because these impressions are, under all ordinary conditions, immediately followed by others upon the same part of the retina, and the new sensation practically obliterates the old one. But, if the instantaneous impression be not followed by a different one, or if it be sufficiently vivid to be perceived, notwithstanding the presence of others, its continuance may be made evident to observation. Thus, in a dark room, if a bright point, like the heated end of a wire, be carried round in a circle with moderate rapidity, the eye follows its movement, as it presents itself successively in different parts of the circle; the light always appearing at one point only, the rest of the space remaining dark. But if the rapidity of the circular movement be greatly increased, the bright point seems to be drawn out more or less into a curved line; and, when the rate of revolution has attained a very high degree of velocity, it becomes transformed into a continuous circle of light, since the impression made upon the retina, when the end of the wire is at one part of the circle, lasts until it has completed a revolution and again returned to the same point. The succession of sparks thrown off rapidly from a knife-grinder's wheel often produce the effect, even by daylight, of an unbroken stream of fire. A circular saw with large teeth, driven by machinery under a high rate of speed, presents apparently a perfectly smooth edge, the outline of which is formed by the moving points of the teeth; and the revolving spokes of a carriage wheel, in rapid motion, become confused upon the retina with each other and with the intervening spaces, and assume the appearance of a uniform glimmering disk.

The absolute duration of visual impressions upon the retina has been the subject of various researches, but it is found that its length cannot be expressed by any single number which would be correct for all cases. A brilliant light leaves, on the whole, an impression which lasts longer than that from a feeble one; but, on the other hand, its relative intensity to the light of surrounding objects diminishes more rapidly, and consequently, when it is in motion, a higher degree of velocity is required to produce the appearance of a uniformly bright line. The experiments employed to determine the length of time, during which a luminous impression remains upon the eye without appreciable diminution of its intensity, have been usually those with revolving disks, the surface of which is variegated in sectors of black and white. The rate of revolution of the disk being known, as well as the width of the different sectors, when the revolving surface presents to the eye the appearance of an absolutely uniform gray tint, the time during which the black or white impressions remain undiminished in strength is readily ascertained. The result obtained, from experiments conducted in this manner, under moderate illumination, gives the duration of perfect visual impressions as one twenty-fourth of a second, and, for the oscillation of a very luminous point following the vibrations of a tuning fork, one-thirtieth of a second. The persistence and apparent continuity of successive visual images, appearing at the same spot, is illustrated in the optical contrivance known as the Thaumatrope, or magic wheel. It consists of an opaque

disk, with a perforation at one spot near its edge, through which another disk is visible, placed immediately behind the first, and capable of re volving rapidly while the first remains stationary. Upon the second disk is a circle of pictures representing the same figure in different positions; and when, by its revolution, these pictures are made to pass in quick succession across the opening of the disk in front, they present the appearance of a single figure in rapid motion. The interval between the perception by the eye of successive pictures is too short to be observed, and the same object appears to take successively the different positions in which it is represented.

Duration of a Luminous Impulse required for the Perception of Visual Impressions.-This point has been investigated by Rood' by means of the light of an electric spark obtained from an induction coil connected by its terminal wires with the inner and outer surfaces of a Leyden jar. On breaking the primary current a discharge takes place between the electrodes, which is of exceedingly short duration. This duration was measured by Prof. Rood with the aid of a mirror revolving upon its transverse axis, by which the light of the electric spark was thrown upon a plate of glass, where it could be examined by the naked eye, or with a magnifying eye-piece, as in Fig. 208.

The light emanating from the spark S, was received by an achromatio lens L, of nine inches focal length. It then fell upon a plane mirror revolving with a uniform velocity of 340 times per second, and, after reflection by the mirror, was brought to a focus upon a glass plate G, where it could be examined by the telescope eye-piece E, magnifying

E

M

Fig 208.

