attitude in his sitters that entertains (however remotely) the possibility of witnessing something supernatural, and this is sufficient to create an adjustment of the powers of observation less fitted to detect trickery than if the performer did not announce himself as the go-between of the supernatural. This is well illustrated in the reports of Mr. Davey's sitters, for a few friends who were told beforehand that they were to witness a sleight-of-hand performance, or were strongly led to believe it such, made much less of a marvel of the performance than those who had not been thus enlightened. It remains to add that not one of the sitters (and they were persons of decidedly more than average intelligence and ability) detected his modus operandi, and a large number concluded that trickery was utterly insufficient to account for the manifestations.
Mr. Davey's performances, as described by many of his sitters, like the descriptions of the performances of many a medium, are marvelous enough to demand the hypothesis of occult agency: "Writing upon slates locked and carefully guarded by witnesses—writing upon slates held by the witnesses firmly against the under surface of the table—writing upon slates held by the witnesses above the table—answers to questions written secretly in locked slates—correct quotations appearing on guarded slates from books chosen by the witnesses at random, and sometimes mentally, the books not touched by the 'medium', . . . messages in languages unknown to the 'medium,' including a message in German, for which only a mental request had been made, and a letter in Japanese in a double slate locked and sealed by the witness, etc. And yet, though 'autographic' fragments of pencil were 'heard' weaving mysterious messages between and under and over slates, and fragments of chalk were seen moving about under a tumbler placed above the table in full view, none of the sitters witnessed that best phenomenon, Mr. Davey writing."
It must not be supposed that the errors of mal-description and lapse of memory thus committed are at all serious in themselves; on the contrary, they are mostly such as would be entirely pardonable in ordinary matters. Mr. Hodgson places them in four classes. In the first, the observer interpolates a fact which really did not happen, but which he was led to believe had occurred. He records that he examined the slate, when he really did not. Or, for similar causes, he substitutes one statement for another closely like it; he says he examined the slate minutely, when he really only did so hastily. Thirdly, he may transpose the order in which the events happened, making the examination of the slate occur at a later period than when it really took place. Lastly, he may omit certain details which he was carefully led to consider trivial, but which really were most important. Such slight lapses