Supercontinent: Ten Billion Years in the Life of Our Planet

by Ted Nield

  In Britain things were less heated. At the British Association meeting in Hull in 1923, British geologists had met to consider the drift idea. According to a report of the ‘lively but inconclusive’ meeting printed in Nature, the main common point of opposition occurred over Wegener’s idea that Pangaea could have broken up and the Atlantic opened during the Quaternary, the most recent geological period. Nobody was very sure at that time when, in years, the Quaternary had begun (actually about 1.6 million years ago) but, while they might have been prepared to concede that continents moved, they didn’t think it likely that they moved that much.

  But while British geologists were arguing over the speed of the process, the basic concept did not seem to give those present much trouble. It helped that many of them had seen the African and South American rocks, but there was neither philosophical outrage nor any feeling at all that public decency had been affronted.

  But there was another problem facing Wegener in the USA that did not affect his Hull audience. British geologists thought of isostasy in terms of the Airy model, with its mountain roots, that presupposed that rocks not only could flow but did. This was not the prevailing view in America. There a different model held sway.

  Models versus reality

  After a seventy-three-second flight, at 11.39am on 28 January 1986, Mission 51L, the Shuttle programme’s oldest vehicle, Challenger, and its seven-strong crew, were blown to oblivion in the Florida sky. The brilliant physicist Richard Feynman quickly and intuitively hit on the immediate cause. Standing on the launch pad in freezing weather had reduced the springiness of rubber O-ring seals between the sections of Challenger’s solid rocket boosters, causing them to fail under pressure during lift-off. In a press conference Feynman demonstrated this simple idea using a section of O-ring, a beaker of iced water and a simple clamp.

  But that engineering failure was only the immediate cause. Subsequent investigations exposed many quality-control shortcomings at NASA, notably a tendency to tweak the safety envelope rather than the engineering, so that systems could be certified ‘safe’ more quickly and so allow NASA to meet its tight launch schedule, dictated largely by the need to put secret defence-related satellites in orbit. In his appendix to the Rogers Commission report, Feynman reserved his most withering criticism for this. NASA, he said, had made the fundamental scientific mistake of confusing their models with reality. ‘Nature,’ Feynman wrote, ‘cannot be fooled.’

  Models encapsulate our view of the world according to scientific principles and hypotheses, though, like supercontinents, even those that really did once exist, they are no more than imaginary constructs. Models are maps of reality, and like all maps, they have errors and it’s never too long before someone finds something that was missed.

  Nevertheless, it is all too easy for scientists to forget this and start thinking that model and reality are the same. As we have seen, in Britain the accepted model of isostasy – the theory that explains why mountains stand high and oceans stand low – was based primarily on the Airy model. This envisaged mountains as the tops of great rocky icebergs with deep ‘roots’, ‘floating’ in the Earth’s mantle. But the other explanation, put forward by Archdeacon Pratt, supposed instead that mountains stood high because they were less dense and had no ‘roots’ reaching deep into the Earth. This idea, though not entirely mistaken, is less in accord with the true internal geometry of the planet than Airy’s. However, Pratt’s model does have one big advantage to mapmakers: it makes the sums easier.

  If you belonged to a hard-pressed survey department in the early twentieth century, desperately churning out accurately triangulated maps of an emerging nation with a booming economy, you also had to correct thousands upon thousands of survey measurements for the effects of local gravity variations. And because nobody had yet invented the computer, this was all done manually, by armies of careful ladies with pencils. The big advantage of the Pratt model is that it makes these laborious calculations much easier. Easier means quicker, and time is money.

  But it appears that the sober and practical US Coast and Geodetic Survey had become so attached to their convenient calculating strategy that some of them actually came to believe it was true, and none more so than the chief geodesist, William Bowie (1872–1940).

  It is often said that the Airy model, with its implicit presumption that rock must flow over long periods of time, actually demands that continents can move. The Pratt model, on the other hand, carries the reverse implication. If you choose to believe it, then the idea of laterally sliding continents becomes worse than unnecessary: it becomes inconceivable.

  In Pratt’s model the base of the ‘crust’ is envisaged as a smooth, uniform surface sitting at the same depth everywhere. Mountains stick up above the general surface level to the extent that their rocks are less dense; the higher the mountains, the less dense the underlying rocks. But lateral motions of the crust would even out these density differences in time, and although the Earth is very old, we still have mountains. To a Pratt believer, therefore, drift seemed to be impossible. The influential Bowie had glimpsed an elegant model and taken her for Mother Nature: just because the maths was simpler.

  Wegener was walking into a tough room.

  Lone drifter

  Not all US geologists were opposed to drift. Today (and indeed also in older literature, when American pro-drifters were trying to make it seem less un-American) continental drift is often referred to as the Taylor-Wegener hypothesis to recognize US Geological Survey geologist Frank Bursley Taylor (1860–1938), who proposed a form of drift in 1910. There were also others. But despite converts (many highly respected), it was slow going. As the Wegener theory’s huge explanatory power won new adherents elsewhere, by and large, and for a wide variety of reasons (some spoken, others not), American geologists stood out against it almost to a man. Just how unanimous this opposition was became evident at a landmark scientific meeting in 1926 organized by Dr Willem Anton Joseph Maria van Waterschoot van der Gracht (1873–1943).

