THE "FLOATERS" ONE SEES WHILE looking at a featureless background are classified as entoptic phenomena, meaning that they arise within the visual system. The floaters are interesting to watch, and they reveal some interesting things about the workings of the visual system.

My experience, which is probably similar to yours, is that when I stare at a featureless background such as a blank wall or a clear sky, my field of view is spotted with small circles and larger hairlike forms. I can never quite bring them into focus. In fact, when I shift my gaze in an effort to see one of them better, the object flits away.

The small circles are easy to ignore. The larger objects are more disturbing. In the field of view of my right eye I see a structure that resembles a stiff rope tied into several large knots with loose ends. This structure is so large that it often interferes with my reading. I can even see it against a complex background. When the background is a page in a book, the object is dark. Against a computer screen, where the background is dark and the symbols are light, the object appears as a fuzzy white region. I can also see it when I close the eye while facing a source of strong illumination such as a bright sky.

One of the best early discussions of floaters is in Treatise on Physiological Optics, by Hermann von Helmholtz. An interpretation of the objects in terms of modern physics was offered in these pages 20 years ago by Harvey E. White and Paul Levatin [see "'Floaters' in the Eye," by Harvey E. White and Paul Levatin; SCIENTIFIC AMERICAN, June, 1962]. Much of what follows builds on the ideas presented in that article.

Although floaters seem to lie on the surface of the cornea, they are actually near the retina. They float in the fluid just in front of the fovea: the small area of acutest vision where the visual receptors are densest. The fovea (or its place in one's field of view) is sometimes called the point of fixation.

Floaters originate with red blood cells that have leaked out of the retina. Once the cells are released they swell into spheres, losing their hemoglobin and thus their red color. The spheres drift through the liquid region in front of the fovea individually or in strings. The individual cells are responsible for the small-circle floaters; the strings give rise to the hairlike structures. The object I see in my right eye seems to be an unusually long string of some blood cells that is wound around itself several times. I could be wrong; several shorter strings tangled together would give the same appearance.

Each swollen cell is some eight micrometers in diameter. A floater, however, is not the direct image of a blood cell and usually is not even the simple shadow of one. Most of the time it is a diffraction pattern projected by the blood cell onto the retina. The pattern consists of bright and dark bands arising from the constructive and destructive interference of the light passing the cell on its way to the photoreceptors of the retina. When the cell is an isolated one, the pattern consists of concentric bright and dark rings. When several cells are linked, the pattern is elongated.

The liquid layer in which the cells float results from partial liquefaction of the vitreous humor, the transparent material

filling most of the eye. Cells relatively distant from the retina cast large patterns that are probably too fuzzy to be distinguished. Cells closer to the retina create smaller, distincter patterns. Many people find that the number of floaters increases with age. Perhaps more blood cells are released by an older retina. Moreover, the liquid layer near the retina often expands with age.

One of the peculiarities of a diffraction pattern from a small object is that the center of the pattern can be bright. In many optical phenomena light can be modeled as either a wave or a ray. With floaters the limits are stricter. If light passed a cell as a collection of rays, the cell would cast a dark shadow on the retina. Since the cell is very small, however, the light behaves not as rays but as waves. They diffract around the edge of the cell, sending light into what should be the shadow region. At the retina the light interferes, creating the bright and dark bands constituting the pattern of a floater. Once past the cell the light can be regarded as being in ray form; at the retina only the wave model serves.

The process is depicted in Figure 4 for two shafts of light passing a cell on opposite sides. The shafts are shown both as rays and as waves that are in phase, that is, in step with one another. The light diffracts around the cell and spreads into the shadow region. The shafts arrive at the retina after traveling equal distances to the center of the shadow region. Since they begin in phase and traverse equal distances, they arrive in phase and therefore interfere constructively, making the center of the shadow region bright.

Slightly off the center the shafts arrive out of phase because they travel different distances to reach the spot. When the path difference amounts to half a wavelength of light, the waves interfere destructively at the retina, leaving darkness at that spot. Symmetry creates a ring of darkness around the center of the pattern.

