§ 1. We have now to quit the experimental branch of our subject. We have been engaged, so far, with cases of thought-transference deliberately sought for and observed within the four walls of a room, both the agent and the percipient being aware of the object in view; and with the further cases where - though the distance between the agent and the percipient was often greater, and the latter had no intimation of what was intended - there was still a deliberate desire on the agent's part to exert a telepathic influence, and a concentration of his mind on that object. For the remainder of our course we shall be entirely occupied with cases where no such desire or idea existed - where the effect produced on the percipient, though we may connect it with the state of the agent, was certainly not an effect which he was aiming at producing. And this change in the character of the facts is accompanied by a marked change in the character of the evidence - a change for which some of the transitional cases in the last chapter have already prepared us. Our conclusions will now have to be drawn from the records of persons who, at the time when the phenomena which they describe took place, were quite unaware that these would ever be used as evidence for telepathy or anything else.

Nor have my colleagues and I any observations of our own to compare with what our witnesses tell us; the facts are known to us only through the medium of their report, and we shall have to decide how far the medium may be a distorting one. Our method of inquiry will thus be the historical method; and success will depend upon the exercise of a wider and less specialised form of common-sense than was required in the experimental work. A great many more points have to be taken into account in weighing human testimony than in arranging the conditions of a crucial trial of thought-transference. There, one precise and simple form of danger had to be guarded against - the possibility of conscious or unconscious physical signs: here, dangers multiform and indeterminate will have to be allowed for. We shall be brought face to face with questions of character, of the general behaviour of human beings in various circumstances, and of the unconscious workings of the human mind; and a quite different sort of logic must come into play, involving often a very complex estimate of probabilities.

So all-important is it for our purpose to form a correct judgment as to the possible sources of error in this new department of evidence, that I have thought it best to devote the present chapter entirely to that subject.

§ 2. First, then, to face the most general objection of all. This may perhaps be stated as follows. All manner of false beliefs have in their day been able to muster a considerable amount of evidence in their support, much of which was certainly not consciously fraudulent. The form of superstition varies with the religious and educational conditions of the time; but within certain limits a diligent collector will be able to obtain evidence for pretty well anything that he chooses. There is, of course, a line - and every age will have its own line - beyond which it would be impossible for anyone who wished to be thought sane and educated to go; for instance, it would be impossible in the present day to obtain anything like respectable contemporary testimony for the transformation of old women into hares and cats. But short of this line there is always a range of ideas and beliefs as to which opinion is divided - which it is perfectly allowable to repudiate, and which science may treat with scorn, but which it is not a sign of abnormal ignorance or stupidity to entertain. And within this range evidence, and even educated evidence, for the beliefs will pretty certainly be forthcoming.

For however much advancing knowledge may have limited the field of superstition, the fund of possibilities in the way of mal-observation, misinterpretation, and exaggeration of facts is still practically inexhaustible; and with such a fund to draw on, the belief, or the mere desire or tendency to believe, in any particular order of phenomena is sure, now and again, to light on facts which can be made to yield the semblance of a proof.

Now, though it is difficult to deny the force of this argument when stated in general terms, I think that it can be shown not seriously to invalidate the evidence which is here relied on as proof of the reality of spontaneous telepathy. For the sake of comparison, it will be worth while to glance at the most striking example that modern times supply of the support of false beliefs by a large array of contemporary evidence - the case of witchcraft.

We may begin by excluding the enormous amount of the witch-evidence which consisted in confessions extracted by torture, terror, or false promises - " the casting evidence in most tryals," as Hutchinson says; and also the large class of cases where the actual facts attested would not be disputed; - as where a woman was condemned because a child who had been with her hung its head on its return home, and rolled over in its cradle in the evening; or because a good many people or cattle had fallen sick in her village; or because she kept a tame frog, presumed to be her "imp"; or because on the very day that she had scolded a carter whose cart knocked up against her house, the self-same cart stuck in a gate, and the men who should have emptied it at night felt too tired to do so.1 Putting these cases aside as irrelevant, anyone who looks carefully into the remaining records will find (1) that the actual testimony on which the alleged facts were believed came exclusively from the uneducated classes; and (2) that the easy acceptance of this evidence by better educated persons was due to the ignorance which was at that time all but universal respecting several great departments of natural phenomena - those of hallucination, trance, hysteria, and mesmerism.