Weighing Light in Sobral: How Britain and Brazil helped change our view of the universe

Weighing Light in Sobral: How Britain and Brazil helped change our view of the universe

by Richard Dunn, Senior Curator for the History of Science, Royal Observatory, Greenwich

Almost a century ago, a city in northern Brazil became famous as the place where Albert Einstein’s theory of general relativity was proved. It was a revolutionary moment for science, but behind it lay painstaking efforts that might have come to nothing without international collaboration and a bit of luck.

In 1919, two British astronomers – Charles Davidson and Andrew Crommelin – made the long journey from the Royal Observatory, Greenwich, to Sobral in northern Brazil to observe a total solar eclipse due to take place on 29 May. In principle, what they were doing was straightforward, but the theory behind it was incredibly complex, as was the task they faced.

According to Einstein’s theories, a massive body like the Sun should cause light passing near it to bend. If this was true, this would mean that the apparent positions of stars close to the Sun would be displaced compared to a time when their light did not pass by the Sun. To try to measure this – to weigh light – and so confirm Einstein’s prediction, astronomers proposed photographing the stars near the Sun during a total eclipse (when they would be visible from Earth). They would then photograph the same stars when the Sun was somewhere else in the sky – around a month later and at night – then compare the photographs and measure any displacement of the stars’ apparent positions. That may sound simple, but it wasn’t. The anticipated deflections would be just a sixtieth of a millimetre on the photographs taken, about the width of the finest human hair. And any deflection due to relativity had to be distinguished from other displacements.

The key to making these extraordinarily fine measurements was to take very sharp photographs during at a place from which the total eclipse would be visible for a very short five minutes. For the 1919 eclipse, this meant taking the equipment to Brazil – as well as to the island of Principe off the African coast, where a second British team observed the same eclipse – and trying to make it work while far from home.

Davidson and Crommelin took two telescopes. The larger one had a lens of 13 inches (33 cm) in diameter, borrowed from an astrographic telescope; the smaller, taken as a backup, had a lens of 4 inches (10.2 cm). These were set up horizontally and were not then moved. Instead, the image to be photographed was fed into them using coelostats – moving mirrors driven by clockwork that tracked their targets so that the images projected onto the photographic plates at the back of the telescopes were kept absolutely still during long exposures of up to 28 seconds. This was essential for producing well focused photographs.

Transporting the equipment was no simple task. The telescopes, coelostats, photographic plates, developing materials and observing huts took up 14 crates, which were transported by sea with the two astronomers. The voyage to Brazil took over two weeks, during which the astronomers suffered their fair share of seasickness. After travelling overland to Sobral, they spent a month setting up their equipment (with a good deal of local help), testing it and carrying out practice runs. This did not always go smoothly, with the unreliable coelostats causing a particular concern. The weather was also challenging – dust got into the coelostat mechanisms, some of the photographic plates failed in the Brazilian heat, and strong winds damaged the observing tent. It took a month of hard work to overcome these problems and get ready for the crucial observations.

The British astronomers were not alone, however. While planning the expedition, they received valuable advice and support from the Brazilian authorities and from Henrique Morize, Director of the Observatório Nacional in Rio de Janeiro. Morize was also in Sobral for the eclipse, heading up a team of Brazilian astronomers studying the solar corona, while a team from the USA was looking at the magnetic and electrical effects of the eclipse.

The Brazilian astronomers provided important support for their British counterparts. Their meteorological station supplied data on temperature, pressure and humidity, which would be crucial for the later analysis of the photographs. Morize also helped the astronomers select their viewing site on the racecourse owned by the Sobral Jockey Club and did his best to ensure that their work would not be disturbed. A few days before the event, he published a newspaper article urging people to remain quiet during the eclipse and not to set off fireworks.

The eclipse day must have been tense and frantic, particularly as early morning cloud threatened to ruin everything. But the clouds eventually shifted and the sky was clear enough during totality to take photographs. During those five minutes, Davidson and Crommelin each worked at a telescope, changing the photographic plates at agreed intervals, using a metronome to beat out the time, with their local interpreter, Dr Leocadio Araujo, calling out every ten beats. Ironically, the intense activity meant that they saw almost nothing of the eclipse itself.

Nor did their work finish there. The two astronomers spent the next week developing photographs at night (when it was cooler). These showed that the photographs taken with the larger telescope were not good enough – the hot Brazilian sun had distorted the coelostat’s mirror, they thought. Happily, those taken with the smaller telescope looked much better. Still, it would take months before the results were ready. In early July, the two astronomers went back to Sobral to take the comparison photographs. They then packed up their equipment and returned to Greenwich to begin painstaking measurements and laborious calculations to work out a value for the anticipated deflection of starlight.

When the results were finally revealed in London on 6 November 1919, it was the photographs taken with the smaller telescope at Sobral that proved decisive – how fortunate it was that they had taken it as a backup. The meeting agreed that the expedition to Brazil had indeed helped prove Einstein’s theories, and the next day the press reported a revolution in science, with the Newtonian system now disproved. Despite the difficulties of transporting delicate equipment across a world recovering from war, Britain and Brazil had successfully worked together on a world-changing experiment.

This post is based on a talk given at the Museu de Astronomia e Ciências Afins, Rio de Janeiro, as part of the UK-Brazil Year of Science and Innovation.

Categorias: Blog