had not left the dining-room.  Herr Parish decides that the same point de repere (the apparent noise of a key in the lock of the front door) ‘acted by way of suggestion on both sisters,’ producing, however, different hallucinations, ’in virtue of the difference of the connected associations.’  One girl associated the sound with her honoured sire, the other with his faithful hound; so one saw a dog, and the other saw an elderly gentleman.  Now, first, if so, this should always be occurring, for we all have different associations of ideas.  Thus, we are in a haunted house; there is a noise of a rattling window; I associate it with a burglar, Brown with a milkman, Miss Jones with a lady in green, Miss Smith with a knight in armour.  That collection of phantasms should then be simultaneously on view, like the dog and old gentleman; all our reports should vary.  But this does not occur.  Most unluckily for Herr Parish, he illustrates his theory by telling a story which happens not to be correctly reported.  At first I thought that a fallacy of memory, or an optical delusion, had betrayed him again, as in his legend of the waistcoat.  But I am now inclined to believe that what really occurred was this:  Herr Parish brought out his book in German, before the Report of the Census of Hallucinations was published.  In his German edition he probably quoted a story which precisely suited his theory of the origin of collective hallucinations.  This anecdote he had found in Prof.  Sidgwick’s Presidential Address of July 1890.[13] As stated by Prof.  Sidgwick, the case just fitted Herr Parish, who refers to it on p. 190, and again on p. 314.  He gives no reference, but his version reads like a traditional variant of Prof.  Sidgwick’s.  Now Prof.  Sidgwick’s version was erroneous, as is proved by the elaborate account of the case in the Report of the Census, which Herr Parish had before him, but neglected when he prepared his English edition.  The story was wrong, alas! in the very point where, for Herr Parish’s purpose, it ought to have been right.  The hallucination is believed not to have been collective, yet Herr Parish uses it to explain collective hallucinations.  Doubtless he overlooked the accurate version in the Report.[14]

The facts, as there reported, were not what he narrates, but as follows: 

Miss C.E. was in the breakfast-room, about 6:30 P.M., in January 1883, and supposed her father to be taking a walk with his dog.  She heard noises, which may have had any other cause, but which she took to be the sounds of a key in the door lock, a stick tapping the tiles of the hall, and the patter of the dog’s feet on the tiles.  She then saw the dog pass the door.  Miss C.E. next entered the hall, where she found nobody; but in the pantry she met her sisters—­Miss E., Miss H.G.E.—­and a working-woman.  Miss E. and the working-woman had been in the hall, and there had heard the sound, which they, like Miss C.E., took for that of a key in the lock.  They