Ole Romer / The speed of light
Does light take time to arrive?
That time, that place
In the 1670s the Paris Observatory held the best telescopes in Europe. The most urgent practical problem for astronomers was knowing where you were at sea, which required a clock in the sky that anyone could read. The moments when Jupiter's moons slipped behind the planet served as that clock.
Why this question
Romer, a young Danish assistant, logged those moments for years, and the predicted times kept drifting from the observed ones.
The drift had a pattern. When Earth was moving away from Jupiter the eclipses ran late; when it was closing in they ran early. The full swing was twenty-two minutes.
A moon does not grow lazy with the seasons. So what was late was not the moon but the news.
What was found
Light has a speed. When Earth sits far from Jupiter the light must cross that extra distance, and the news arrives correspondingly late.
Romer read those twenty-two minutes as the time light needs to cross the diameter of Earth's orbit. A speed follows from that. His figure came out somewhat low, but the order of magnitude was right.
The number is not the point. This was the first observation to show that **light is not infinitely fast**.
Before this, most held that light arrives instantaneously. Galileo had tried stationing men on two hills with shuttered lanterns, but human reaction time swallowed the result. Any distance available on the ground was far too short.
Romer issued a prediction first: the next eclipse would run about ten minutes behind schedule. It did. He did not explain a result after the fact — he called it in advance and was right.
Even so, Cassini, who ran the observatory, and others resisted for a long time. Half a century later the apparent tilt of starlight caused by Earth's own motion gave the same conclusion by another route, and only then did it settle.