We find, however, our clearest examples of the central initiation of hallucinations, when we turn to cases where excitation from the outer world is plainly absent. This class includes phantasms seen in complete darkness,1 and also hallucinations of pain, and probably the large majority of auditory hallucinations, which have so far been disregarded. Here the alternative is simple. The initiation must either occur in the brain, or be due to some morbid or abnormal condition of the outer sensory apparatus. We have already duly noted the latter mode as a frequent one. But the fact that certain hallucinations have been undoubtedly due to injury of the external organ does not establish, or even strongly suggest, the existence of a similar condition in cases where it defies detection.2 As a rule, where the abnormal condition has been made out, hallucinations have not been its only result. The ulceration of the cornea which initiates visual hallucinations has begun by affecting the vision of real objects. Illusions, or false perceptions of colour, usually precede the appearance of more distinct phantasms.3 So, in cases of more transient abnormality - such as the illusions hypnagogiques mentioned in the last chapter -other signs precede the hallucination.

The observer, whose eyes are heavy with sleep, begins by seeing luminous points and streaks, which shift and change in remarkable ways; and it is from these as nuclei that the subsequent pictures develop. Mr. James Britten, of Isleworth, tells me that, as a boy, he often saw in the dark a distant, tiny point of light, which approached and became an eye, then turned into a face, and then, coming nearer, " developed into a mass of very horrible faces," quite unlike any that he had ever seen or imagined. Similarly one of the seers of "Faces in the Dark " (St. James's Gazette, February 10th, 15th, and 20th, 1882) described the frequent vision of a shower of golden spangles, which changed into a flock of sheep. Now, since our physiological knowledge leaves no doubt that the points, streaks and spangles are due to the condition of the retina, it is safe in such cases to conclude that this condition has initiated the hallucination. But it is not equally safe to conclude that the process must be the same for cases where the points, streaks and spangles are absent. I do not forget that even a normal eye is subject to affections which escape attention until a special effort is made to realise them.

1Le Progres Medical, 1878, p. 38. It is probable that an attentive regard is a condition for the establishment of points de repere. In the case of M. Binet's "subjects," a certain peculiarity in the fixed regard, which might act in this way, is strongly suggested by the following fact. In some cases, after a screen had been interposed between the patient's eyes and the imaginary object, she continued to see not only that object (say, a mouse), but a real object (say, a hat) on which it had been placed. Thus the hat assumed the property - shared by the imaginary mouse, but unshared by any other real objects - of remaining as a percept in spite of an opaque barrier. Kahlbaum describes a patient who saw the form of his deceased child only when he fixed his eyes steadily on a point. ("Die Sinnesdelirien " in the Allgemeine Zeitschrift fur Psychiatric, vol. xxiii., p. 7; and see also below, p. 316, second sentence.) But in the general run of spontaneous hallucinations, there is no reason for supposing that the regard has any of this exceptional intentness.

2Revue Medicate, 1825, vol. i., p. 34.

3Traite des Maladies Mentales, p. 357.

4Baillarger, Des Hallucinations, p. 312; Ball, Lecons sur les Maladies Mentales, p. 73.

1 Sir J. F. W. Herschel mentions that some of his own hallucinations could be seen with open eyes only if the darkness was complete. .(Op. cit., p. 407.) Some of Nicolai's visions could only been seen when the eyes were shut; and this was also a feature of a very interesting case recorded by Dr. Pick (in the Vienna Jahrbuch fur Psychiatric for 1880, vol. ii, p. 60), where nevertheless the figures seen gave an impression of complete externality, and were often addressed by the patient. Sehule records a similar instance. (Handbuch der Geisteskrankheiten, p. 128.) In the Gazette Medicale de Paris for March 21, 1885, Professors Bernheim and Charpentier describe some experiments made in a dark room, where - points de repere being necessarily absent - the visual hallucinations of hypnotic "subjects" proved not to be modifiable by optical instruments in the way above described.

2For statements of the opposite view see (as well as M. Binet's papers above referred to) Professor Ball, in L'Encephale, 1882, p. 6, and Maladies Mentales, p. 111, etc.; and the classical paper of Dr. Regis on unilateral hallucinations, L'Encephale, 1881, p. 41.

3Dr. Max Simon in the Lyon Medical, vol. xxxv., p. 439; Voisin, Op. cit., p. 70.