# Waves: sound, light, colour and the wave-particle puzzle: Bending light and making colour: refraction, rainbows, blue skies and mixing

Source page: https://www.materialinsights.org/learn/science/waves-sound-light/p4/

[All courses](https://www.materialinsights.org/learn/) / [Understanding Science](https://www.materialinsights.org/learn/science/) / [Module 7: Waves: sound, light, colour and the wave-particle puzzle](https://www.materialinsights.org/learn/science/waves-sound-light/) Part 4 of 6

## Bending light and making colour: refraction, rainbows, blue skies and mixing

A straw looks broken in a glass, a rainbow follows the rain, the sky is blue at noon and red at dusk, and a screen makes yellow from red and green. All four come from one fact, that light's speed and the response of matter depend on wavelength. This part builds them in order, from refraction to the eye's three detectors.

### 4.1 Light bends when it enters a medium where it travels more slowly, and Snell's law gives the bend

Put a coin in an empty steel tumbler and lower your head until the rim just hides it. Keep still while someone pours in water, and the coin comes back into view. Light from the coin changes direction at the water's surface, so it now clears the rim and reaches your eye. This is refraction.

The cause is the change of speed from the last part. Picture a line of marching soldiers walking from a road onto soft sand at a slant. The soldier who reaches the sand first slows down, the others catch up, and the whole line swings round. A wave front in light does the same: its frequency stays fixed, its speed drops, so its wavelength shortens and the front pivots towards the normal. Going into a faster medium it pivots away.

The refractive index n = (speed in vacuum) / (speed in the medium). The rule is n1 x sin(angle1) = n2 x sin(angle2). The Persian scholar Ibn Sahl, at the court in Baghdad, set it out in 984 in a treatise on burning mirrors and lenses; Willebrord Snellius found it again in 1621 and Descartes published it in 1637.

Bending explains apparent depth: the water in a stepwell 1.2 m deep, seen from straight above, looks about 0.9 m deep (1.2 m / 1.33, calculated). It explains lenses: a convex lens is thicker in the middle so rays from a distant point meet at a focus, as in a camera or a magnifier (see 'Refraction and lenses'). At steep angles, light inside a dense medium cannot escape at all; this total internal reflection traps light in optical fibres (see 'Light guiding by total internal reflection').

Data Worked estimate: a ray striking at 45 degrees from the normal, entering three media

| Medium | Index n | Speed (3.00 x 10^8 / n) | Angle after bending (sin = 0.7071 / n) | Critical angle (sin = 1 / n) |
|---|---|---|---|---|
| Water | 1.33 | 2.26 x 10^8 m/s (calculated) | 32.1 degrees (calculated) | 48.8 degrees (calculated) |
| Glass | 1.50 | 2.00 x 10^8 m/s (calculated) | 28.1 degrees (calculated) | 41.8 degrees (calculated) |
| Diamond | 2.42 | 1.24 x 10^8 m/s (calculated) | 17.0 degrees (calculated) | 24.4 degrees (calculated) |

Calculated. Typical indices for yellow light. The critical angle applies to light going from the medium into air.
 Source: Hewitt, Lyons, Suchocki and Yeh, Conceptual Integrated Science, 2nd edition (Pearson, 2013), ch. 8, 'Waves, Sound and Light'  Common misreading

#### The speed of light never changes

What is constant is the speed in a vacuum. In water the wave travels at about three quarters of that, in glass two thirds, and in diamond about 41%. Light in a medium is a combination of the original wave and the wave re-radiated by the material's electrons. It returns to full speed on leaving.

