History of continental drift

The idea that continents move is now so fundamental to geology that it is difficult to appreciate how revolutionary — and how fiercely contested — it once was. For half a century after Alfred Wegener first proposed that the continents had once been joined and had since drifted apart, the hypothesis was rejected by the majority of the geological establishment, particularly in the English-speaking world, not because the evidence was weak but because no one could explain how continents could plough through solid oceanic crust. The eventual acceptance of continental drift, recast within the broader framework of plate tectonics, required not just new evidence but new ways of thinking about the Earth — and its history offers one of the most instructive case studies in the sociology of scientific revolutions.

Wegener and the evidence for drift

Alfred Wegener, a German meteorologist and polar explorer, first presented his continental drift hypothesis in a lecture to the Geological Association in Frankfurt on 6 January 1912, and published it in expanded form in his 1915 book Die Entstehung der Kontinente und Ozeane (The Origin of Continents and Oceans), which went through four editions by 1929.¹ Wegener was not the first to notice the apparent fit of the Atlantic coastlines — Francis Bacon had remarked on it in 1620, and Antonio Snider-Pellegrini had proposed a joined origin for the continents in 1858 — but he was the first to assemble a comprehensive body of evidence from multiple independent disciplines and to argue systematically that the continents had been united in a single supercontinent, which he named Pangaea (from the Greek for “all lands”), that had begun to fragment in the Mesozoic era.^{1, 10}

Wegener’s evidence was drawn from four principal domains. First, the geometric fit of the continents, particularly the complementary shapes of the eastern coast of South America and the western coast of Africa, which he demonstrated were not merely superficial resemblances but close matches when the continental shelves rather than the present-day shorelines were compared.¹ Second, the distribution of identical fossil species on continents now separated by thousands of kilometres of ocean: the Permian reptile Mesosaurus, found only in Brazil and South Africa; the Permian–Triassic seed fern Glossopteris, distributed across South America, Africa, India, Australia, and Antarctica; and the Triassic reptile Lystrosaurus, present on every southern continent.¹ Third, matching geological structures — mountain belts, rock types, and stratigraphic sequences — that aligned across ocean basins when the continents were reassembled: the Caledonian–Appalachian mountain chain, for instance, could be traced from Scandinavia through Scotland, across to Newfoundland and down the eastern seaboard of North America.¹ Fourth, paleoclimatic evidence, particularly the distribution of Late Paleozoic glacial deposits (tillites and striated pavements) across continents that now lie in tropical and subtropical latitudes, and coal deposits (indicating tropical swamp conditions) in regions that are now polar or subpolar. These patterns made no sense if the continents had always occupied their present positions but were readily explained if the southern continents had been clustered near the South Pole during the Carboniferous and Permian periods.¹

The rejection and the mechanism problem

Despite the strength of Wegener’s observational evidence, his hypothesis met overwhelming opposition, particularly from British and American geophysicists and geologists. The reasons for this rejection have been the subject of extensive historical analysis, most notably by Naomi Oreskes, whose 1999 study The Rejection of Continental Drift remains the definitive account.⁶

The central objection was mechanistic. Wegener proposed that the continents, composed of lighter sialic (granitic) rock, moved through the denser simatic (basaltic) oceanic crust under the influence of centrifugal force from Earth’s rotation and tidal forces from the Sun and Moon. Harold Jeffreys, the leading British geophysicist of the era, demonstrated conclusively that these forces were many orders of magnitude too weak to move continents through rigid oceanic rock.⁶ This was not a minor quibble: without a physically plausible mechanism, the hypothesis struck most physicists as not merely unproven but impossible. Geologists who might have found the distributional evidence persuasive were reluctant to accept a theory that appeared to violate basic physics, and the alternative explanation — that continents were connected by land bridges that had since sunk beneath the oceans — seemed to many a more conservative interpretation of the fossil data, even though it was itself unsupported by evidence and geophysically implausible.⁶

