The history
From Ptolemy's tables of bent light to glass threads under the ocean, how people learned exactly how much light turns when it crosses into something new.
A straw in a glass of water looks broken, and people have wondered why for thousands of years. Ptolemy measured the bend, Ibn Sahl in Baghdad found the true rule, and it was found again and again in Europe until it was printed in 1637. Then came the harder question: why does light bend at all? The answer, that light slows down in glass and water, led to glass fibres that carry the internet and to strange materials that bend light the wrong way.
984
Correct law of refraction used
Ibn Sahl, Baghdad
1611
Total internal reflection described
Johannes Kepler, Dioptrice
1637
Sine law printed
René Descartes, La Dioptrique
1841
Light guided along a water jet
Jean-Daniel Colladon, Geneva
1850
Light shown to be slower in water
Léon Foucault, Paris
1954
Image carried by a flexible fibre bundle
Harold Hopkins and Narinder Singh Kapany, London
1970
Fibre with loss below 20 dB/km
Maurer, Keck and Schultz, Corning
2001
Negative refraction observed
Shelby, Smith and Schultz, San Diego
150 – 1021
Ptolemy writes down tables of bent light, Ibn Sahl finds the correct rule in Baghdad, and Ibn al-Haytham studies refraction by experiment in Cairo.
150
c. 150 AD
In Book V of his Optics, Ptolemy measured how much a ray of light turns when it passes from air into water, from air into glass and from water into glass. He set the angles out in tables, every 10 degrees. His numbers follow a neat pattern that historians think he partly smoothed to fit his own rule.
Why it mattered. It was the first careful attempt to measure refraction, and his tables were copied for over a thousand years.
984
In a book called On Burning Mirrors and Lenses, Ibn Sahl used the correct law of refraction to design lenses that focus sunlight to a single point. The book survived only in scattered pieces in libraries in Damascus and Tehran. In 1990 the historian Roshdi Rashed put it back together and showed that Ibn Sahl had the law.
Why it mattered. The law we call Snell's was first used in Baghdad, more than six centuries before Snell.
1021
1011–1021
In his Book of Optics, Ibn al-Haytham built a brass instrument to measure rays passing into water and glass. He explained that light bends because it moves more easily in one material than another, and studied how the eye and lenses use refraction. He came close to the exact law but did not state it.
Why it mattered. His book, translated into Latin, taught European scientists to test ideas about light with experiments.
1601 – 1637
Harriot, Kepler, Snellius and Descartes each chase the rule. Kepler finds total internal reflection, and Descartes finally prints the sine law.
1602
c. 1601–1602 (unpublished)
Harriot, an English mathematician, measured refraction in many liquids and found the sine law. He wrote letters about refraction to Kepler, but he never shared the rule itself. It was found in his papers long after his death in 1621.
Why it mattered. A lesson every scientist learns: a discovery you do not publish may be credited to someone else.
1611
In his book Dioptrice, written to explain Galileo's new telescope, Kepler described how light inside glass, meeting the surface at a steep slant, is not let out at all but is reflected back completely. He found this without knowing the exact law, which he never managed to find.
Why it mattered. Total internal reflection is what keeps light trapped inside every optical fibre today.
1621
1621 (unpublished)
Snellius, a professor at Leiden who also measured the size of the Earth, found a rule for refraction written with lengths rather than sines, but it is the same law. He died in 1626 without publishing it. Christiaan Huygens later read his manuscript and made his name known.
Why it mattered. Because Huygens spread the story, the law carries Snell's name in English, though he was not the first.
1637
Descartes published the sine law in La Dioptrique, one of the essays printed with his Discourse on Method. In another essay, Les Météores, he used it to trace thousands of rays through a drop of water and showed that a rainbow appears about 42 degrees from the point opposite the Sun.
Why it mattered. It was the first time the law appeared in print, which is why the French call it the Snell–Descartes law.
