§ 3. I turn now to the second of the two theories above propounded - the theory that one percipient catches the hallucination from another by a process of thought-transference. This is certainly the explanation that would suggest itself in telepathic cases where one of the percipients has previously had no relations, or only slight relations, with the distant agent. But clearly the most conclusive evidence for the theory of infection would be derived from cases involving no distant "agent" at all; cases which in their inception are pathologic, not telepathic - purely subjective delusions on the part of some one present - but which proceed to communicate themselves to some other person or persons. If it can be shown that this self-propagation is an occasional property of hallucinations as such, there will be no difficulty in extending the same explanation to cases where the hallucination is in its inception due to a distant agent. If B's purely subjective hallucination may affect C, it is only what we should a priori expect that B's telepathic hallucination might affect C: such communi-cability would merely be one more of those points of resemblance, which we have already seen to be so numerous, between the purely subjective and telepathic classes.

And as collective hallucinations even of subjective or non-veridical origin (i.e., not due to the critical situation of some distant agent) would constitute in themselves a form of thought-transference, no excuse is needed for examining them here at some length.

What evidence, then, do we find that hallucinations of the senses, as such, may be infectious? It must be allowed at starting that no property of the sort has ever been attributed to them by psychologists of repute:1 the doctrine would be as new to science as every other variety of telepathic affection. This, however, is easily accounted for. We have already seen that psychologists have never made hallucinations, or at any rate transient hallucinations of the sane, the subject of careful collection and tabulation; and it is among the sane rather than the insane2 that we should expect any phenomenon of thought-transference to present itself. It is therefore not surprising that the rare and sporadic evidence for collective hallucinations should have escaped notice. But if, on the one hand, collective hallucinations have not been recognised by science, on the other hand phenomena have sometimes been described by that title which have no sort of claim to it. It is here that the real importance of distinguishing illusions from hallucinations lies; and I cannot well proceed without first making this distinction plain.

Illusion consists either in perceiving a totally wrong object in place of the right one, as when Don Quixote's imagination transformed the windmills into giants; or in investing the right object with wrong attributes, as when the stone lion on Northumberland House was seen to wag its tail.3 Either sort of illusion may easily be collective. The error is not in the actual sensory impression, which is given by the real object and is common to all present, but in the subsequent act of judgment by which the nature of the object is determined; and in this act of judgment one person has every opportunity of being influenced by another. In the attitude of trying to imagine what further attributes will fit in naturally with those which the senses perceive, and will with them compose some known object, the mind is almost at the mercy of external suggestion. We see this constantly exemplified in cases where a group of people are puzzling as to the nature of some barely visible object, or of some imperfectly heard sound: as soon as someone expresses an opinion, someone else is pretty sure to endorse it, and to see or hear the thing in the suggested sense, though on nearer approach this may prove to have been incorrect.

Even in cases where we feel as if we were right beyond the possibility of mistake, it often needs an effort to realise how little is given us, and how much we ourselves supply. A few slight sensory signs will introduce to the mind a whole array of attributes that have been associated with them on other occasions; the whole is then taken to be a single and immediate perception of the object; and since the actual sensory signs may be common to several different groups of attributes - i.e., to several different objects - it may easily happen that they suggest some group which is not the object actually present. For instance, the slight sensory signs which Scott would normally have interpreted as the folds of coats and plaids hanging in a dimly lit hall, were interpreted by him, at a moment when the idea of Byron was running strongly in his head, as the figure of the deceased poet.1 Here the idea which happened to be dominant at the moment was what determined the false judgment; and such a dominant idea may, of course, often operate upon many minds at once; as when, in a conflagration at the Crystal Palace, a sympathetic crowd watched the struggles of an agonised chimpanzee - alias a piece of tattered blind - in the roof; or when a horrified crew recognised in a piece of old wreck, which was floating on the waves, the form and peculiar limping gait of a drowned comrade.2 The case of the proverbial crowd and the stone lion's tail is somewhat different; for there the object was clearly seen, and recognised for what it was.

But we are all of us well exercised in imagining familiar objects as moving in position and changing in contour; and the power of evoking mental pictures is often, I think, strong enough to enable us slightly to modify our visual impressions; while such devices as half-closing our eyes, or shutting them alternately in quick succession, or moving or inclining the head, will increase the illusion. It is not surprising, then, that a strong effort to see a thing in a way in which others are professing to see it, should, for a brief period, introduce illusory elements into what seems to be a clear and complete view of the object.

1 This was written before the appearance of Dr. E. von Hartmann's tract on Spiritism (lately translated by Mr. C. C. Massey), in which he treats the apparitions seen at seances as collective hallucinations; but he regards the influence exercised on the sitters by the medium as to some extent exceptional in kind.

2 As an instance of the insusceptibility of the insane to abnormal influences, it is worth noting that they are peculiarly difficult to hypnotise.

3 I have never discovered on what authority this anecdote rests; but such an illusion is, I believe, quite possible.

1 An interesting ease was given by Mr. W. H. Pollock, in the Christmas number, for 1884, of the Illustrated Sporting and Dramatic News, under the title "The Ghost at the Lyceum." Mr. Pollock has assured me that the description is "an absolute record of fact, without a word of garnish"; and his recollection of the incident, and of the bewilderment that it caused, was quite confirmed by his companion's account, as reported to me independently by a common friend. Seated in a box at the theatre, Mr. Pollock and a friend saw, during several hours (with intermissions when the lights were turned up), the vivid appearance of a decapitated head, with a fine profile and a grey Vandyke beard, resting on the lap of a lady in the stalls. At the time, they rejected the idea that this could have been an optical effect due to the folds of the lady's garments - as they noticed that she moved more than once in the course of the evening, while the face remained the same. Mr. Pollock seems to have been unaware that, as a possible example of collective hallucination, the vision had a very high scientific interest; or ho would scarcely, even for "sporting and dramatic" purposes, have taken refuge in so meaningless a desiguation as "ghost." It may be, however, that the case was after all one not of hallucination, but only of illusion.

2 Dr. Hack Tuke, Influence of the Mind upon the Body, 2nd edition, p. 59; Wundt, op. cit., vol. ii., p. 368.