What more natural, therefore, than that those who saw the absurdity of these pretensions should regard further inquiry or suspension of judgment as a concession to ignorant credulity? "Irving Bishop," it seemed fair to argue, "is a professed 'thought-reader'; Irving Bishop's tricks are, at best, mere feats of muscular and tactile sensibility; ergo whoever believes that there is such a thing as ' thought-reading ' is on a par with the crowd who are mystified by Irving Bishop".

§ 4. If, then, the ground of experiment had remained unchanged -if the old " willing-game " had merely continued to appear in various forms - no definite advance could have been made. But on the path of the old experiments, a quite new phenomenon now presented itself, which no one could have confidently anticipated, but for which the suggestions drawn from the most advanced phenomena of the "willing-game" had to some extent prepared the way. It was discovered that not only transferences of impression could take place without contact, but that there was no necessity for the result aimed at to involve movements; the fact of the transference might be shown, not - as in the "willing-game" - by the subject's ability to do something, but by his ability to discern and describe an object thought of by the "wilier." Both parties could thus remain perfectly still; which was really a more important condition than even the absence of contact. In this form of experiment, muscle-reading and all the subtler forms of unconscious guidance are completely excluded; and the dangers which remain are such as can, with sufficient care, be clearly defined and safely guarded against.

Indications of a visual kind - for instance, by the involuntary direction of glances - have no scope if the object which the percipient is to name is not present or visible in the room. There is, of course, an obvious danger in low whispering, or even soundless movements of the lips; while the faintest accent of approval or disapproval in question or comment may give a hint as to whether the effort is tending in the right direction, and thus guide to the mark by successive approximations. Any exhibition of the kind before a promiscuous company is nearly sure to be vitiated by the latter source of error. But when the experiments are carried on in a limited circle of persons known to each other, and amenable to scientific control, it is not hard for those engaged to set a watch on their own and on each other's lips; and questions and comments can be entirely forbidden.

I have been speaking of the danger of involuntary guidance. There is, of course, another danger to be considered - that of voluntary guidance -of actual collusion between the agent and percipient. Contact being excluded, such guidance would have to be by signals; and it is impossible to lay down any precise limit to the degree of perfection that a plan of signalling may reach. The long and short signs of the Morse code admit of many varieties of application; and though the channels of sight and touch may be cut off, it is difficult entirely to cut off that of hearing. Shufflings of the feet, coughs, irregularities of breathing, all offer available material. But though the precise line of possibilities in this direction cannot be drawn, we are at any rate able to suggest cases where the line would be clearly overpassed. For instance, if the idea to be transferred from the agent to the percipient is inexpressible in less than twenty words; and if hearing is the only sensory channel left open; and if it is carefully observed that there are no coughs or shufflings, and that the agent's breathing appears regular, then one seems justified in saying that the necessary information could not be conveyed by a code without a very considerable expenditure of time, and a very abnormally acute sense of hearing on the percipient's part.

There is no relation whatever between a private experiment performed under such conditions as these, and the feats of a conjurer, like Mr. Maskelyne, who commands secret apparatus, and whose every word and gesture may be observed and interpreted by a concealed confederate.

It would be rash, however, to represent as crucial any apparent transferences of thought between persons not absolutely separated, where the good faith of at least one of the two is not accepted as beyond question, and where the genuineness of the result is left to depend on the perfection with which third parties have arranged conditions and guarded against signs. The conditions of a crucial result, for one's own mind, are either (1) that the agent or the percipient shall be oneself; or (2) that the agent or percipient shall be someone whose experience, as recorded by himself, is indistinguishable in certainty from one's own; or (3) that there shall be several agents or percipients, in the case of each of whom the improbability of deceit, or of such imbecility as would take the place of deceit, is so great that the combination of improbabilities amounts to a moral impossibility. The third mode of attaining conviction is the most practically important. For it is not to be expected of most people that, within a short time, they will either themselves be, or have intimate friends who are, successful agents or percipients; and they are justified, therefore, in demanding that the evidence to which they might fairly refuse credence if it depended on the veracity and intelligence of one or two persons, of however unblemished a reputation, shall be multiplied for their benefit.

Whatever be the experimenter's assurance as to the perfection of his conditions, it is in the nature of things impossible that strangers, who only read and have not seen, should be infected by it. They cannot be absolutely certain that this, that, or the other stick might not break; then enough sticks must be collected and tied together to make a faggot of a strength which shall defy suspicion.1 As regards the experiments of which I am about to present a sketch, it is not necessary to my argument that any individual's honesty shall be completely assumed, in the sense of being used as a certain basis for conclusions. The proof must depend on the number of persons, reputed honest and intelligent, to whom dishonesty or imbecility must be attributed if the conclusions are wrong, i.e., it must be a cumulative proof. Not that my colleagues and I have any doubt as to the bona fides of every case here recorded. But even where our grounds of certainty are most obvious, they cannot be made entirely obvious to those to whom we and our more intimate associates are personally unknown; while outside this inner circle our confidence depends on points that can scarcely even be suggested to others - on views of character gradually built up out of a number of small and often indefinable items of conversation and demeanour.