In filling up the lacuna in Mr. Tylor's anthropological work, in asking questions as to the proportion between phantasms of the living which coincide with a crisis in the experience of the person seen, and those which do not, it is obviously necessary to reject all evidence of people who were ill, or anxious, or overworked, or in poignant grief at the time of the hallucination. It will be seen later that neither grief nor amatory passion (dominating the association of our ideas as they do) beget many phantasms. Our business, however, is with the false perceptions of persons trustworthy, as far as we know, sane, healthy, not usually visionary, and in an unperturbed state of mind. There remains a normal cause of subjective hallucinations, expectancy. This appears to be a real cause of hallucination or, at least, of illusion. Waiting for the sound of a carriage you may hear it often before it comes, you taking other sounds for that which you desire. Again, in an inquiry embracing 17,000 people, the S.P.R. collected thirteen cases of an hallucinatory appearance of one person to another who was _expecting_ his arrival. Once more, it is very conceivable that a trifle, the accidental opening of a door, a noise of a familiar kind in an unfamiliar place, may touch the brain into originating an hallucination of a person passing through the door, or of the place where the sound now heard used once to be familiar. Expectancy, again, and nervousness, might doubtless cause an hallucination to a person who felt uncomfortable in a house with a name to be 'haunted,' though, as we have seen, the effect is far less common than the cause. All these sorts of causes are undoubtedly more apt to be prevalent among superstitious savages than among educated Europeans. And it stands to reason that savages, where one man 'thinks he sees something,' will be readier than we are to think they 'see something' too. Yet collective hallucinations, which are shared by several persons at once, are especially puzzling. Even if they occur when all are in a strained condition of expectancy, it is odd that all see them in _the same way_.[12] Examples will occur later. When there is no excitement, the mystery is increased. We may note that, among the expectant multitudes who looked on while Bernadette was viewing the Blessed Virgin at Lourdes, not one person, however superstitious or hysterical, pretended to share the vision. Again, only one person, and he on doubtful evidence, is asserted to have shared, once, the visions of Jeanne d'Arc. In both cases all the conditions said to produce collective hallucination were present in the highest degree. Yet no collective hallucination occurred.