In September 1676, a young Dane stood before the Royal Academy of Sciences in Paris and made a public bet with the sky. Ole Rømer announced that the eclipse of Io — the innermost of Jupiter's four great moons — expected on 9 November would begin about ten minutes later than the best astronomical tables predicted. When the November night came, observers at the Paris Observatory duly watched Io slip into Jupiter's shadow late, much as promised. Weeks later Rømer stood up again and delivered his explanation: light, which almost every authority from Aristotle to Descartes had declared instantaneous, in fact takes time to travel. The eclipse looked late because the Earth had drifted farther from Jupiter since the summer, and the news of the eclipse had farther to come.
Io was well suited to the role of cosmic stopwatch. Galileo had discovered it, along with three companions, in 1610, and it whips around Jupiter every 42 and a half hours, vanishing into the planet's shadow with clockwork regularity — hundreds of observable eclipses a decade. Galileo himself had once tried to time light's passage with shuttered lanterns on neighbouring hilltops and got nowhere; the baseline was absurdly short. Jupiter offered a longer racetrack. The eclipse timetable also had serious practical value: Giovanni Domenico Cassini, director of the Paris Observatory, promoted the moons of Jupiter as a celestial clock that surveyors could use to fix longitude, one of the great unsolved problems of the age. Rømer, born in Aarhus in 1644, had impressed the visiting French astronomer Jean Picard during a survey at Tycho Brahe's ruined island observatory of Uraniborg, and was carried off to Paris in 1672 to join the effort.
Poring over years of timings, Rømer saw a pattern the tables could not explain. The eclipses ran progressively early in the months when the Earth was closing on Jupiter, and progressively late as it swung away. His inference was that the light itself was in transit: he reckoned it needed about 22 minutes to cross the diameter of the Earth's orbit — the modern figure is close to 17 — and, since the orbit's size was then only roughly known, he prudently published no speed in miles or leagues. Christiaan Huygens did that sum, arriving at a figure equivalent to about 220,000 km per second: some three-quarters of today's exact value of 299,792.458 km per second, a constant now so fundamental that the metre itself has been defined from it since 1983.
The doubting director
The discovery's chief opponent worked in the same building. Cassini had himself briefly floated a light-delay explanation for the anomalies before abandoning it, and he pressed the counter-arguments hard: perhaps Io's orbit was simply irregular, and why did the other three moons show no such tidy effect? His authority was immense, and scepticism outlived both men. Broad acceptance came only in 1729, when the English astronomer James Bradley discovered the aberration of starlight — a tiny seasonal tilt in the apparent positions of stars that made sense only if light moved at a fast but finite speed, and which yielded a value agreeing handsomely with the eclipse method. Rømer had died in 1710, nineteen years short of his vindication.
Body heat
Rømer, by then, had long since gone home and become one of history's more improbable civic titans. Back in Copenhagen from 1681, he served as professor of astronomy, royal mathematician, master of the mint, the city's chief of police and eventually its mayor, introducing street lighting along the way and inventing the transit instrument that became a staple of observatories. Around 1701 he devised one of the first temperature scales anchored to two fixed calibration points, with zero pinned near the freezing point of brine, water freezing at 7.5 degrees and boiling at 60. In 1708 a young instrument-maker named Daniel Gabriel Fahrenheit visited him in Copenhagen, watched him calibrate thermometers, and went away to refine the scheme — rescaling the numbers to shed the awkward fractions. The scale on which an American thermometer reads healthy body heat as 98.6 degrees descends, at its root, from Rømer's.
Most of his papers burned in Copenhagen's great fire of 1728, so we know much of his work only through the hands of others. Two legacies escaped the flames comfortably: the fever thermometer at your bedside, and the speed of light itself — first caught in the act by a man watching a distant moon go dark.
Quiz nuggets
- In 1676 the Danish astronomer Ole Rømer gave the first demonstration that light travels at a finite speed, using eclipses of Jupiter's moon Io.
- Rømer estimated light took about 22 minutes to cross the diameter of Earth's orbit; the modern figure is close to 17 minutes.
- Cassini, Rømer's director in Paris, rejected the finding, and James Bradley's 1729 discovery of stellar aberration finally settled the question.
- Rømer later became Copenhagen's chief of police and mayor, and devised the temperature scale that Fahrenheit adapted after visiting him in 1708.
- The speed of light is exactly 299,792.458 km per second, and the metre has been defined from it since 1983.