The philosophical problem of partial selves cannot be here enlarged on. For a discussion of the subject from the point of view of cerebral localisation, as well as for further quotations from Mr. Newnham's record, I may refer the reader to Mr. Myer's paper on "Automatic Writing," in vol. iii. of the Proceedings of the S.P.R.

A further noteworthy point is that so often the questions and not the answers in the agent's mind should have been telepathically discerned; but we may perhaps conceive that the impulse first conveyed set the percipient's independent activity to work, and so put an end for the moment to the receptive condition. The power to reproduce the actual word thought of is sufficiently shown in the cases where names were given (15 and 87), and in some of the Masonic answers; and the following examples belong to the same class.

48. What name shall we give to our new dog? A. Yesterday was not a fair trial.

49. Why was not yesterday a fair trial? A. Dog.

And again: -

108. What do I mean by chaffing C. about a lilac tree? A. Temper and imagination.

109. You are thinking of somebody else. Please reply to my question. A. Lilacs.

Here a single image or word seems to have made its mark on the percipient's mind, without calling any originative activity into play; and we thus get the naked reproduction. In these last examples we again notice the feature of deferred impression. The influence only gradually became effective, the immediate answer being irrelevant to the question. We may suppose, therefore, that the first effect took place below the threshold of consciousness.

§13.I may now proceed to some further results which were obtained with percipients of less abnormal sensibility, and which demand, therefore, a careful application of the theory of probabilities. For the development of the motor form of experiment in this direction, we have again to thank M. Richet; who here, as in the case of the card-guessing, has brought the calculus to bear effectively on various sets of results many of which, if looked at in separation, would have no significance.1 The fact that the "subjects" of his trials were persons who had betrayed no special aptitude for "mental suggestion," made it clearly desirable that the bodily action required should be of the very simplest sort. The formation of words by a planchette-writer requires, of course, a very complex set of muscular co-ordinations: all that M. Richet sought to obtain was a single movement or twitch. In the earlier trials an object was hidden, and the percipient endeavoured to discover it by means of a sort of divining-rod - the idea being that he involuntarily twitched the rod at the right moment under the influence of "mental suggestion" from the agent, who was watching his movements.

But where the subject of communication is of such an extremely simple kind, very elaborate precautions would be needed to guard against unconscious hints. Indications from the expression or attitude of the "agent" may be prevented by blindfolding the "percipient," and in other ways; but if the two are in close proximity, it is harder to exclude such signs as may be given by involuntary movements, or by changes of breathing. M. Richet's later experiments were ingeniously contrived so as to obviate this objection.

1I have given a fuller description and criticism of M. Richet's investigations in vol. ii. of the Proceedings of the S.P.R.

The place of a planchette was taken by a table, and M. Richet prefaces his account by a succinct statement of the orthodox view as to "table-turning. " Rejecting altogether the three theories which attribute the phenomena to wholesale fraud, to spirits, and to an unknown force, he regards the gyrations and oscillations of seance-tables as due wholly to the unconscious muscular contractions of the sitters. It thus occurred to him to employ a table as an indicator of the movements that might be produced, by "mental suggestion." The plan of the experiments was as follows. Three persons (C, D, and E) took their seats in a semicircle, at a little table on which their hands rested. One of these three was always a "medium" - a term used by M. Richet to denote a person liable to exhibit intelligent movements in which consciousness and will apparently take no part. Attached to the table was a simple electrical apparatus, the effect of which was to ring a bell whenever the current was broken by the tilting of the table. Behind the backs of the sitters at the table was another table, on which was a large alphabet, completely screened from the view of C, D, and E, even had they turned round and endeavoured to see it.

In front of this alphabet sat A, whose duty was to follow the letters slowly and steadily with a pen, returning at once to the beginning as soon as he arrived at the end. At A's side sat B, with a note-book; his duty was to write down the letter at which A's pen happened to be pointing whenever the bell rang. This happened whenever one of the sitters at the table made the simple movement necessary to tilt it. Under these conditions, A and B are apparently mere automata. C, D, and E are little more, being unconscious of tilting the table, which appears to them to tilt itself; but even if they tilted it consciously, and with a conscious desire to dictate words, they have no means of ascertaining at what letter A's pen is pointing at any particular moment; and they might tilt for ever without producing more than an endless series of incoherent letters. Things being arranged thus, a sixth operator, F, stationed himself apart both from the tilting table and from the alphabet, and concentrated his thought on some word of his own choosing, which he had not communicated to the others. The three sitters at the first table engaged in conversation, sang, or told stories; but at intervals the table tilted, the bell rang, and B wrote down the letter which A's pen was opposite to at that moment.