It is probable that the young child of Israel, if he had been attempting any mischief, did not succeed in it. Had there been any trouble, it is likely enough that he would have involved Allen in the grief. Then Allen would have been in a, perhaps, unprecedented position. He could have established an alibi, as far as the Jew's affairs went, by proving that he had been at Blocksby's at the hour when the boy would truthfully have sworn that he had let him into Isaacs' chambers. And, as far as the charge against him at Blocksby's went, the evidence of the young Jew would have gone to prove that he was at Isaacs', where he had no business to be, when we saw him at Blocksby's. But, unhappily, each alibi would have been almost equally compromising. The difficulty never arose, but the reason why Allen refused to give any account of what he had been doing, and where he had been, at four o'clock on that Saturday afternoon--a refusal that told so heavily against him--is now sufficiently clear. His statement would, we may believe, never have been corroborated by the youthful Hebrew, who certainly had his own excellent reasons for silence, and who probably had carefully established an _alibi_ of his own elsewhere. The true account of Allen's appearance, or apparition, at Blocksby's, when I and Tarras, Wentworth and the attendant recognised him, and Miss Breton did _not_, is thus part of the History of the Unexplained. Allen might have appealed to precedents in the annals of the Psychical Society, where they exist in scores, and are technically styled "collective hallucinations." But neither a jury, nor a judge, perhaps, would accept the testimony of experts in Psychical Research if offered in a criminal trial, nor acquit a wraith. Possibly this scepticism has never yet injured the cause of an innocent man. Yet I know, in my own personal experience, and have heard from others, from men of age, sagacity, and acquaintance with the greatest affairs, instances in which people have been distinctly seen by sane, healthy, and honourable witnesses, in places and circumstances where it was (as we say) "physically impossible" that they should have been, and where they certainly were not themselves aware of having been. That is why human testimony seems to me to establish no more, in certain circumstances, than a highly probable working hypothesis--a hypothesis on which, of course, we are bound to act.