§ 1. Transient hallucinations of the sane (a department of mental phenomena hitherto but little studied) comprise two classes: (1) hallucinations of purely subjective origin; and (2) hallucinations of telepathic origin - i.e., "phantasms of the living " which have an objective basis in the exceptional condition of the person whom they recall or represent. Comparing the two classes, we should expect to find a large amount of resemblance, and a certain amount of difference, between them .......... 319-320

§ 2. Certain marked resemblances at once present themselves; as that (generally speaking) neither sort of phenomenon is observably connected with any morbid state; and that each sort of phenomenon is rare - occurring to a comparatively small number of persons, and to most of these only once or twice in a lifetime .......... 320-321

§ 3. But in pressing the comparison further, we are met by the fact that the dividing line between the two classes is not clear; and it is important to realise certain grounds of ambiguity, which often prevent us from assigning an experience with certainty to this class or that . ...... 322-323

§ 4. Various groups of hallucinations are passed in review; - "after images "; phantasmal objects which are the result of a special train of thought; phantasms of inanimate objects, and of animals, and non-vocal auditory phantasms; visual representations of fragments of human forms; auditory impressions of meaningless sentences, or of groaning, and the like; and visions of the "swarming" type. Nearly all specimens of these types may safely be referred to the purely subjective class ....... 323-325

It is when we come to visual hallucinations representing complete and natural-looking human forms, and auditory hallucinations of distinct and intelligible words (though here again there is every reason to suppose the majority of the cases to be purely subjective), that the ambiguous cases are principally to be found; the ground of ambiguity being that either (1) the person represented has been in an only slightly unusual state; or .(2) a person in a normal state has been represented in hallucination to more than one percipient at different times; or (3) an abnormal state of the person represented has coincided with the representation loosely, but not exactly; or (4) the percipient has been in a condition of anxiety, awe, or expectancy, which might be regarded as the independent cause of his experience ..... 325-327

§ 5. The evidence that mere anxiety may produce sensory hallucination is sufficient greatly to weaken, as evidence for telepathy, any case where that condition has been present ...... 327-330

§ 6. The same may be said of the form of awe which is connected with the near sense of death; and (except in a few "collective" cases) abnormal experiences which have followed death have been excluded from the telepathic evidence, if the fact of the death was known to the percipient. As to the included cases that have followed death by an appreciable interval, reasons are given for preferring the hypothesis of deferred development to that of post mortem influence -though the latter hypothesis would be quite compatible with the psychical conception of telepathy . . . . . . . . . 330-332

§ 7. There is definite evidence to show that mere expectancy may produce hallucination .......... 332-334

One type which is probably so explicable being the delusive impression of seeing or hearing a person whose arrival is expected .... 334-335

§ 8. There is, however, a group of arrival-cases where the impending arrival was unknown or unsuspected by the percipient; or where the phantasm has included some special detail of appearance which points to a telepathic origin ...........335-336