- [San Diego State University, A. T. Young, Discovery of the law of refraction](https://aty.sdsu.edu/explain/optics/discovery.html (Ibn Sahl's work of 984 uses the true law; Snel 1621; Descartes first to publish, 1637)) · web
- Hewitt, Lyons, Suchocki and Yeh, Conceptual Integrated Science, 2nd edition (Pearson, 2013) · ch. 8, 'Waves, Sound and Light'

### 4.2 Glass and water slow violet light slightly more than red, so a prism spreads white light and raindrops build a rainbow

The index of refraction depends a little on wavelength, because electrons respond differently to different frequencies. For water, the index is about 1.331 for red and 1.344 for violet. Violet is slowed more and so bends more. A prism turns this small difference into a fan of colours: dispersion. In the mid-1660s Newton passed sunlight through one prism, then sent one colour through a second prism and found it did not change. A second prism turned upside down brought the colours back to white. The glass was not adding colour; white light is a mixture that the prism separates. He reported this to the Royal Society in 1672 and set it out fully in his Opticks of 1704.

A rainbow is dispersion in countless raindrops, with one extra step. A sunbeam enters a drop and bends (violet most), reflects once off the back, and bends again on leaving. Rays leave the drop in many directions, but they pile up near one angle, about 138 degrees from their original direction, giving a bright ring around the antisolar point, the spot opposite the Sun. Red lies at about 42 degrees from that point and violet at about 40 degrees, so red is on the outside of the arc.

A fainter secondary bow, between about 50 and 53 degrees from the antisolar point, has reversed colours because of a second reflection inside each drop, and the sky between the bows is darker (Alexander's dark band). The Sun must be lower than 42 degrees for the primary bow to be above the horizon, which is why a high midday Sun gives none.

Data Worked estimate: how high does a rainbow rise? (top of the bow = 42 degrees minus the Sun's elevation)

| Time or place | Sun's elevation | Top of the bow above the horizon |
|---|---|---|
| Early morning after rain | 5 degrees | 37 degrees (calculated) |
| Mid morning | 20 degrees | 22 degrees (calculated) |
| Late afternoon in the monsoon | 35 degrees | 7 degrees (calculated) |
| The Sun at 42 degrees | 42 degrees | 0 degrees (calculated), none visible |
| Delhi at noon on 21 June (90 - /28.6 - 23.4/) | about 85 degrees (calculated) | below the horizon by about 43 degrees (calculated): no bow |

Calculated. Sun elevation at noon = 90 degrees minus the difference between latitude and the Sun's declination.
 Source: Astronomy magazine, How to read rainbows (November 2024), https://www.astronomy.com/science/how-to-read-rainbows/ (primary bow 42 degrees from the antisolar point; secondary bow reversed and about a tenth as bright; Alexander's dark band)  Common misreading

#### A rainbow sits at a particular place in the sky

A rainbow is a set of directions, not an object at a distance. Every drop that lies 42 degrees from the antisolar point sends red towards you, whether it is 3 m away in the spray of a garden hose or 3 km away in a monsoon shower. The bow you see can be built from both at once, and nothing in its look tells you how far away it is.

- [Astronomy magazine, How to read rainbows (November 2024)](https://www.astronomy.com/science/how-to-read-rainbows/ (primary bow 42 degrees from the antisolar point; secondary bow reversed and about a tenth as bright; Alexander's dark band)) · web
- [Light, Colour, Vision, Secondary rainbow](https://lightcolourvision.org/rainbows/secondary-rainbow (secondary bow at approx. 50.4 to 53.4 degrees; violet outside, red inside; never as bright as the primary)) · web
- [Museum of the History of Science, Oxford, Newton's prism experiment](https://www.cabinet.ox.ac.uk/node/4307 (optics and colour from around 1665; paper of February 1672; Opticks of 1704)) · web
- Hewitt, Lyons, Suchocki and Yeh, Conceptual Integrated Science, 2nd edition (Pearson, 2013) · ch. 8, 'Waves, Sound and Light'

### 4.3 Air scatters short wavelengths far more than long ones, so the sky is blue, the low Sun is red and clouds are white

Sunlight crossing the air shakes the electrons in nitrogen and oxygen molecules, which are far smaller than a wavelength of light. These electrons re-radiate the energy in all directions: scattering. A faster-shaken charge radiates much more strongly, in proportion to the fourth power of its frequency. So the scattered share of each colour is proportional to 1 / wavelength⁴. Blue, at 450 nm, scatters about 4.4 times as strongly as red at 650 nm, and violet nearly ten times as strongly as red. Blue light is thrown sideways and reaches your eye from every part of the sky.