  Three years earlier Royal Dutch Shell had fired this brilliant, Amsterdam-born geologist. Van der Gracht was by no means the last geologist to find that being sacked by a multinational was the making of him, and a distinguished career in the United States as a pioneer in prospecting by seismic interpretation lay before him. He was also destined to play a pivotal role in the history of continental drift theory. In 1917 he had co-founded the American Association of Petroleum Geologists (AAPG), today the world’s largest Earth-science society, and in 1926 convened one of that body’s meetings. The conference, in New York City, had a single purpose: to discuss theories of continental drift.

  For the AAPG, which was then still a fledgling organization, this was a decisive moment. Its co-founder had chosen a controversial topic because it was, he believed, the biggest scientific game in town, with massive implications for oil prospecting. What better launch for a prestigious series of published symposium proceedings than to treat such a revolutionary subject?

  In the mythology of the subject this meeting has often been portrayed in something of the manner of the famous confrontation at the British Association in Oxford in 1860 between Darwin’s defender and champion, Thomas Henry Huxley, and Bishop Samuel Wilberforce. Although the earlier debate is always held up as a triumphant rout of the forces of fundamentalism, the New York meeting on drift is portrayed as one man’s lonely defence of Wegener’s theory against seemingly insuperable opposition. Neither of these histories is really accurate.

  Wegener was not present in New York, and it should not be forgotten that his theory was by no means the only drift theory in town. Other theories about how continents might have drifted were also on the agenda; it’s just that Wegener’s happened to be the main target. Frank Bursley Taylor, for example, also presented his theory. The American survey geologist suggested that two supercontinents, formerly at or near the Earth’s poles, gradually drifted towards the Equator, pushing up mountains along their leading edges and splitting und
er tension along their poleward, trailing edges. However, although Taylor’s theory had been around for about two years longer than Wegener’s, only Wegener had written so persuasively and at such length about it, and almost none of the other speakers addressed themselves to Taylor.

  In fact, with Taylor concentrating on his own ideas, the only person present who was at all in favour of Wegener was the convener himself. Almost without exception, Wegener was condemned, and even those who did not join in the condemnation merely urged further research. Reservation of judgement, however, was not to be the order of the day.

  The interesting thing about discussion meetings is, of course, discussion. If unanimity should unexpectedly break out, people reading the resulting proceedings will wonder why the meeting took place at all. So, in order to save his publication, and introduce at least a semblance of balance, van der Gracht solicited a number of supporting contributions from people who had not been present in New York. He commissioned contributions from Alfred Wegener himself, from the Irishman John Joly of Trinity College, Dublin, his fellow Dutchman Gustaaf Molengraaf, of Delft’s Institute of Technology, and Glasgow University’s professor John W. Gregory.

  These men were known to be broadly open to drift theory. Joly (1857–1933) was a highly original thinker with his own theory. His contribution, though, barely covered two sides, and concentrated on generalities. Molengraaf pronounced himself a drift enthusiast, but quibbled about Wegener’s insistence that Pangaea’s fragmentation had all been westerly. Gregory was also lukewarm. He was not opposed to drift per se, but saw no reason why the present distribution of land and sea could not equally well be attributed, à la Suess, to vertical rather than horizontal movements.

  These endorsements, though hardly ringing, did help, but not very much. So the convener decided to exert his privilege not only to have the first word, reproducing a seventy-five-page set of opening remarks, but also the last, book-ending the proceedings with a twenty-nine-page summing-up, in which he critically examined (and often demolished) various objections. In the end van der Gracht himself wrote 43 per cent of the entire volume. The proceedings were saved (and are now a sought-after collector’s item); but in its list of contributors the transatlantic divide over continental drift theory was made flesh. The supercontinent of science had undergone a geological rift.

  Du Toit’s proof

  In the pragmatic, fieldwork-obsessed world of early-twentieth-century American geology, where hard work and first-hand familiarity with the rocks were the only justification for any excursion into theoretical territory, the accusation of being an ‘armchair geologist’ was a grave one. It was levelled at, for example, Suess, whose great insights were indeed the product of much book learning and relatively little fieldwork (at least in his mature years). Both Suess and Wegener, in pointing to the geological correlations between the modern fragments of Gondwanaland, were generalizing from observations made by others about the congruence of facing Atlantic coastlines, Wegener doing little more than taking Suess’s data and interpreting them differently.

  American scientists might not have liked the method, but, some asked, what was the truth of this assertion? How alike were these sequences of rocks? Would it not be scientific to go there and check it out? The logic seemed faultless, but there was a snag. Proceeding in this way would be engaging in deductive research – setting out explicitly to test a ruling theory – and that was not the American way. Given that, who in the States would be prepared to put up the funds needed to send someone halfway across the globe to test a theory that most geologists in America had already dismissed?

