Fizeau's Toothed-Wheel Speed of Light Measurement Explained
For most of history, the speed of light was either assumed to be infinite or inferred from the sky. The first evidence that it was finite came from the timing of Jupiter's moons, and the best numbers of the early 1800s still came from astronomy. In 1849 the French physicist Hippolyte Fizeau brought the measurement down to Earth. He sent a beam of light from a house in Suresnes, west of Paris, to a mirror on the hill of Montmartre 8,633 metres away and back again, chopping the beam with a spinning wheel of 720 teeth. At a rotation rate of 12.6 turns per second, light that left through a gap came back to find a tooth in the way, and the returning beam vanished. From that number Fizeau calculated a speed of about 313,000 kilometres per second, within 5 percent of the modern value. It was the first measurement of the speed of light that did not depend on astronomy, and it could be checked by anyone with the right equipment and a long enough line of sight.
The puzzle: is light instantaneous?
Many ancient and medieval thinkers held that light did not travel at all, but appeared everywhere at once. Galileo proposed testing this with two people on hilltops uncovering lanterns, but the delay he was looking for, a few microseconds, was far too short for human reflexes.
The breakthrough came in 1676, when Ole Rømer, working at the Paris Observatory, noticed that eclipses of Jupiter's moon Io ran progressively late when Earth was moving away from Jupiter and early when it was approaching. He attributed the discrepancy to the extra time light needed to cross the changing distance. Our article on Rømer's Io eclipses explains the method. In 1729 James Bradley found a second astronomical route through stellar aberration, the slight tilt in the apparent direction of stars caused by Earth's orbital motion relative to the finite speed of light.
By the 19th century, physicists were broadly agreed that light travelled at roughly 300,000 km/s. But both methods depended on astronomical distances, such as the radius of Earth's orbit, that were themselves uncertain. A direct measurement over a known distance on the ground would cut that dependency.
Who was Fizeau?
Armand Hippolyte Louis Fizeau was born in Paris in 1819 to a wealthy family. He trained initially in medicine before turning to physics and optics, and worked closely with Léon Foucault in the 1840s on early photography and on the interference of light and heat. In 1848 he worked independently on the Doppler effect for light, which is why the shift of spectral lines from a moving source is sometimes called the Doppler–Fizeau effect in France.
By the late 1840s, the idea of measuring light's speed on Earth was in the air. François Arago had proposed in 1838 a rotating-mirror experiment to compare the speed of light in air and water, which would decide between particle and wave theories of light. Fizeau and Foucault both took up the challenge, initially together and then as rivals. Fizeau chose a different mechanism, a toothed wheel, for his first measurement in air.
The design: light, telescopes and a spinning wheel
The apparatus had four main parts.
A light source. Fizeau used an intense lamp, with its light formed into a narrow beam using lenses and a partially reflecting glass plate set at an angle. The plate sent the beam forward toward the distant station while still letting the observer look back along the same line.
The toothed wheel. The wheel had 720 teeth, with gaps of the same width between them, around its rim. It was placed at the focus of the outgoing beam, where the beam was narrowest, so that a single tooth could block it completely. A clockwork mechanism driven by weights, with a gear train and a counter, spun the wheel at speeds that could be raised steadily and measured.
Two telescopes. One telescope at Suresnes sent the chopped beam out as a near-parallel bundle of light. A second telescope at Montmartre collected it and brought it to a focus on a mirror, which sent it straight back along the same path. Using telescopes at both ends kept the beam concentrated over the 8.6-kilometre distance, which open lamplight could never do.
The observer's eye. Looking through the partially reflecting plate, Fizeau watched the returning light as a point of brightness that passed back through the wheel.
The two stations, both elevated, gave a clear line of sight across the city. Fizeau's father had a house at Suresnes, which served as the sending station.
How the method works
Picture the beam passing out through a gap at the edge of the wheel. It travels to Montmartre, reflects and returns, arriving back at the wheel after a round trip of about 17.3 kilometres. What it meets on its return depends on how far the wheel has turned in the meantime.
- Wheel stationary or slow: the gap has barely moved, so the light passes back through it and the observer sees a bright point.
- Faster: the gap moves partway on, and the returning light is partly blocked. The image dims.
- First eclipse: the wheel has turned by exactly half of one tooth-and-gap period, so a tooth sits where the gap was. The returning light is blocked completely.
- Twice that speed: the next gap has arrived. The light returns.
