In two other cases (A) ladies, in their usual health, saw their husbands in their rooms, when, in fact, they were in the drawing- room or study. Here then are eight cases of non-coincidental hallucination, some of people awake, some of people probably on the verge of sleep, which are wholly without 'coincidence,' wholly unveridical. None of the 'percipients' was addicted to seeing 'visions about.' {199} On the other side, though the writer knows several people who have 'seen ghosts' in haunted houses, and other odd phenomena, he knows nobody, at first hand, who has seen a 'veridical hallucination,' or rather, knows only one, a very young one indeed. Thus, between these personally collected statistics of spectral 'sells' on one part, and the world-wide diffusion of belief in 'coincidental' hallucination on the other, the human mind is left in a balance which mathematics, and the Calculus of Probabilities (especially if one does not understand it) fail to affect. Meanwhile, we still do not know what causes these solitary hallucinations of the sane. They can hardly come from diseased organs of sense, for these would not confine themselves to a single mistaken message of great vivacity. And why should either the 'sensory centre' or the 'central terminus' just once in a lifetime develop this uncanny activity, and represent to us a person to whom we may be wholly indifferent? The explanation is less difficult when the person represented is a husband or child, but even then, why does the activity occur once, and only once, and _not_ in a moment of anxiety? The coincidental hallucinations are laid to the door of 'telepathy,' to 'a telepathic impact from the mind of an absent agent,' who is dying, or in some other state of rare or exciting experience, perhaps being married, as in Col. Meadows Taylor's case. This is a theory as old as Lavaterus, and was proclaimed by Mayo in the middle of the century; while, substituting 'angels' for human agents, Frazer of Tiree used it, in 1700, to explain second sight. Nay, it is the Norse theory of a 'sending' by a sorcerer, as we read in the Icelandic sagas. But, admitting that telepathy may be a cause of hallucinations, we often find the effect where the cause is not alleged to exist. Nobody, perhaps, will explain our nine empty hallucinations by 'telepathy,' yet, from the supposed effects of telepathy they were indistinguishable. Are all such cases of casual hallucination in the sane to be explained by telepathy, by an impact of force from a distant brain on the central terminus of our own brains? At all events, a casual hallucination of the presence of an absent friend need obviously cause us very little anxiety. We need not adopt the hypothesis of the Maoris.