Ole Rømer's Speed of Light from Io's Eclipses Explained
In 1676 the Danish astronomer Ole Rømer told the French Academy of Sciences that light does not travel instantaneously. His evidence came not from the laboratory but from the motion of the heavens. For years, Rømer had been timing the eclipses of Jupiter's moon Io at the Paris Observatory—moments when Io passed behind Jupiter, disappearing from view in a planetary shadow. He noticed something odd: the timing of these eclipses drifted depending on whether Earth was moving toward or away from Jupiter. The drift accumulated over months, and by the time of the November 9, 1676 eclipse he predicted a delay of about 10 minutes relative to the published tables, and the eclipse came as predicted. Rømer concluded that the drift was caused by light taking time to cross the changing distance between Earth and Jupiter. An account of his announcement stated that light takes approximately 22 minutes to traverse the diameter of Earth's orbit. This was the first quantitative estimate of light's finite travel time, obtained not with a laboratory apparatus but with patient observation of the heavens and mathematical deduction.
The puzzle: does light move instantaneously?
In the 1670s, many natural philosophers, including Descartes, had held that light travels instantaneously. A few had speculated that light might move at a finite speed, and Galileo had tried to test it with lanterns on distant hills, but no one had decisive evidence. Most experiments with light were local, offering no insight into its speed over large distances. Astronomical observations offered the only way to measure over large scales, but astronomers had not yet made systematic enough observations to detect an effect of light's finite travel time.
Rømer was meticulous. At the Paris Observatory, observing Jupiter's moons was routine work connected to the practical problem of determining longitude. But the data revealed something unexpected.
The observation: timing Io's disappearances
Io — the innermost of Jupiter's four large moons discovered by Galileo; also called Jupiter I — orbits Jupiter roughly every 42.5 hours and passes behind Jupiter once per orbit, creating an eclipse observable from Earth. Rømer could time these disappearances using the Paris Observatory's clocks and telescopes.
Rømer noticed that the times of Io's eclipses, recorded over many months, did not follow a simple, perfectly periodic schedule. The time between successive eclipses varied slightly depending on whether Earth was moving toward or away from Jupiter. When Earth was moving toward Jupiter (Earth and Jupiter growing closer), the eclipse times seemed to occur earlier than expected. When Earth was moving away from Jupiter, the eclipses seemed to occur later.
The effect on any single eclipse was tiny, only seconds, but it accumulated over many orbits into minutes, and Rømer's records, which included roughly 140 observations of Io eclipses between 1671 and 1677, were extensive enough to catch it. Other astronomers had recorded Io eclipses, but Rømer interpreted the long-term drift as a consequence of light's travel time.
The deduction: light takes time to cross space
In 1676, Rømer presented his hypothesis to the French Academy: the variation in eclipse times was caused by light's finite travel speed. When Earth was farthest from Jupiter, the light carrying the image of Io's disappearance had to travel a greater distance, so it arrived later. When Earth was closest, the light had less distance to travel, arriving earlier. The difference between the farthest and closest positions was twice Earth's orbital radius—the diameter of Earth's orbit around the Sun.
Rømer calculated that the difference in light travel time between Earth's farthest and closest approach to Jupiter was approximately 22 minutes. From this, he could estimate the speed of light. He knew the diameter of Earth's orbit (approximately 300 million kilometers by modern measurements, though 17th-century estimates were cruder). Dividing this distance by the time light takes to cross it:
Speed of light ≈ (2 × Earth's orbital radius) / (22 minutes)
Rømer himself never published a speed in units of distance per time; he stated the delay. Christiaan Huygens combined Rømer's 22 minutes with the contemporary estimate of the Earth–Sun distance in his Treatise on Light (1690) and arrived at a figure equivalent to roughly 212,000–220,000 kilometers per second. The modern value is 299,792 kilometers per second in vacuum. That estimate was low by about 25–30 percent, but it was in the right ballpark—a quantitative value rather than pure speculation. (Some modern sources quote slightly different values for Huygens's figure, depending on the units and Earth–Sun distance assumed.)
Source: "Démonstration touchant le mouvement de la lumière trouvé par M. Römer," Journal des sçavans, 7 December 1676, which reports the 22-minute figure for light to cross a distance equal to the diameter of Earth's orbit.
