Under the same head may fairly be reckoned such cases as Mr. A. Wedgwood's (p. 316, note), where the phantasmal object, though not reproducing anything that had recently been before the eyes, was the immediate and indisputable result of a very special train of thought.

Comparable, again, are representations of an object or a sound which depend on the fact, not that it has been seen or heard, but that it is about to be seen or heard - that the percipient is expecting it; but as in almost all the visual experiences of this sort the object has been a living human form, the cases will fall more conveniently into a later group.

In the next group the object seen, or the sound heard, is non-human in character, but is no longer traceable to any special previous occupation of the mind or the senses. Such cases are common in insanity and in disease, and the hallucination is then often of a grotesque or horrible sort. They also occur, though forming a decided minority, among the waking hallucinations of sane and healthy persons; but seem then to be rarely grotesque or horrible.* The most grotesque case that I have received is a vision of dwarfish gnomes dancing on the wall; but this was seen by a young child. A star, a firework bursting into stars, a firefly, a crown, landscape-vignettes, a statue, the end of a draped coffin coming in through the door, a bright oval surrounding the words, "Wednesday, October 15, Death" - these are the principal phantasms of inanimate objects in my collection. Another known type, described by Sir J. F. W. Herschel, is a geometrical pattern, which sometimes takes very complicated forms. I have also three cases where the hallucination was of a dog, and another where it was of running cats, indefinite in colour and form (this last, however, occurring only when the brain was exhausted). Out of my 302 visual cases (p. 253), only 20 belong to this non-human type.

The non-vocal auditory cases are also comparatively few in number - 41 out of the total of 187. They comprise tappings, tickings, knocks, and crashes;1 the sound of footsteps or of a door opening; 7 cases of the ringing of bells, 2 of the striking of clocks, and 7 of music. Such types seem, on the very face of them, to be altogether remote from telepathy; and though we shall find further on that this is not quite universally the case - that there are instances of strong and unique hallucinations of light or noise which have too markedly coincided with some external crisis for the hypothesis of telepathic origin to be ignored - the vast majority of these non-human phantasms may be safely pronounced purely subjective affections.

1 Kraft-Ebing, Die Sinnesdelirien, p. 16. An exception should be made for certain endemic hallucinations (see p. 478, note).

The same may be said of another smaller group of visual hallucinations which represent fragments of human forms. Thus, two of my informants on waking from sleep, and a third when awake and up, have distinctly seen an imaginary hand and arm;2 another sometimes sees a little finger in the air; another when recovering from illness, had a vision of decapitated heads; another has suffered at times from an appearance of eyes and part of a face floating by; and with these may be classed, as auditory parallels, cases where what has been heard has been a sound of groaning, or indistinguishable sounds of talking, or short meaningless sentences - a class of which I have some half-dozen specimens.

We must regard as a separate type the cases where faces or forms appear either in rapid succession or in a multitudinous way. Several varieties of this experience were described in the last chapter (pp. 299 - 300). I have one other example, where crowds of people and animals made their appearance every night for months; but the percipient in this case was in weak health. We have seen that this type is probably connected with the Licht-staub of the retina, and its subjectivity will not be questioned. It scarcely belongs to the family of transient hallucinations at all, since - alone among the waking hallucinations of persons in apparently normal health - these swarming and changing visions are liable to last for a considerable time.

And finally we come to the visual cases representing complete and (more often than not), quite natural-looking human forms, usually alone but occasionally in company, and occasionally also with the addition of some independent object, such as a carriage or a coffin; and to the parallel auditory cases where distinct and intelligible words are heard, which are not (as in the first group) mere echoes of vividly-impressed phrases. These phenomena - which comprise the great majority of the whole number of transient hallucinations of the sane - fall at once into two great classes; that where the figure or voice is recognised, and that where it is unrecognised. Of both these classes, as of the previous ones, it may at once be said that the majority of the instances included in them are in no way available to prove influence from another mind. Where the figure or voice has been recognised, the absent person whose presence was suggested has generally been in a quite normal state at the time; and where the figure or voice has been unrecognised, no Crisis affecting any person nearly connected with the percipient lias coincided with the hallucination.