APPARATUS for measuring the duration of an electric spark-S. Position of the spark. L. Achromatic lens, M. Revolving mirror. G. Glass plate for receiv

ten diameters. From the known rate of revolution of the mirror, and its distance from the glass plate G, the necessary rate of movement of a reflected beam upon the plate was determined. If the spark, used in these experiments, lasted long enough for its reflected image to move over an appreciable distance, this image would appear to the eye to be drawn out in the direction of the movement, owing to the persistence of its visual impression as described above. But with the mirror revolving at this speed no such deformation was perceptible, the spark

ing the image of the spark. E. Telescope image appearing of precisely the

eye-piece.

same form as if the mirror were

stationary; showing that the duration of the light could not be greater than .000002 (500.000) of a second.

The American Journal of Science and Arts. New Haven, September, 1871.

In a continuation of the experiments, there was interposed between the spark and the mirror a blackened glass plate, ruled with parallel transparent lines of a millimetre in width, and separated from each other by the same distance. The image of this plate, when illuminated by the spark, would appear upon the glass G, so long as the mirror were stationary, as a series of equal alternating black and white lines. With the mirror in motion, if the illumination lasted long enough for the image to be shifted a distance equal to the combined width of a black and white line, these lines would become undistinguishable from each other, as in the case of the revolving disk with black and white sectors. Thus the continuance of the visible lines, under a given rate of motion, proved that the duration of the electric spark was less than a certain calculable period. Their disappearance as distinct objects indicated that the limits of this duration had been reached; and that it was long enough to allow of the shifting of two adjacent lines. The result showed that the duration of the shortest measurable spark was but little over .00000004 (25,000 0) of a second.

With a spark of this duration, distinct vision of motionless objects was perfectly possible. The letters on a printed page were plainly to be seen, and even the phenomena of polarization of light distinctly observ. able. It is accordingly sufficient to produce a complete retinal impres sion.

These experiments do not indicate the time required for the necessary nervous action in the perception of light. They only show that a luminous impulse having the above duration is sufficient to cause a distinct sensation. But the time which is requisite for the sensation to be perceived is very much longer. From the results given in a preceding chapter (p. 431) it appears that the transmission of a luminous impression through the optic nerve, would undoubtedly require at least 1000 of a second, and its perception in the brain considerably more. It follows from this that, at the instant when the image of the electric spark is seen, in the experiment of Prof. Rood, it has, in fact, already disappeared; the interval which elapses between its actual occurrence and its perception by the observer being very much greater than the duration of the spark itself.

The facts detailed above explain the cause of a peculiar optical effect, which has often been observed under the use of the electric spark; namely, that bodies in rapid motion, if illuminated by an instantaneous discharge, appear to the observer as if at rest. A disk, painted with black and white sectors, if set in revolution under continuous light, appears of a uniform gray; or, if the sectors be painted of the rainbow colors, their tints are mingled and the disk appears white. But if such a disk, revolving in a dark room, be illuminated by the electric spark, it becomes visible for an instant, with its different sectors as distinct from each other as if they were at rest. A jet of water discharged from an orifice at the bottom of a vessel, though transparent in the immediate neighborhood of the orifice, is turbid lower down; and by instan

taneous illumination the turbid portion is seen to be composed of separate drops, which appear to be motionless. A flash of lightning has a similar effect in exhibiting objects which are in motion as if they were quiescent. The passage of a cannon ball or a rifle bullet by daylight is imperceptible; because, as an opaque object, it does not remain long enough at any one point to efface the persistent impression of the objects visible behind it, and the sight of these objects accordingly does not appear to have suffered any interruption. But if such a missile should happen to be passing in front of the observer in the night time during a thunder storm, at the moment of a flash, it would be visible equally with the other parts of the landscape, and would appear as a motionless object suspended in the air.

The momentary closure of the eyes in winking, for the same reason, does not cause any noticeable interference with sight, and is not even observed by the individual; since the visual impression of external objects appears to be continuous during the short interval occupied by the movement of the lids.

