_M. Binet's theory accounts indeed for a multitude of cases, but certainly not for all._ The prism does not always double the false appearance,[134] nor does the latter always disappear when the eyes are closed. Dr. Hack Tuke[135] gives several examples in sane people of well-exteriorized hallucinations which did not respond to Binet's tests; and Mr. Edmund Gurney[136] gives a number of reasons why intensity in a cortical process may be expected to result from local pathological activity just as much as its peculiar nature does. For Binet, an abnormally or exclusively active part of the cortex gives the _nature_ of what shall appear, whilst a peripheral sense-organ alone can give the _intensity_ sufficient to make it appear projected into real space. But since this intensity is after all but a matter of degree, one does not see why, under rare conditions, the degree in question _might_ not be attained by inner causes exclusively. In that case we should have certain hallucinations centrally initiated alongside of the peripherally initiated hallucinations, which are the only sort that M. Binet's theory allows. _It seems probable on the whole, therefore, that centrally initiated hallucinations can exist._ How often they do exist is another question. The existence of hallucinations which affect more than one sense is an argument for central initiation. For grant that the thing seen may have its starting point in the outer world, the voice which it is heard to utter must be due to an influence from the visual region, i.e. must be of central origin. Sporadic cases of hallucination, visiting people only once in a lifetime (which seem to be by far the most frequent type), are on any theory hard to understand in detail. They are often extraordinarily complete; and the fact that many of them are reported as _veridical_, that is, as coinciding with real events, such as accidents, deaths, etc., of the persons seen, is an additional complication of the phenomenon. The first really scientific study of hallucination in all its possible bearings, on the basis of a large mass of empirical material, was begun by Mr. Edmund Gurney and is continued by other members of the Society for Psychical Research; and the 'Census' is now being applied to several countries under the auspices of the International Congress of Experimental Psychology. It is to be hoped that out of these combined labors something solid will eventually grow. The facts shade off into the phenomena of motor automatism, trance, etc.; and nothing but a wide comparative study can give really instructive results.[137]