Why not violet? Sunlight holds less violet than blue, the upper air absorbs some, and the eye is less sensitive to it. The mix we see is blue.

At sunset the light slants through the atmosphere, along a path some 30 to 40 times longer than from overhead. Most blue is scattered out of the beam on the way, leaving red and orange. Droplets in clouds are larger than a wavelength and scatter all colours alike, so clouds look white.

In September 1921, sailing home from England to Bombay on the SS Narkunda, C. V. Raman looked at the Mediterranean and doubted the textbook answer. His work on how light scatters in liquids led, in February 1928, to the Raman effect, in which a small share of the scattered light comes out at a changed frequency, and to the Nobel Prize in Physics in 1930.

Data Worked estimate: how much more is each colour scattered than red? (1 / wavelength⁴, relative to red at 700 nm)

| Colour | Wavelength | Working | Relative scattering |
|---|---|---|---|
| Red | 700 nm | 1 | 1.0 |
| Orange | 600 nm | (700 / 600)⁴ | 1.9 (calculated) |
| Green | 530 nm | (700 / 530)⁴ | 3.0 (calculated) |
| Blue | 470 nm | (700 / 470)⁴ | 4.9 (calculated) |
| Violet | 400 nm | (700 / 400)⁴ | 9.4 (calculated) |

Calculated. The Raman details are from the Public Domain Review account, see sources.
 Source: Scientific American, Why is the sky blue?, https://www.scientificamerican.com/article/why-is-the-sky-blue (scattered intensity proportional to the fourth power of frequency; blue scattered almost 10 times as efficiently as red; reddish setting Sun)  Common misreading

#### The sea is blue because it reflects the sky

Lord Rayleigh thought so. Raman showed that the deep sea's blue is a phenomenon in itself: water molecules scatter blue most, while water also absorbs red and orange, so the light that returns from depth is blue. Deep clear lakes under a grey sky stay blue, which reflection alone could not explain.

- [Scientific American, Why is the sky blue?](https://www.scientificamerican.com/article/why-is-the-sky-blue (scattered intensity proportional to the fourth power of frequency; blue scattered almost 10 times as efficiently as red; reddish setting Sun)) · web
- [The Public Domain Review, A Distinct Phenomenon in Itself: C. V. Raman's Discovery of Why the Sea is Blue (1921)](https://publicdomainreview.org/collection/raman-sea/ (SS Narkunda, September 1921; sea colour not merely reflected skylight; Nobel Prize 1930)) · web
- [American Physical Society, This Month in Physics History: February 1928, discovery of the Raman effect](https://aps.org/publications/apsnews/200902/physicshistory.cfm (returned to Bombay on the S.S. Narkunda, leaving England in September 1921; February 1928 observations; 1930 Nobel Prize)) · web
- Hewitt, Lyons, Suchocki and Yeh, Conceptual Integrated Science, 2nd edition (Pearson, 2013) · ch. 8, 'Waves, Sound and Light'

### 4.4 The eye has three kinds of colour detector, so lights mix by adding and pigments mix by taking away

Your eye does not measure the wavelengths one by one. It has three kinds of cone cell, tuned broadly to the blue, green and yellow-green parts of the spectrum, and colour is the brain's reading of three signals. Red light plus green light excites the green and yellow-green cones in about the same ratio as true yellow light does, so a screen made of red and green dots looks yellow even though no yellow wavelength is present. A mix of red and blue gives magenta, a colour that no single wavelength produces. This is why a screen of red, green and blue dots can display almost any hue: this is additive mixing. Pairs that add to white are called complementary: yellow and blue, red and cyan, green and magenta.

