Of much greater importance than errors of observation or inference are those due to defects either in narration or memory. Deliberate deception amongst educated persons is no doubt comparatively rare, though it would perhaps be unwise to hold out any pecuniary inducement for the production of evidence. But there are those, like Colonel Capadose in Mr. Henry James' story _The Liar_, who tell ghost stories for art's sake, and on a slender basis of fact build up a large superstructure of fiction. And there are many more who, with a natural and almost pardonable desire to appear as the hero, or at least the _raconteur_, of a good story, or from the mere love of the marvellous, allow themselves to exaggerate the coincidences, adjust the dates, elaborate the details, or otherwise improve the too bare facts of an actual experience. This kind of embellishment, however, is probably more frequent in second-hand accounts, where the narrator speaks with less sense of responsibility, and, it may be added, of reality. Again, a common form of inaccuracy is to quote as the experience of a friend one of those weird stories which are passed on from mouth to mouth in ordinary society--the inconvertible currency of psychical research. We all know these old friends--at a distance, for no one has ever succeeded in making their nearer acquaintance. There is the ghost at No. 50 B---- Square; the driver of the dream-hearse, recognised a year later in a lift, which fell straightway, with all its passengers, to the bottom of the hotel; the Form which accompanies the priest, or Quaker, or godly merchant to save him from robbery on his lonely nocturnal journeyings; the young lady who took part in some _tableaux vivants_ whilst her body was lying cold in death--and all the rest of the phantom throng. Only a few months ago I heard one of them--it was the ghost of the lift--from the son of a doctor, who assured me that the incident occurred to one of his father's patients, and gave me the name of the foreign hotel which had been the scene of the disaster.[74] Sometimes a story is improved by the narrator that it may the better serve for instruction and edification. This tendency is especially liable to distort the evidence in cases connected with death. It must be remembered that though we may view a coincident hallucination, for instance, as merely an instance of an idea transferred from a _living_ mind, to the percipient it frequently represents the spirit of the dead. From a certain class of witnesses the account of such an incident is as little to be trusted as the text of an apocryphal gospel. It inevitably becomes a _Tendenz-schrift_, which reflects not the facts as they occurred, but the narrator's conception of what the facts ought to have been.