The lost supercontinent

Four researchers, one idea, and the measurement that settled it

A Alfred Wegener was a meteorologist by training, which is part of the reason his idea was received as it was. In 1912 he proposed that the continents had once formed a single mass and had since moved apart, and he assembled his case from several disciplines at once: the fit of the Atlantic coastlines, which had been noticed for centuries; identical fossils of land animals and plants on continents now separated by oceans; matching sequences of rock across those separations; and evidence of ancient glaciation in places that are now tropical. Each strand was known. The synthesis was his, and so was the willingness to follow it to a conclusion that specialists in each of the contributing fields found unacceptable.

B The reception was hostile, and it is worth being precise about why, since the episode is often told as a simple case of a stubborn establishment. Wegener had no mechanism. He suggested that continents ploughed through the ocean floor, driven by forces associated with the earth's rotation, and the physicists who examined the proposal calculated that the forces available were smaller than required by several orders of magnitude and that the crust was far too rigid to permit it. They were right. The evidence for movement was strong; the account of how it happened was demonstrably wrong; and a good many geologists concluded, not unreasonably, that a correct-looking conclusion supported by an impossible mechanism was probably a coincidence. The opposition was also geographical in a way that is often overlooked: the idea was taken far more seriously in the southern hemisphere, where the evidence for it lay under the feet of the people assessing it, than in North America, where it did not.

C Alexander du Toit, working in South Africa, made the empirical case considerably harder to dismiss. He argued that the correspondences between the southern continents were far too detailed to be accidental, matched rock formations, mountain ranges and glacial deposits across South America, Africa, India and Australia, and proposed that these had formed a southern landmass distinct from a northern one. His work was in effect a demonstration that the pattern was real whether or not anybody could explain it, and it made the geographical distribution of the evidence — overwhelmingly southern — an argument in itself.

D Arthur Holmes supplied the missing physics, and did so before it could be tested. He suggested in the late 1920s that heat from radioactive decay in the earth's interior would drive slow convection currents in the mantle, and that continents might be carried on such currents rather than ploughing through anything. The proposal was speculative and he presented it as such. It was also, in outline, correct, and it removed the objection that had defeated Wegener; but with no means of examining the sea floor, it remained an idea that could be neither confirmed nor refuted, and it changed few minds at the time. Holmes himself gave the proposal a cautious paragraph in a textbook and moved on to the work for which he was then better known, the use of radioactivity to date rocks, which incidentally supplied the vast timescale that any account of drifting continents requires.

E The evidence that settled the matter came from the ocean, and it came because navies wanted to know what was down there. Systematic surveying after the Second World War revealed a continuous mountain range running down the middle of the Atlantic, with a valley along its crest, and sediments that were thin near the ridge and thicker further away — the opposite of what an ocean floor of great and uniform age should show. The floor, it appeared, was younger in the middle. Nothing in the prevailing view of a permanent, ancient ocean basin could accommodate that, and by the early 1960s the idea that the sea floor was spreading outwards from the ridges was being discussed seriously by people who had no interest in continental drift at all.

F Confirmation arrived in 1963 in a form that convinced almost everyone at once. Fred Vine and Drummond Matthews pointed out that if new crust forms at the ridge and spreads outwards, and if the earth's magnetic field reverses periodically, then the rock should record those reversals as parallel stripes of alternating magnetisation, symmetrical about the ridge. The stripes were there in the magnetic surveys already collected. The prediction was specific, the data had been gathered before anyone made it, and the match was exact.

G The framework assembled from these pieces within a few years now explains earthquakes, volcanoes, mountain building and the distribution of fossils with a single mechanism, and it is taught as settled fact. The history is a better lesson than the conclusion. Wegener was right about the movement and wrong about the cause; his critics were right about the physics and wrong about the outcome; and the question was decided not by argument but by a measurement of the sea floor that nobody in the original dispute could have made.