- Three times the first speed: the next tooth. Another eclipse.
| Rotation rate | What the returning light meets | Observed |
|---|---|---|
| Below 12.6 rev/s | Mostly the same gap | Bright, dimming as speed rises |
| 12.6 rev/s | Next tooth | First eclipse |
| 25.2 rev/s | Next gap | Bright again |
| 37.8 rev/s | Following tooth | Second eclipse |
As the wheel spins faster, the returning light alternately meets teeth and gaps.
The method turns a time too short to measure directly into a rotation rate that can be counted with ordinary clockwork.
The calculation, step by step
Let N be the number of teeth (720), d the one-way distance (8,633 m) and f the rotation rate at the first eclipse (12.6 revolutions per second).
Step 1: the angle the wheel must turn. One full tooth-and-gap period is 1/720 of a revolution. The first eclipse happens when the wheel turns half of that, so 1/(2N) = 1/1,440 of a revolution.
Step 2: the time that takes. At f revolutions per second, a turn of 1/1,440 revolution takes:
t = 1 / (2N × f) = 1 / (1,440 × 12.6) s = 1 / 18,144 s ≈ 55.1 microseconds
Step 3: the distance the light covers in that time. The round trip is 2d = 17,266 m.
Step 4: the speed.
c = 2d / t = 2d × 2N × f = 4 N d f
c = 4 × 720 × 8,633 m × 12.6 s⁻¹ = 313,274,304 m/s
That is about 313,300 km/s. The modern value for light in vacuum is 299,792,458 m/s, so Fizeau's figure is about 4.5 percent high.
Notice how small the critical time is: light crosses the 17-kilometre round trip in about 55 millionths of a second. Fizeau never measured that interval directly. He measured a wheel's speed, a count of teeth and a distance, and let geometry do the rest.
How Fizeau reported it
Fizeau announced the result to the Paris Academy of Sciences at its session of Monday, 23 July 1849, in a short note titled "Sur une expérience relative à la vitesse de propagation de la lumière." It occupies only about three pages (pages 90–92) of volume 29 of the Academy's Comptes rendus. The brevity is striking. Fizeau described the stations, the distance of 8,633 metres established by triangulation, the wheel and its teeth, and the rotation rate at which the first eclipse occurred, and gave the resulting speed. He also noted that the light reappeared at higher speeds and vanished again, the alternating pattern that confirmed he was seeing the effect he expected rather than an artefact.
The note did not include an extended error analysis of the kind a modern paper would require. Fizeau expressed his result in leagues per second, using the French league of 25 to a degree of latitude. Converted to modern units, it is usually quoted as roughly 313,000 to 315,000 kilometres per second; the spread comes from rounding of the rotation rate and the conversion of the old league. Historians rely on that short note, on later accounts by Cornu and others, and on the surviving descriptions of the apparatus.
Why a toothed wheel, and why so many teeth?
The design involves a trade-off between three quantities: the number of teeth, the baseline and the rotation rate. The formula c = 4Ndf shows that for a given speed of light, the product N × d × f is fixed. A short baseline demands either an enormous number of teeth or an impossibly fast wheel. Fizeau's 8.6-kilometre path was long enough that a wheel with several hundred teeth needed to spin only a dozen or so times a second — well within the reach of mid-nineteenth-century clockwork.
There were practical limits on the other side as well. Teeth cannot be made arbitrarily fine, because the beam focused on the rim has a finite width; if a tooth is narrower than the focused spot, it can never block the beam completely and there is no clean eclipse. And a longer baseline means more light is lost to haze, so the returning point of light becomes faint. Fizeau's choice of 720 teeth and the Suresnes–Montmartre line balanced these constraints. Cornu's later experiment showed how far the method could be pushed with a longer path, brighter illumination and electrical timing.
Fizeau's wider career
Fizeau's work on light was part of a broader career in experimental optics. In April 1845, with Léon Foucault, he made one of the earliest detailed photographs of the Sun, a daguerreotype that showed sunspots. In 1868 he suggested that interference of light from two separated parts of a telescope aperture could be used to measure the angular diameters of stars, an idea that later developed into stellar interferometry and that Michelson put into practice in the twentieth century. He also studied the thermal expansion of crystals with interference methods.
Fizeau was elected to the Academy of Sciences in 1860 and received the Royal Society's Rumford Medal in 1866. He died in 1896. His name appears among the 72 scientists inscribed on the Eiffel Tower.
