But though numerous, the instances are sporadic; they appear as isolated marvels, which even those who experienced them regarded as such, and not as evidences to any widely-believed reality. So much is this the case that to many persons with whom we have conversed on the subject we find that the very idea of such phenomena is practically new; and that "apparitions," whether delusions or realities, have always been considered by them as apparitions of the dead.1 And if this is true of the more striking telepathic cases, a fortiori is it true of the less striking. The class of apparitions and impressions which have corresponded with the death of the "agent " has only been vaguely recognised; the class which have corresponded with a state of passing excitement or danger can hardly be said to have been recognised at all. Even persons with whose general way of thinking they might seem compatible are apt to be repelled by their apparent uselessness, and certainly are not wont to exhibit any a priori belief in their reality; while to others who have encountered them, they have appeared in the objectionable light of a puzzle, without analogies and without a place in the recognised order of Nature.

1 It is amusing sometimes to encounter arbitrary fragments of scepticism, combined with a belief in the "supernatural" character of many of the coincidences which we are endeavouring to account for as natural. Thus a gentleman contributes a case to Knowledge (May l6th, 1884) and concludes his letter thus : "Personally, I do not believe in apparitions, nor in anything akin thereto; hut coincidences such as you record from week to week must have happened to most of us, and obtuse indeed must the individual be who docs not think that there is something supernatural sometimes even in coincidences".

But though I think that it is not hard to distinguish the evidence on which we rely from the evidence for various forms of popular superstition, and to show that, as a matter of fact, telepathy is not a popular superstition, I am far from denying a certain degree of force to the line of objection above suggested. Ignorance, credulity, and a predisposition to believe in a particular order of marvels, are not the only sources of unconscious falsification in human testimony; and it by no means follows, because these particular elements of error are absent, that a bond-fide first-hand narrative of contemporary facts is trustworthy. And having briefly considered certain dangers and objections from which we think that our telepathic evidence is free, I proceed now to consider certain others to which it is to a certain extent exposed, and to explain the means by which we have endeavoured to obviate or reduce them.

§ 4. It will be best to enumerate, one by one, the general sources of error which may affect the testimony of honest and fairly-educated persons, to events that are both unusual and of a sort unrecognised by contemporary science. We shall thus be able to observe in detail how far each is likely to have affected the evidence here brought forward.

The most obvious danger may seem to lie in errors of observation and inference. And first as to errors of observation. The phenomena Since the above remarks were written, I am glad to find them implicitly confirmed by a very high authority on myth and folk-lore, Mr. Andrew Lang. In the Nineteenth Century for April, 1885, he showed very clearly and amusingly how the same types of "ghost-story" are found in the most distant places, and in the most diverse stages of culture - whether owing to some common basis of fact, or to the same pervading love of the mysterious, or (as is sometimes undoubtedly the case) to the survival of remnants of primitive superstitions in the midst of an advancing civilisation. But though most of his instances are drawn from barbarous countries, he "has not encountered, among savages, more than one example" pointing to a belief in what we call telepathic impressions; and even that one is a very doubtful example. There is, as I have said, a certain amount of sporadic evidence that the phenomena have been noticed at many different times and places; but of any pervading belief, such as would cause people to be on the qui vive for them and would ensure a perpetual supply of spurious evidence, neither we nor apparently Mr. Lang can find any indication whatever, with which these have to do are naturally objective phenomena.

It is only in reference to the objective world that observation can be proved to be accurate or faulty; the faulty observation is that which interprets real things in a way that does not correspond with reality. Now misinterpretation of this sort may undoubtedly produce spurious telepathic cases; and wherever we can suppose it to have been possible, we are bound to exclude the case from our evidence. Thus we have a group of narratives of the following type, suggesting a mistake of identity. Mrs. Campbell, of Dunstaffhage, Oban, wrote, in June, 1884:-

1Next to these, the best-recognised class are undoubtedly the 'premonitory apparitions of "second-sight".

"Two years ago one of our tenant farmers was very ill, and my brother asked me to inquire how he was, on my way back from a walk I was going to take with a cousin of mine. We went, but on passing the old man's house I forgot to go in, and soon we arrived at our avenue, when my cousin reminded me of not having asked for the sick man. I thought of returning, when I distinctly saw the old man, followed by his favourite dog, cross a field in front of us, and go into his house, and I remarked to my cousin, who also had seen the old man and his dog, that as he was so well that he was able to walk about, there was not much use in going to inquire for him, so we went on home. But on arriving there, my brother came to tell us that the old man's son had just been to say that his father had just died".

Here it is possible, and therefore for evidential purposes necessary, to suppose that the figure seen was a neighbour, or perhaps the old man's son.1

[Two further examples are here omitted. - Ed.] In another case, two gentlemen crossed Piccadilly under the impression that they saw a friend, who, as it turned out, died in India on that day. But it is needless to multiply instances; in all of them the figure seen has been out of doors, and at some yards' distance; and these being the very circumstances in which we know that spurious recognitions often take place, there is nothing surprising in an occasional coincidence of the sort described. Similarly, a person may hear a call, perhaps of his own Christian name, outside his house, and may mistake the voice for that of a friend; and, "in due course," as our informants sometimes say, the news of that friend's death may arrive. But it is only to an inconsiderable fraction of the evidence here presented that such explanations could by any possibility be applied. The large majority of the alleged experiences are, on the face of them, subjective phenomena, in the sense that they are independent of any real objects in the environment, and of any mistakes possible in connection with such objects, and are due to a peculiar affection of the percipient's own mind.