evoked not by conscious effort, but from some unconscious region of my mind. And an act of reflection is needed to enable me to be sure that it is not a reality. The act of reflection in this case is of course so habitual and easy that it generally passes unnoticed; but a dream may easily slide into a waking hallucination. I may dream of Smith, and after waking I may still seem for a few moments to see him standing beside me. In such a case the dream actually manifests itself as a sensory hallucination; there is the dreamimage; and for a few moments it deceives even the waking senses.

clear. Suppose that I have a friend Smith | dreams are hallucinations. It is a figure whom I expect to see. I see some other man in the twilight and take him for Smith. This is a mere mistake; but it probably involves something of illusion; that is to say, that my mental interpretation of the vague figure actually seen contains certain elements drawn from my recollection of Smith. I go into the house and see Smith, as I fancy, sitting in a chair by the fire. On going closer I find that what I saw was only a coat thrown over the back of the chair. This is a fullblown illusion, and it possibly contains something of hallucination also. Part of the form of Smith, perhaps, was actually invented, was actually externalized, by my mind, was not merely the result of unconscious selection amongst the confused lines of the coat and chair. I then sit down and think of Smith. If I have good visualizing memory I can fancy Smith sitting in the chair can draw a sketch of him as he would look in the chair, correcting my drawing from time to time by reference to the picture of him in my "mind's eye." But this is not a hallucination. I am not deceived by my self-summoned picture. It is called into being by the conscious part of my mind, and I know perfectly well that it is only my imagination.

And now suppose that I suddenly see Smith walk into the room as I think. I start up to greet him, but the figure passes on and walks out through the wall. This is a hallucination; it is a percept, or thing seen (I am here for simplicity's sake taking sight as the representative sense), which lacks the objective basis which it suggests; that is to say, which does not really tell me truly that Smith is there in the room, and would be seen by other persons as well as myself. And note at the same time that it has required a distinct though of course a momentary - act of reflection on my part to assure me that this figure was not actually Smith. This act of reflection was not needed when I had merely summoned up a mind's-eye picture of Smith. That was not a hallucination, it was a figure which my conscious self summoned up, and I knew (in a certain sense) why it came and how it got there. But the unexpected figure of Smith coming in at the door was summoned up by some unconscious part of myself; it took me by surprise, it was a hallucination.

Once more. Suppose that I go to sleep and dream that I see Smith. Is this a hallucination? The answer must be: Yes,

Well, then, hallucinations are images sensations of sight, sound, taste, smell, touch-which are not due to any object in the world about us, and are not set going by our conscious mind, but by some working of the brain of which we, our recognized habitual selves, are not aware. And, having got thus far with our defini. tion, we see at once both why hallucinations have in times past been neglected even by philosophers, or treated as mere meaningless disturbances of our rational being; and also why, with the gradual rise of a more searching psychology, they come to have a profound interest of their own.

The reason is that they are messages whose obvious superficial meaning is false or nonsensical; but from which, nevertheless, an indication may be drawn of the nature of processes within us which we cannot get at in any other way.

The value in diagnosis of the indica. tions given by the hallucinations of the insane has long been recognized. With the hallucinations of insanity or delirium, however, we have here nothing to do; our present inquiry is restricted to sane persons, most of them, as we shall soon see, in perfectly normal health. Now until lately it was hardly thought possible for a sane and healthy person to undergo a hallucination. Hallucinations were vaguely confounded with nightmares; and if any one said that he had "seen a ghost," the recognized joke was to bid him "cure it with a pill " and avoid late suppers. Now late suppers will certainly produce nightmares, vague, dreamy oppressions of circulation or breathing, etc.; but, oddly enough, we cannot find among several hundreds of recent first-hand cases, which we have collected and studied, a single one where over-eating seems to have been the exciting cause of any definite hallucinatory figure or voice. Starvation, indeed, does produce hallucinations; so that

if my reader should "see a ghost," and wish to ascribe it to his own interior condition, he may at least console himself by supposing that he has eaten too little instead of too much.

between the conscious and the unconscious mind, etc., be henceforth conducted without reference to what the study of hallucination has taught us.

Still more recently, a further discovery, or rather re-discovery of an ancient phenomenon, has shown still further possibilities of instruction. In a paper on "Some recent Experiments in Crystal-gazing," in Part XIV. of the " Proceedings of the So

But the fact is that until a few years ago hardly anything was known as to these casual hallucinations of the sane. The same scanty anecdotes were repeated over and over again; and it hardly occurred to any one that the content of the hallucina-ciety for Psychical Research (Trübner), tory pictures might be a valuable key to mental processes impossible to reach by other means. Two independent researches were then made which have given quite a new aspect to the study. In the first place the French hypnotists (Liébeault, Richet, Bernheim, etc.) showed again, as the older mesmerists had shown long ago, that it was possible to create in certain healthy subjects vivid and prolonged hallucinations by suggestion in the hypnotic state, - such suggestion taking effect either immediately, or at any subsequent date which the operator may choose to assign. That is, the hypnotizer can either say to his subject, "See, there is B.! Go and shake hands with him; or he can say, "At noon next Tuesday, B. will enter your room, and you will shake hands with him," and in each case the subject will see B. at the time and in the attitude thus previously fixed for him. In this way hallucinations can be manufactured in any quantity; and we can analyze the elements of which they are composed, noting how much of the detail is due to the hypnotizer's suggestion, and how much to the subject's own mind.

