But there is another view. I have noted four ways in which the machinery may be set in motion; but there is a fifth possible way. The excitation may come downwards from C - from the seats of ideation and memory. And clearly this sort of excitation will have a dominance of its own. It will have its own psychical counterpart - an idea or a memory; and when it sets the sensory machinery in motion, that machinery will not now produce or combine a group of sensations determined by its own activity; but will merely embody, or as we might say execute, the idea or memory imposed on it. Here, then, the only machinery which is in any sense constructive is situated in the higher ideational tracts. But as long as the nervous activity is confined to the ideational tracts, though there is construction, there is no hallucination. That word is never used to describe the mere image or memory of an object; and in serving as a basis to such an image or memory, however vivid, the cells at C are merely performing their normal functions. It is only when the activity escapes downwards, with such force as strongly to stimulate the cells at the lower centre, that sensation floods the image, and we get the delusive percept or hallucination. The force of this downward current may exhibit all degrees.

It is probable that even for the barest idea or memory of an object there is some slight downward escape, causing a slight reverberation at B; and where, as in rare morbid cases,1 the escape is wholly barred, all power of calling up visual images is lost. With every increase in the force of the escape, there will be a rise of sensory quality, and a nearer approach to absolute hallucination; and every stage will thus be accounted for, from the picture "in the mind's eye" to the phantom completely externalised in space. On this view, the provenance of the phenomenon clearly cannot be assigned to a single locality: the hallucination is constructed at one place, but it only becomes hallucination at another. We do not dispute whether a photograph comes into existence in the camera or in the developing-room.

1Krafft-Ebing, Op. cit., p. 11; Despine, Etude Scientifique sur le Somnambulisme, p. 328; Tamburini in the Revue Scientifique, 1881, p. 139; Wundt, Op. cit., vol. ii., p. 350. This point, however, can hardly be said to be established. The mere subjective fact of the reference to the external organ would not prove (as Tamburini seems to assume) that the organ had been actually excited by a refluent current. Nor (for the reason given on p. 292) can a proof be found in the fact that pressure on the side of one eyeball doubles the phantom - a result which must be attributed either to the doubling of some point de repere, or to the association of the sense of lateral pressure with the doubling of real objects. Nor does the fact that "hypnagogic illusions," hallucinations which consist in the surviving of dream-images into waking moments, and hypnotic hallucinations, can give rise to after-images (as noted by H. Meyer, Gruithuisen, and MM. Binet and Fere, respectively), imply more than the brief continuance of excitation at the central cells.

A stronger case, at first sight, is that described by Dr. Pick, of Prague, where only the upper halves of imaginary figures were seen with the right eye; and where it was ascertained that the field of vision was defective over nearly the whole of the upper half of the right retina, to which, of course, the lower half of the figure would have corresponded, but in which the ophthalmoscope revealed no abnormality. But even this is not conclusive as to a refluent current; for Br. Pick may be right in supposing that some lesion of the nerve-fibres connected with this portion of the retina had led to functional inactivity of the corresponding part of the centre itself. I may add, for what it is worth, an observation of Sir J. F. W. Hersehel on his own hallucinations seen in the dark, that "the forms were not modified by slight pressure [of the eye], but their degree of visibility was much and capriciously varied by that cause".

§ 9. Here, then, are the two possibilities: (1) that hallucinations are produced by an independent activity of the specific sensory cells - the sensations which arise there being perceived as objects when the nervous current passes upwards to the higher parts of the brain; (2) that the part played by the specific sensory cells is only a response to what may be called ideational excitation, propagated downwards from the higher tracts where the image has been formed.

In attempting to decide between these possibilities, not much assistance is to be had from direct pathological and physiological observations. These have been mainly directed to an end rather the converse of mine - to utilising the facts of hallucination for fixing the locality of the centres, by inspection of the brains of persons who have been in life markedly hallucinated; and even so not very successfully; for cerebral pathology, as Ball trenchantly remarks, has a way of lending itself to the demonstration of whatever one wants.2 We are thus thrown back on less direct sometimes quite neutral in character.1 So of optical hallucinations. With the sane, a large number consist in the casual vision - an after-image2 as we might say - of a near relative or familiar associate. A friend, who has had considerable experience of the persistence of a dream-image into the first few moments of waking, tells me that "sometimes these images return during the day, or continue appearing, always suddenly, for several days after." The Rev. Robertson Wilson, writing from the United Presbyterian Manse, Strath Devon House, Dollar, tells me that at a time when he was in disordered health owing to overwork, he used to take long excursions, and especially interested himself in the inscriptions in country churchyards.

1In the Archives de Neurologie, vol. vi., p. 352, there is cited "un cas de suppression brusque et isolee de la vision mentale des signes et des objets," where the suppression extended to dreams. "Je reve seulement paroles, tandis que je possedais auparavant dans mes revers la perception visuelle".

2Lesions rarely confine themselves neatly to specific areas. We find Dr. Luys, the chief advocate of the optic thalami as the primary seat of hallucinations, admitting the constant spread of lesions from the thalami to the cortex (Gazette des Hopitaux, Dec, 1880, p. 46); and Dr. W. J. Mickle (Journal of Mental Science, Oct., 1881, p. 382) considers - as the result of a number of very careful necropsies - that in cases of hallucination "thalamic disease plays a less important part than cortical." But, on the other hand, he did not find that the lesions were definitely associate with the spots on the cortex which Ferrier and the advocates of restricted cortical localisation arguments, derived from the nature of the hallucinations themselves. And, I think, the mistake has again been in imagining that one or other of two alternatives must be exclusively adopted - that either the lower or the higher origin of hallucinations is the universal one. All, I think, that can be fairly said is that, while the first mode of origin is a probable one for some cases, the second mode is a certain one for others.