Slightly farther from the center is a bright band. Here shafts of light arrive from opposite sides of the cell approximately in phase and so interfere constructively. Additional bright and dark bands lie farther from the center. A floater consists primarily of the central bright spot and the first of the bands.

Y. P. Hwu of Wytheville Community College in Virginia achieved a similar diffraction pattern from blood cells by putting diluted blood on a transparent slide, covering it with a cover slide and illuminating the preparation with laser light. The laser beam was expanded by the objective lens of a microscope (of 10-diameter magnifying power) positioned just in front of the slide. Diffraction patterns of the blood cells appeared on a screen placed 50 centimeters from the slide.

The floaters in my field of view seem to drift when I tilt my head. The drift can be deceiving because the perceptive apparatus reinverts an image that is upside down at the retina. If a floater drifts downward past the retina because of gravity, the brain applies the same interpretation, making the floater seem to move upward.

I can make floaters move if I tilt my head or suddenly glance down and then up. The motion of the eye shifts the vitreous humor and the liquid layer near the retina. The resulting flow of liquid carries the floaters past the retina. If the flow is truly upward, the floaters appear to drift downward.

White and Levatin investigated the appearance of floaters when the viewer lay on his back and stared upward. The position makes the floating cells drift closer to the retina, where they settle into the foveal depression. The motion decreases the distance between a cell and the surface on which its diffraction pattern forms. As a result the diffraction pattern shrinks because it depends on the difference in path lengths traveled by the light rays.

For example, the first dark band arises when rays from opposite sides of the cell travel distances that differ by half a wavelength of light. When the cell lies closer to the retina, this condition is met by a spot closer to the center of the pattern. The dark band appears smaller in the field of view.

Because the patterns are smaller the reclining observer can more easily detect the ones from individual cells. They become sharper, and more bands are apparent. Eventually a cell moves so close to the fovea that the pattern loses resolution because several bands fall across adjacent photoreceptors.

Floaters can be examined with the aid of a pinhole held up to the eye. The pinhole not only serves to keep ambient light from entering the eye but also provides conditions in which the diffraction patterns from a cell are more distinct. So far this discussion has been proceeding on the basis that the light passing a cell is entirely in phase, but that is not the case when the light is from many sources. If the light comes through a pinhole, however, the shafts passing a cell are more nearly in phase. The light therefore creates a sharper diffraction pattern.

A pinhole can be made by putting a sheet of aluminum foil over the hole in the end of a spool of thread and punching a small hole in the foil with a pin. More uniform and permanent pinholes can be made by following the instructions I gave in this department last November for the construction of pinhole cameras. You can also buy a pinhole mounted on a small flashlight from the Edmund Scientific Co. (101 East Gloucester Pike, Barrington, N.J. 08007). If you make your own pinhole, you will need a diffuse light source on the opposite side of it. A hand-held slide viewer works well, as does any small flashlight. Do not use a laser. Although laser light is ideally suited for experiments involving optical interference, it can easily damage your retina. The light is so coherent (the waves are almost exactly in phase) that the beam can dissipate enough energy in a small region of your retina to destroy the photoreceptors, even though you may feel no pain. The floaters derive ultimately from the blood vessels of the retina, an intricate network that actually lies in front of t he photoreceptors. One might suppose the vessels would cast shadows on the visual field. They do, but you rarely perceive them. Any constant visual pattern is ignored by the visual system after a few seconds. A classical explanation is that the system becomes fatigued by a 5, constant pattern, which therefore fades from perception. A stationary external object remains perceptible only because the eye constantly moves in small jerks that prevent the formation of a constant image on the retina. Since the circulatory system is fixed to the retina, its shadows remain constant and are therefore imperceptible.

You can see the network under special circumstances. Some morning when you wake up in a sunlit room, look with one

eye at a featureless wall or the ceiling. You will see a faint colorless outline of the vessels on your retina, but only for a few seconds. Then the visual system begins to ignore the pattern and it fades from perception. You might be able -to see it again if you close your eye - briefly and then open it. The network will resemble a tree that sprouts from the "blind spot," the point on the retina where the larger blood vessels come through it from the rear.