Why the result was about 5 percent off
The largest source of error was not the distance or the wheel but judging the eclipse. Because the beam has a finite width and the teeth have finite size, the returning light does not switch off sharply at a single speed. It fades, reaches a minimum and then brightens. The observer has to decide by eye when the light is at its dimmest, and the faint glow of stray light makes that harder. A small error in the eclipse speed translates directly into a proportional error in c.
The difference between light's speed in air and in vacuum is far too small to explain the discrepancy. The refractive index of air near the ground is about 1.0003, so light in air travels only about 0.03 percent slower than in vacuum. That would make an accurate air measurement come out slightly low, not 4.5 percent high.
Why the measurement was groundbreaking
Before 1849, every value for the speed of light depended on astronomical distances. Fizeau's method used only a ruler-measured baseline on Earth, a counted number of teeth and a clockwork speed. Any well-equipped laboratory could, in principle, repeat it. It also confirmed independently that the astronomical values were in the right range, and it placed the speed of light among the quantities that could be measured and steadily improved, like the density of water or the acceleration due to gravity.
Follow-up work: Foucault, Cornu, Michelson
Foucault's rotating mirror. In 1850, Léon Foucault used a rapidly spinning mirror rather than a toothed wheel to compare the speed of light in air and in water over a short path inside a laboratory. He showed that light travels more slowly in water, as the wave theory predicted and the particle theory did not. Fizeau and Louis Breguet reported a similar result at almost the same time. In 1862 Foucault used the rotating-mirror method to measure the speed of light in air, obtaining about 298,000 km/s, much closer to the modern value than Fizeau's figure.
Cornu's refinement of the toothed wheel. Between 1872 and 1876, Alfred Cornu returned to Fizeau's method with major improvements: a path between the Paris Observatory and the tower at Montlhéry, south of Paris, nearly three times longer than Fizeau's, observation of many higher-order eclipses rather than only the first, and an electrical circuit that recorded each turn of the wheel on a chronograph against an observatory clock. Cornu's value, about 300,400 km/s, was within 0.2 percent of the modern figure.
Michelson. Beginning in 1878–1879, Albert Michelson refined Foucault's rotating-mirror technique and pushed the precision further over the following decades. The precision measurements of light's speed and the interferometry he developed led to the Michelson–Morley experiment and helped prepare the ground for relativity.
Fizeau's other famous experiment
In 1851 Fizeau sent light through water flowing in opposite directions in two tubes and looked for a shift in the interference pattern. If light were simply carried along by the water, its speed would increase by the full speed of the water. Instead, Fizeau found it increased by only a fraction of that speed, in line with a "partial drag" coefficient predicted earlier by Augustin-Jean Fresnel. The result puzzled physicists for decades. Its full explanation came from special relativity, in which velocities do not simply add, and Einstein cited Fizeau's moving-water experiment as one of the results that influenced his thinking.
From measurement to definition
Over the following century, the speed of light was measured with increasing precision using rotating mirrors, electronic shutters, microwave cavities and lasers. By 1975 it was known to be 299,792,458 m/s with an uncertainty of about 4 parts per billion. In 1983 the General Conference on Weights and Measures took the final step: it defined the metre as the distance light travels in vacuum in 1/299,792,458 of a second. Since then the speed of light has been exact by definition, and experiments that once measured c now, in effect, measure lengths.
Common misconceptions
The wheel was a stopwatch. Not exactly. The wheel did not time the light directly. It worked as a shutter whose rotation rate could be measured, and the geometry of teeth and gaps converted that rate into a time.
The wheel had 360 teeth and 360 gaps. It had 720 teeth, with 720 gaps between them. The factor of 2 in the formula comes from the fact that the first eclipse occurs after half a tooth-and-gap period, not a full one.
Fizeau's error came from measuring in air. Air slows light by only about 0.03 percent, and in the wrong direction to explain a high result. The error came mainly from judging the point of darkest eclipse by eye.
Fizeau was the first to measure the speed of light. Rømer had shown light's speed was finite in 1676, and Bradley had found another astronomical route in 1729. Fizeau made the first measurement over a known distance on Earth.
The experiment was done in a laboratory. The light path ran across several kilometres of open air between two buildings, which exposed it to haze and turbulence and limited the times it could be done.
Modern relevance
The principle of Fizeau's wheel survives in modern technology. Optical choppers, rotating slotted discs that switch a light beam on and off at a known frequency, are still common laboratory equipment. Laser rangefinders and lidar time short light pulses over a round trip, exactly the quantity Fizeau inferred from his wheel; satellite laser ranging measures the distance to orbiting reflectors to within millimetres. And the finite speed of light is something engineers must design around every day, from GPS satellites to the latency of data sent through fibre-optic cables. Our thought experiment on what would change if light were slower explores how much everyday life depends on that number.
