The Society, as represented in the two bulky volumes entitled " Phantasms of the Living," edited by Edmund Gurney, Frederic W. H. Myers, and Frank Podmore, does not claim that the cases which they have presented, drawn from dreams, would be sufficient to prove the truth of telepathy. They confess that they are on doubtful ground, and say:

For (1) the details of the reality, when known, will be very apt to be read back into the dream, through the general tendency to make vague things distinct; and (2) the great multitude of dreams may seem to afford almost limitless scope for accidental correspondences of a dream with an actual occurrence resembling the one dreamt of. Any answer to this last objection must depend on statistics, which, until lately, there has been no attempt to obtain; and though an answer of a sort can be given, it is not such a one as wrould justify us in basing a theory of telepathy on the facts of dreams alone.

They acknowledge that, dreams being often somewhat dim and shapeless things, " subsequent knowledge of events may easily have the effect of giving body and definiteness to the recollection of a dream." They concede that "millions of people dream every night, and in dreams, if anywhere, the range of possibilities seems infinite." But when they come to present the subsequent cases, their reasoning upon them is in many instances unscientific in its destitution of rigor. For example, in cases of partial fulfilment where a person dreamed of death, and the dream did not occur until a number of hours after the death, they call that a deferment of percipiencc. They say that the impression when it first arrived "was unable to compete at the moment with the vivid sensory impressions and the crowd of ideas and images that had belonged to normal senses and waking life, and that it may thus remain latent until darkness and quiet give a chance for its development." The same sort of reasoning might be applied to account for the fact that such information is not universally communicated. It is flying about loose in the heavens and in the earth; but, not being able to compete with the crowd of images in any except few cases, does not generally materialize.

When they come to cases where the dreams contain the general feature of conversation between the dreamer and the agent, they say, " This is, of course, a clear instance of something superadded by the dreamer's own activity"; and when the circumstances of the death do not concur with it, they claim a fulfilment, and attribute a failure to agree to a death imagery superadded by the independent activity of the dreamer.

Where a woman dreams twice of death and it is fulfilled, and she also has the candor to state that on another occasion she dreamed of a death and nothing came of it, they say:

The absence of any ascertained coincidence on the third occasion might be represented as an argument for regarding the correspondence on the two previous occasions as accidental, but it woidd be a very weak one; since even if the dream had recurred a thousand times, the chances against the accidental occurrences of two such coincidences would still remain enormous.

Many of the cases they cite depend upon vague memory, and others do not supply adequate particulars.

Their general method of writing about these dreams and of the whole theory of telepathy is that of an affectionate mother lingering over her child, and wherever coddling is necessary doing it con an/ore There arc two radical defects to be seen in the entire method: First, not a twentieth part of the care is taken in the investigation of the cases and their authentication wrhich wrould be required for a case of ordinary importance in a court of justice; secondly, the use of the so-called doctrine of chances is so ludicrous as to be practically a burlesque of science. They sent to 5360 persons taken at random, asking them to state whether they had ever had a dream of the death of some person known to them, which dream was an exceptionally vivid one, and of which the distressing impression lasted an hour after arising in the morning, at any time within the twelve years 1874 to 1885 inclusive. Of these 173 answered " Yes." It would be difficult to believe, if it were not published to the world on the authority of the society, that any one should conclude that that number could furnish a basis upon which to ascertain an average to be applied to the whole population; yet they do so, and say that it is as satisfactory as the proof that a similar number of persons taken at random would afford on the average number of cases of short sight or colorblindness.

Short sight and color-blindness are physical conditions, depending upon physical causes ; dreams are evanescent, irregular, depending upon phenomenal causes; and the dream-images of a single family in a single week may amount to hundreds of thousands, of which any one under the operations of laws not subject to statistics may be vividly remembered.

But of the whole number of 173 who had vivid dreams of death, there were only 24 where the event fell within 12 hours of the dream. By an application of the law of chance they endeavor to maintain that there would not be more than one such coincidence in that time, and that therefore "twenty-four is twenty-four times larger than the doctrine of chance would have allowed us to expect." As well might the law of chance be applied to the determination of the number of thoughts on any given subject that would naturally arise in one or more minds in a given period.

As shown in Chapter II, the "law of chance" is not capable of application to such subjects. Events are continually occurring, whether attention is directed to them or not. Of all possible occurrences, the time, place, and manner of death are most uncertain. Human lives revolve about a few central points—home, business, health, friends, travel, religion, country. Dream-images are about persons and things. That there can be millions of images portrayed in the gallery of dreams, and that the great majority deal with these pivotal points of human life and human thought, taken in connection with the fact that all the events of human history, past, current, and future, revolve about these same points, make it absolutely certain that the number of coincidences must be vast. It is, in fact, smaller rather than larger than might reasonably be expected.