Pigments work the other way. A leaf looks green because chlorophyll absorbs red and blue and returns mostly green (see 'Photosynthesis'). Yellow paint absorbs blue and returns red, yellow and green. Cyan paint absorbs red and returns blue and green. Mix them and only green survives both: this is subtractive mixing, and the primaries are cyan, magenta and yellow, with black added in printing. Under a coloured light an object shows only what that light has: a green leaf in pure red light looks black.

Data Worked estimate: what mixing does

| Mixing | What happens | Result |
|---|---|---|
| Red light + green light | both cone signals rise | yellow |
| Red light + blue light | red and blue cones rise | magenta |
| Green light + blue light | green and blue cones rise | cyan |
| Red + green + blue light | all three rise equally | white |
| Cyan ink + yellow ink | cyan removes red, yellow removes blue | green |
| Magenta ink + yellow ink | magenta removes green, yellow removes blue | red |
| Cyan + magenta ink | removes red and green | blue |
| Cyan + magenta + yellow ink | removes nearly all | dark grey or black |

Standard colour mixing; the leaf example is from plant pigment behaviour.
 Source: Hewitt, Lyons, Suchocki and Yeh, Conceptual Integrated Science, 2nd edition (Pearson, 2013), ch. 8, 'Waves, Sound and Light'  Common misreading

#### The primary colours are red, yellow and blue

That is the art-class rule of thumb. It cannot make bright cyan or magenta. Printers use cyan, magenta and yellow, and screens use red, green and blue. Primaries are whichever three colours let you match the greatest range of others by the mixing method in use.

- Hewitt, Lyons, Suchocki and Yeh, Conceptual Integrated Science, 2nd edition (Pearson, 2013) · ch. 8, 'Waves, Sound and Light'

### See it in the table
[H₂O Water](https://www.materialinsights.org/commodities/water/)[St Steel](https://www.materialinsights.org/commodities/steel/)[O2 Oxygen](https://www.materialinsights.org/chemicals/oxygen/)

### Check your understanding

A ray hits water at 45 degrees from the normal. Water has index 1.33. About what angle does the ray make with the normal in the water?
 sin(angle) = sin 45° / 1.33 = 0.7071 / 1.33 = 0.53, so about 32 degrees. The ray bends towards the normal on entering the slower medium.

Why can't a rainbow be seen near midday in Delhi in June?
 The bow is centred on the point opposite the Sun and lies 42 degrees around it. With the Sun about 85 degrees high that circle is under the ground.

Which statement about why the sky is blue is correct?
 Scattering by molecules is proportional to 1/wavelength⁴, so blue is scattered about four times as strongly as red and reaches the eye from all directions.

[← Part 3: Light: its speed, its spectrum, and what happens when it meets matter](https://www.materialinsights.org/learn/science/waves-sound-light/p3/)[Mark read · Part 5: Waves meeting waves: interference, diffraction, Young's experiment and polarisation →](https://www.materialinsights.org/learn/science/waves-sound-light/p5/)

## Other sections

- [Module overview](https://www.materialinsights.org/learn/science/waves-sound-light/index.md)
- [What a wave is: a pattern that travels while the material stays](https://www.materialinsights.org/learn/science/waves-sound-light/p1/index.md)
- [Sound: pressure waves, loudness, resonance and moving sources](https://www.materialinsights.org/learn/science/waves-sound-light/p2/index.md)
- [Light: its speed, its spectrum, and what happens when it meets matter](https://www.materialinsights.org/learn/science/waves-sound-light/p3/index.md)
- [Waves meeting waves: interference, diffraction, Young's experiment and polarisation](https://www.materialinsights.org/learn/science/waves-sound-light/p5/index.md)
- [The photoelectric effect and the wave-particle puzzle](https://www.materialinsights.org/learn/science/waves-sound-light/p6/index.md)
- [Key terms, quiz and sources](https://www.materialinsights.org/learn/science/waves-sound-light/review/index.md)
