05. The Speed of Light — Late Moons

A clock in the sky that ran slow

Jupiter's moon Io passes into its planet's shadow with great regularity, which made it, in the seventeenth century, one of the best clocks available. In 1676 Ole Rømer, working at the Paris Observatory, noticed that the eclipses did not arrive on schedule across the year. When the Earth was moving away from Jupiter the eclipses came late; when moving towards it they came early, and the pattern followed the geometry of the two orbits. Rømer's explanation was that light takes time to cross space, so the delay measured the extra distance the light had to travel. He announced that light needed roughly twenty-two minutes to cross the diameter of the Earth's orbit, and used it to predict a specific eclipse would run late — which it did. That prediction, verified in public, is what made the argument stick.

The speed he never stated

Rømer published a light-time, not a speed, and the distinction is the whole historical dispute. Converting his figure into a velocity requires knowing the diameter of the Earth's orbit, which nobody knew well in 1676. Christiaan Huygens performed that conversion in 1690, and the number usually attributed to Rømer in textbooks is therefore partly Huygens' arithmetic and partly a seventeenth-century guess at the size of the solar system. Terrestrial measurement arrived with Hippolyte Fizeau in 1849, who sent a beam through the gap in a spinning toothed wheel to a distant mirror and back, adjusting the speed until a returning pulse was blocked by the next tooth. Albert Michelson spent decades refining rotating-mirror versions of the same idea into the twentieth century. In 1983 the process ended: the metre was redefined by fixing the speed of light at exactly 299,792,458 metres per second.

When a measurement becomes a definition

Once a quantity is known better than the unit it is expressed in, measuring it more precisely stops meaning anything. The speed of light is now exact by decree, and every future improvement in the experiment refines the metre instead. This inversion — the constant fixed, the unit floating — is the pattern the last chapters of this book are built on, and 1983 is where it started.

Example: Arguing about the length of a metre stick is useful until the stick is what defines a metre, at which point the argument is about your ruler.

Verdict. Rømer established that light has a finite speed and measured the time it takes to cross Earth's orbit; the velocity credited to him required Huygens and a poorly known orbital diameter, and since 1983 the number has been a definition rather than a result.

Overturned if: a manuscript shows Rømer stating a velocity in his own units, which would restore the textbook attribution.

Sources. 2026 · "Rømer's determination of the speed of light" — Wikipedia · 2009 · "Ole Rømer and the Speed of Light," Optics & Photonics News — Optica