But there remains a fourth way of setting the machinery in motion. The first way, in normal sensation and perception, was the natural action of the organ of sense, stimulated by a material object. The second way was by the stimulus of a vice in the organ of sense. The third way was a vicious activity in a sensory centre. All three stimuli reach the 'central terminus' of the brain, and are there created into perceptions, the first real and normal, the second a hallucination from an organ of sense, _from without_, the third a hallucination from a sensory centre, _from within_. The fourth way is illustrated when the machinery is set a-going from the 'central terminus' itself, 'from the higher parts of the brain, from the seats of ideation and memory'. Now, as long as these parts only produce and retain ideas or memories in the usual way, we think, or we remember, but we have no hallucination. But when the activity starting from the central terminus 'escapes downwards,' in sufficient force, it reaches the 'lower centre' and the organ of sense, and then the idea, or memory, stands visibly before us as a hallucination. This, omitting many technical details, and much that is matter of more dispute than common, is a statement, rough, and as popular as possible, of the ideas expressed in Mr. Gurney's remarkable essay on hallucinations. {186} Here, then, we have a rude working notion of various ways in which hallucinations may be produced. But there are many degrees in being hallucinated, or enphantosme, as the old French has it. If we are interested in the most popular kind of hallucinations, ghosts and wraiths, we first discard like Le Loyer, the evidence of many kinds of witnesses, diversely but undeniably hallucinated. A man whose eyes are so vicious as habitually to give him false information is not accepted as a witness, nor a man whose brain is drugged with alcohol, nor a man whose 'central terminus' is abandoned to religious excitement, to remorse, to grief, to anxiety, to an apprehension of secret enemies, nor even to a habit of being hallucinated, though, like Nicolai, he knows that his visionary friends are unreal. Thus we would not listen credulously to a ghost story out of his own experience from a man whose eyes were untrustworthy, nor from a short-sighted man who had recognised a dead or dying friend on the street, nor from a drunkard. A tale of a vision of a religious character from Pascal, or from a Red Indian boy during his Medicine Fast, or even from a colonel of dragoons who fell at Prestonpans, might be interesting, but would not be evidence for our special purpose. The ghosts beheld by conscience-stricken murderers, by sorrowing widowers, by spiritualists in dark rooms, haunted by humbugs, or those seen by lunatics, or by children, or by timid people in lonely old houses, or by people who, though sane at the time, go mad twenty years later, or by sane people habitually