Nevertheless, I do not think that it is necessary to appeal to occultism for an explanation of the case. In the first place, the theory of fraud was not actually disproved by the investigators. They say they took all kinds of precautions, but it has not been proved that Mrs. Piper might not have previously found out at least much of what she communicated. There is no necessity to assume that all successes can be explained in this way. Dessoir certainly further admits that some of the successes might be accounted for by the medium carefully putting together all that she heard during the stance. When, for example, the medium posed as the spirit of Hyslop's dead father and said to the son, "You are not the strongest man," and the son remarked, " He warned me hundreds of times that I was not as strong as other people," Max Dessoir thinks that this can be explained by the fact that any one could see that Hyslop was not very robust. At all events, some of the cases could quite well be explained in this way.

When we further come to consider that in a whole series of cases the statements were wrong, there seems to be very little necessity indeed for our assuming that Mrs. Piper possesses unknown powers.

The fact that memory is not infallible also fills me with mistrust. In any case, some one was present at Hyslop's experiments who was acquainted with the episodes related by the medium, and of whom, at least, it was not proved that he was ignorant of their occurrence. We may add that letters subsequently received showed that some of the communications made by Mrs. Piper were correct. As we shall see later on, some of these cases can be very easily explained as mere toincidences.

4. Suggestive gestures and other signs, by means of which the subject learns what he or she is expected to do, play an important part in those sources of error that lie outside the person of the subject. When dealing with the question of the training of hypnotics I explained how often the latter perceive signals in the most unimportant signs. We can readily understand that such insignificant signs given by the experimenter or anybody else present constitute an important source of error. It is essentially to hypnotism that we owe the discovery of this source of error, since the training of hypnotics clearly showed that even insignificant signs act suggestively on the subject experimented on. When discussing super-normal thought-transference, I gave a detailed explanation of the way in which the most diverse organs of sense may help the subject to make use of such signs. Consequently, whenever there is the possibility of the subject being influenced by signs, it is the first duty of the experimenter to impose such conditions that the subject will be unable to perceive such signs.

As another of the chief sources of error I must, moreover, add the circumstance, that if the experimenter or any other person present believes, or begins to believe, in the reality of the phenomena, they very easily become unintentional accessories of the medium. It is just the same as with table-turning, in which the believers push harder than the unbelievers and so set the table in motion.

It must also be pointed out, that by means of signs things may be communicated which, at the same time, are nonexistent in the primary consciousness of the agent who communicates them. We have already seen this in automatic writing (p. 246). But if, as Lehmann and Hansen have pointed out, unintentional whispering plays so great a part, it is quite conceivable that some people, perhaps those who are psychically predisposed to the influence of processes going on in the secondary consciousness, may be induced to whisper by such processes. I am led to this assumption by the reflection that involuntary whistling is a process analogous to automatic writing. If subconscious processes can be communicated by means of automatic writing, it is presumable that the same thing can be done by whispering. In any case, I have recently learned from some experiments I made - they did not produce anything new in this respect - that thought-reading is possible even when the subject is not conscious of what has to be done. I requested a hypnotized subject to write down the problem I was to solve on a piece of paper.

In one case I was to take a duster from its place and put it in a cigar-box. The subject wrote this down on a piece of paper which she folded while I stood apart and therefore had no knowledge of the nature of the task. I then suggested to the subject, who was still in hypnosis, that on waking she was to fix her thoughts intently on the task but was not to remember its nature consciously. At the same time I suggested to her - I had often made the ordinary thought-reading experiments with her - that she was not to notice whether she made any movement or was guiding me. I then woke her up. In spite of the fact that I did not know what the task was, I performed it correctly, guided by her. Since the task was written on the paper the experiment was easily controlled. The subject neither knew that she moved nor was she conscious of the processes that made her do so. This case is only intended to show that processes going on in the secondary consciousness can cause the subject to make signals. The experiment was quite a simple one. Finally, if this sort of thing does happen, the assumption that there are people who can make known the processes of their secondary consciousness unintentionally and without noticing it would not be contrary to our scientific ideas or the views we already hold.

This is also a very important source of error in clairvoyance, and may have played a part in Hyslop's experiments, for, at all events, it was never proved that he had not previously heard of the things that were then told him about his relatives - i.e., that the matters in question were not hidden in his secondary consciousness. Eberhard Wolff even attempts to utilize the case of Mrs. Piper as a proof that unintentional whispering does not explain the phenomena of telepathy. I do not consider that the conclusions hitherto arrived at finally decide whether subconscious processes may not be communicated by involuntary whispering. 4. Fictitious interpretation. It should not be forgotten that the vast majority of prophecies are as indefinite as the sayings of the Delphian Oracle, but that the subsequent interpretation of them leads the uncritical to believe that even clairvoyance is an established fact. We come across a similar phenomenon in cases of thought-transference. As Lehmann and Hansen have shown, the drawings said to be reproduced by telepathy are often so indefinite that they may be taken to represent anything you please. They have shown, for example, that the same drawing will do for a candlestick or a cat.