§ 5. And now to turn to the parallel features of original or fantastic construction in the telepathic class of hallucinations. The question as to the existence and interpretation of these features is of such special importance to my general argument that, at the risk of wearying the reader, I must make its bearings plain.

I must recur for a moment to the breach which has been more than once noted (pp. 92-3, 134, 174-5) as dividing telepathic phantasms from the less concrete forms of telepathy, and especially from the results of experiments in thought-transference. To resume the two marked points of difference: in the experiments, (1) the "subject" never perceives the transferred image as an actual sight or sound - there is never an external hallucination; and (2) the image always represents the precise object which is consciously occupying the agent's mind; whereas in the case of the spontaneous phantasms the percept (1) appears to be external, and (2) represents something which is certainly not consciously occupying the agent's mind - to wit, his own form or voice.1 Here is an undoubted crux. When the agent is painfully concentrating his attention on a card, with the object of getting the impression of it transferred to the percipient, one might have imagined that he was in a stronger and more hopeful condition for producing a telepathic effect than when his thoughts are wandering at random, and are perhaps not occupied with the percipient at all; and yet the effect on the percipient seems often to be of a far more dominant and startling kind in the latter case than in the former.

We may observe further that just the two forms of telepathic impression which, in the ordinary course of experiment, do occasionally reach the pitch of hallucination - the transferences of pains and of tastes - are conspicuous by their absence from the records of spontaneous cases, the latter being totally unrepresented and the former very rare: while, on the other hand, the spontaneous class abounds in specimens where the psychical condition of the agent - whether sleeping, swooning, or dying - has apparently lapsed to the very verge of nothingness. Spontaneous telepathy would thus seem to depend as little on the agent's intensity of feeling (for what can be more intensely felt than acute physical pain?) as on his intensity of concentration; or, if some actual intensity of experience be imagined for him even at moments when it is least apparent (Chapter V (Specimens Of The Various Types Of Spontaneous Telepathy)., § 10), it will at any rate not be imagined as abnormal preoccupation with his own physical attributes.

1 I am here speaking of the common type of spontaneous cases. There are other cases, both visual and auditory, which conform more nearly to the experimental type, in so far as something that is occupying the senses or the mind of the agent is distinctly represented in the phantasmal appearance (see, e.g., cases 157, 158, 220, 221, 267, 2C8). But such representation is decidedly exceptional, and in visual cases applies usually to only a single feature in the appearance; e.g., in case 213 below we might trace the perception of the checked shawl to the agent's conscious thought, but not the perception of her complete figure.

Now it is manifest that these differences may be reduced, in proportion as the extent of the impression actually transferred from the agent to the percipient in the phantasmal cases can be conceived to be small, and the part which the percipient's own creative energy supplies can be conceived to be large. If we are at liberty to assume that even a dim and shadowy idea, when once it obtains a lodgment in the mind, may body itself forth as a sensory phantasm, clearly all that we shall have to suppose transferred from the one mind to the other is a dim and shadowy idea. We shall thus shift, so to speak, the responsibility for the hallucination to the percipient's mind; which we shall conceive as actively generating and projecting it under a peculiar form of impulse, instead of passively receiving a full-fledged percept from the agent's mind, where nothing in the least resembling such a percept had any conscious place.

An unknown quantity - the peculiar form of impulse - has, no doubt, here to be assumed; for unless the transferred idea involves a certain impulsive force which causes the mind to react on it, and to project it as a hallucination, why does it not remain as a mere idea? It might, indeed, be not unreasonably replied that it does remain as a mere idea in many cases, as we saw in the 6th chapter; and that there may be many other cases where it never reaches the stage of even a conscious idea - never forces itself on the attention at all, - and where, therefore, we never hear anything about it (see p. 77). We have already had numerous experimental instances of what may be called "underground telepathy": in Mr. Newnham's cases, for instance (pp. 50-7), no one would ever have known that there had been a transference of ideas, but for the fact - which was not in any way vital to the transference- that Mrs. Newnham had her hand on a planchette at the time. And if we could conceive that a great deal of spontaneous telepathy takes place similarly underground and unnoticed, then we might regard the sensory phantasms as a sort of accidental group - as just the cases which here and there get above ground, owing to some exceptional favouring condition in the percipient.

