How You See

Chapter 1 Anatomy and optics

The parts of the eye

Fifteen named parts, one job: put a sharp image of the world on the retina. Tap any part to see what it does, then follow the light.

14 min read · 7 interactive figures · 3 questions

A horizontal section through a right eye, seen from above; light enters from the left. Tap a part, or a name in the list, to see what it does. The Quiz button hides the labels and asks you to find each part.

The main parts of the eye are the cornea, the clear front window; the iris and the pupil, the colored ring and the hole it controls; the lens, which fine-tunes focus; the retina, the light-sensitive lining at the back; and the optic nerve, which carries the retina's signals to the brain[1]. Around them are the tough white sclera, the blood-rich choroid, the muscles of the ciliary body and two clear fluids that keep the eye's shape. The diagram above names fifteen parts. Everything in the front half of the eye serves one purpose: putting a sharp image on the retina.

This chapter is about how that happens, and about how strange the result is. The image is tiny, upside down and, for most people over about 45, blurred for anything close. It also has a hole in it. You notice none of this, and the reasons why are the subject of the rest of the site.

Three layers and two chambers

The eyeball is a hollow sphere built from three layers of tissue[1]. Anatomists call them tunics. established

  • The fibrous tunic is the outer coat. Most of it is the sclera, the white of the eye, which covers about five-sixths of the eyeball. At the front it becomes the transparent cornea. A thin membrane, the conjunctiva, covers the visible sclera and lines the eyelids.
  • The vascular tunic is the middle layer. At the back it is the choroid, a sheet of blood vessels that feeds the outer retina. Toward the front it thickens into the ciliary body, a ring of muscle that holds the lens on fine zonule fibers, and then into the iris, the colored ring of smooth muscle around the pupil.
  • The neural tunic is the retina: nervous tissue, about half a millimeter thick, that turns light into electrical signals and begins to process them[2].

The lens divides the inside into two cavities. In front of it, between lens and cornea, is the watery aqueous humor. Behind it, filling most of the eye, is the clear gel of the vitreous humor[1]. Both keep the eye pressurized and in shape, and both are transparent, which is the main thing the light needs from them.

Two small regions of the retina matter more than their size suggests. The fovea, a pit less than a quarter of a millimeter across at the center of the yellowish macula, is packed with cones and free of blood vessels: it is where you point your gaze, and the only place you see fine detail[2]. The optic disc, about 2 by 1.5 millimeters, is where the nerve fibers leave the eye, and it has no photoreceptors at all[2]. established

Following the light

Trace one ray from a word on this page. It crosses the cornea, the aqueous humor, the pupil, the lens and the vitreous humor, and reaches the retina. Then it does something that surprises everyone who meets it for the first time: before it reaches a single photoreceptor, it passes through the retina's own nerve fibers, blood vessels and the bodies of its neurons. The rods and cones lie in the outermost layer, against the pigment epithelium and the choroid, facing away from the light[2][1]. established

The signal then travels back the way the light came. Photoreceptors pass it to bipolar cells, bipolar cells to retinal ganglion cells, and the ganglion cells' axons run across the inner surface of the retina to the optic disc, where they gather into the optic nerve[1]. That inside-out design is why the eye has a blind spot (below), and the fovea is the exception: there, the inner layers are pushed aside so that light reaches the cones unobstructed[2].

How the eye focuses

Light from a point on an object spreads out in every direction. To see the point sharply, the eye has to bend the part of that spray that enters the pupil so that it meets again at one point on the retina. Two clear surfaces do the bending. The larger share is done by the cornea, where light passes from air into tissue and slows the most: about two-thirds of the eye's focusing power is the cornea's, and the lens provides the rest[3]. established

Question 1 Prediction: what does the lens do for near objects?#

The cornea's power is fixed. The lens is not. To focus on something close, the eye needs more power, and it gets it by changing the shape of the lens, a process called accommodation[3].