Oreskes has argued that the rejection was also shaped by methodological culture. American geologists, trained in an inductivist tradition that valued detailed fieldwork and local stratigraphy over grand theorizing, were suspicious of Wegener’s synthetic, cross-disciplinary approach, which they saw as cherry-picking evidence from fields in which he was not an expert.⁶ Wegener’s status as a meteorologist rather than a geologist did not help: he was seen as an outsider proposing a radical reinterpretation of a field he had not contributed to through original fieldwork. The national dimensions of the debate were also significant: the hypothesis found more support in the Southern Hemisphere, where the geological evidence for former continental connections was most obvious, and among South African, Australian, and South American geologists who could see the matching formations in their own outcrops.^{2, 10}

Du Toit's support and Holmes's convection currents

Among the few prominent geologists who championed continental drift during the decades of rejection, two figures stand out. Alexander du Toit, a South African geologist with extensive field experience across the southern continents, published Our Wandering Continents in 1937, presenting evidence for drift that was in many respects more detailed and more rigorously argued than Wegener’s.² Du Toit refined the concept of Pangaea by proposing that it had comprised two large landmasses: Laurasia in the north and Gondwana in the south, separated by the Tethys Sea — a reconstruction that remains the accepted framework today. His intimate knowledge of South African, South American, and Indian geology allowed him to document stratigraphic correlations across the southern continents with a precision that was difficult to dismiss, though dismissed it largely was by the Anglo-American establishment.²

Arthur Holmes, a British geologist better known for his pioneering work on radiometric dating and the age of the Earth, proposed in 1931 that thermal convection currents in the mantle could provide the mechanism Wegener had lacked.³ Holmes envisioned the mantle not as a rigid solid but as a material capable of slow, viscous flow over geological timescales, driven by the heat of radioactive decay. In his model, rising convection currents beneath a continent would diverge at the surface, dragging the continent apart and creating a new ocean basin in the gap.³ This was a genuine advance: it replaced the untenable idea of continents ploughing through oceanic crust with the physically more plausible concept of continents being carried passively on the backs of convective cells. Holmes published a widely read version of this model in his textbook Principles of Physical Geology (1944), ensuring that the idea of mantle convection entered the geological mainstream even as continental drift itself remained controversial.^{3, 10}

The paleomagnetic breakthrough

The evidence that ultimately broke the impasse came not from traditional geology but from the study of paleomagnetism — the record of Earth’s ancient magnetic field preserved in the iron-bearing minerals of rocks. In the 1950s, Keith Runcorn, Edward Irving, and their colleagues at Cambridge and the Australian National University showed that the direction and inclination of remnant magnetization in rocks of different ages from the same continent traced out a path — the apparent polar wander path — that recorded the continent’s movement relative to the magnetic pole over geological time.^{9, 12} Critically, the apparent polar wander paths derived from rocks on different continents diverged when calculated in their present positions but converged when the continents were reassembled in their Pangaean configuration. This was powerful independent evidence that the continents had moved, derived from a physical method that did not depend on biological, climatic, or geological interpretation.¹²

Irving’s 1956 paper demonstrated that paleomagnetic data from India were irreconcilable with a fixed continent but consistent with northward drift from a southern position, matching the independent prediction from Gondwanan glacial evidence.¹² The paleomagnetic evidence did not immediately convince all skeptics — questions about the reliability of remnant magnetization and the possibility of magnetic overprinting generated legitimate debate — but it shifted the burden of proof. By the late 1950s, a growing number of geophysicists accepted that some form of continental motion had occurred, even if the mechanism remained uncertain.⁹