Why this mattered: proving light is not instantaneous
Rømer's discovery resolved an ancient question: light does not travel infinitely fast. It has a measurable, finite speed. This simple fact had profound implications. If light takes time to travel, then observations of distant objects—the sun, stars, planets—show them as they were in the past, not as they are at the moment of observation. This retroactivity was philosophically unsettling to some but scientifically liberating. It meant that observations of distant objects were observations of history, and by comparing observations taken at different times, astronomers could infer past states of the cosmos.
Rømer's method was also conceptually important. He showed that you did not need a laboratory apparatus to measure light's speed. You could use the cosmos itself as a laboratory, if you made careful observations and had the mathematical insight to recognize the pattern hidden in the data. Astronomical observations could yield fundamental physical constants.
The speed of light that Rømer estimated was close enough that later astronomers recognized it as genuine progress. Even though Rømer's value was corrected over the years (as better measurements of Earth's orbit became available), the underlying principle was sound.
Common misconceptions about Rømer's method
Many students mistakenly think Rømer directly measured the time light takes to travel from Jupiter to Earth. In fact, he measured the difference in eclipse timing between Earth's two extreme positions (closest and farthest from Jupiter) and from that inferred the speed. The beauty of the method is that by using Earth's orbital motion, Rømer created a natural experiment where the light-travel distance changed in a measurable way, revealing the effect of that change on observation timing.
A second misconception is that Rømer's value for the speed of light was immediately accepted as correct. In fact, some astronomers doubted his interpretation. Giovanni Cassini, Rømer's colleague at the Paris Observatory, initially resisted the idea that light had finite speed. The acceptance grew slowly as more observations accumulated and as Rømer's prediction—that the November 9, 1676 eclipse would be 10 minutes late—proved accurate.
A third error is confusing Rømer's method with parallax measurements or stellar aberration. These are different astronomical techniques for measuring distances and detecting Earth's motion. Rømer was concerned with the time light takes to cross a known distance; parallax measures the shift in an object's apparent position from different vantage points, and aberration measures the apparent shift in a star's position due to Earth's orbital motion and light's finite speed. All three involve light's interaction with Earth's motion, but the physics is distinct.
The legacy: from eclipses to lasers
Rømer's discovery that light has finite speed opened the door to measuring it in other ways. Fizeau, with his toothed-wheel apparatus in 1849, made the first direct terrestrial measurement. Foucault followed with a rotating mirror. Modern methods use lasers, electronic circuits, and atomic clocks, and since 1983 the speed of light has been fixed by definition. But the underlying insight—that light speed can be measured, that it is a fundamental constant of nature—originated with Rømer's patient observation of Io's eclipses.
Rømer's work also pioneered the use of Jupiter's moons as cosmic clocks. Later astronomers used the moons' orbital periods and eclipse timings for navigation and timekeeping. The precision demanded by these applications drove improvements in both observational astronomy and theoretical mechanics, contributing to the broader advancement of science.
In the 20th century, Einstein elevated the speed of light to a central role in physics: the maximum speed at which any signal or object can travel, and a constant in all reference frames. But before Einstein, before Newton, before modern physics existed, Ole Rømer discovered that light has a speed and estimated it from the heavens. This was profound: a fundamental property of the universe revealed by patient observation and mathematical insight.
Who Ole Rømer was, and why he was in Paris
Ole Christensen Rømer (1644–1710) was born in Aarhus, Denmark, and studied at the University of Copenhagen under the mathematician and physician Rasmus Bartholin, the scientist who described double refraction in Iceland spar. In 1671 the French astronomer Jean Picard travelled to Denmark to fix the exact position of Tycho Brahe's old observatory, Uraniborg, on the island of Hven. Knowing Uraniborg's longitude relative to Paris would let astronomers combine Tycho's sixteenth-century observations with new ones. Rømer worked as Picard's assistant, and the two observed eclipses of Jupiter's moons from Hven while Giovanni Domenico Cassini observed the same events from Paris. Comparing the local clock times of the same eclipse gave the difference in longitude.
Picard was impressed enough to bring the young Dane back to Paris in 1672. There Rømer joined the new Royal Observatory, founded under Louis XIV, worked with Cassini, and tutored the Dauphin. He also designed instruments and helped engineer the fountains at Versailles. The Jupiter work was routine observatory business, connected to the most pressing applied problem of the age: finding longitude.