  By good fortune, Reginald Daly, Harvard University’s Sturgis Hooper Professor of Geology, had been on a nine-month fact-finding mission to Africa in 1922, the year Wegener’s book was published for the first time in America. The main focus of the expedition had been South Africa’s massive Bushveld Igneous Province, and during the trip Daly had met Alex du Toit. The party had also included Molengraaf and Frederick Wright (1877–1953), a geologist from the Carnegie Institution of Washington, which funded the trip.

  Returning that autumn a convinced ‘drifter’, Daly began to hatch a plan with his friend in that well-endowed institution. The result was an illuminating exchange of letters between Wright and the Carnegie’s Director, John Merriam. In the first, Wright told Merriam about how the geological world was now in the thrall of an amazing new theory. He told his boss about the British Association for the Advancement of Science, and how its meeting in Hull had listed some crucial tests that should be done to verify the claims made for the geology of facing shores. The correlation of the Karoo rocks of South Africa with their South American equivalents particularly cried out for attention. There was only one man for the job, Alex du Toit. The Institution should pay for him to go to Brazil and verify it. In high hopes he sent off the proposal.

  Merriam turned it down. After the initial shock of disappointment, however, Wright found that the boss had not closed the door completely. Merriam was not about to put to his Trustees a request to spend the Carnegie Institution’s money on any overtly deductive search after facts. No, the proposal had to be more circumspect, more pluralistic, more American. To judge by Wright’s second (successful) attempt, it also had to be less exciting.

  Where version one had begun, like a piece of contemporary journalism, ‘The hypothesis of continental drift … is arousing the keen interest of geologists the world over’, version two opened: ‘In accordance with your request for information I take pleasure in presenting, herewith, an outline of some of the results that may be expected from a comparative study by A L du Toit of the Karoo, Gondwana and other equivalent formations of the southern hemisphere …’

  When Merriam finally gave the project the green light, Wright copied both versions to du Toit and explained the situation. Sure, they all knew why he was going to Brazil: to test Wegener’s hypothesis. They just couldn’t say so. This was important information for du Toit because, whatever his own intentions, he would now have to write up his results the American way, using multiple working hypotheses. This thoroughly deductive expedition had to be turned on its head and made to look like induction in action. The stated main purpose of the trip would be to gather facts: to add these good, solid bricks to the great edifice of science, and then to offer many different explanations of them, drift being merely one theory among equals. Wright’s letter arrived in Pretoria the very day du Toit set off on the Trans-Karoo railway for Cape Town, where he would catch a boat for Brazil.

  It would take du Toit two years to complete his monograph, and it would be four years before it finally appeared in print, in 1927. It is a curious document, and the tension between the dictates of American-style induction and du Toit’s own pro-drift convictions shows in almost every line. Du Toit succeeded in confirming the transatlantic correlations of Suess, expanded by Wegener, and he added some more of his own; yet to his dismay, all his work hardly seemed to help the cause of drift. In fact, he often found his painstaking fieldwork dismissed as irrelevant or inconclusive; he was even tarred with the same brush as Suess and branded an armchair geologist for his pains – he, a man who had mapped 50,000 square miles!

  Non posse

  The reason for du Toit’s failure lay in the nature of the evidence itself, of the ‘proofs’, as Skerl would have translated it. Both du Toit and Wegener were aware of this problem, which was that by their nature, geological correlations did not compel drift. In fact, du Toit committed a PR blunder of his own by admitting in his monograph that the similarities between the fossils of Africa and South America ‘can generally be explained equally well, even if less neatly, by the orthodox view that assumes the existence of extended land bridges …’. This was putting weapons in the hands of his enemies; but his point was that fossils on their own could not decide anything because fossils are remains of living things and living things can move. Instead, he concentrated on the rocks themselves.

  Sedimentary rocks change their characte
r from place to place, depending on the environment that lays them down. Beach sediments change as the bay merges with, say, a river estuary. What du Toit showed was that sediments of the same age in South America and Africa often showed greater changes within their outcrops in those continents than they did across the wide expanse of the South Atlantic. This, for him, was far more compelling evidence than the fossil similarities that the two continents were once close together and hence probably joined along their now distant coastlines.

  But this evidence was also circumstantial. Geophysicists persisted with their non posse, ‘it’s impossible’, line of argument (it is a well-known fact that a scientist always finds evidence from his own field most convincing). If you wanted to win over physicists, it was little use drawing their attention to fossils and sediments. For that group, drift didn’t happen because it couldn’t happen.

  Writers of popular science are often accused of ‘hardening up’ stories to make them simpler, clearer or more exciting. This sometimes gets them into trouble with scientists, who tend to favour caution. Charles Ray was writing The World of Wonder, my father’s unwieldy science encyclopaedia, in the early 1930s. Wegener, referred to throughout as ‘the late Professor Wegener’, had not long perished on the Greenland icecap. Continental drift theory had been around for eighteen years but was still highly controversial. Yet undaunted, after an excellent summary of the theory, Ray writes: ‘It seems a startling theory to think of the continents sliding or drifting over their foundations; but distinguished geologists say there is nothing at all impossible in the theory from a mechanical point of view …’