Seafloor spreading and the Vine–Matthews hypothesis

The decisive mechanism came from the ocean floor. In 1962, Harry Hess of Princeton published a paper he self-deprecatingly described as “an essay in geopoetry,” proposing that new oceanic crust was continuously created at mid-ocean ridges by the upwelling of mantle material, spread laterally away from the ridge axis, and eventually consumed by descending into oceanic trenches.⁴ In Hess’s model, the ocean floor was not a permanent feature but a conveyor belt, continuously produced and recycled. Continents did not plough through oceanic crust; they were carried along by the spreading ocean floor, like logs frozen into a moving ice sheet. This was Holmes’s convection model made specific and testable, and it dissolved the mechanism problem that had dogged continental drift for fifty years.⁴

The test came in 1963. Frederick Vine and Drummond Matthews, working at Cambridge, proposed that the pattern of magnetic anomalies discovered on the ocean floor — alternating stripes of higher and lower magnetic intensity, running parallel to and symmetric about mid-ocean ridges — was the result of periodic reversals of Earth’s magnetic field recorded in basaltic lava as it erupted at the ridge, cooled through the Curie temperature, and was carried away from the axis by seafloor spreading.⁵ Each stripe represented a period of normal or reversed polarity, and the symmetry about the ridge axis reflected the bilateral spreading process. This elegant hypothesis, which Vine confirmed with additional data in a landmark 1966 Science paper, converted an abstract model into a quantitative, falsifiable prediction that could be tested against the accumulating magnetic surveys of the ocean floor.¹³

Heirtzler and colleagues extended the magnetic anomaly timescale across the world’s ocean basins in 1968, demonstrating that the same sequence of magnetic reversals could be identified in every ocean and that spreading rates could be calculated with precision.¹⁴ J. Tuzo Wilson’s identification of transform faults — a new class of fault unique to a spreading ocean floor — provided further geometric confirmation of the plate model, and his recognition that mid-ocean ridges, transform faults, and subduction zones formed a connected network of mobile boundaries laid the conceptual groundwork for the plate tectonic synthesis.^{15, 16}

The plate tectonic revolution

The transformation from continental drift to plate tectonics occurred with remarkable speed between 1965 and 1968. The key conceptual advance was the recognition that the Earth’s surface is divided not into continents and oceans but into rigid lithospheric plates, each comprising both continental and oceanic crust, that move as coherent units relative to one another. W. Jason Morgan formalized this concept in 1968, describing plate motions as rotations on a sphere and deriving the geometric relationships between spreading ridges, subduction zones, and transform faults.⁷ In the same year, Bryan Isacks, Jack Oliver, and Lynn Sykes demonstrated that the global distribution of earthquakes defined the boundaries of these plates with remarkable precision, and that the focal mechanisms of earthquakes at different plate boundary types matched the predictions of the plate model exactly.⁸

The convergence of evidence from magnetics, seismology, heat flow, bathymetry, and geology was overwhelming, and the scientific community shifted from majority rejection to near-universal acceptance within a few years — a pace of paradigm change unusual in the earth sciences. By the early 1970s, plate tectonics had become the unifying theory of geology, providing a single framework that explained mountain building, volcanism, earthquake distribution, the formation and destruction of ocean basins, the distribution of fossils and climatic indicators, and the long-term evolution of Earth’s geography.¹¹

The history of continental drift is frequently cited as a cautionary tale about the sociology of science: a correct hypothesis rejected for decades because it lacked a mechanism, championed by outsiders, and ultimately vindicated by evidence from an unexpected direction. But the story is more nuanced than this narrative suggests. Wegener’s hypothesis was rejected not simply because of prejudice but because a key component — the driving mechanism — was genuinely missing, and the proposed alternatives were demonstrably inadequate.⁶ The geophysicists who objected were not wrong that centrifugal and tidal forces could not move continents; they were wrong only in concluding that no force could. The resolution required not just more evidence but a fundamentally new conception of the ocean floor and the mantle, one that could not have been reached without the technological advances — magnetometers, seismometer arrays, deep-sea drilling, submersibles — that transformed earth science in the postwar period.¹¹ Continental drift was not merely confirmed; it was subsumed into a larger, more powerful theory that Wegener himself could not have imagined, but to which his original insight remains indispensable.