The local sensibility of the retina is diminished by continued visual impressions. This diminution of the retinal sensibility appears to be continuous from the very commencement of a visual impression, so that it may be made perceptible within a few seconds. In the experiment of exhibiting the image of the retinal blood vessels by changing the posi tion of their shadows (page 622) these shadows are visible for an instant with extreme sharpness. But they begin to fade almost at once and after a short interval become imperceptible. They can only be seen for a considerable time, by keeping the light in motion, so that the shadows fall alternately upon different parts of the retina. The portions of the retina which are in full illumination have their sensibility so rapidly diminished, that the shadow, if motionless, is no longer visible by contrast. Those which are in shadow, on the other hand, become comparatively more sensitive by repose; and when the shifting of the light brings them again into illumination, they not only receive more stimulus than the adjacent parts, but are also more impressible to its influence.

If one eye be covered by a dark glass, and the other be used exclusively, for an hour or two, in reading or writing, at the end of that time the difference in retinal sensibility of the two eyes will be very apparent. A single faintly luminous object in a dark room may then be almost imperceptible to the eye which has been in use, while it will appear to the other quite brilliant. If the application of the eye have not been carried beyond the bounds of moderation, this difference is transitory; and by reversing the conditions, that is, covering the eye previously in use, and reading or writing by aid of the other, that which was before the most sensitive to light becomes less so, and that which was previously fatigued recovers its sensibility.

The alternate diminution and recovery of the retinal sensibility, by excitement and repose, is directly connected with the phenomena of negative images. If the eye be steadily fixed for a short time upon a

white spot in the middle of a black ground, and then suddenly directed toward a blank wall of a uniform white or light gray color, a dark spot will appear at its centre, of the same apparent size and figure with the white one previously observed. This is the "negative image" of the retinal impression. That part of the retina which was first impressed by the rays from the white spot becomes less sensitive to light; and another white surface, looked at immediately afterward, appears darker than usual. On the other hand, those parts which were exposed only to the dark ground, that is, to the comparative absence of light, are more sensitive than before; and the surface of the white wall, outside the central spot, appears brighter than usual. It is not necessary that the contrast in hue between the different parts of a retinal image should be as strong as that of black and white, in order to produce this effect. Any decided difference in illumination will be sufficient. It is not even essential to look at a different background, to observe the appearance in question. If a piece of furniture of dark wood be placed against a blank wall of white or gray surface, and looked at steadily for a short time, on shifting the eyes to a different part of the same wall, the figure of the chair or table will appear, with all its details of outline, expressed in a lighter tint than that of the surrounding parts.

The above effect may be also produced in a still more simple manner. Let a black ruler, about one inch wide, be laid upon a sheet of white paper, and looked at steadily for thirty or forty seconds. If the ruler be now removed by a sudden motion, the eye remaining fixed, its image will appear as a bright band upon the paper, fading gradually as the sensibility of the retina becomes equalized in its different parts.

If the figure which is thus examined be a colored one, its negative image, subsequently produced, will present a complementary hue to that of the original object. A strip of red paper placed upon the white sheet, and then suddenly removed, leaves a negative image which is bluish-green; and a green one leaves an image which has a decided tinge of red. This shows that the sensibility of the retina may be increased or diminished separately for the different colored rays of the luminous beam. While looking at a red object, the retina becomes less sensitive to the red rays, but more so for those at the opposite end of the spectrum, and vice versâ; so that, on looking subsequently at a white object, the negative image exhibits a tint corresponding to the rays for which the retina has remained most sensitive. That this is the mechanism of the production of complementary colors in negative images becomes evident on simplifying the experiment. If the black ruler be laid upon a book bound in blue cloth, on taking it away the band which remains in its place is of a more intense blue than the rest. If a red book be used for the same purpose, the negative image of the ruler presents a remarkably pure red color, while the remainder of the surface appears of a dull brown.

The variable sensibility of the retina, according to its exposure, affords an explanation of the well-known fact, that under some condi