Page:The Naturalisation of the Supernatural.pdf/293

Rh been committed to writing only after hearing the descriptions of others; so that features discerned or believed to be common become more definite in recollection, and discrepancies tend to disappear. In short, the image which remains in the memories of the percipients is apt to resemble a composite photograph, in which all the common features are emphasised, and details found only in individual cases are blurred or faintly indicated.

But even in the accounts forwarded to us, mostly written some years after the events, when there has been ample time for the several experiences to have been talked over and smoothed into uniformity, it frequently happens that we can discern marked discrepancies in the description of the figures. In many cases the figures seen are admittedly different. In the Case, for instance, a fragment of which has been already quoted at the end of the last chapter, (No. 55), the most frequent apparition was a figure, sex uncertain, clothed in black with, according to most witnesses, some white about the head and shoulders. But one inmate of the house saw the figure of a man in his shirt-sleeves; the apparition of a white dog was also seen by several persons. Here then in this one case we have four distinct kinds of apparition.

The impression left upon the mind after a careful survey of the best attested narratives is that the authentic ghost rarely appears in recognisable, perhaps not even in constant shape; that his connection with tragedies is obscure and uncertain. He