JOVANA
Explore Library Glossary Getting Started Three Levels Fields How it works Mission
Join the mission
Back to the library
Physics 1919

A Determination of the Deflection of Light by the Sun's Gravitational Field

Frank Dyson, Arthur Eddington & Charles Davidson

Starlight grazing the Sun bends 1.75″ — just as Einstein said, twice what Newton allowed.

Choose your version
In depth · the introduction

In 1919, astronomers waited for the Moon to blot out the Sun — and caught starlight bending by exactly the amount Einstein had dared to predict.

The big idea

Einstein's general relativity says that mass bends the shape of space and time around it, and that light, taking the straightest available path, must follow that curve. So a ray of starlight skimming past the Sun should get bent slightly, and a star seen right next to the Sun should appear nudged a little away from it.

Einstein put a number on it: a star at the very edge of the Sun should shift outward by about 1.75 arcseconds — a hair's width of an angle, but twice as much as anyone could explain with Newton's older physics. The whole test came down to measuring that tiny shift, and seeing whether it matched Einstein's number or the smaller Newtonian one.

How it came about

The trouble is that you cannot see stars next to the Sun — its glare drowns them. The one exception is a total solar eclipse, when the Moon covers the Sun for a few minutes and stars pop out in a darkened daytime sky. The eclipse of 29 May 1919 was perfect: the hidden Sun would sit in a thick cluster of bright stars, the Hyades.

Britain sent two expeditions to be safe against clouds — one to Sobral in Brazil, one to the island of Príncipe off West Africa, where Arthur Eddington went himself. They photographed the stars around the eclipsed Sun, then compared with photographs of the same stars taken at night months later. The stars near the Sun had shifted outward. When Frank Dyson announced the numbers in London that November, the result made headlines around the world and turned Einstein, almost overnight, into the most famous scientist alive. Honesty requires a footnote: the data were noisy, and one blurred instrument that disagreed was set aside — a choice that later drew accusations of bias, though re-analysis has since confirmed the result.

Why it mattered

For two centuries Newton's gravity had been the unquestioned law of the cosmos. The 1919 eclipse was the first hard evidence that it was not the final word — that near a heavy enough body, Einstein's stranger picture of curved spacetime described reality better. It turned general relativity from beautiful mathematics into tested physics, and opened the century of black holes, the expanding universe, gravitational waves and GPS — all of which lean on the theory those eclipse plates confirmed.

An everyday picture

Imagine the fabric of space as a taut rubber sheet. Set a heavy ball — the Sun — in the middle, and the sheet sags into a dip around it. Now roll a marble straight across the sheet, well clear of the ball: it veers slightly as it crosses the slope of the dip, even though nothing pushed it sideways. Starlight does the same near the Sun — not pulled by a force in the old sense, but following the valley that the Sun's mass carves into space itself. The closer the marble passes to the ball, the steeper the slope and the sharper the bend — which is exactly why the deflection is largest for light grazing the Sun's edge.

A plot of how far light bends versus how close it passes the Sun. Two curves drop as the light passes farther out; Einstein's is always double Newton's. The 1919 eclipse measurements, plotted with error bars at the Sun's edge, land on Einstein's curve.

Where it sits

This is the experiment that confirmed the theory laid out in this Library's entry on Einstein's general relativity (1916). It is the hinge between Newton's Principia (1687), whose gravity it dethroned at the edges, and the modern science of gravity that followed — the gravitational waves caught by LIGO in 2016, the physics of black holes, and the lensed galaxies that let astronomers weigh dark matter. Every one of those rests on the curved spacetime that a few photographic plates first caught the Sun bending light around.

The original document
Original source text
F. W. Dyson, A. S. Eddington & C. Davidson · Phil. Trans. R. Soc. Lond. A 220 (1920): 291–333
The question
The purpose of the expeditions was to determine what effect, if any, is produced by a gravitational field on the path of a ray of light traversing it.
Three alternatives
The report frames the test as a choice between three outcomes for a star seen close to the Sun's edge:
(1) The path is uninfluenced by gravitation. (2) The energy or mass of light is subject to gravitation in the same way as ordinary matter. If the law of gravitation is strictly the Newtonian law, this leads to an apparent displacement of a star close to the sun's limb amounting to 0″·87 outwards. (3) The course of a ray of light is in accordance with Einstein's generalised relativity theory. This leads to an apparent displacement of a star at the limb amounting to 1″·75 outwards.
The method
During the few minutes of totality on 29 May 1919, when the Moon hid the Sun's disc, each expedition photographed the field of stars surrounding it — a particularly rich part of the Hyades. Months later the same star field was photographed at night, with the Sun far away. Superposing the plates and measuring how far each star had shifted outward gives the deflection directly. Two sites were used against the risk of cloud: Sobral in northern Brazil, and the island of Príncipe in the Gulf of Guinea, where Eddington himself observed.
[ … ]
The measurements
Corrected to the Sun's limb, the well-exposed Sobral 4-inch plates gave 1.98″ ± 0.12″, and Eddington's Príncipe plates gave 1.61″ ± 0.30″ — both near Einstein's 1.75″ and far from the Newtonian 0.87″. A third set, the Sobral astrographic plates, gave a much smaller figure but had been blurred by the Sun's heat distorting the mirror, and was set aside. (Probable errors as quoted in the report.)
Conclusion
Thus the results of the expeditions to Sobral and Principe can leave little doubt that a deflection of light takes place in the neighbourhood of the sun and that it is of the amount demanded by Einstein's generalised theory of relativity, as attributable to the sun's gravitational field.
Royal Observatory, Greenwich · read to the Royal Society, 6 November 1919