The network is often called Purkinje's tree because it was first described by Johannes Purkinje in 1823. Purkinje, a pioneer in physiological optics, was able to outline the circulatory system of his eyes with the aid of a small light source. You can do the same with a pinhole. You might have to experiment with the size of the pinhole, however, to enhance your perception of the network.

Slowly move the pinhole around in front of one eye while keeping the other eye closed. The light goes past the circulatory network in the illuminated eye at constantly shifting angles. The shadow of the network therefore moves across different photoreceptors, with the result that the pattern remains perceptible.

The light source on the other side of the pinhole can be any relatively bright, diffuse source. An illuminated screen of the kind normally employed for viewing photographic slides works well. Even the clear sky will do if you can somehow shield your eye so that it receives light only through the pinhole.

William Oldendorf suggests using a penlight lightly touching a closed eyelid. The light should be positioned on the closed upper eyelid at the outside corner of the eye, just above the junction of the lids. Do not apply pressure to the eye and definitely do not touch the eyeball itself. The light enters through the eyelid and casts a shadow of the circulatory system onto the retina To keep the image from standing still and fading out move the penlight around in a small circle. Vary the frequency of the movement until you get the clearest impression of the system.

Tom M. Cornsweet states in his book Visual Perception that a sharp image of the circulatory system can be seen when the light is provided by optical fibers. Such fibers can be bought from the Edmund Scientific Co., along with a small light source to attach to one end of a bundle of them. The fiber light source works better than a penlight because the fibers are very thin and therefore cast sharper shadows.

Cornsweet also points out that the best view of the retinal blood vessels can be got if the light is predominantly blue (with wavelengths approximating 415 nanometers). Light at those wavelengths, obtained with a blue filter, is absorbed by the hemoglobin in the red blood cells. If the light consists of only those predominantly blue wavelengths, the blood vessels cast darker shadows. Edmund Scientific sells color filters that serve the purpose.

The standard explanation of why the retinal circulatory system is not normally visible does not really explain the cause of the perceptual fading of a constant signal. Recent work by A. E. Drysdale of the University of Reading suggests that several factors may be involved. The general lack of contrast in the shadows cast by the network may be partly responsible. The fading might also be due to a mechanism in the visual cortex that specifically inhibits perception of the network.

When I stare at a featureless background, I occasionally see another entoptic phenomenon associated with the retinal circulatory system. Tiny specks dart across most of my field of view in a seemingly random manner and then disappear. As with the floaters I cannot bring these objects into focus. Indeed, none of them pass through the foveal region in my field of view, which is marked by a slightly grainy texture.

If I hold my gaze as steady as possible, however, by fixing it on some small, distant object, the specks pass repeatedly along the same paths. I do not always see the specks when I look at the sky. I am not sure what it is I do that succeeds in bringing them out, but it probably entails not only fixing my gaze but also relaxing the lenses of my eyes so that they are focused for distance.

The specks are due to the circulation of blood through the retina. If you carefully monitor your pulse as you view them, you will find they appear in correlation with the pulse. The specks are luminous and have relatively dark tails. They always move in single file and none ever overtakes another. Their speed varies; they move faster during the systolic (contraction) phase of the heartbeat and slower during the diastolic (dilation) phase. The specks are thought to be due to the flow of white blood cells through the retinal blood vessels.

You might find the specks more pronounced after physical exercise, since the pulse rate and blood pressure are then increased. The specks might be more visible if you suddenly bend over, thereby changing the blood pressure in your eyes. You can also vary the visibility of the specks by pressing lightly on an eyelid. This action increases the pressure on the retina and therefore increases the pressure in the retinal circulation. A slightly stronger pressing of the eye exerts enough pressure on the retina to make the blood flow decrease, almost eliminating the specks. (If you try this experiment, you may also see the luminous figures called phosphenes that I described last May. They are an entirely different entoptic phenomenon.)

I see the specks best against a clear blue sky. The sky is free of interfering features, and the blue light enhances the speck effect. You can create similar conditions by looking toward a bright featureless background through a blue filter. A filter that passes only light with wavelengths of about 415 nanometers is ideal because this light is absorbed by the hemoglobin in the red blood cells. The light is not absorbed by the white blood cells, so that the motion of the white cells is in sharper contrast. Since the specks arise from the circulation of blood through the retina, they follow the circulation pathways (rather than darting randomly as was my first impression). No specks pass through the foveal region because there are no blood vessels there.

