There is no case that I have watched longer than Miss A.'s;--none where I have more absolute assurance of the scrupulous probity of the principal sensitive herself and of the group who share the experiments;--but none also which leaves me more often baffled as to the unseen source of the information given. There is a knowledge both of the past and of the future, which seems capriciously limited, and is mingled with mistakes, yet on the other hand is of a nature which it is difficult to refer to any individual human mind, incarnate or discarnate. We meet here some of the first indications of a possibility that discarnate spirits communicating with us have occasional access to certain sources of knowledge which even to themselves are inscrutably remote and obscure. The written diagnoses and prognoses given by the so-called "Semirus," often without Miss A.'s even seeing the patient or hearing the nature of his malady, have become more and more remarkable. Miss A. and her friends do not wish these private matters to be printed, and I cannot therefore insist upon the phenomena here. Yet in view of the amount of telæsthesia which Miss A.'s various automatisms reveal, it should first be noted that human organisms seem especially pervious to such _vue à distance_. "Semirus," "Gelalius," etc., are obvious pseudonyms; and neither Semirus' prescriptions nor Gelalius' cosmogony contain enough of indication to enable us to grasp their origin.[191] From the communications of these remote personages I go on to certain messages avowedly coming from persons more recently departed, and into which something more of definite personality seems to enter. One element of this kind is _handwriting_; there are many cases where resemblance of handwriting is one of the evidential points alleged. Now proof of identity from resemblance of handwriting may conceivably be very strong. But in estimating it we must bear two points in mind. The first is that (like the resemblances of so-called "spirit-photographs" to deceased friends) it is often very loosely asserted. One needs, if not an expert's opinion, at least a careful personal scrutiny of the three scripts--the automatist's voluntary and his automatic script, and the deceased person's script--before one can feel sure that the resemblance is in more than some general scrawliness. This refers to the cases where the automatist has provably never seen the deceased person's handwriting. Where he _has_ seen that handwriting, we have to remember (in the second place) that a hypnotised subject can frequently imitate any known handwriting far more closely than in his waking state; and that consequently we are bound to credit the subliminal self with a mimetic faculty which may come out in these messages without any supraliminal guidance whatever on the automatist's part. In _Proceedings_ S.P.R., vol. v. pp. 549-65 [864 A], is an account of a series of experiments by Professor Rossi-Pagnoni at Pesaro, into which