To adopt it here would drive us to describe the diseased Nicolai - when he saw phantoms in the room, but had his mind specially directed to the fact that they were internally caused - as less hallucinated than a healthy person in the unreflective exercise of normal vision. I prefer to keep to the ordinary language which would describe Nicolai's phantoms as the real specific case of hallucination. And I should consider their distinctive characteristic to be something quite apart from the question whether or not they were actually mistaken for real figures - namely, their marked resemblance to real figures, and the consequent necessity for the exercise of memory and reflection to prevent so mistaking them. The definition of a sensory hallucination would thus be a percept which lacks, but which can only by distinct reflection be recognised as lacking, the objective basis which it suggests.1 It may be objected that this definition would include illusions. The objection could be obviated at the cost of a little clumsiness; but it seems sufficient to observe that illusions are merely the sprinkling of fragments of genuine hallucination on a background of true perception. And the definition seems otherwise satisfactory.

For while it clearly separates hallucinations from true preceptions, it equally clearly separates them from the phenomena with which they have been frequently identified - the remembered images or mental pictures which are not perceptions at all. It serves, for instance, to distinguish, on the lines of common sense and common language, between the images of "daydreams" and those of night-dreams. In both cases vivid images arise, to which no objective reality corresponds; and in neither case is any distinct process of reflection applied to the discovery of this fact. But the self-evoked waking-vision is excluded from the class of hallucinations, as above defined, by the point that its lack of objective basis can be and is recognised without any such process of reflection. We have not, like Nicolai, to consider and remember, before we can decide that the friends whose faces we picture are not really in the room. We feel that our mind is active and not merely receptive - that it is the mind's eye and not the bodily sense which is at work, and that the mind can evoke, transfigure, and banish its own creatures; without attending to this fact, we have it as part of our whole conscious state.

Dreams of the sensory sort on the other hand are pure cases of hallucination, forcing themselves on us whether we will or no, and with an impression of objective reality which is uncontradicted by any knowledge, reflective or instinctive, that they are the creatures of our brain.

But, though the definition may be sufficient for mere purposes of classification, it takes us but a very little way towards understanding the real nature of the phenomena. It says nothing of their origin; and, though it distinguishes them from mere normal acts of imagination or memory, it leaves quite undetermined the faculty or faculties actually concerned in them. And when we pass on to these further points, we find ourselves in a most perplexed field, where doctors seem to be as much at variance as philosophers, The debate, most ardently carried on in France, has produced a multitude of views; but not one of the rival theorists seems ever to have convinced any of the others. The contradictions might even seem to lie in the facts themselves; for what single guiding clue shall be found to phenomena of which some occur only in the light, others only in the dark; some are connected with hyperesthesia of the senses, others with blindness and deafness; some are developed, others dispersed, by fixity of gaze; some are promoted by silence and solitude, others by the stir of the streets; some are clearly relevant, others as clearly irrelevant, to the percipient's mental and moral characteristics? Still progress has been made, to this extent at any rate, that it is now comparatively easy to see where the disputed points lie, and to attack them with precision.

1Objective basis is to be taken as a short way of naming the possibility of being shared by all persons with normal senses. But in that case, how far will the definition fairly allow for collective hallucinations? This question will be considered in Chap. xviii. I may mention that, owing to the special difficulties which collective hallucinations present, I have not included them in the statistics given above, p. 263, and below, pp. 325, 330.