Why Jupiter's moons were a clock
Galileo had proposed around 1612 that the eclipses of Jupiter's moons could serve as a clock visible from anywhere on Earth. If an astronomer had a table predicting when Io would vanish into Jupiter's shadow as seen from Paris, then an observer elsewhere could note the local time of the same event. The difference between the two times gives the difference in longitude, at one hour per 15 degrees. The method worked well on land, and it was used to redraw the map of France with much greater accuracy. At sea it was nearly useless, because a long telescope could not be held steady on a rolling deck.
Cassini published improved tables of the moons in 1668. To make tables like that, astronomers had to assume Io's orbit was perfectly regular. The trouble began when the observed eclipse times kept drifting away from the tables in a way that followed the seasons, or more precisely, followed where Earth was in its orbit relative to Jupiter.
The 1676 prediction and the Academy announcement
In September 1676, Rømer told the Academy that the eclipse of Io expected on 9 November would come about ten minutes later than the tables, based on an extrapolation from earlier observations. When the eclipse was observed on that date, the emergence was indeed late by roughly the predicted amount. Rømer explained the delay as the time light needed to cross the extra distance between Earth and Jupiter, which had grown since the earlier observations.
The account usually cited is not a paper Rømer signed. An unsigned account titled "Démonstration touchant le mouvement de la lumière" appeared in the Journal des sçavans on 7 December 1676, based on his presentation. It stated that light takes about 22 minutes to cross the diameter of Earth's orbit. An English translation appeared in the Royal Society's Philosophical Transactions in 1677.
Ingress and egress: the clever part of the method
Rømer noticed that the intervals between successive eclipses were systematically shorter while Earth approached Jupiter and longer while it receded. Each individual difference is small, on the order of seconds per orbit, but the effect accumulates across dozens of orbits. Summing the delays over the months in which Earth moves from near Jupiter to far from it gives a total of many minutes.
That is why patience mattered. A single eclipse tells you nothing about the speed of light. A run of many eclipses, timed against a reliable pendulum clock of the kind Christiaan Huygens had recently developed, reveals a systematic drift. Rømer's own records included roughly 140 observations of Io eclipses.
Cassini's objection and competing explanations
Cassini had himself briefly entertained a finite speed of light in 1676, then rejected it. His objection was reasonable: if the delays were caused by light, the same pattern should appear in the eclipses of Jupiter's other moons. The data for the outer moons did not show it cleanly. Rømer and his supporters could not yet answer that, because the outer moons' orbits are perturbed by one another in ways no one could calculate. The orbital resonance linking Io, Europa and Ganymede was worked out only by Pierre-Simon Laplace more than a century later.
Cassini's nephew Giacomo Filippo Maraldi continued to argue against the interpretation into the early 1700s. The debate was therefore genuinely open. Within France the finite-speed idea was not generally adopted for decades, while in England, Isaac Newton accepted it and wrote in his Opticks (1704) that light takes about seven or eight minutes to travel from the Sun to Earth. Edmond Halley also supported it.
Confirmation: Bradley's aberration of starlight
Decisive independent support came in 1729 from the English astronomer James Bradley. While trying to measure stellar parallax, he found instead that stars appear to shift in small ellipses over the course of a year by about 20 arcseconds. He explained this as aberration: because Earth moves, starlight appears to arrive at a slight angle, much as vertical rain appears slanted to someone running. The size of the effect depends on the ratio of Earth's orbital speed to the speed of light. Bradley's analysis implied that light takes about 8 minutes 12 seconds to cross from the Sun to Earth, close to the modern figure of about 8 minutes 19 seconds.
Because Bradley's method was completely different and did not depend on Jupiter's moons, it settled the question that light has a finite speed. Later, the French astronomer Jean-Baptiste Delambre analysed about 150 years of Jupiter satellite observations and arrived at a light-time from the Sun of a little over eight minutes, showing that Rømer's approach, given enough data, converged on the right answer.
Where the 22 minutes went wrong
The modern light-travel time across Earth's orbit is about 16 minutes 40 seconds, not 22. Two things explain the gap. The timing of eclipses is uncertain because Io does not vanish instantly; it fades as it moves into the shadow, and the moment of disappearance depends on the telescope and on the observer's eye. And the approach is sensitive to Jupiter's own motion and to the irregularities of Io's orbit, which were not fully understood. The scale of the Solar System was a separate problem. In 1672 Cassini and Jean Richer had measured the parallax of Mars and derived an Earth–Sun distance of about 140 million kilometres, which was actually quite good. Error in the speed estimate came more from the time delay than from the distance.
