These sounds, in their rudimentary attempt at showing intelligence, are about on a level with the exploits of the "Poltergeist," where coals are thrown about, water spilt, and so forth. Poltergeist phenomena, however, seldom coincide with the ordinary phenomena of a haunt. We have one remarkable case (_Journal_ S.P.R., vol. ix. p. 280-84 [868 B]) where Poltergeist phenomena coincide with a death, and a few cases where they are supposed to follow on a death; but, as a rule, where figures appear there are no movements; and where there are movements no apparition is seen. If alleged Poltergeist phenomena are always fraudulent, there would be nothing to be surprised at here. If, as I suspect, they are sometimes genuine, their dissociation from visual hallucinations may sometimes afford us a hint of value. But after Poltergeists have been set aside,--after a severe line has been drawn excluding all those cases (in themselves singular enough) where the main phenomena observed consist of non-articulate sounds,--there remains a great mass of evidence to haunting,--that is, broadly speaking, to the fact that there are many houses in which more than one person has independently seen phantasmal figures, which usually, though not always, bear at least some resemblance to each other.[157] The facts thus baldly stated are beyond dispute. Their true interpretation is a very difficult matter. Mrs. Sidgwick gives four hypotheses, which I must quote at length as the first serious attempt ever made (so far as I know) to collect and face the difficulties of this problem, so often, but so loosely, discussed through all historical times. (From _Proceedings_ S.P.R., vol. iii. pp. 146-8.) "I will, therefore, proceed briefly to state and discuss the only four theories that have occurred to me. "The two which I will take first in order assume that the apparitions are due to the agency or presence of the spirits of deceased men.