Banishment was very easy; it only needed a relaxed tension." And I may add that in one of the cases equally the origin of a large number of the undoubted psycho-sensorial hallucinations, I cannot recognise this exception; and to me the class in question is of interest, not as distinguished from the psycho-sensorial family, but as a true species of that genus, presenting the sensorial element reduced to its very lowest terms. I regard it as the first stage of a graduated series - the embryonic instance of the investiture of an image or representation with a sensory or presentative character. In proportion as the sensorial element in hallucination is attenuated and dim, or full and distinct, will the perception appear internal or external; and these cases are simply the most internal sort, between which and the most external sort there exists various degrees of partial externalisation. This view has surely everything to recommend it. We can but take the percenter's own account - that he has a distinct impression of words; and that this impression has an actuality which clearly separates it from the mere image or memory of words.

How can this separation be conceived, except by recognising the presence of a genuine, though faint, sensorial element?1 The question is of importance to my argument; for to admit a genuine sensory element in the most "internal" species of hallucination - which all agree to be centrally initiated - will practically be to admit a similar initiation for other psycho-sensorial hallucinations.

1Demonomanie des Sorciers, Paris, 1582, pp. 11-12. For a wonder Bodin gives the case at first-hand; and there is no reason to doubt its truth.

2It is odd to find involuntariness not infrequently taken as the distinctive abnormality in hallucinations (Falret, Op. cit., p. 281; Buchez and Be Castelnau in the French debates of 1855-6); and the odder, inasmuch as not only may hallucinations be voluntary, but the mental pictures and memories, from which they are to be distinguished, are often of course involuntary.

3The Duality of the Mind, p. 124.

4See also Cardan, De Varietate Rerum (Basle, 1557) p. 314, who says he has always been shy of mentioning the peculiarity - "Cum volo, video quae volo, oculis, non vi mentis." Kahlbaum (Op. cit., p. 33), and Maisonneuve (Recherches et Observations sur l'Epilepsie, Paris, 1803, p. 295), give each an instance; and Sir H. Holland (Chapters on Mental Physiology, p. 47) says that he has met with several cases. (See also a case mentioned by Dr. Abercrombie's (Inquiries Concerning the Intellectual Powers, 1838, p. 363). One of the seers of "Faces in the Dark" reported that he could produce the vision of the spangles and sheep at will. His case differs, however, of persistent dream-images mentioned in the last chapter, my informant, Lieutenant-Colonel Hartley, of Hartley, near Dartford, remarks, "I can always produce this phenomenon, if I know that I am dreaming, by opening my eyes, which wakes me, but the dream-image persists." I confess that I should have been tempted to regard the voluntary cases as conclusive evidence of central initiation, had I not found so high an authority as Professor Ball explicitly claiming them as hallucinations provoked by an "abnormal sensation." 1 He does not tell us what the abnormal sensation is, or what causes it.

He contents himself with pointing out that hallucinations are very like dreams; that some dreams arc (and, therefore, apparently, all dreams must be) provoked by external stimulation - say a knock at the door; and that we can sometimes direct the course of a dream at will: ergo, it is easy to see how some people may start a hallucination at will. It would be more to the purpose if he would introduce us to a dreamer who can designedly start a prearranged dream by knocking at his own door.

§ 6. There is only one other point, in relation to the question of origin, that needs special attention; and that concerns hallucinations of what may be called the lowest or most rudimentary grade. There is a class of experience which all the writers who describe it2 agree to treat as a quite unique type, and of which frequent examples have been observed among religious mystics, and persons who believe themselves to be in direct communication with spiritual guides. They describe a voice which is yet soundless, which utters the "language of the soul " inside them, and which they hear by means of a "sixth sense," and without any apparent participation of the ear. "I should hear the voice just as well if I were stone deaf," such a person will say; "my ear has nothing to do with it."3 Owing to the absence of a definable sensory quality, Baillarger distinguished this class as psychic4 hallucinations, in opposition to psycho-sensorial; and M. Binet himself, who insists on an external or peripheral basis for all other hallucinations, is inclined to treat these as exceptional, and to grant them an origin from within. As one who holds that that is from those given in the text.

For, in the first place, his vision was one of old standing; and, in the second place, his retina must have been pretty constantly in the abnormal state. I should thus ascribe the phenomenon to a concentration of attention on actual visual sensations, which fell by habit into the familiar lines. It would be interesting to know whether, after the spangles had appeared, it was possible to check their development into sheep.

1Op. cit.I p. 122. This error (as it seems to me) depends on what has become a very common misreading of the term "psycho-sensorial." The theory so designated is often described as "theorie mixte" - the "mixture" being of the " imagination " and of some prior " abnormal sensation," which sets the imagination to work. But hallucinations (as we have seen) are psycho-sensorial in virtue not of their antecedents but of their content - because they are things actually seen and actually heard. The "sensorial" clement in them is not the incentive or the raw material of the abnormal activity, but its product.

2I find that I must except Dr. Dagonet (Op. cit., p. 02); who, however, is certainly wrong in regarding the phonomena in question as necessarily indicative of insanity.

3Koppe, Op. cit., p. 34.

4The term must, of course, not be confounded with our special sense of the word "psychical," explained on p. 4, note.