"There is first the popular view, that the apparition is something belonging to the external world--that, like ordinary matter, it occupies and moves through space, and would be in the room whether the percipient were there to see it or not. This hypothesis involves us in many difficulties, of which one serious one--that of accounting for the clothes of the ghost--has often been urged, and never, I think, satisfactorily answered. Nevertheless, I am bound to admit that there is some little evidence tending to suggest this theory. For instance, in the account,[158] of which I have given an abstract, of the weeping lady who has appeared so frequently in a certain house, the following passage occurs:--'They went after it (the figure) together into the drawing-room; it then came out, and went down the aforesaid passage (leading to the kitchen), but was the next minute seen by another Miss [M.] ... come up the outside steps from the kitchen. On this particular day, Captain [M.'s] married daughter happened to be at an upstairs window ... and independently saw the figure continue her course across the lawn and into the orchard.' A considerable amount of clear evidence to the appearance of ghosts to independent observers in successive points in space would certainly afford a strong argument for their having a definite relation to space; but in estimating evidence of this kind it would be necessary to know how far the observer's attention had been drawn to the point in question. If it had been a real woman whom the Miss [M.'s] were observing, we should have inferred, with perfect certainty, from our knowledge that she could not be in two places at once, that she had been successively, in a certain order, in the places where she was seen by the three observers. If they had noted the moments at which they saw her, and comparing notes afterwards, found that according to these notes they had all seen her at the same time, or in some other order to that inferred, we should still feel absolute confidence in our inference, and should conclude that there must be something wrong about the watches or the notes. From association of ideas, it would be perfectly natural to make the same inference in the case of a ghost which looks exactly like a woman. But in the case of the ghost the inference would not be legitimate, because, unless the particular theory of ghosts which we are discussing be true, there is no reason, so far as we know, why it should not appear in two or more places at once. Hence, in the case of the ghost, a well-founded assurance that the appearances were successive would require a careful observation of the times, which, so far as I know, has never been made. On the whole, therefore, I must dismiss the popular theory as not having, in my opinion, even a _primâ facie_ ground for serious consideration.