or any other marvel on the score of a few third-hand reports or vague personal experiences. Another turns away from the facts in whatever strength accumulated, on the ground that they are à priori impossible or unprovable. Both are equally remote from the rational scepticism which alone is the proper attitude for approaching psychical investigation. Apart from such an attitude of mind, no treatment of the subject, whether constructive or critical, can be of any value; and here Mr. Peirce and I are wholly at one. But, in an inquiry so novel and difficult, it is likely that two persons, even though they both begin as rational sceptics, will develop differences of opinion; and it is at least equally likely that they will both make mistakes. Thus, some of Mr. Peirce’s strictures depend (as I shall hope to show) on distinct errors and misconceptions, while others appear to me to be unreasonable and overstrained. On the other hand, he has pointed out some errors on my part; and in so doing, and generally in enabling me to make the present apologia, he has done me a valuable service.
Mr. Peirce prefaces his detailed criticisms with a more general remark which cannot be quite passed over. Referring to Phantasms of the Living, Chap. XIII, he objects to the “enormous odds ciphered out in favor of the hypothesis of ghosts,”—more correctly, to the enormous improbability that a certain series of coincidences were due to chance alone—as calculated to “captivate the ignorant,” but to “repel thinking men, who know that no human certitude reaches such figures as trillions or even billions to one.” It is as well to be accurate, even at the risk of repelling “thinking men.” But most thinking men, whose thoughts have been directed to the subject of probabilities, will, I imagine, support me in dissenting from Mr. Peirce’s view. There are many cases of practically absolute certitude, where the actual degree of certitude can be measured. For instance, if dice turned up sixes a hundred times running, which could any day be made to happen, the mathematical probability that the dice were not both evenly weighted and honestly thrown would reach a figure higher than those which have offended Mr. Peirce.
To proceed now at once to his numbered list of objections.
1st. Case 199. The discovery that this incident occurred as long ago as April, 1873, was only made after the work was printed off. (That it was made so late was partly due to a very rare accident—a misspelling of a name in the Register of Deaths at Somerset House. Much time was wasted in the search there, before it occurred to me to apply to the Coroner.) The date has been rectified in the “Additions and Corrections”; and it was careless of me not to remember, when this was done, that the case had been included in the list in Chap. XIII, so as to have added a warning in reference to that list. But, of course, the limitation of the list to cases occurring in a period of twelve years, starting from Jan. 1, 1874, was purely arbitrary. Had a period of thirteen years, starting from Jan. 1, 1873, been selected instead, the numerical argument would not have suffered appreciably, if at all.
Case 355. The inclusion of this case was a bad blunder, for which I take the fullest blame. My eye was misled by the date in the first line of the account; but that, of course, is no excuse.
2nd. This objection seems to me fallacious. We can scarcely doubt that our number of cases would have been increased had we prosecuted our search during the whole period (1874–85), instead of during the last quarter of it only. Had we done so, I should still have been perfectly justified in representing the size of the group of persons to whom we had had access by the number of them all alive at any one time—say half a million—though the half-million would not have throughout consisted of the same individuals. The reason why this would have been legitimate is that in the calculation the whole population is similarly treated. Of course a much larger number of persons are alive during some portion of a period of twelve years than are simultaneously alive on the day of it when the census is taken. And if the group of half a million were increased so as to allow for persons becoming adults during the period, and thus joining the group (so to speak) at one end while others died off it at the other, the size of the whole population would have to be reckoned in a similar way; and the two increases would balance each other in the calculation, which would only be made more complex without being made more correct. Thus any case of percipience within the given period (where the evidence which reaches us is on a par with first-hand (see Vol. I, p. 148)) may be legitimately included, even though the percipient be dead, if it is practically certain that we should equally have obtained it direct from the percipient, had he or she survived. This applies to three of the four cases which Mr. Peirce cites (his number 237, is, I suppose, a mistake for 238). Cases 170 and 695 were obtained through private channels, and Case 238, though our first knowledge of it was due to a published account, would have been at once procured at first-hand from the percipient had we been at work in 1876. The receipt of Case 214, however, was due to a newspaper-appeal of our own, which it is not certain that the deceased percipient would have independently seen and acted on, had it been published during her lifetime; and as, moreover, it is only by straining a definition (as I have pointed out) that this case can be regarded as on a par with first-hand, it would be best to drop it from the list.
3rd. Case 184. Mr. Peirce says that the percipient “seems to have hallucinations nearly every day.” He has had only one other hallucination in his life. This occurred many years ago, in his boyhood, and represented a vague, unrecognized figure. But the list is confined to cases where the appearance was recognized; and the only subjective hallucinations which have to be considered per contra are those presenting the same characteristic. The other experiences from the same informant, Nos. 21, 38, 56, have, in the first place, been coincidental, and have a fair claim to be considered telepathic; and, in the second place, have not been hallucinations at all. They have conveyed no impression of external reality, but are distinctly described as impressions and “mind’s-eye” visions, parallel to those which a good visualizer can summon up at will. Thus Mr. Peirce’s objection is doubly out of place.
Case 175. The percipient draws a distinct line between the experience which he here describes and those which he has had without any coincidence. In the latter he “quite believes he was asleep,”—i.e., there is no ground for regarding them as hallucinations at all, in the sense in which I throughout employ the word.
As regards Cases 173 and 298, Mr. Peirce’s use of the plural “other hallucinations” is misleading. Each of the two percipients has had one other hallucination, and neither of these was of a nature to affect the legitimacy of including their cases in the list. The narrator of Case 173 had once seen an unrecognized figure, which seems curiously to have corresponded in aspect with a person who, unknown to her, had recently died in the room in which it appeared; but it has been impossible to obtain corroboratory evidence of this incident. The other hallucination of the narrator of Case 298 was not visual.
4th. The percipient in Case 29 was in perfect health. (Query—Is it Case 28 that Mr. Peirce means, where the percipient “had a headache”? If so, does he really consider that such a condition at the end of a day’s work amounts to not being “in good health”?)
Case 201. The percipient says, “I had been in ill-health for some years, but at that time was stronger than I ever was in my life, the warm climate suiting me—so well that I felt a strength and enjoyment of life for its own sake, which was a delight to me.” Many of us would be glad enough to be “not in good health” on these terms.
Case 202. The percipient had been ordered to rest and do no work. But hers was not a condition which would have prevented me from counting her hallucination against my argument, as a purely subjective specimen, had she happened to be included in the census, and had no coincidental event occurred.
Case 214. The percipient’s illness succeeded the vision.
In Case 174, the percipient, Miss P., was still “far from well,” having recently had a distinct attack of illness; and in Case 702, the percipient, Mr. G., was weak but convalescent after fever. My information on the subject of hallucinations does not lead me to suppose that there was anything in Mr. G.’s state especially favorable to an experience of the sort; as to Miss P., I cannot tell. Unless their state was so favorable,—indeed, unless visual hallucinations,