Opticians and physics textbooks treat the cornea and lens together as a single thin lens, a fixed distance in front of the retina[3]. Its power P, in diopters (D, the reciprocal of the focal length in meters), must satisfy the thin-lens equation for an object at distance do and the retina at di:

P = 1/do + 1/di(1)

With di = 2.00 cm, a typical figure, focusing on a distant hill (1/do ≈ 0) takes 50 D, and focusing on a page at 25 cm takes 54 D[3]. The lens adds the difference, 4 D. In the figure below, the eye starts on a letter F half a meter away, where it needs only 2 D of accommodation. established

Figure 1 The model eye focusing on an F half a meter away; the lens adds 2.0 D. Drag the F (or use the slider) toward the eye and watch the lens thicken. The red rays from the top of the F always meet on the retina until the F passes the near point (blue tick); closer than that, they meet behind it and the inset shows the blur.#

Why reading glasses arrive in your forties

The extra power the lens can add is called the amplitude of accommodation, and it shrinks steadily through life as the lens stiffens. Optometrists have long estimated the average amplitude with Hofstetter's formula of 1950, in diopters[5]:

Amax ≈ 18.5 − 0.3 × age(2)

The nearest distance the eye can focus, the near point, is 1/Amax for an eye that sees distant things sharply without effort. At 20, that is 1/12.5 D = 8 cm. Try it: bring a finger toward your nose until its fingerprint blurs.

Question 2 Prediction: at what age does reading distance blur?#
Figure 2 The same eye at age 50 with the F at 25 cm, a phone held close. Focusing there takes 4 D; the lens is at its limit of 3.5 D, so the image forms behind the retina. Lower the age to bring the near point in; raise it and watch the near point run away.#

By formula (2), the near point is about 6.5 cm at age 10, 10.5 cm at 30, 15 cm at 40, 29 cm at 50 and two meters at 60. The table compares them. This loss of near focus with age is presbyopia, "old eye", and it is corrected with reading glasses: a converging lens that supplies the power the lens no longer can[3]. established

AgeAmplitudeNear point
1015.5 D6.5 cm
2012.5 D8 cm
309.5 D10.5 cm
406.5 D15 cm
503.5 D29 cm
600.5 D2 m

Nearsighted and farsighted eyes

An eye that is too long for its optics, or too powerful for its length, focuses distant light in front of the retina[7]: it is myopic, or nearsighted. Relaxing the lens cannot help, because relaxing is all the lens can do to lower its power. Far things blur; near things, which need more power, come into focus on their own. The farthest distance a myopic eye sees sharply, its far point, is 1/|R| for a prescription of R diopters: 33 cm for a −3 D eye. An eye with too little power is hyperopic, or farsighted: it can still see far things by accommodating, but it spends part of its amplitude doing so, and near things go first. A diverging (negative) lens corrects myopia and a converging (positive) one corrects hyperopia[7]. established

Figure 3 A −3 D (myopic) eye, aged 30, looking at an F two meters away: the rays meet in front of the retina, and its far point (blue tick) sits at 33 cm. Tick Wear the correcting lens to put the image back on the retina. Then try a positive (farsighted) prescription and watch the near point move out.#

What the pupil adds

The iris narrows the pupil in bright light and widens it in dim light[1]. That controls how much light gets in, and it also changes how badly an out-of-focus eye blurs. The rays from one point fill a cone whose base is the pupil, so when they meet in front of or behind the retina they strike it as a disc. By similar triangles, the disc's angular size in radians is the pupil diameter a (in meters) times the focusing error ΔP (in diopters):

β = a × |ΔP|(3)

Halve the pupil and the blur halves. By the same geometry, an eye that has lost some focusing range should read more easily in bright light, when the pupil is small, and squinting sharpens things for the same reason: a small aperture deepens the range of distances that look acceptably sharp, the depth of field. A pinhole is the extreme case.

Figure 4 A 60-year-old eye with the F at 40 cm: the lens has only 0.5 D to give, so the image falls behind the retina. With a 6 mm pupil the blur disc is large. Narrow the pupil and watch the bundle of rays, the blur on the retina and the F in the inset all shrink together, although the focus itself does not change.#

The image is upside down. So what?