An important step had thus been made in the study of the mechanism of experimental hallucination. There still remained the need of some wider knowledge as to what hallucinations spontaneously occur. It is to the late Mr. Edmund Gurney that we owe the first systematic attempt to supply this information on a large scale. He set on foot the first census of hallucination in 1885, and succeeded, after much trouble, in getting five thousand seven hundred and five persons, selected at random, to answer questions somewhat resembling those which I shall presently describe. With the resulting information to go upon, the study of the hallucinations of the sane has left the anecdotical and entered on the scientific stage. A multitude of psychological questions are opened up; nor can any discussion on the nature of memory, the association of ideas, the scheme of images by which thought is carried on, the relation

we find the rational interpretation of many a discredited story-from the Dark Ages or the East. Crystal-gazing, in fact, is simply an empirical method of inducing artificial hallucination. If a person gifted with the right kind of visual memory - or whatever the faculty be looks intently into some clear object, undisturbed by reflections, he will gradually see scenes or figures shaping themselves therein. These figures are plainly analogous to figures seen in dreams; they seem generally to proceed from some unconscious stratum of the gazer's own mind; they rarely depict anything which he might not conceivably have dreamt. But at any rate there the figures are; they are the hallucinations experimentally produced; the gazer can watch their behavior -sometimes even through a magnifying-glass-and become, as it were, the conscious spectator of the automatic working of his own mind. Little is as yet known as to the conditions which tend to produce these figures; but there seems thus far to be no evidence that they are morbid phenomena, but rather to the contrary, that they come in times of healthy tranquillity, and are put a stop to by illness or fatigue.

These self-induced hallucinations, however, lie outside of our present subject. I mention them here in order to illustrate the growing change in our attitude towards hallucinations. We are ceasing to look on them exclusively as signs of injury or disturbance; we are beginning to regard them as messages transmitted upwards from the unconscious to the conscious self.

Enough, perhaps, has been said to show that there may be a good deal of knowledge to be gained from the study of these singular bye-products of the human mind. Let us see in what way the census attempts to gather it.

Professor William James, Cambridge, Mass., will send to any one willing to aid, a Paper A., affording space for twenty-five answers, yes or no, to the following ques. tion: "Have you ever, when believing yourself to be completely awake, had a

vivid impression of seeing or being much exaggerated, and that emotional haltouched by a living being or inanimate lucinations (so to term them) form a small object, or of hearing a voice; which im- proportion of the total numbers. And pression, so far as you could discover, here we approach the most curious point was not due to any external physical in the whole inquiry; the evidence, cause ?" namely, that the percipient's hallucination is often due not to his own state, but to the state of some other person. The next question on Paper B. runs as follows: "Was the impression that of some one whom you were in the habit of seeing, and do you know what he or she was doing at the time?" Now in a proportion of cases which, as it stands at present, is far too large for chance to explain, the answer to this question would have to be, "the person whose figure I saw was dying at the time, although I was in no way aware of it."

This question has been carefully framed so as to exclude, as far as possible, both dreams and mere illusions, or misinterpretations of real sights and sounds; and to include all hallucinations, except those of taste or smell, which are rare and difficult to distinguish from mere illusions. It will be observed, moreover, that reports of sounds other than voices are not asked for; the reason being that it is difficult to be sure that such sounds have not some physical, but undetected cause. The first point which we wish to make out is what percentage of sane adults have had any kind of hallucination. It is therefore just as important to collect negative answers as affirmative. The question should be put indifferently to any acquaintance of the collector's; he should not single out those whom he knows to have had some hallucination. Such persons should indeed be asked for their experiences, but a mark should be put to their names in the census-paper to indicate that the collector | knew before he asked them that their answer would be yes. With a little care in this and other points, which I need not here explain in detail, it is possible to get a very fair sample of the experience of the community at large. There were good reasons for thinking that even Mr. Gurney's fifty-seven hundred formed a fair sample; and the number of replies which we now hope to collect should be five or ten times larger.

