On the morning of 29 May 1919, on the tiny volcanic island of Príncipe in the Gulf of Guinea, Arthur Eddington watched a tropical storm threaten to ruin the most important five minutes of his life. He had travelled thousands of miles to photograph a total eclipse of the Sun, not for its beauty but for a single number: how far the light of distant stars would appear to bend as it grazed the Sun's edge. On that number rested the fate of a theory almost nobody in Britain yet understood.
The theory was Albert Einstein's general relativity, completed in 1915 in the middle of a war that had cut German science off from the Allied world. Einstein proposed that gravity is not a force pulling across empty space but a curvature of spacetime itself, and that mass tells space how to bend. One of its boldest predictions was that starlight passing close to a massive body would be deflected — and crucially, deflected by twice the amount that Isaac Newton's older physics allowed. Newton's corpuscles of light, tugged by the Sun's gravity, gave a deflection of about 0.87 arcseconds. Einstein's curved spacetime predicted roughly 1.75 arcseconds, twice as much. The two theories disagreed by a factor no telescope could mistake — if only you could see stars right next to the Sun.
Waiting for the Moon
You cannot, of course, because the Sun drowns them out. The one moment you can is during a total solar eclipse, when the Moon blots out the disc and the star field around it springs into view. The eclipse of May 1919 was ideal: the Sun would sit in front of the bright Hyades star cluster, giving plenty of stars to measure. Frank Dyson, the Astronomer Royal, organised two expeditions to be safe — Eddington's to Príncipe off West Africa, and a second, led by Andrew Crommelin, to Sobral in northern Brazil.
There was a deeper poetry to the choice of Eddington. A devout Quaker and pacifist, he had refused to fight in the First World War and nearly been sent to a labour camp; it was partly to keep so gifted an astronomer at his work that the eclipse expedition was arranged. Here was an Englishman crossing the world to test, and possibly to vindicate, the work of a German at a moment when the two nations had just finished slaughtering each other. Eddington saw the mission as a gesture of scientific brotherhood above the trenches.
The clouds over Príncipe parted just in time. Eddington's team captured a handful of usable plates; the Sobral party, under clearer skies, got more. Then came months of painstaking measurement, comparing the eclipse photographs against reference plates of the same stars taken at night, hunting for the tiny outward shift of star positions near the darkened Sun. The Sobral telescope's best instrument gave a deflection close to Einstein's 1.75; Eddington's Príncipe plates, though fewer and messier, agreed. A third, poorer set of Sobral plates was set aside as unreliable — a decision later critics questioned as convenient, though modern reanalysis has broadly supported the conclusion.
Revolution in Science
On 6 November 1919, at a joint meeting of the Royal Society and the Royal Astronomical Society in London, under a portrait of Newton, Dyson announced that the measurements favoured Einstein. The theory that had upended Newton's two-century reign had passed its first great observational test. The next morning The Times of London ran the headline "Revolution in Science — New Theory of the Universe — Newtonian Ideas Overthrown." The New York Times followed with talk of "lights all askew in the heavens."
Overnight, Einstein became the most famous scientist alive, a global celebrity in a way no physicist had been before — his tangled hair and gentle gaze turning into the twentieth century's shorthand for genius. He is said to have been asked how he would have reacted had the eclipse contradicted him, and to have replied that he would have felt sorry for the dear Lord, because the theory was correct anyway.
The measurements themselves were, by later standards, crude, with error bars uncomfortably large; sceptics have argued the 1919 data alone did not truly settle the matter. But repeated eclipse expeditions through the following decades, and later radio measurements of quasars, confirmed the deflection to ever finer precision. Gravitational lensing — the bending of light by mass that Eddington glimpsed at the Sun's rim — is now a routine tool of astronomy, used to weigh galaxies and map invisible dark matter.
The storm-broken sky over an equatorial island had done what no lecture hall could: it turned an abstract equation into a picture of stars nudged out of place, and made bent starlight the most persuasive argument in physics.
Quiz nuggets
- The eclipse expeditions of 29 May 1919 were led by Arthur Eddington (to Príncipe) and organised with a second team at Sobral, Brazil.
- They tested Einstein's general relativity by measuring how much starlight bent near the Sun.
- Einstein predicted a deflection of about 1.75 arcseconds — twice the value allowed by Newtonian gravity.
- Results were announced on 6 November 1919; The Times headline read "Revolution in Science."
- The bending of light by gravity is now known as gravitational lensing, a standard tool of modern astronomy.