"When the impressions are very vivid" (Dr. Roget is speaking of visual impressions), "another phenomenon often takes place,--namely, their _subsequent recurrence after a certain interval, during which they are not felt, and quite independently of any renewed application of the cause which had originally excited them."_ (I mark by italics the words which more precisely coincide with Julius Faber's explanations.) "If, for example, we look steadfastly at the sun for a second or two, and then immediately close our eyes, the image, or spectrum, of the sun remains for a long time present to the mind, as if the light were still acting on the retina. It then gradually fades and disappears; but if we continue to keep the eyes shut, the same impression will, after a certain time, recur, and again vanish: and this phenomenon will be repeated at intervals, the sensation becoming fainter at each renewal. It is probable that these reappearances of the image, after the light which produced the original impression has been withdrawn, are occasioned by spontaneous affections of the retina itself which are conveyed to the sensorium. In other cases, where the impressions are less strong, the physical changes producing these changes are perhaps confined to the sensorium." It may be said that there is this difference between the spectrum of the sun and such a phantom as that which perplexed Allen Fenwick,--namely, that the sun has been actually beheld before its visionary appearance can be reproduced, and that Allen Fenwick only imagines he has seen the apparition which repeats itself to his fancy. "But there are grounds for the suspicion" (says Dr. Hibbert, "Philosophy of Apparitions," p. 250), "that when ideas of vision are vivified to the height of sensation, a corresponding affection of the optic nerve accompanies the illusion." Muller ("Physiology of the Senses," p. 1392, Baley's translation) states the same opinion still more strongly; and Sir David Brewster, quoted by Dr. Hibbert (p. 251) says: "In examining these mental impressions, I have found that they follow the motions of the eyeball exactly like the spectral impressions of luminous objects, and that they resemble them also in their apparent immobility when the eye is displaced by an external force. If this result (which I state with much diffidence, from having only my own experience in its favour) shall be found generally true by others, it will follow that the objects of mental contemplation may be seen as distinctly as external objects, and will occupy the same local position in the axis of vision, as if they had been formed by the agency of light." Hence the impression of an image once conveyed to the senses, no matter how, whether by actual or illusory vision, is liable to renewal, "independently of any renewed application of the cause which had originally excited it," and the image can be seen in that renewal "as