Io’s eclipses as a clock
Io disappears when it passes through Jupiter’s shadow and becomes visible again afterward. Astronomers recorded these eclipses and used their recurring period to predict subsequent events. Working on observations in Paris, Rømer found that this distant clock appeared alternately early and late.
Eclipse times predicted from a fixed period differed from observations, and the discrepancy accumulated as Earth changed position in its orbit.
The changing distance to Jupiter
As Earth approaches Jupiter, the path travelled by the light becomes shorter; as Earth recedes, it becomes longer. Rømer connected early and late eclipse observations with this change: light from the same kind of distant event takes different times to reach the observer.
The 1676 report estimated this delay using the scale of Earth’s orbit, discussing about 22 minutes for light to cross its diameter. Light propagation became a measurable astronomical interval.
The report published in December 1676
The Journal des sçavans published an account of Rømer’s work on December 7, 1676. It brought together satellite motion, Earth’s motion and the time at which a signal was received, explaining that light takes time to travel from Jupiter to Earth.
An English translation in Philosophical Transactions in 1677 preserved the orbital diagram and timing argument. Readers could compare the path travelled by light with eclipse observations at different positions of Earth and Jupiter.
A prediction tested by observation
Rømer’s method used recurring astronomical events to measure a propagation delay, avoiding the need for an immense experimental path on Earth. The interval between an event and its observation became part of the calculation.
The original report stated that Io’s emergence observed on November 9, 1676, was about ten minutes later than a prediction derived from August observations. Rømer had predicted the delay at the beginning of September. The observation supplied a further test of light taking time to travel.