The hallucinations so far dealt with belong to classes numerically strong, and the narratives quoted could be paralleled over and over again from our records by other narratives equally well attested. And this fact furnishes in itself a strong presumption of the substantial accuracy of the accounts given. For as there is little in the kind of incident described--the bare occurrence of a hallucination coincidentally with an external event or with another hallucination--to suggest the work of the imagination, there is little warrant for ascribing this consensus of testimony among the narratives to any other cause than a common foundation in fact. The episodes consist, indeed, of such simple elements as to leave small room for embellishment. Moreover, by those who accept the theory of telepathy an additional argument for the authenticity of these narratives may be found in the consideration that in that theory they receive a simple and sufficient explanation. But we meet occasionally with accounts of hallucinatory experiences which do not fall readily under any of the comparatively simple categories already discussed. The mere difficulty of explaining the genesis of hallucinations of such aberrant types would not, in the present stage of our knowledge, be an argument against their authenticity. But it serves to rob them of the support which they might otherwise have received from their affiliation with better known forms of hallucination; whilst the recent first-hand evidence actually available is not sufficient in itself to substantiate them. Whilst, therefore, such cases should be duly recorded and may legitimately be discussed, it seems best to await the receipt of further evidence before a final judgment is passed upon them. But in some instances there is a further reason why the question should at most be held unproven. Some of the features which distinguish these cases from ordinary telepathic hallucinations, whilst occurring rarely in well-attested recent narratives, are to be found more commonly in remote, uncorroborated, and traditional stories. This circumstance is, of course, a strong argument against their genuineness, since it proves that the imagination tends to create such features. But it is not a conclusive argument. The imagination may itself have been inspired in the first instance by fact; it may have copied, not bettered, nature. That the legendary epics of the older world have invented winged dragons is clearly not an argument that can weigh against positive evidence for the existence in a still more remote past of pterodactyls. _Reciprocal Cases._