And do we, on the other hand, find, by empirical observation of the phenomena around us, anything which indicates the existence of a supernormal perceptivity such as theory would suggest? It is known to readers of the Society for Psychical Research _Proceedings_ that we do find such indications, scattered at first, and appearing unsought-for amid the phenomena of mesmeric or somnambulic states; but now to some slight extent isolated into distinctness, and brought under experimental control. To some slight extent only, I repeat; for the experiments thus far made, although completely convincing to those who, like myself, have witnessed many of them, under very varied conditions, have nevertheless not yet passed into that desired stage at which one may be able to repeat them before any observer, at any moment. At present they are proved by the same kind of evidence as certain rare pathological phenomena (I do not of course mean that telepathy is itself in any way a morbid product)--phenomena such as those surprising rises and falls of the human temperature which are unpredictable, sporadic, and transitory, and must rest for their evidence on the good faith and accuracy of comparatively few observers. Yet these telepathic experiments have a very hopeful side. Experience has already shown that the phenomena may be developed at any moment, between quite normal persons, and with no bad effects of any sort whatever. Only we cannot tell except by actual trial, and trial of a patient and careful kind, between _which_ persons, out of all mankind, these telepathic messages can be made to run. What we desire, then, what we ask of all who sympathize with our efforts, is neither premature praise nor equally premature theorizing, but active co-operation in our endeavor to improve and extend our experiments in thought-transference. We want to get our telepathic transmissions distant, definite, and reproducible. It is desirable to get them _at long distances_,--not because it is really more marvellous that thought should thus travel a million miles than that it should travel a millimetre,--but for the merely practical reason that at long distances it is easy to avoid two main sources of error, namely, _hyperaesthesia_, which may be quite unconscious, and _fraudulent codes_, which may be hard to detect. Most, nay, probably all, of the so-called experiments in thought-transference which have been offered by "thought-readers," etc., from the public platform, have really had nothing at all to do with thought-transference, have depended either on abnormal delicacy of tactile and other sensory perception, or on the adroit use of preconcerted signals. It is only when the observer has complete control of the conditions (which he never has in any public exhibition), that it is worth while to conduct experiments between two persons in the same room.