produce a given effect. On waking, the daylight is at first dazzling, but soon the retina can bear the stimulus. An increase of intensity does not cause an exactly proportional increase of stimulation, for we find the more the light is intensified the less we notice a fresh increment of light until a degree of intensity is arrived at, when no further addition can be detected, and the light becomes blinding. The less the absolute intensity of two lights the better we distinguish any difference that may exist between them.

Duration.—The effect lasts for an appreciable time after the stimulus has been removed, particularly if the light be very intense. This can be observed when a brilliant point is in rapid motion; instead of a point a streak of light is seen. Thus, part

at least of the trail of falling stars is caused by the persistence of the stimulation, and a luminous body rapidly rotated gives the impression of a circle of fire.

When the stimulus is very intense, such as an electric light, or when we look at a bright object like the globe of a lamp steadily for some time, the effect persists, and after the eyes are shut we see a faint image of the object. This is called the positive after image. If the retina be exposed to a bright light until it be fatigued, and then suddenly turning we gaze at a white wall, the bright part of the positive after image is replaced by a dark figure which is termed the negative after image.

A strong stimulus applied to the retina spreads from the part upon which the bright image falls to those in its immediate neighborhood, so that the bright object looks larger. This phenomenon is called irradiation. It helps to explain many of the peculiarities of vision.

EXCITATION OF NERVE IMPULSE.

The question now arises, How do the retina, or rather their outer layers, convert light into a nerve stimulus? It would appear quite out of the question that the 394 to 760 billions of waves of light per second could mechanically excite the nerve terminals as the waves of sound are believed to excite the endings of the auditory nerve. We know that light has a very distinct action

on many chemical combinations, such as reducing salts of silver and gold, etc. We therefore imagine that the light waves may set up, in the outer layer of the retina, certain intermolecular motions or chemical changes, the result of which is that the nerve fibres are stimulated to activity and transmit an impulse to the brain. The light possibly produces a change in the outer layer of the retina which in some respects may be compared to that which occurs on a sensitive photographic plate. In some respects only, because, while the chemical change on the sensitive plate persists so as to give rise to a permanent photograph, in the eye it only lasts for the brief moment during which we can recognize the positive after image. The chemical change in muscle may be compared to the explosion of gunpowder, in giving rise to force, but not in the result produced in the materials. For in muscle the chemical change causing the contraction is rapidly repaired, while in the powder permanent alteration of the substance is produced. In the retina a new sensitive plate is at once produced by the restoration of the normal condition of the molecules, and similarly its explosive qualities are at once restored to the muscle.

The view that the layer of rods and cones undergoes a chemical change on exposure to light which suffices to excite the optic nerve, receives support from the observation that a color of a red or purplish hue exists in the outer part of the rods, and that this color changes when exposed to the light. But this so-called visual purple has not an inseparable connection with vision, since it is absent when the retina is most sensitive, i. e., the fovea centralis, where there are no rods, and further, frogs with blanched eyes seem to see quite well. Certain rays of light have a distinct thermic influence, and hence the possibility exists that the nerve impulse is started in the retina by some delicate thermic stimulus.

Against the chemical and thermic origin of the retinal stimulation may be urged the fact that the rays of the spectrum which are most efficient in exciting chemical and thermic variations (ultra violet and ultra red respectively) do not excite any nerve impulse in the retina.

The pigmented epithelial cells of the retina have been observed

to change their shape slightly, and definitely to alter the position of the pigment granules they contain when exposed to light. When we remember how sensitive to light the protoplasm of many unicellular infusoria is, we cannot be surprised that the protoplasm of the retinal epithelium is affected by it. In the pigment cells of the frog's skin we are familiar with a change in shape and in the arrangement of their pigment granules in response to different light stimuli. We know further that in the nervous centres nerve impulses often originate in protoplasm under the influence of slight changes in temperature or nutrition. It would hardly be too much to assume, then, that the retinal epithelium has some important share in the transformation of light into a nerve stimulus. The arguments pointing to the

Epithelial cells of the retina. a, Seen from the outer surface; b, seen from the side, as in a section of the retina; c, shows some rods projecting into the pigmented protoplasm.

rods and cones as the essential part of the retina apply equally well to the pigmented epithelium, for they are so dove-tailed one into the other that practically they form but one layer. They are not known to be connected with the nerve fibres, but they may still be influenced by the light, and communicate the effect to the contiguous nerve terminals, which appear to be elaborately adapted to the appreciation of subtle forms of stimulation.

COLOR PERCEPTIONS.

If a beam of white sunlight be allowed to pass through an angular piece of glass it is decomposed into a number of colors which may be seen by looking through the prism, or may be thrown on a screen, like that of a camera. These colors, which

trum.

look like a thin slice of a rainbow, are together called the specThe white solar light is thus shown to be a compound of rays of several colors which possess different degrees of refrangibility, and hence are separated on their way through the prism. The violet rays are the most bent, and the red the least, so that these form the two extremes of the visible spectrum. The difference of color depends upon the different lengths of the waves, the vibrations of violet (762 billions per sec.) being much more rapid than those of red (394 billions per sec.). Beyond the visible spectrum at the red end there are other rays which, though they look black to the eye, are capable of transmitting heat. This thermic power is best developed in these ultra-red rays and fades gradually toward the middle of the spectrum. Outside the violet are ultra-violet rays, which, though non-exciting to the retina, are very active in inducing many chemical changes. Only those ether vibrations which have a medium length can stimulate the retina.

If two different colors be mixed before reaching the retina, or be applied to it in very rapid succession one after the other, an impression is produced which differs from both the colors when looked at separately; thus, violet and red give the impression of purple, a color not in the spectrum. If all the colors of the spectrum in the same proportion and with the same brightness fall upon the retina, the result is white light. This we know from the common experience of ordinary white light, which is really a mixture of all the colors of the spectrum, and we can see it with a "color top" painted to imitate the colors of the spectrum. When the top is spinning, the colors meet the eye in such rapid succession that the stimulus of each falls on the retina before that of the others has faded away, and thus many colors are practically applied to the retina at the same time, and the top looks nearly white.

It has been found that certain pairs of colors taken from the spectrum when mixed in a certain proportion produce white. These are complementary to one another. The complementary colors are:

Red and peacock-blue.
Orange and deep blue.

Yellow and indigo.
Greenish-yellow and violet.

If colors which lie nearer to each other in the spectrum than these complementary colors be mixed, the result is some color which is to be found in the spectrum between the two mixed.

The perception of the vast variety of shades of color that we can distinguish can only be explained by means of this color mixing. We may suppose (with Hering) that there are three varieties of material in the retina, each of which gives rise to antagonistic or complementary color sensations according as they undergo increased or decreased molecular activity, these antagonistic states being produced by the complementary colors. Thus, one substance gives the sensation of black or white, another red or green, another yellow or blue, according as they are in exalted

Diagram of the three Primary Sensations. I = red; 2 =

The letters below are the initials of the colors of the spectrum.

The height of the shaded part gives extent to which the several primary sensations are excited by different kinds of light in the spectrum.

or diminished activity. A varying degree of these stimulations. can be easily shown to give many differences of shade.

Or we may assume that there are three primary colors which overlap one another in the spectrum so as to produce all the various tints. These are red, green and violet; the arrangement of which may be diagrammatically explained (Fig. 235).

We must in this case further assume (Young, Helmholtz) that there are in the retina three special sets of nerve terminals, each of which can only be stimulated by red, green, or violet respectively, and the innumerable shades of color we see depend upon