"Do I believe in ghosts?" asks Mr Andrew Lang. "One can only answer: 'How do you define a ghost?' I do believe, with all students of human nature, in hallucinations of one, or of several, or even of all the senses. But as to whether such hallucinations among the sane are ever caused by physical influence from the minds of others, alive or dead, not communicated through the ordinary channels of sense, my mind is in a balance of doubt. It is a question of evidence." If the evidence of "hauntings" were measurable by bulk alone, no phase of occultism would be more completely demonstrated. It is only when we come to examine the quality of the available data that we realise how formidable a task it is we have undertaken. In nothing, perhaps, have credulity and superstition been allowed so wide a scope; nowhere is it more difficult to winnow the grain of reliable testimony from the chaff of mythology and invention. For we must remember that the belief in ghosts is as old as the hills themselves. It is common to all countries and to all nations, and in the literature of every language are to be found tales of the supernatural scarcely less plausible than many which assail our ears to-day. What I now set myself to investigate is that class of phenomena seemingly attached to various localities and comprising, besides apparitions, sights and sounds of various kinds and degrees. According to Mr E. T. Bennett, for twenty years assistant secretary of the Psychical Research Society, the records of the Society contain descriptions of "a large number of cases in which the evidence of the reality of phenomena incapable of ordinary explanation is absolutely conclusive." When the sounds are intelligible, or a sentence is spelt out in response to the inquiry of the auditor, the _raison d'être_ of the manifestation is more or less obvious. But there is evidence of a large number of so-called "hauntings" where steps are heard, or noises which convey no intelligible information. Sometimes, also, we are told that simultaneously with the death of a friend bangs have been heard, which, but for the coincidence of their occurrence in association with a death, are without meaning. M. Flammarion cites several cases of this sort. The following will serve as an illustration:--[2] [Footnote 2: It will be found on page 178 of "L'Inconnu et les Problemes Psychiques."]