it from presentiments, which in their turn are rarely true except where the immediate future is concerned.  To sum up, in the present state of our experience, we observe that what the psychometers and clairvoyants foretell us possesses a certain value and some chance of proving correct only in so far as they put into words our own forebodings, forebodings which again may be quite unknown to us and which they discover deep down in our subconsciousness.  They confine themselves—­I speak of the genuine mediums—­to bringing to light and revealing to us our unconscious and personal intuition of an event that is hanging over us.  But, when they venture to predict a general event, such as the result of a war, an epidemic, an earthquake, which does not interest ourselves exclusively or which is too remote to come within the somewhat limited scope of our intuition, they almost invariably deceive themselves and us.

It is very difficult to fathom the nature of this intuition.  Does it relate to events partly or wholly realized, but still in a latent state and perceived before the knowledge of them reaches us through the normal channels of the mind or brain?  Does our ever-watchful instinct of self-preservation notice causes or traces which escape our ever-inattentive and slumbering reason?  Are we to believe in a sort of autosuggestion that induces us to realize things which we have been foretold or of which we have had presentiments?  This is not the place to examine so complex a problem, which brings us into contact with all the mysteries of subconsciousness and the preexistence of the future.

There remains another point to which it is well to draw attention in order to avoid misunderstanding and disappointment.  Experience shows us that the medium perceives the person in question quite clearly, in his present and usual state, but not necessarily in the exact accidental state of the moment.  She will tell you, for instance, that she sees him ailing slightly, lying in a deck-chair in a garden of such and such a kind, surrounded by certain flowers and petting a dog of a certain size and breed.  On enquiring, you will find that all these details are strictly correct, with one exception, that at that precise moment this person, who ordinarily spends his time in the garden, was inside his house or calling on a neighbour.  Mistakes in time therefore are comparatively frequent and simultaneity between action and vision comparatively rare.  In short, the habitual often masks the accidental action.  This, I insist, is a point of which we must not lose sight, lest we ask of psychometry more than it is obviously able to give us.

2