Here I may close this preliminary survey of the possibilities of error which must be constantly kept in view in the investigation of alleged telepathic cases, and which must be either excluded by evidence or carefully allowed for. Both the dangers and the safeguards will, of course, be better realised when we come to the details of particular cases. It does not seem necessary to give a similar synopsis of the evidential flaws and weaknesses which are not in any sense errors. Some of these may be apparent on the very face of the evidence; as when the percipient expressly states that his impression was of an undefined sort, or was of a sort which he had experienced on other occasions without the correspondence of any real event, or that the coincidence of dates, though close, was not exactly ascertained. Others may appear when we take all the circumstances into consideration, although the percipient may fail to admit them; for instance, a person who is in decided anxiety about an absent relative or friend may be regarded as to some extent predisposed to subjective impressions which suggest his presence, so that the accidental coincidence of such an impression with some actual crisis that is apprehended may be regarded as not violently improbable.

All such topics, however, will find a more convenient place in the sequel.

§ 16. And now with regard to the cases that have been included in the evidential part of the present work. A certain separation has been attempted. In the main body of the book, no cases are given which are not first-hand, or of the particular second-hand sort which (as explained on p. 119) is on a par with the first-hand1; or in which the prima facie probability that the facts stated are substantially correct is not tolerably strong. But the Supplement includes a good many second-hand accounts2; as well as first-hand accounts where the evidence, from lack of corroboration or other causes, falls short of the standard previously attained.1 Our principle in selecting cases for the Supplement has been to take only those which - supposing telepathy to be established as a fact in Nature - would reasonably be regarded as examples of it. Their existence adds force to the proof of telepathy; but we should not have put them forward as an adequate proof by themselves. This separation, however, does not apportion the evidential weight of the two divisions with rigid precision. For, given a certain amount of assurance that the facts are correctly reported, the value of the facts in the argument for telepathy will vary according to the class to which they belong.

There are strong classes and weak classes. Now the body of the work includes specimens of purely emotional impressions, and of dreams - classes which we shall find by their very nature to be weak; and more weight might reasonably be attached to some case in the Supplement, even though less completely attested, if it belonged to the strongest class, which we shall find to be the class of waking visual phantasms. And even within the limits of a single class, it is impossible to evaluate the cases with exactness. A phantasm of sight or sound which does not at the moment suggest the appearance or voice of an absent friend, may still - if unique in the percipient's experience, and if the coincidence of time with the friend's death is exact - have about an equal claim to be considered telepathic with a distinctly recognised phantasm, the coincidence of which with the death (though it may have been exact) cannot with certainty be brought closer than three or four days.

1Cases 256, 257, are in part exceptions, but see remarks thereon.

2We have seen that there is one sort of second-hand evidence which must rank as on a par with first-hand. On the same principle there is one sort of third-hand evidence which must rank as on a par with second-hand. A few third-hand accounts of this type have been admitted to the Supplement; and one or two others by special exception.

Then as regards the mere accuracy of the records - though it has been possible to draw up a sort of table of degrees, such a table, of course, affords no final criterion. It is a guide in the dissection of testimony; it directs attention to important structural points; but it takes no account of the living qualities, the character, training, and habits of thought of witnesses. We have included no cases where the witnesses were not, to the best of our belief, honest in intention, and possessed of sufficient intelligence to be competent reporters of definite facts with which they had been closely connected. But the report, say, of a sceptical lawyer or a man of science, who had totally disbelieved in the whole class of phenomena until convinced by his own experience, is naturally stronger evidence than the report of a lady who, whether owing to natural proclivities or to want of scientific training, has no sense of any a priori objections to the telepathic hypothesis. The report of a person who has seen the phantasm of a friend at the time of his death, but considers that the coincidence may have been accidental, is stronger evidence than the report of a person who would regard such a supposition as irreverent.