Gurney held the view that the main distinction of kind between his "alert" and his "deep" stage of hypnosis was to be found in the domain of memory, while memory also afforded the means for distinguishing the hypnotic state as a whole from the normal one. As a general rule (though with numerous exceptions), the events of ordinary life are remembered in the trance, while the trance events are forgotten on waking, but tend to recur to the memory on rehypnotisation. But the most interesting part of his observations consisted in showing alternations of memory in the alert and deep stages of the trance itself;--the ideas impressed in the one sort of state being almost always forgotten in the other, and as invariably again remembered when the former state recurs. (_Proceedings_ S.P.R. vol. ii., pp. 61 _et seq._ [523 A].) On experimenting further, he met with a stage in which there was a distinct third train of memory, independent of the others;--and this, of course, suggests a further doubt as to there being any fixed number of stages in the trance. The later experiments of Mrs. Sidgwick [523 B] on the same subject, in which eight or nine distinct trains of memory were found--each recurring when the corresponding stage of depth of the trance was reached--seem to show conclusively that the number, may vary almost indefinitely. We have already seen that in cases of alternating personalities the number of personalities similarly varies, and the student who now follows or repeats Gurney's experiments, with the increased knowledge of split personalities which recent years have brought, cannot fail to be struck with the analogies between Gurney's artificial light and deep states,--with their separate chains of memory,--and those morbid alternating personalities, with their complex mnemonic cleavages and lacunæ, with which we dealt in Chapter II. The hypnotic stages are in fact secondary or alternating personalities of very shallow type, but for that very reason all the better adapted for teaching us from what kinds of subliminal disaggregation the more serious splits in personality take their rise. And beneath and between these awakenings into limited, partial alertness lies that profound hypnotic trance which one can best describe as a scientific or purposive rearrangement of the elements of sleep;--a rearrangement in which what is helpful is intensified, what is merely hindering or isolating is removed or reduced. A man's ordinary sleep is at once unstable and irresponsive. You can wake him with a pin-prick, but if you talk to him he will not hear or answer you, until you rouse him with the mere noise. That is sleep as the needs of our timorous ancestors determined that it should be. Hypnotic sleep, on the contrary, is at once stable and responsive; strong in its resistance to such stimuli as it chooses to ignore; ready in its accessibility to such appeals as it chooses to answer.