The takeaway
Fizeau's toothed wheel turned the speed of light from an astronomical inference into a laboratory measurement. With 720 teeth, an 8,633-metre baseline and a first eclipse at 12.6 turns per second, the simple formula c = 4Ndf gave him 313,000 km/s, about 4.5 percent high. The idea was more important than the number. Within three decades Foucault, Cornu and Michelson had pushed the precision to a few parts in ten thousand, and a little over a century later the speed of light was fixed by definition. A wheel spinning in a house in Suresnes was the first step on that path.
Sources
- Fizeau's measurement of the speed of light in air — Wikipedia — Stations, distance, teeth count, rotation rates, result, Cornu's refinement
- Resolution 1 of the 17th CGPM (1983) — BIPM — Definition of the metre in terms of the speed of light
- Fizeau, H. (1849). "Sur une expérience relative à la vitesse de propagation de la lumière." Comptes rendus hebdomadaires des séances de l'Académie des sciences, 29, 90–92.
- Cornu, A. (1876). "Détermination de la vitesse de la lumière d'après des expériences exécutées en 1874 entre l'Observatoire et Montlhéry." Annales de l'Observatoire de Paris.
- Sorbonne Université heritage collections, notes on Fizeau's toothed-wheel apparatus.
- Experimental Determination of the Velocity of Light — A. A. Michelson, via Project Gutenberg — Michelson's own account of the rotating-mirror method
FAQ
Q: Why did Fizeau use two telescopes? A: Over 8.6 kilometres an ordinary lamp beam would spread out and fade. The telescope at Suresnes sent out a concentrated beam, and the one at Montmartre gathered it and focused it on a mirror that returned it along the same path, so enough light came back to be seen through the spinning wheel.
Q: How fast was the wheel spinning at the first eclipse? A: 12.6 revolutions per second, or about 756 revolutions per minute. With 720 teeth, a turn of half a tooth-and-gap period took about 55 microseconds, which matched the time light needed for the 17.3-kilometre round trip.
Q: How do you calculate the speed of light from Fizeau's numbers? A: Use c = 4Ndf, where N is the number of teeth (720), d is the one-way distance (8,633 m) and f is the rotation rate at the first eclipse (12.6 per second). The result is 313,274,304 m/s.
Q: Why was Fizeau's value about 5 percent too high? A: Mainly because the eclipse is gradual and had to be judged by eye. The returning light fades and brightens over a range of speeds, and picking the darkest point introduces error. Light being slightly slower in air than in vacuum is too small an effect to account for it and would push the result the other way.
Q: Did other scientists replicate Fizeau's experiment? A: Yes. Alfred Cornu repeated and improved the toothed-wheel method between 1872 and 1876 and obtained about 300,400 km/s, within 0.2 percent of the modern value. Léon Foucault's rotating-mirror method gave about 298,000 km/s in 1862, and Michelson refined that approach further.
Q: Is the speed of light still measured today? A: Not in the old sense. Since 1983 the metre has been defined as the distance light travels in vacuum in 1/299,792,458 of a second, so the speed of light in vacuum is exactly 299,792,458 m/s by definition. Precision experiments now use it to realise lengths.
Read next:
Follow new articles
Add SciHub101 to your feed reader to see new science and technology stories.
About the Author
SciHub101 is an independent science and technology publication founded and edited by Buğra Sözeri, a software engineer and app developer based in Türkiye. AI tools assist with drafting and research. Source documentation in the existing archive is uneven, and articles are being reviewed and corrected. The Editorial Team byline identifies editorial responsibility; it does not certify specialist or independent human review. Where we compare products, anything built by our own team is labelled as such — see the editorial guidelines and About page for how corrections and disclosures work.
Related Articles
Complete Science Experiments Guide 2026 — Every Landmark Experiment by Field
Every landmark science experiment in one hub: Pasteur's swan-neck flask, Galvani's frog legs, Eratosthenes, Michelson–Morley, Hafele–Keating and more than 140 others, grouped by field and era.

Hess's Balloon Flights: How Cosmic Rays Were Found (1912)
In 1912 Victor Hess carried electroscopes to 5,350 m in a balloon and found radiation rising with height. Here is how the flights worked and what they proved.