Rømer's later career
Rømer returned to Copenhagen in 1681 as professor of astronomy and royal mathematician. He became a busy public official, serving as chief of police of Copenhagen from 1705, and he was involved in reforming the Danish system of weights and measures. He devised a thermometer scale, later adapted by Daniel Gabriel Fahrenheit, who visited him in 1708. He also built a transit instrument and a meridian circle for precise timing of star positions. Most of his observational notebooks were destroyed in the great Copenhagen fire of 1728, so historians rely heavily on the work of his student Peder Horrebow and on the French records.
Try it yourself with Jupiter's moons
Jupiter's Galilean moons are visible in a small telescope or steady binoculars. Eclipse times are published in astronomical almanacs and planetarium software such as Stellarium. Amateur astronomers have repeated Rømer's measurement by timing many eclipses or occultations of Io across several months, then comparing early-season timings, when Earth is near Jupiter, with late-season timings, when it is farther away. With careful timing to within several seconds, they recover a delay consistent with light crossing the extra distance, and a speed of light within a few percent of the true value. A simpler classroom version uses published ephemeris data instead of personal observations: students plot predicted versus listed eclipse times against the Earth–Jupiter distance and read the speed of light off the slope.
The takeaway
Rømer's discovery that light has finite speed came not from laboratory experiments but from watching a moon's regularly recurring disappearance and noticing when the eclipse timings did not perfectly match the predictions. By recognizing that the variations depended on Earth's distance from Jupiter, Rømer deduced that light must take time to travel across space. The roughly 22 minutes reported for light to cross Earth's orbit was the first quantitative estimate of light's travel time, from which Huygens derived a speed. The method was ingenious: use Earth's orbital motion as a ruler, measure the effect on light's arrival time, and from that deduce its speed. Three centuries later the speed of light is no longer measured at all in the strict sense: since 1983 it has been fixed by definition at exactly 299,792,458 metres per second, and the metre is defined from it.
FAQ
Q: How did Rømer know the exact distance across Earth's orbit to calculate light speed?
A: Rømer used astronomical measurements of Earth's orbital distance available in the 17th century. Estimates of the Earth–Sun distance, such as the roughly 140 million kilometres derived from the 1672 Mars parallax observations, were fairly good. The speed derived from Rømer's delay came out about 25 to 30 percent low, mainly because the 22-minute delay is larger than the true light-crossing time of about 16.6 minutes.
Q: Why did other astronomers initially doubt Rømer's interpretation?
A: The idea that light had finite speed contradicted the intuition that light is instantaneous. Cassini, for example, objected that the effect should appear in the eclipses of Jupiter's other moons, and the data for those moons were not clean. Rømer's 1676 prediction that the November eclipse would be late provided strong evidence for his interpretation—an accurate prediction that confirmed his hypothesis.
Q: Did Rømer's method measure light speed in vacuum or in air?
A: Rømer's method concerned light traveling through interplanetary space, so it relates to the speed of light in vacuum (c = 299,792,458 m/s). The low value derived from his delay owed more to the uncertain eclipse-timing delay than to the Earth–Sun distance, which was already fairly good.
Q: Why does Io eclipse Jupiter instead of other moons?
A: Io, Europa and Ganymede are eclipsed on every orbit, and Callisto is eclipsed during part of Jupiter's orbit but can pass above or below the shadow at other times. Io has the shortest orbital period (~1.77 days, or 42.5 hours) and orbits closest to Jupiter, making its eclipses frequent and relatively easy to observe from Earth. Rømer focused on Io, but the other moons show similar patterns.
Sources
- "Démonstration touchant le mouvement de la lumière trouvé par M. Römer." Journal des sçavans, 7 December 1676.
- Cohen, I. B. (1940). "Roemer and the First Determination of the Velocity of Light (1676)." Isis, 31(2), 327–379.
- Huygens, C. (1690). Traité de la lumière. Leiden.
- Bradley, J. (1729). "An Account of a New Discovered Motion of the Fix'd Stars." Philosophical Transactions, 35, 637–661.
- BIPM. The International System of Units (SI Brochure), 9th ed. (definition of the metre).
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


What If Jupiter Became a Star?
Jupiter is sometimes called a 'failed star' — but it would need to be 80 times more massive to ignite hydrogen fusion. If we somehow gave it enough mass, our solar system would gain a faint red dwarf companion, and Earth would barely notice.
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.