1662 – 1916
Least time, colours, waves and wavelets. Experiments show that light really is slower in water, water jets trap light, and physicists work out how atoms slow it down.
1662
1 January 1662 (letter)
Fermat guessed that light always takes the path that needs the least time, and that it goes more slowly in water and glass than in air. In a letter to Marin Cureau de la Chambre, he showed that this idea gives exactly the sine law. Like a lifeguard who runs further on sand to swim less, light bends to save time.
Why it mattered. It explained the law with one simple idea, and predicted that light is slower in denser materials.
1672
1666 (published 1672)
Newton let a beam of sunlight into a dark room through a hole and passed it through a glass prism. White light spread into a band of colours, because each colour is refracted by a slightly different amount. He published it in 1672 in the Royal Society's journal.
Why it mattered. It showed that a material's refractive index depends on colour, the reason prisms and rainbows split light.
1678
1678 (published 1690)
Huygens, a Dutch scientist working in Paris, imagined every point on a wave of light sending out tiny wavelets. If the wavelets move more slowly in glass, the front of the wave swings round as it enters, like a line of marchers stepping from road onto mud. He read his idea to the Academy in 1678 and printed it in his Treatise on Light in 1690.
Why it mattered. It explained Snell's law with waves, and it is still the clearest picture of why light bends.
1823
c. 1821–1823
Snell's law says where light goes, but not how much of it reflects instead. Fresnel, treating light as a sideways-wobbling wave, wrote equations for how much light is reflected and how much passes through at each angle, for each direction of wobble.
Why it mattered. His equations explain why a window is also a faint mirror, and why polarising sunglasses cut the glare off water.
1842
1841 (described 1842)
To light up a lecture on water jets, Colladon shone sunlight into a tank so that it came out with the stream. The light stayed inside the curving jet, bouncing off its inner surface, and lit up the spot where the water landed. John Tyndall showed a similar jet in London in 1854 and made it famous.
Why it mattered. It was the first light guide, the ancestor of every optical fibre.
1850
27 April 1850
Foucault sent two beams to a fast-spinning mirror, one through air and one through a tube of water. The water beam came back later, so light was slower in water. Fizeau and Breguet found the same seven weeks later. Descartes and Newton had thought light went faster in water.
Why it mattered. It settled the argument: the bend comes from light slowing down, just as Fermat and Huygens said.
1897
1895–1897
Bose built his own senders and detectors of very short radio waves, now called millimetre waves. With prisms, lenses and wire grids he showed they reflect, refract and totally reflect just like light. In 1897 the Royal Society published his tests of total reflection between two prisms with a thin air gap.
Why it mattered. It showed that Snell's law holds for invisible electromagnetic waves too, and made Bose a pioneer of microwave research.
1915
1912–1916
Light does not really slow down by bumping into atoms. The passing wave makes the electrons in each atom wobble, and each wobbling electron sends out a little wave of its own, slightly delayed. Ewald and Oseen proved that these waves add up to cancel the original wave and build a new one that moves more slowly.
Why it mattered. It is the modern answer to why the refractive index exists at all.
1954 – 1979
Bundles of glass threads carry pictures, Charles Kao shows that pure glass could carry messages, and fibre loss falls a thousandfold.
1954
2 January 1954
Hopkins and his student Kapany, who was born in Punjab and went to school in Dehradun, lined up thousands of thin glass fibres into a flexible bundle. Each fibre carried one dot of light by total internal reflection, so the bundle carried a whole image round bends. Their paper in Nature named it the fibrescope, and the Dutch scientist Abraham van Heel described coated fibres in the same issue.
Why it mattered. It led to flexible endoscopes that let doctors look inside the body, and in 1960 Kapany coined the term fibre optics.
1966
1966 (published July 1966)
Glass fibres of the time lost almost all their light within a few metres, around 1,000 dB per kilometre. Kao and Hockham worked out that the loss came from impurities, not from glass itself, and that pure glass could get below 20 dB per kilometre, enough for telephone lines.