Another entoptic effect can be observed if you view a brightly lighted sheet of white paper while looking alternately through blue and yellow filters. After a switch from yellow to blue a small dark ring surrounds the foveal point in your field of view. Within the ring is a darker spot. Not everyone can perceive these structures and no one sees them for longer than a few seconds. The phenomenon is called Maxwell's spot after James Clerk Maxwell, who is better known for his theories on the nature and propagation of electromagnetic waves. Filters of other colors will work, but one of the filters should pass more blue light than the other. If purple light is passed by the filter replacing the blue one, the spot is pink.

Although the origin of Maxwell's spot is a subject of debate, it appears to be due to the pigmentation of the macula lutea, or yellow spot, that lies across the foveal depression. The location of the macula lutea is not precise, but the spot is most pigmented in that depression. The molecules of the pigment absorb in the blue range; therefore when blue light traverses the macula on its way to the underlying photoreceptors, the light is absorbed. For the few seconds necessary for the eye to adapt to the blue light, Maxwell's spot is apparent. Then it fades. It can be restored briefly if the eye is adapted to light of a different color before it is presented with blue again.

If you examine entoptic phenomena with a pinhole, you are likely to see a variety of other structures. Just after I blink I see bright spots with an ill-defined gray surround. In addition bright and dark horizontal lines gently descend across the entire view. Both effects are caused by the transparent fluid left on the cornea by a blink. The fluid (droplets or a film) refracts the light passing through it. In a droplet the refraction concentrates light toward the center of what should be the geometric shadow of the droplet. Hence the center of the image is bright and the perimeter is relatively dull. The uneven liquid film retreating across the cornea after a blink creates lines of concentrated light along with dimmer ones.

Several other entoptic features were described by Helmholtz, but their interpretation is not complete. You might see small, relatively bright specks that resemble air bubbles. They probably are not the normal floaters; they seem to lack the floater's characteristic pattern of concentric circles. They might be objects in the medium between the cornea and the lens.

You might also see stationary dim and dark lines radiating from the center of your field of view. Helmholtz suggested that these lines result from the radial structure of the lens of the eye. Dark specks, due to small opaque regions in the lens, may also be present. Stationary bright patches with small arms apparently result from seams within the membrane on the front of the lens. The radial structure may originate when the membrane separates from the cornea at an early stage of fetal development. Occasionally I see bright, wrinkled patches floating across my field of view. They appear to be patches of membrane floating in the eye just as the floaters do, but I do not know their source.

A new entoptic phenomenon was recently described by Christopher W. Tyler. Before you open your eyes in the morning you may notice a peculiar spot in your field of view that is either much darker than the rest of the field or (apparently more rarely) brighter than the field. The spot subtends no more than about a degree in the field of view. It is more likely to be noticed near the center of the field but not necessarily in the foveal region. It may suddenly fade and then reappear later, seemingly without cause. Sometimes spots appear as pairs, sometimes as circles and sometimes as elongated and irregular shapes. Generally they are distincter if you do not move your eyes.

Tyler calls such a spot a perceptual microaneurysm. It has much the same appearance that a physiological microaneurysm in the retina has to someone who has suffered from one. A microaneurysm is a bulging outward of a blood vessel. It is seen by a physician examining the retina as a tiny black spot. The cause of the perceptual type of black spot is not known.

A novel kind of stereogram has been sent to me by Theodore C. Pickett of Sante Fe, N.M. He photographed a woodpile in a stereographic pair in the conventional manner. Then he mounted the transparencies backward. He did not exchange the transparencies left and right.

Pickett's result appears in the illustration above. You might be able to view the setup without a stereoscope if you are able to stare at the picture until you fuse the separate images. I can do this with some difficulty. I try to adjust the convergence of my eyes as if I were looking at something farther away than the photographs. Then I readjust my focus (essentially without thought) until they overlap. (Extra images lie on each side of the overlapping pair, but I ignore them.) As the overlapping pictures slide one over the other in my field of view they suddenly snap together to provide a three-dimensional image.

