Again, the frequency of coincident dreams of death offers a favourable opportunity for estimating the probabilities of their occurrence by chance. The problem is simplified in one direction by the consideration that death is at all events a unique event in the history of the agent. If we can ascertain the proportion which "true" bear to "not-true" dreams of death, we can calculate by means of the tables of mortality the probabilities for some other cause than chance. The problem was actually attempted by Mr. Gurney, who found that coincident dreams of death in the collection published in _Phantasms of the Living_ were twenty-four times as numerous as chance would allow.[89] Theoretically, dreams are of considerable interest as throwing light upon the nature of waking impressions; for it should be observed that dreams are of many kinds and of many degrees of vividness. Some in the vagueness and ideality of the impressions resemble closely the waking experiences recorded in the preceding chapter. Others in their extreme clearness and semi-externalisation approach nearly to the level of hallucinations. But whilst few persons above the level of the savage believe that their dream percepts correspond to actual external objects visibly present, there are some who think that the hallucinatory image of a dying friend which they see with their eyes open, and taking a place in the external order of things, must, just because they see it with open eyes, form a part of that external order. And if the percipient himself is not under any such misconception, the journalist who sneers at him for believing in "ghosts" is so, by his own confession. If once it is recognised that between dreams and hallucinations there is no essential difference, the chief obstacle to the acceptance, by two different classes of minds, of telepathy as the explanation of coincident hallucinations will have disappeared. It will become clear, on the one hand, that a belief in the significance of such hallucinations does not necessarily carry with it a belief in "ghosts"; and on the other, that the fact of an apparition taking its place as a fully externalised percept does not imply any substantial basis for the percept. In dealing with dreams we will discuss first those which resemble most closely the experimental results and the cases considered in the last chapter, and proceed from these to dreams which include a definite representation of the agent. Finally, cases of somewhat aberrant type and clairvoyant dreams will be considered. _Simultaneous Dreams._ No. 52.--From MISS INA BIDDER. "RAVENSBURY PARK, MITCHAM, _June_ 10_th_, 1890.