Dispersion of light
Hold a glass prism up to a beam of sunlight in a darkened room and something strange happens: the plain white beam walks in one side and a glowing ribbon of colour walks out the other — violet, indigo, blue, green, yellow, orange, red, laid out in order like a tiny permanent rainbow on the wall. Nothing was added to the light and nothing was painted on the wall. The colour was hiding inside the "white" all along, and the prism simply sorted it out. This splitting of white light into its component colours is called dispersion, and it is also the secret behind the rainbow that appears in the sky after rain, when sunlight passes through countless raindrops instead of a single wedge of glass.
White light is a mixture of seven colours (VIBGYOR) that each bend by a slightly different amount when refracted through a prism's non-parallel surfaces, so they fan out into a visible spectrum — the same effect, produced by raindrops instead of glass, that makes a rainbow.
The core idea
A prism is not a slab. A rectangular glass slab has two flat refracting surfaces that are parallel to each other; a ray entering the slab is bent once on entry and bent back by an equal and opposite amount on exit, so it leaves travelling in its original direction, only shifted a little sideways (a "lateral displacement"). A prism, by contrast, is a transparent block — commonly a triangular wedge of glass — whose two refracting surfaces are slanted and not parallel. Because the exit surface is tilted relative to the entry surface, the second bending does not cancel the first. Light leaving a prism comes out tilted, or deviated, from its original path.
White light is a mixture, not a single colour. Ordinary sunlight looks colourless, but it is actually a blend of seven colours travelling together: violet, indigo, blue, green, yellow, orange and red, remembered by the acronym VIBGYOR. You cannot see this while the colours travel as one bundle through air. It only becomes visible once something bends each colour by a different amount, peeling the bundle apart.
Each colour bends by a different amount. When light crosses from air into glass, it slows down and bends towards the "normal" (refraction) — but not every colour slows down by the same fraction. Violet light is slowed the most by glass and so bends the most; red light is slowed the least and bends the least; the other five colours fall in between, in the order VIBGYOR. This colour-dependent bending is the entire mechanism behind dispersion — no other ingredient is needed.
Put the two together and you get a spectrum. Inside a prism, the non-parallel surfaces let this small difference in bending accumulate instead of cancelling out, so by the time the colours emerge they have fanned out into a visible band, violet deviated most and red deviated least. This ordered band of colours is called a spectrum, and the act of producing it is dispersion.
Why it works
Dispersion happens because two separate facts combine, and understanding it means understanding both together rather than memorising a rule. The first fact is that different colours of light travel at slightly different speeds inside glass — violet slows down the most, red the least, when they cross from air into the denser glass. This colour-dependent slowing is what makes each colour bend by a slightly different angle at a refracting surface; it is a property of the glass and of light itself, not something special about a prism's shape. The second fact is about geometry: a prism's two refracting surfaces are not parallel to each other, unlike a slab's. This matters because refraction happens twice in a row, once on entry and once on exit. In a slab, the two parallel surfaces undo each other's bending — the ray tilts one way going in and tilts back by exactly the same amount coming out, so every colour ends up travelling in the original direction again (just shifted sideways), and any tiny colour-splitting that started on entry is cancelled on exit. That is why you do not see a rainbow through a windowpane. In a prism, the tilted exit surface does not cancel the entry bending — it adds to it. Each colour, having already been bent a little differently on entry, is now bent again on exit, in a way that increases rather than removes the separation between colours. Violet, bent the most at each surface, ends up deviated furthest from the original path; red, bent the least, ends up deviated the least. The gap between them, invisible in the original white beam, has now become a visible band of colour ordered from violet to red — the spectrum. Take away either ingredient — colour-dependent bending, or non-parallel surfaces — and dispersion vanishes: a slab with colour-dependent bending recombines the colours, while a prism made of a hypothetical material where every colour bent identically would deviate the beam but never split it into colours.
