607 A. The distinction between after-images and memory-images, although sometimes neglected by careless writers, is a marked one; since after-images, properly so called, are a form of entoptic vision, due to the actual condition of the retina at the moment; while memory-images are a form of that "mind's-eye" vision (central, cerebral, internal, subjective) which lacks as yet a recognised scientific name. Nevertheless, a transitional phenomenon is found in illusions hypnagogiques, the vivid pictures which with many persons rise before the "inward eye" at the moment of falling asleep, or even in waking hours. These may closely resemble postponed after-images; or again, they may assume the more generalised character of memory-images; or (and this perhaps is the commonest case) they may show combinations as novel and fantastic as any which deliberate imagination could summon up. A good example of this transition from after-images to memory-images is afforded by the following careful account, which I owe to the kindness of Dr. Th. Flournoy, Professor of Philosophy in the University of Geneva, and now quote from an article of my own in Proceedings S.P.R., vol. viii. p. 453: -

1 This curious case may be compared with that of Pedrono, cited by Dr. Krohn in his useful historical sketch of "Pseudo-Chromesthesia," American Journal of Psychology, vol. v., part i., p. 25: "These colour impressions he describes as sudden and spontaneous. The sounds are translated into colour before he can stop to think whether the voice is high or low." In other experiments it has been found that the colour was seen before the meaning of the word which determined the colour was consciously observed.

August 1892. Although not habitually subject to hypnagogic hallucinations, I have experienced some ten or twelve, of the visual type, and concerned with objects which had engaged my prolonged attention during the day.

In 1875, after my first day of anatomical dissection, I had the hallucination of an aponeurosis, spread out widely. In 1879, after a long afternoon spent at chess, the vision of a chessboard came to me before I went to sleep. Several times in 1878, after a day of microscopical work, the image of a preparation came before me as a hypnagogic hallucination. These three cases struck me forcibly, because at that date I had never heard any mention of such phenomena. Since then I have had a few more experiences, under the following conditions: -

(1) When after a long interval I occupy myself earnestly with some given subject I see that evening a corresponding hypnagogic hallucination; - as of chessboard, geometrical figures, microscopical preparation; - which does not repeat itself if I continue to attend to the same matter on subsequent days. (2) But, on the other hand, I sometimes see in the daytime, two or three days later, a very distinct image of the object in question. This image does not amount to a hallucination; it is not externalised, nor is it as vivid as a perception; - but, as I am a very bad visualiser, this vivid and precise image is broadly distinguished from my habitual images, which are weak, fragmentary, and indistinct. (3) Neither the hypnagogic image nor the diurnal image are exact reproductions of the given object at a given moment. They are typical images of aponeurosis or chessboard - not reproductions of some specific aspect which has strongly impressed me. Yet they are concrete and precise.

I regard these hypnagogic images, and all memory-images, however recent and intense, as radically different from the " after-images " of the eye. These last have a quality suigeneris, a "sensational co-efficient" which makes them seem to me to exist outside me, if I see them with open eyes; or to belong to my eye, as though stuck inside my eyelid, if I see them with eyes closed. All other images seem to me to be seen with the mind's eye. I class after-images with external perceptions; all other images, whether as vivid as my hypnagogic hallucinations or as faint as the ordinary visual furniture of my mind, I place in a quite different category.

In this case two points come clearly out - namely, (1) the resurgence of a decadent image with something like its original brilliancy; (2) the process of generalisation which this image has subliminally undergone. There are other well-known instances of the same kind: Newton and the spectrum; Baillarger and the gauze which he had been using in anatomical preparations; Pouchet and the microscopical objects.1 In all these cases the resurgent images appear to have been of a generalised type, - but generalised, so to say, diagrammatically, not blurred or uncertain, but as one would prepare them for a demonstration.2 For this reason it is certainly better not to class them as "after-images," reserving that term for cases of purely physiological reproductions of external images. And in fact we find that these hallucinations hypnagogiques depart further and further from mere reproductions of objects seen. They assume all kinds of grotesque forms, and show remarkable inventiveness in producing hundreds of faces which the percipient has never seen before. 610 A. The following remarks by Mrs. Verrall (which I quote from Proceedings S.P.R., vol. viii. p. 480) are of interest on the whole question of the externalisation of images, and also as to the transformation into visual terms of a tactile impression subliminally received: -

Since recording my experiments in crystal-vision I have been trying some further experiments of a different kind, with a view to testing my power of visualisation.

I find that I am able to call up a voluntary picture of an object or a scene with extreme distinctness; indeed, in the case of a simple object, the visual image, as far as I can discover, differs from the actual object only in having no solidity - that is, it casts no shadow and appears to be all on the same plane. The form is as well defined and the colour appears as vivid in the visualisation as in the real thing. I have endeavoured to ascertain whether form and colour are equally well reproduced in the visualisation, and as far as my own impression goes there is no difference; I never think of a coloured object without its colour, but I have not been able to produce complementary after-images from gazing at an imaginary colour. I may say that it is only after a prolonged gazing at a real colour that I can see the complementary after-image, and the colour of the after-image is never anything but very faint.

My visualisations usually, as I have said, have no solidity, but I am able at any time to invest them with solidity by imagining that I see the real thing and not a memory or imagination picture. In that case they cast shadows, and I am able to realise distance. I have much more control over them when once they are there than I have over crystal-visions.

1 See W. James's Principles of Psychology, vol. ii. p. 66, etc.; Binet's La Psychologic du Raisonnement, p. 48, etc.

2 It is probable that something of the same kind takes place in hypnotic hallucinations, where, say, a dog seen by suggestion is not necessarily any special remembered dog. On the other hand, if the order is to see some special object, it will be seen more accurately than it is remembered; - as is curiously illustrated by the frequent dissatisfaction of a female subject when told to see her own portrait on a blank card. The idealised memory-picture, which modifies even her perception of her own image in the looking-glass, is ruthlessly displaced by the subliminal fidelity to truth. "J'ai bien des taches de rousseur," said a subject of Binet's, " mais je n'en ai pas tant que 9a".

At one time, some three years ago, I tried a longish series of experiments with cards with a view to seeing whether I could educate my sense of touch sufficiently to distinguish the cards after passing my thumbs once swiftly over the face of each card.1 I had some success, but the reason I record the experiment here is this: At first, while my attention was consciously directed to my fingers, I was aware that I could detect differences in smoothness of surface, which I learnt gradually to interpret; but after a couple of hundred trials, when I grew more expert and more familiar with the experiment, I lost all consciousness of the means which enabled me to guess, and "saw " pictures of the cards which "determined" the particular guess. This experiment seems to show that conclusions arrived at by other means are presented to my mind in the form of visual images, and suggests that sudden visual impressions, spontaneous as well as induced, may in my case be projected visually after they have been produced in some other way.