910. Now when we proceed from these apparitions of the living to apparitions of the departed, we find very much the same types persisting still. We find symbolical visions of departed persons, and of scenes among which they seem to dwell. We find externalised apparitions or phantasms of departed persons, - indicating that some point in the percipient's brain has been stimulated by his own or by some other spirit. And finally, as has already been said, we find that in certain cases of possession these two kinds of influence are simultaneously carried to an extreme. The percipient automatist of earlier stages becomes no longer a percipient but an automatist pure and simple, - so far as his body is concerned, - for his whole brain - not one point alone - seems now to be stimulated and controlled by an extraneous spirit, and he is not himself aware of what his body writes or utters. And meantime his spirit, partially set free from the body, may be purely percipient; - may be enjoying that other spiritual form of communication more completely than in any type of vision which our description had hitherto reached.

911. This point attained, another analogy, already mentioned, will be at once recalled. There is another class of phenomena, besides telepathy, of which this definition of possession at once reminds us. We have dealt much with secondary personalities, - with severances and alternations affecting a man's own spirit, in varying relation with his organism. Felida X.'s developed secondary personality, for instance (231 A), might be defined as another fragment - or another synthesis - of Felida's spirit acting upon her organism in much the same way as the original fragment - or the primary synthesis - of her spirit was wont to act upon it.

Plainly, this analogy is close enough to be likely to lead to practical confusion. On what grounds can we base our distinctions? What justifies us in saying that Felida X.'s organism was controlled only by another modification of her own personality, but that Mrs. Piper's is controlled by George Pelham (959)? May there not be any amount of self-suggestion, colouring with the fictitious hue of all kinds of identities what is in reality no more than an allotropic form of the entranced person himself? Is even the possession by the new personality of some fragments of fresh knowledge any proof of spirit-control? May not that knowledge be gained (as by Leonie B., see 230 A and 568 A) clairvoyantly or telepathically, with no intervention of any spirit other than of living men?

Yes, indeed, we must reply, there is here a danger of confusion, there is a lack of any well-defined dividing line. While we must decide on general rules, we must also keep our minds open to possible exceptions.

On the negative side, indeed, general rules will carry us a good way. We must not allow ourselves to ascribe to spirit-control cases where no new knowledge is shown in the trance state. And this rule has at once an important consequence, - a consequence which profoundly modifies the antique idea of possession. I know of no evidence, - reaching in any way our habitual standard, - either for angelic, for diabolical, or for hostile possession.