Worked example
A beam of white sunlight is shone through a glass prism in a darkened room and a screen is placed beyond the prism to catch the emerging light. A friend argues that since a rectangular glass slab also refracts light, it should produce the same coloured band on a screen if you shine light through it instead. Decide whether the friend is right, and describe what actually appears on each screen.
Step 1. Start with the prism. The beam entering it is white light, a mixture of the seven VIBGYOR colours travelling together along the same path.
Step 2. As the beam crosses the prism's first (entry) surface, each colour refracts by a slightly different angle, violet the most and red the least, because each is slowed by a slightly different amount inside glass.
Step 3. Because the prism's second (exit) surface is not parallel to the first, the second refraction does not cancel the first — it adds to the separation already begun. The colours leave the prism travelling along visibly different directions, violet deviated most and red deviated least, with indigo, blue, green, yellow and orange spread out between them.
Step 4. On the screen beyond the prism, this produces a continuous coloured band in VIBGYOR order — the spectrum of white light, direct proof that sunlight is a mixture of seven colours rather than one.
Step 5. Now consider the slab. Refraction still happens at both of its surfaces, and each colour still bends by a slightly different amount at the first surface — but because the two surfaces are parallel, the second refraction exactly reverses the first for every colour. Each colour arrives back travelling parallel to its original direction (only shifted sideways slightly), so all seven recombine before reaching the screen.
The friend is wrong: only the prism produces a coloured band on the screen; the slab's screen shows a single spot of ordinary white light, laterally displaced but not split into colours.
Competitive stretch
A student wonders whether cutting the prism's triangular cross-section into a rectangle (turning it into a slab) but keeping the same glass would still show any colour at the very edges of the emergent beam. In principle, the two refractions still bend each colour by very slightly different amounts internally, but because the surfaces are exactly parallel, this splitting is always undone before the light exits — there is no residual colour fringe from the bulk of the slab itself. Any faint colour fringes seen at the sharp edges of real objects come from a different effect (edge diffraction, outside this topic), not from the slab's internal refraction.
Common traps
- Assuming any transparent block disperses light — only a shape with non-parallel refracting surfaces (like a prism) does; a slab's parallel surfaces undo the colour-splitting before the light exits.
- Thinking white light is a single, pure colour — it is a mixture of seven colours (VIBGYOR) that only become visible as separate colours once they are refracted unequally.
- Reversing the VIBGYOR bending order — violet bends the most and red the least; getting this backward flips the whole spectrum in your explanation.
- Reaching for a formula or angle calculation for dispersion at this level — the required explanation is entirely qualitative: "different colours bend by different amounts," with no numeric law to apply.
Ask the tutor: Curious why you never see a rainbow when you look directly towards the sun, or want more everyday examples of dispersion beyond prisms and raindrops? Ask the AI tutor.
Why does a prism split light into colours but a plain glass window does not?
A window pane is essentially a slab with parallel surfaces. Light entering it does bend a tiny bit differently for each colour at the first surface, but the second, parallel surface bends every colour back by the same amount it was bent on entry, so all the colours arrive travelling in their original direction again and recombine into white light. A prism's two surfaces are tilted relative to each other, so the second bending adds to the first instead of cancelling it, letting the colour separation build up into a visible spectrum.
Is a rainbow just sunlight passing straight through a raindrop like a prism?
Not quite straight through. Sunlight entering a spherical raindrop is first refracted (colours starting to separate, just as at a prism's first surface), then it reflects off the inside back surface of the droplet, and finally refracts a second time as it exits, separating the colours further. So a raindrop behaves like a prism combined with a mirror. Millions of droplets across the sky each do this, and you see only the one colour from each droplet that reaches your eye at just the right angle, building up the arc of the rainbow.
Do we need to remember an equation to answer dispersion questions in exams?
No. At this level the expected explanation is entirely qualitative — you should be able to state that white light is a mixture of seven colours (VIBGYOR), explain that each colour bends by a different amount because a prism's non-parallel surfaces do not cancel the refraction, and identify violet as bending the most and red the least. No refractive-index values or angle calculations are required for this topic; those belong to later, more advanced treatments of refraction.