In the nineteenth century, Lord Kelvin and other physicists concluded that the Earth was cooling.2 The heat generated during the formation of the Earth was slowly leaking into space, and the physicists knew of no source of heat to replace that being lost. They inferred that the Earth was relatively young and would eventually conduct the balance of its heat away, leaving it a frozen lump of rock floating through the solar system.
Grim though this idea was, it held some appeal to geologists grappling with the origin of features of the Earth's surface. Historically, geology has been a science of careful observation: Practitioners look at rocks, as many as they can, and they record what they see in close, sometimes exhaustive detail. These details can then be used to assess the age and origin of rocks, and to recognize ancient ocean floors or the cores of eroded mountain ranges. Whether the perception of geology as a mostly descriptive science is flattering or pejorative is largely in the eye of the beholder; in the 1900s the description at least had the virtue of being accurate. Even then geologists were not exactly theory-deprived; James Hutton's and Charles Lyell's theory of uniformitarianism, which argued that processes operating in the present could, over long time-scales, account for the formation of geological features, was successful in explaining the origin of streambeds, canyons, sedimentary rocks, and many other features. But explanations for other fundamental, large-scale geological phenomena were absent or flawed. Why were continents continents and oceans oceans? How did mountain belts form? No one knew.
If theories were somewhat lacking, it was not for a scarcity of observations. If anything, widespread mapping in the nineteenth century provided too many: Geologists were drowning in observations from the Alps, the American West, and dozens of other localities being mapped in detail for the first time. Reports from the field revealed new problems and contradictions far more quickly than theories could be modified to account for them. Two examples important to the continental drift debate concerned the distribution of fossils and ancient glacial deposits across the continents. Geologists found that similar fossil assemblages sometimes occurred in locations as widely separated as South America and South Africa.3 These homologies were deeply problematic. According to Darwin's theory of evolution, the development of identical species and similar assemblages in such distant areas was impossible, suggesting that the areas must once have been connected. How this could be when they were now separated by thousands of kilometers of ocean was an open question. Likewise, glacial deposits dating from the Permian had been found scattered across the globe.4 How glaciers could have formed at the same time over such a wide range of latitude was not at all obvious.
Geologists working during the mid-to-late nineteenth century thus faced a bewildering array of questions, prominent among them these four: What are the differences between continental and oceanic crust? How do mountain belts form, both within and at the edges of continents? How can similar fossil assemblages form at sites separated by the width of an ocean? How did the scattered Permian glacial deposits come to exist? A successful theory of the surface of the Earth must address (or at minimum not contravene) the lines of evidence underpinning each of these questions. Formulating such a theory proved difficult.
To return, then, to the physicists, the appeal of the theory of the cooling Earth was that it provided a way to address these questions and their attendant observations. A cooling Earth was also a contracting Earth, and thermal contraction, geologists argued, could produce many of the features existing at the surface of the Earth.5 As the Earth shrunk, its surface compressed to form the areas of high topography represented in the modern continents. Low-lying areas would contract more quickly, forming ever-deepening ocean basins. At the edges of continents, the difference in contraction allowed the accumulation and uplift of sediments to form coastal mountain ranges. Geological theories based on contraction, as formulated by Eduard Suess and James Dwight Dana, were widely praised for their explanatory power and popular within the geological community well into the twentieth century.6
They were nevertheless flawed theories. To begin with, the two leading versions of contraction disagreed on points as fundamental as the permanence of the continents and ocean basins.7 Dana thought that continents formed early in the history of the Earth and had remained much the same ever since. Suess argued that continents and ocean basins were transient and subject to periodic upheavals, with continents sinking and the ocean floor rising to take their place. Each formulation ignored a key piece of evidence: Dana's theory offered no explanation for fossil homologies, whereas Suess's contradicted developing ideas about the long-term stability of continental crust. (This second point will be discussed in more detail in the next section.) Dana's theory ultimately proved the more successful, with its most pressing difficulty addressed by the addition of land bridges to explain fossil homologies.8 Land bridges were narrow tracts of continental crust linking sites containing identical fossil species. Land bridges would appear, permitting species to migrate between distant sites for a time, then eventually collapse.
Beyond these discrepancies, the contraction theories shared a more fundamental problem: Field observations and mathematical analyses made in the latter half of the nineteenth century suggested that thermal contraction could produce neither the pattern nor extent of deformation seen in existing mountain belts.9 Detailed mapping of mountain ranges showed that high elevations resulted not from the vertical motions Dana promoted but rather from shortening of huge tracts of continental crust.10 In the Alps, Albert Heim and Marcel Bertrand mapped stacks of thrust sheets that implied shortening of the original crust by a factor of four or more. Thermal contraction was unable to account for such large amounts of compression.
Shortly before the beginning of the twentieth century, a pair of discoveries further complicated the case for contraction theory. The combined impact of radioactivity and the theory of isostasy would force geologists to reconsider their conception of the surface of the Earth.
Go To Radioactivity and Isostasy