Refraction of light
Dip a straight pencil into a glass of water at an angle and look at it from the side — it suddenly appears snapped in two at the water's surface, the underwater half shifted sideways. Nothing actually broke. Look into a swimming pool and the tiled floor seems much closer to the surface than it really is. Watch the night sky and stars seem to flicker and shimmer instead of glowing steadily. All three puzzles have the same one-line answer: light bends whenever it crosses at an angle from one transparent medium into another, because it changes speed as it crosses. That bending is called refraction of light, and this article builds the vocabulary and the bending rule that explain it.
Light bends at the boundary between two transparent media whenever it crosses obliquely, because its speed changes there — towards the normal on entering a denser medium, away from the normal on entering a rarer one.
The core idea
Why light bends at all. Light travels at different speeds in different transparent media: fastest in vacuum and air, slower in water, slower still in glass. At Class 8 level only this ranking matters — there is no numeric refractive index and no Snell's-law calculation to do. When a ray meets the boundary between two such media travelling straight along the normal, it keeps going in the same direction, just slower or faster. But when it meets the boundary at an angle, one edge of the beam changes speed a fraction of a second before the rest, and this uneven change of speed swings the whole ray around — the direction changes. This bending of light as it crosses the boundary of two media is refraction.
The vocabulary of a refraction diagram. Every refraction problem is described using six labelled parts of the diagram, all defined at the exact point where the ray crosses the boundary:
- Incident ray — the ray travelling towards the boundary, in the first medium.
- Point of incidence — the point on the boundary where the incident ray strikes.
- Normal — an imaginary line drawn perpendicular (at ) to the boundary at the point of incidence; it is not part of either ray, only a reference line.
- Refracted ray — the ray that continues on into the second medium after bending.
- Angle of incidence () — the angle between the incident ray and the normal, measured in the first medium.
- Angle of refraction () — the angle between the refracted ray and the normal, measured in the second medium.
The bending rule. This is the one rule worth memorising cold, because every qualitative question in this chapter reduces to it: light travelling from a rarer medium (such as air) into a denser medium (such as water or glass) bends towards the normal, so is smaller than . Light travelling from a denser medium into a rarer one bends away from the normal, so is larger than . A ray that strikes the boundary exactly along the normal, with , passes straight through undeviated — refraction only bends oblique rays, never a ray hitting the surface head-on.
Why it works
Refraction can be pictured without any formula by imagining what happens to a wide beam of light — or, more concretely, a set of car wheels or a line of marching soldiers — rolling from a hard road obliquely onto soft sand. This reasoning shows why an oblique speed change must turn the whole beam, not just slow it down evenly.
Step 1. Picture a wide beam of light as many parallel rays travelling side by side, like a row of wheels on an axle rolling towards the road-to-sand boundary at an angle rather than straight on. Because the boundary is crossed obliquely, one side of the beam (one wheel) reaches the sand before the other side does.
Step 2. The wheel (or ray) that enters the sand first slows down first, while the other wheel is still on the hard road moving at full speed. For a short interval, one side of the axle is moving slower than the other side.
Step 3. Whenever one end of a connected line moves slower than the other, the line swings, pivoting toward the slower side — exactly as a cart's axle swings when one wheel bogs down in mud while the other keeps rolling. This is the mechanical reason the whole ray changes direction, not just its speed.
Step 4. Once the entire beam has crossed into the slower medium, both sides move at the same, now slower, speed again, so the ray travels straight once more — but along a new direction, bent towards the normal. Run the reasoning in reverse, sand back onto road (denser to rarer), and the side that speeds up first swings the beam away from the normal instead — matching the bending rule exactly.
Worked example
A student shines a torch beam from air onto the still surface of a swimming pool at an angle, so the beam enters the water obliquely, and then asks why the pool floor looks shallower than it really is.
Step 1. Identify the two media. The beam starts in air (rarer) and enters water (denser), crossing the surface obliquely rather than straight down.
Step 2. Apply the bending rule. Going from a rarer medium to a denser one, the ray bends towards the normal at the water's surface, so the angle of refraction is smaller than the angle of incidence .
Step 3. Trace what the eye does. Light from the pool floor bends towards the normal as it leaves the water surface... no — light travelling up from the floor to an eye above the water actually goes water-to-air (denser to rarer), so it bends away from the normal on emerging. The eye, assuming light always travels in a straight line, extends the emerging ray backwards in a straight line and locates an apparent source higher up than the real floor.
Step 4. Compare the apparent and real positions. Because the backward-traced straight line meets a point above the true floor, the floor appears raised — that is, the pool looks shallower than its actual depth.
Answer: the swimming pool's floor looks shallower because light rays from the floor bend away from the normal on leaving the water, and the eye traces them straight back to a falsely raised apparent position.
Competitive stretch
The same reasoning, applied through many thin layers of air of gradually changing density, explains two sky effects. Starlight passes through an atmosphere whose density (and bending power) constantly fluctuates as pockets of warm and cool air drift across the path, so its apparent position jitters and brightness flickers — the twinkling of stars. Near sunrise and sunset, sunlight from a Sun still slightly below the geometric horizon is bent by the denser air near the ground enough to curve into the eye, so the Sun becomes visible a little before it geometrically rises and stays visible a little after it geometrically sets — both are refraction, applied gradually through a medium whose density changes continuously rather than at one sharp boundary.
Common traps
- Thinking refraction only happens on entering a denser medium — it also happens, bending away from the normal, whenever light leaves a denser medium for a rarer one.
- Confusing the normal with the boundary surface itself — the normal is the perpendicular reference line at to the surface, not along it.
- Believing every ray bends by some fixed amount — a ray striking the boundary exactly along the normal () does not bend at all.
- Mixing up refraction with reflection — refraction is light crossing into a new medium and changing direction due to a speed change; reflection is light bouncing back into the same medium off a surface.
Ask the tutor: Want to walk through labelling a refraction ray diagram yourself, or see how atmospheric refraction explains the flattened shape of the setting Sun?
Why does the pencil look bent only at the water's surface, and not along its whole underwater length?
Because refraction only happens exactly where a ray crosses a boundary between two media. Light from every point along the submerged part of the pencil crosses the same single water-to-air boundary at the surface and bends there, then travels straight through air to your eye. The part of the pencil above the water is seen directly, with no boundary crossed and hence no bending. Since the underwater part's apparent position is shifted while the above-water part is not, the two halves stop lining up exactly at the surface — which is why the "break" always appears right at the waterline, never in the middle of the underwater section.
Do stars actually move when they twinkle, or is it only an illusion?
It is an illusion caused entirely by our atmosphere, not by the star itself. Starlight travels through many kilometres of air whose density keeps changing slightly as pockets of warmer and cooler air drift past, each one refracting the ray by a slightly different, constantly changing amount. The star's true position and brightness never change; what changes is how much the ray is bent on its way down to your eye at each instant, which makes the star's apparent position and brightness flicker. Planets twinkle far less because they appear as tiny discs rather than points, so the flickering from many directions partly averages out.
Is the coin-in-a-bowl effect the same as a magnifying glass?
No — they use refraction in two different ways. A magnifying glass is a curved lens shaped so that refraction at its curved surfaces bends many rays to converge and form an enlarged image; the shape of the glass is deliberately chosen to do this. The coin-in-a-bowl effect uses a single flat water surface with no special shape — it does not magnify the coin, it only shifts its apparent depth upward because the eye wrongly assumes the emerging ray travelled in a straight line all the way from the coin. The coin looks closer to the surface, not bigger.