Why it mattered. It started the race to make optical fibre, and won Kao the Nobel Prize in Physics in 2009.
1970
August 1970
The Corning team made fibre from very pure silica glass built up from vapour, with a core a little denser than the glass around it. Keck measured a loss of about 17 dB per kilometre and wrote "Whoopee!" in his notebook. By 1972 they had reached 4 dB per kilometre.
Why it mattered. It beat Kao's target and made fibre-optic communication practical.
By the numbers
Loss per kilometre of the best fibres, falling about five-thousandfold in 13 years. Every 10 dB means a tenfold loss of light.
1967 – 2001
A Soviet physicist imagines materials with a negative refractive index, and three decades later a lab in California builds one.
1967
1967 (in English 1968)
Veselago asked what would happen in a material where both the electric and the magnetic response were negative. He showed that light would bend the wrong way at the surface, to the same side of the normal, as if the refractive index in Snell's law were negative. No such material existed, so the idea sat quietly for thirty years.
Why it mattered. It opened the door to materials whose optical properties come from their structure, not their chemistry.
2001
6 April 2001
Following designs by John Pendry of Imperial College, who argued in 2000 that such a material could make a perfect lens, the San Diego team built a prism from tiny copper rings and wires on circuit boards. Microwaves passing through it bent to the negative side, exactly as Veselago had predicted.
Why it mattered. It was the first experiment to show negative refraction, and started the field of metamaterials, including early invisibility cloaks.
Did you know?
The law is named after Snellius, but Ibn Sahl used it in Baghdad 637 years earlier. In France it is called the Snell–Descartes law.
Light in ordinary glass travels at about 200,000 km per second, about two thirds of its speed in empty space.
Diamond has a refractive index of about 2.42, so light inside it is trapped unless it meets a face within about 24 degrees of straight on. That is why diamonds sparkle.
A fish looking up sees the whole sky squeezed into a bright circle about 97 degrees wide, called Snell's window.
The best fibres lose only about 0.2 dB per kilometre, so half the light is still there after about 15 km of glass.
Descartes and Newton both thought light moves faster in water. Foucault's 1850 experiment showed they were wrong.
The people
c. 940 – c. 1000 · Mathematician · Baghdad (Iraq)
Used the correct law of refraction in 984 to design perfect burning lenses.
c. 965 – c. 1040 · Physicist and mathematician · Basra (Iraq), worked in Cairo
Tested refraction by experiment in his Book of Optics, which shaped optics for centuries.
c. 1560 – 1621 · Mathematician and astronomer · England
Found the sine law around 1602 but never published it.
1571 – 1630 · Astronomer and mathematician · Germany
Described total internal reflection and explained how lenses and telescopes work.
1580 – 1626 · Mathematician and astronomer · Netherlands
Found the law in 1621; it bears his name thanks to Huygens.
1596 – 1650 · Philosopher and mathematician · France
First to publish the sine law, and used it to explain the rainbow.
1607 – 1665 · Lawyer and mathematician · France
Derived the law from the idea that light takes the quickest path.
1629 – 1695 · Physicist and astronomer · Netherlands
Explained refraction with wavelets that travel more slowly in glass.
1858 – 1937 · Physicist and biologist · India (born in Bengal)
Showed in Calcutta that millimetre radio waves refract and totally reflect like light.
1926 – 2020 · Physicist and inventor · India (born in Moga, Punjab)
Built the first image-carrying fibre bundle with Harold Hopkins and named the field fibre optics.
1933 – 2018 · Electrical engineer · China (born in Shanghai), worked in Britain and Hong Kong
Showed that pure glass fibre could carry messages over long distances; Nobel Prize 2009.
Where it happened
Sources
Dates marked “c.” are approximate, and historians sometimes disagree about who was first. If you spot a mistake, tell us.
↑↓ move↵ open/ or ⌘K search anywhere