Eclipses through history

The most-cited episode from antiquity is the so-called eclipse of Thales. According to the Greek historian Herodotus, on 28 May 585 BC a total solar eclipse interrupted a battle between the armies of Lydia and Media on the banks of the river Halys, in present-day Turkey. The sudden darkness in the middle of the fighting frightened both sides so much that they laid down their arms at once and negotiated peace. Tradition credits the philosopher Thales of Miletus with having predicted that eclipse, which — if the legend is to be believed — would make it the earliest documented astronomical prediction in Western history; modern historians doubt Thales had the tools to calculate the exact date with the astronomy available in his time, but the episode has endured as the classic example of an eclipse that, quite literally, stopped a war.

Almost 2,500 years later, another eclipse would change the course of physics. On 29 May 1919, two British scientific expeditions — one to the island of Príncipe, off the coast of West Africa, led by Arthur Eddington, and another to Sobral, in Brazil — used totality to photograph stars visible near the edge of the eclipsed Sun. If the light from those stars bent as it passed close to the Sun, as Albert Einstein's still-recent theory of general relativity predicted, their apparent positions would shift slightly compared with how they looked at night, with the Sun nowhere near them. The measurements confirmed that bending, and the news travelled around the world: it was the moment Einstein went from being a respected physicist within his field to a global celebrity.

Long before anyone understood the geometry of an eclipse, Babylonian astronomers were already predicting them with remarkable accuracy. For centuries they kept systematic, detailed eclipse records on cuneiform clay tablets, within an astronomical and astrological tradition that reaches back at least to the second millennium BC. By accumulating so many observations, they identified that eclipses tend to recur roughly every 18 years and 11 days — what we now call the saros cycle — and even without being able to explain why that cycle existed, they learned to use it to anticipate when the next one was likely, without needing any model of the solar system.

China preserves some of the oldest known eclipse records: inscriptions on oracle bones and tortoise shells from the Shang dynasty, over 3,000 years old, already document solar eclipse observations. Chinese tradition also preserves a legend — of uncertain date — about two court astronomers, Hsi and Ho, said to have been punished for failing to predict an eclipse; whether or not it's literally true, the story reflects just how much predicting an eclipse was seen as a matter of state, tied to the emperor's mandate over the celestial order.

Greek astronomy took the next step, looking not just for the periodicity but for the geometric cause: it understood that a solar eclipse was the Moon's shadow and a lunar eclipse the Earth's shadow, and on that foundation, centuries later, Claudius Ptolemy developed mathematical models in the 2nd century capable of predicting eclipses with remarkable accuracy for the time, even though they were built on a geocentric system that modern astronomy would replace centuries afterwards.

On 18 August 1868, a total eclipse visible from South and Southeast Asia let astronomers analyse the light spectrum of the solar chromosphere — the thin layer of the Sun's atmosphere normally hidden by the glare of the photosphere — in detail for the first time. Working independently, the Frenchman Pierre Janssen and the Briton Norman Lockyer detected a yellow spectral line that matched no known chemical element on Earth. Lockyer, together with the chemist Edward Frankland, proposed it was a new element, which they named helium, after Helios, the Sun: it was the first time a chemical element had been identified on a celestial body before being found on Earth, where it wouldn't be isolated until 1895.

The other great example is the eclipse of 29 May 1919, already mentioned above for its historical role: scientifically, its importance lies in testing a very specific, measurable prediction of general relativity — how much a star's light should bend as it passed close to the Sun's mass — and confirming it with data, not just mathematics. Both episodes share something: only during the totality of an eclipse does the sky darken enough to observe, alongside the solar disc, both the faint chromosphere and the stars near the Sun's edge, something impossible at any other time without highly specialised instruments.

Spain has been the stage for some decisive eclipses in the history of solar astronomy. On 18 July 1860, the path of totality crossed the north of the peninsula, and the British astronomer Warren de la Rue photographed the eclipse from Rivabellosa (Burgos) while the Italian Jesuit Angelo Secchi did the same, independently, from Desierto de las Palmas (Castellón). By comparing their photographs, taken simultaneously from points several hundred kilometres apart, they were able to show that the reddish prominences visible next to the eclipsed Sun were real features of the Sun itself, not an optical effect of the Moon's atmosphere as some argued at the time: it was one of the earliest uses of photography as scientific evidence during an eclipse.

On 30 August 1905, another total eclipse crossed Spain again, this time drawing numerous international scientific expeditions to makeshift observatories along the path, equipped with spectroscopes dedicated to studying the corona and chromosphere. For more than a century it remained the last total eclipse to touch mainland Spain: the next time the Moon's shadow touched Spanish soil was on 2 October 1959, and only over the Canary Islands. On 12 August 2026, with totality crossing the peninsula once again, Spain came under the Moon's shadow 121 years later.