When photographs are mounted in a stereoscope in the usual way, the observer assigns depth to objects in them by examining their relative positions and sizes. Presumably he makes the same analysis of depth that he would make in looking at the original scene. Although his experience and various clues to depth may help, the critical clue seems to be the angular position of an object as seen by each eye.

Suppose two objects lie in the field of view. One object (call it A) is farther away than the other (B), but neither is notably distant. When the eyes view A, the lines of sight converge with a certain angle between them at A. When B is viewed, the lines of sight converge on B with a larger angle because B is closer. From the difference in the two angles the observer attributes depth to the two objects.

When the scene. is photographed stereoscopically, the position of the camera is shifted between exposures to match the separation between the eyes. The photographs are then placed in a stereoscope, which facilitates the fusion of the two images. The scenes mimic the original scene. Although the photographs are flat representations of the scene, the slightly different views of the same objects cause the observer to assign depth as before. As before, the lines from the eyes to the apparent object A subtend a smaller angle than the lines from the eyes to the apparent object B. The observer therefore attributes greater depth to A.

Many investigators have experimented with a left-and-right exchange of photographs, which results in an inversion of depth so that truly distant objects seem to be nearby and vice versa. Pickett's stratagem plays a similar trick on the observer. The objects are reversed left and right in each photograph. The lines to the apparent objects A and B now converge with a larger angle for A than for B. As a result the truly distant object is interpreted as being relatively nearby and the truly nearby object seems to be distant.

Photographs made and mounted following Pickett's directions can create a perplexing view of the world. Some of the strangeness results from a conflict between the observer's experience and what he sees. Depth is assigned not only according to the apparent angles of objects but also according to the relative size of the images on the retina. If two identical objects are viewed when one is closer to the observer than the other, depth is easily assigned because the more distant object creates a smaller image on the retina. When a stereographic pair is viewed in Pickett's setup, this clue conflicts with the strong clue from the analysis of angles. The angular analysis usually prevails, but the conflict in clues !; creates an eerie sight for the observer.

The assignment of depth according to angles dominates in normal vision until the objects are quite distant (perhaps 450 meters away). Then either experience or the relative size of the retinal images determines the assigned depths. Usually objects farther away than about 450 meters are perceived by the observer as lying on a flat plane at approximately that distance.

The most startling of Pickett's photographs include both distant and nearby objects. When a stereographic pair containing such objects is viewed in Pickett's reverse mounting, the distant objects no longer appear to lie on a relatively flat and distant plane. They seem to be closer and to have greater disparity of depth. The truly nearby objects are relegated to the distant flat plane and appear to have much less disparity in depth.

If you would like to experiment with Pickett's procedure, you might try photographic scenes with objects that have a considerable range of depth. A view over a valley with both nearby and distant trees might be ideal. To make such a picture, however, you will have to follow certain special procedures. Increase the separation between the two positions of the camera by about 17 feet for each mile of range to the most distant of the objects you will be photographing. Exclude from your prospective picture anything that is nearer than 25 percent of that maximum distance, otherwise the stereographic pair will yield a double image of the nearer object. Finally, arrange for some distant object to be in the center of each photograph of the stereographic pair.

In my November column I described a method of "pinspeck" photography devised by Adam Lloyd Cohen. It turns out that the principle underlying pinspeck photography was first, and quite independently, demonstrated by Ronald Cowart of the University of Texas and described in his doctoral dissertation as "the inverse pinhole camera." After further development by Alfonso Zermeno, James M. Hevesi and Lee M. Marsh the device was patented in 1978. The patent is held by the University of Texas and all patent rights are currently licensed to Texas Medical Instruments, Inc., of San Antonio.

Bibliography

ENTOPTIC IMAGES AND RELATED PHENOMENA. G. S. Brindley in Physiology of the Retina and Visual Pathway. The Williams & Wilkins Company, 1970.

THE VISIBILITY OF RETINAL BLOOD VESSELS. A. E. Drysdale in Vision Research, Vol. 15, No. 7, pages 813-818; July, 1975.

The Society for Amateur Scientists (SAS) is a nonprofit research and educational organization dedicated to helping people enrich their lives by following their passion to take part in scientific adventures of all kinds.