On 1 November 1930, his fiftieth birthday, Alfred Wegener was at the small research station at Eismitte in the geographic centre of the Greenland ice cap. The temperature outside was about minus sixty Celsius. He had spent the previous month hauling sledges of supplies to the station so that two of his men could try to overwinter there. The supplies were now delivered. Wegener and his Inuit guide Rasmus Villumsen began the four-hundred-kilometre journey back to the western coast. They never arrived. Wegener's body was found six months later in a sleeping bag under a rough cairn that Villumsen had built for him. Villumsen, and Wegener's diary, were never found.

Wegener was a meteorologist. He had a doctorate in astronomy. By the standards of the German universities of his time, he had no business writing about geology. This had not stopped him.

On the sixth of January 1912, eighteen years before he died, Wegener stood before the German Geological Society in Frankfurt and presented a paper proposing that the continents of the Earth had once been joined and were now drifting apart. The match between the coastlines of Africa and South America had been noticed for centuries; what Wegener added was a body of evidence from across half a dozen unrelated disciplines. The same plant fossils on either side of the Atlantic, in rocks of the same age. The same glacial scars running off South Africa, India, Australia and South America, all pointing to a single ice cap that had once covered them. Coal seams in Antarctica, where no tropical forest could grow today. Mountain ranges that, when the continents were rotated together on a globe, lined up across what is now ocean. Wegener called the original supercontinent Pangea, the all-Earth.

In 1915 he published the first edition of The Origin of Continents and Oceans, expanding through three further editions over the next fifteen years. In 1926 the American Association of Petroleum Geologists held a symposium in New York to evaluate his theory. Of the fourteen contributors, most were hostile. The chief objection was not that the evidence was bad. It was that Wegener had no mechanism. He could not explain what force could possibly drag a granite continent thousands of miles through solid ocean floor. Without a force, there was no science. There was only a Sunday afternoon's puzzle about coastlines.

The geological establishment gave the theory a name. They used the word the way other professions used heresy.

Wegener kept publishing. He corresponded with the South African geologist Alexander du Toit, who walked the Karoo basin and confirmed that its rocks matched South America's at every layer. He corresponded with the Harvard geologist Reginald Daly, who privately agreed but could not always say so publicly. Wegener went to Greenland a fourth time, in 1930, to gather one more set of measurements about the polar climate. He did not come back.

For the next thirty years his theory was a footnote in geology textbooks, often used as a cautionary tale about the danger of scientific overreach.

Then in 1962 Harry Hess of Princeton proposed that the seafloor itself was spreading outward from mid-ocean ridges, carried by convection in the mantle. The next year, Frederick Vine and Drummond Matthews at Cambridge, working independently of the Canadian geologist Lawrence Morley, showed that the basalt of the seafloor recorded the Earth's magnetic field reversals in symmetrical bands either side of the ridges, like growth rings turned outward. There was the missing mechanism. The ocean floor was new. The continents were riding on it. Within five years the theory of plate tectonics had been formalised, and the entire structure of the Earth's surface had become Wegenerist.

Wegener had been correct on every meaningful point and had been dead since 1930.

*

On the twenty-third of April this year, Nature Communications published a paper titled "Necking of the active Turkana Rift Zone and the priming of eastern Africa for continental breakup." The lead author is Christian Rowan, a graduate student at Columbia University's Lamont-Doherty Earth Observatory. His team analysed the seismic structure of the crust beneath a five-hundred-kilometre stretch of rift cutting across northern Kenya and southern Ethiopia. They found that the crust there has been thinned from about thirty-five kilometres at the margins to roughly thirteen kilometres along the rift's axis, and that the African and Somali plates are pulling apart at four point seven millimetres a year.

Necking is the word geologists use when stretched lithosphere starts to look like the thinned waist of a hot glass tube. It is the last phase before rupture. The Turkana Rift is not yet an ocean. It will become one. In a few million years, eastern Africa will detach into a separate continent, and a new sea will fill the gap.

This is exactly what Wegener said could happen.

The rift system doing the work stretches from the Afar Depression in Ethiopia south through Kenya and Tanzania to Mozambique. The same slow-motion process that is now thinning the African crust has been opening lake basins along this corridor for about forty-five million years. Olduvai Gorge in Tanzania is one of those basins. Hadar in Ethiopia, where Donald Johanson's team found Lucy in November 1974, is another. Koobi Fora on the eastern shore of Lake Turkana, where the Leakey teams recovered most of the early Homo skulls, sits inside the same rifting system. Every famous early-hominid fossil was preserved in lake sediments and volcanic ash sequences laid down by the very break-up that Wegener was laughed at for proposing. The thing that was supposed to be impossible is the geological process that made our own bones legible to us.

In a tent on the Greenland ice cap in November 1930, Alfred Wegener was, by the measure that mattered, the most correct person in his field.

He died not knowing it.