The

It might have been expected that relatives watching by a deathbed, or anxiously awaiting the news of a death, might experience some imaginary sound or sight. But no ordinary explanation will meet the unquestionable fact that many trustworthy men and women have experienced the sole hallucination of their lives in the shape of the figure of a friend, at the moment when that friend, about whom they felt no anxiety whatever, was actually dying in some distant place. This, as some of my readers may know, is the main_thesis which the testimony collected in "Phantasms of the Living " tends to prove; and during the three years which have passed since the publication of that work the evidence for that thesis, in this and other countries, has become materially stronger. force of evidence of this kind is cumulative; and inasmuch as the detailed cases are tedious reading, and the whole conception of telepathy, or influence exercised at a distance by one mind upon another, is strange and repugnant to many minds, it will be necessary to go on patiently gathering fresh evidence for a long time before we can expect its weight to be generally admitted. But I beg of the reader to observe that in advocating and carrying out After asking for an account of the ac- this present census we are offering to tual experience, Paper B. proceeds to in- those who differ from us the only possible quire whether the percipient- the person method of conclusively disproving our who experienced the hallucination was own view. Suppose that fifty thousand in grief or anxiety at the time. Grief and answers, or more, are collected from Enanxiety are popularly supposed to be gland, France, America, etc., and that strongly predisposing causes of hallucina-among those answers we find few or no tion; and no doubt they are so to some extent. But the result of our collections thus far, — both of Mr. Gurney's census and of many other inquiries made in different ways has been to show that the influence of these moral causes has been

When, however, these answers yes or no have been collected, the greater part of the work still remains to be done. It remains to elicit the real meaning of the affirmative answers; and for this purpose, a Paper B. is submitted to each informant who has answered yes to the question on Paper A.

veridical or truth-telling hallucinations sights or sounds which in some way coincide with some actual event, like a death, occurring at a distance, but a great multitude of falsehood-telling images; figures of friends whom the percipient supposes

to be dying, but who are really in their is the only course worthy of fair-minded ordinary state, and the like then it may men in an age of science.

become plain that we must explain away The next question on Paper B. brings as the effect of chance even the close and us to a point of singular significance. detailed coincidences of which "Phan-"Were there other persons present with tasms of the Living" affords many speci- you at the time? and, if so, did they in mens. If the inquiry is pushed far enough, any way share the experience?" Now it must either refute or confirm our theory hitherto hallucinations, strictly speaking, in a decisive manner. Other points of have been supposed as a matter of course interest there will be on which the census to be confined to the one mind which will probably suggest as many problems creates them. Of course, insane deluas it solves. But on this point of coincisions, of persecution and the like, are fredence, if only the inquiry goes far enough, the mere doctrine of chances must afford a conclusive reply.

quently propagated by suggestion from one insane person to another. But who would think of asking whether a stranger Those of us who believe in these truth- coming into the room while Nicolai was telling or veridical hallucinations have at watching his phantasmal figures would least, therefore, done all that we could to have observed any greyish people passing put our view to the test. We formed that through the apartment? The delusion view on the strength of evidence collected depending on the state of Nicolai's brain in a less systematic mode than the census must obviously be confined to the sufferer offers, but greatly exceeding in amount himself. Well, we have discovered a good all previously existing first-hand evidence many cases in which, contrary to all apas to the hallucinations of the sane. We parent probability, the same phantasmal tested this evidence as well as we could; figure has been observed, or voice heard travelling many hundred miles in order to simultaneously, distinctly, and without obtain personal knowledge of our inform- traceable suggestion - by more than one ants. We then published the evidence in percipient at the same moment. Look at full detail, endeavoring to bring out its this fact how you will, it is one of the weak as well as its strong points. Mr. greatest puzzles which psychology has Gurney then laboriously carried out his ever encountered. We cannot wonder census, in order to ascertain whether there that persons who have had such an expewas such a multitude of merely delusiverience as this should altogether repudiate hallucinations in the world that the coin- the idea of a hallucination · should assert cidences which we had discovered could be explicable by chance. The figures resulting from his census told strongly I might say conclusively — against the explanation by chance. But it was still his wish - which is now being carried out so largely to extend this basis of inquiry, that the result, on one side or the other, might come out with the clearness of a mathematical operation.

The public may, I think, be confident that the census will be fairly conducted. The name of Professor Sidgwick, whom the Congress has set over the task in England, does not need my comments. M. Marillier, who is managing the census in France, is necessarily less known to my readers; so I may say without offence that he was selected simply for his scientific competence, and that he is at present unconvinced of the existence of any veridical hallucinations at all, and inclined to press the explanations of chance and defective testimony to the utmost.

Whatever the truth may ultimately prove to be, surely the patient dispassionate collection of actual contemporary facts

that what they saw must have been in some sense a reality. And in the present state of our knowledge we cannot answer such remonstrances. We cannot bring forward cases where hallucinations which were provably the mere result of morbid states have been communicated without suggestion from one person to another. And, if the word hallucination be objected to, it may be dropped altogether. Its use has been avoided in the census-papers which I am describing, in order to avoid even the appearance of prejudging any question which the inquiry raises.

As an illustration of the kind of difficulty which meets us here, I will give a brief sketch of a case, not of an emotional or exciting kind, communicated to us in dependently by the two percipients, who have never talked of the matter and scarcely met since the month of the incident, and whose accounts coincide with remarkable closeness, considering that one account was written down nineteen years, and the other twenty-three years, after the incident. It is worth noting, by the way, that it is impossible to generalize as to