How We First Timed Light

Step 1 of 5

Everyone assumed light took no time at all.

Galileo had tried to measure it with two men, two lanterns and two hilltops, and found only that light was faster than anyone could time by hand. Most people concluded it was simply instantaneous.

Step 2 of 5

Jupiter's moon Io is a very good clock.

Io goes round Jupiter every 42.5 hours and disappears into Jupiter's shadow each time. That is a bright, sharp, repeating event, visible through a small telescope, and utterly regular — the best clock in the sky.

Step 3 of 5

Except it was not regular. It drifted.

Rømer kept records for years and found the eclipses ran progressively early for half the year and progressively late for the other half. Nothing about Io was changing. What was changing was where Earth was standing.

Step 4 of 5

The lateness is the extra distance.

If light takes time, its journey is longer when Earth is on the far side of its orbit. The extra distance is the whole width of that orbit, and the light arrives that much later. The total drift across the year was about sixteen minutes.

Step 5 of 5

Divide the distance by the lateness.

That is the whole calculation. Rømer's own answer was around 220,000 kilometres per second — low, because the size of Earth's orbit was not yet well known. The method was right. The real figure is 299,792.

He was ignored for fifty years. The clinching confirmation came in 1729, from a completely different method, long after he died.