Look at the inset in any of the figures: the F on the retina is upside down and reversed left to right. Every converging lens does this to a real image[3]. It leads to a famous puzzle.

Question 3 Prediction: why does an inverted image look upright?#

The puzzle assumes that somewhere in the brain the retinal image is shown again, the right way up, to something that looks at it. Nothing like that happens. "Up" in vision is not a property of the image. It is a relationship: things whose images fall on the lower retina are the things you raise your eyes, or your hand, to reach. Many introductory texts say the brain "inverts the image again"[3]; it is a convenient shorthand, but taken literally it is a myth. documented

The best test is one man's week. In 1896 and 1897 the psychologist George Stratton wore, over one eye with the other covered, a lens system that turned the retinal image the right way up, so that the world looked upside down. In his longer trial he wore it for eight days. At first everything looked inverted and unreal, and he reached the wrong way for things. Day by day the new arrangement felt more natural and his movements fitted it again, and by the last days things looked upright to him for long stretches, especially when he was busy[8]. Orientation is learned from the agreement between seeing and acting, which is the point. documented

The blind spot

Because the retina is inside out, its output cable has to pass back through it. At the optic disc, the ganglion cell axons and the retina's blood vessels leave the eye, and there is no room for rods or cones[1][2]. Light that falls there is not detected. Every eye has this blind spot, and it is not small: the optic disc lies on the nose side of the fovea, about 17 degrees from it according to Webvision[2], so in your field of view the blind spot sits roughly 15 degrees out on the side toward your temple (13 to 18 degrees in different people)[9], a region several degrees across. The French physicist Edme Mariotte described it in the 1660s, to the surprise of those who expected the optic nerve's entry to be the most sensitive point of all[10]. established

You can find yours. Close your left eye, look at the cross with your right, and slide the dot outward. Hold the screen at the distance you set. Positions are shown as a percentage of the way across until you calibrate the screen with a bank card, because a browser does not know how big its pixels are; after that the figure shows degrees, as accurate as your viewing distance.

Figure 5 The blind-spot test for the right eye, with the dot placed where it is usually still visible. With your left eye closed and your right eye on the cross, drag the dot (or the slider) slowly outward until it vanishes, then on until it reappears, and mark both. On a phone, hold the screen close (the figure suggests a distance) or turn it sideways. Calibrate with a card to read the marks in degrees. The full-size version is on the blind spot test page.#

Two things should strike you. The first is how large the blind region is: marking where the dot vanishes and where it returns usually spans several degrees. The second is what you see in its place, which is not a black hole or a gap but the plain background.

What fills the gap

Why is the blind spot invisible in everyday life? With both eyes open, each eye covers the other's blind spot, which lie on opposite sides. But with one eye closed you still do not see a hole. Researchers have debated two answers. One is that the brain actively fills in the missing region from its surroundings, color, texture and even moving noise included. The other is that it ignores the region: no signal arrives, so nothing is reported, and "nothing" is not experienced as black. There is evidence for both under different conditions, including recordings from monkey visual cortex in which neurons representing the blind spot respond to a bar that crosses it, as if the bar were complete[11]. active

Try the strongest version. Switch the target to a broken line: a straight line with a gap. When the gap falls in your blind spot, most people see the line continue straight through.

Figure 6 The same test with a gap in a line instead of a dot, placed near where the blind spot usually is. With your left eye closed and your right eye on the cross, move the gap until it falls in your blind spot: the line usually looks unbroken. Switch to Pattern for a striped field with a blank patch, which the stripes seem to cover.#

Here is the lesson of this chapter in one demonstration. The light reaching your brain from that region carries nothing. What you see there is a guess, and usually a good one.

From image to signal

So far the eye has behaved like a camera with an unusual sensor: inverted, self-focusing, with a blind patch near the middle. Next comes the sensor. The retina does not record the image as a camera chip does. It samples it with two kinds of cell, and one of those kinds comes in three versions whose answers overlap. Color, night vision and color blindness all follow from that design, in the chapter on rods and cones.