The question why an idea that has been telepathically transferred should give rise to a hallucination, might thus be answered by saying that it does so only in the proportion in which purely subjective ideas develop purely subjective hallucinations - the specific cause being as unknown in the one case as in the other. All this, however, is highly hypothetical; and we need not shrink from the provisional hypothesis that a telepathically conveyed idea which is to some extent charged with emotion really has a certain peculiar tendency to develop into hallucination. Our very ignorance in some measure justifies this assumption; for unable as for the most part we are to connect transient hallucinations of the sane with any antecedent condition at all, we need the less scruple to admit a novel condition when we find one. And in connection with such an onward passage of the impression, it is very pertinent to recall the particular impulsive quality which we found to attach to some of the experimental transferences - where the impulse took effect not in sensory but in muscular disturbances (see pp. 59-63, and 67-8).1

To return, however, to the actual content of the percepts, and their relation to the ideas from which they spring. If once it be granted that the telepathic phantasm need not be the literal embodiment of any clearly defined idea or image, but may be worked out from a suggestion of a vague kind, we should certainly be prepared for variety and independence in the working out. The mind is no mere collection of separate compartments, into which new ideas will fit and then rest in a passive way; but an organism of interacting parts, where any change or any intruding element may set in motion whole trains of images and associations. We know what small and dim suggestions will sometimes set large tracts of mental machinery to work; we may therefore well credit the vaguer or subconscious order of telepathic impressions with such a power. What more natural, then, than that these further images and associations should be embodied in the sensory projection? We have already noted the process in dreams; we found the telepathic impression operating not to suspend or fetter, but simply to invoke and inform, the spontaneous activities of the dreaming condition.

We have since caught the percipient mind at work, so to speak, in the gradual stages which waking hallucinations often present; and we have noted the originative activity which often goes to the shaping of purely subjective specimens. What is now suggested is that the waking mind may unconsciously react, as in a dream, on the nucleus of a "transferred impression," and, in the act of externalising the percept, may invest it with its own atmosphere and imagery.1 We shall thus have a ready explanation for many degrees of distinctness and indi-vidualisation, and many diversities of character, in what is perceived. Suppose the same kind of real event - say the peaceful death of an aged parent - to occur in twenty cases, and in each of them to produce a real and unique sort of disturbance in some absent person's mind; then, if that disturbance clothed itself in some sensory form - or, as I should say, if it reached the point of causing hallucination - the hallucination might take twenty different forms.

One percipient may hear his parent's voice; another may imagine the touch of his hand upon his head; a third may see him in his wonted dress and aspect; a fourth may see him as he might appear when dying; a fifth may see him in some transfigured aspect; a sixth may see a figure or hear a voice resembling his, but not recognise it, or recognise it only in recollection; and others may invest the disturbing idea with every sort of visible symbolism, derived from their mind's habitual furniture and their wonted trains of thought.

1 I may note further how the very fact of an unrecognised telepathic phantasm bears witness to the suggested specific tendency of telepathic impressions to pass into hallucinations. We have had instances, in former chapters, of vague but strong impressions, not connected by the percipient at the time with any particular person, which afterwards seemed to have been due to the condition of a particular person. Now one would have supposed, if an undistinctive telepathic idea was to assume more definiteness, that it must be by becoming a distinctive idea, an idea of the particular agent. But what we find is that the undistinctive telepathic idea sometimes becomes more definite in a quite different way - namely, by becoming an undistinctive percept, an unrecognised form. It is hurried on into hallucination, so to speak, without having first declared itself as an idea; it assumes the definiteness of visible shape, while yet its content or message remains indefinite.