possibility of its own clairvoyant perception, or active absorption of some kind, of facts lying indefinitely beyond its supraliminal purview.