Special relativity had redefined motion and time in 1905, but left gravity out. Einstein worked on that gap for a decade and published the field equations in 1915: mass curves spacetime, and bodies follow the shortest path through that curved geometry. Newton's two-hundred-year-old force of attraction became a consequence of geometry.

The theory offered a testable prediction: starlight passing near the Sun should bend by twice the Newtonian amount. The only way to see it was a total solar eclipse. Arthur Eddington and Frank Dyson sent two teams — one to Sobral in Brazil, the other to the island of Príncipe off West Africa. The star positions on the plates taken on 29 May 1919 were compared with the positions of the same stars in the night sky; the deflection matched Einstein's value.

The announcement in London that November turned Einstein overnight into a globally known figure; that a British team had confirmed a German theory so soon after the First World War added political weight to the news. The measurement's error bars were debated then and are debated now — some of the Sobral plates were discarded for a focus problem. But over the following century radio interferometry, the timing corrections in GPS, gravitational lensing and the gravitational waves measured in 2015 confirmed the theory to far greater precision. 1919 endures less for its result than as the example of staking a theory's fate on a single measurement.