by John McPhee
Wegener’s hypothesis in general is of the foot-loose type, in that it takes considerable liberty with our globe, and is less bound by restrictions or tied down by awkward, ugly facts than most of its rival theories. Its appeal seems to lie in the fact that it plays a game in which there are few restrictive rules and no sharply drawn code of conduct. So a lot of things go easily. But taking the situation as it now is, we must either modify radically most of the present rules of the geological game or else pass the hypothesis by. The best characterization of the hypothesis which I have heard was the remark made at the 1922 meeting of the Geological Society of America at Ann Arbor. It was this: “If we are to believe Wegener’s hypothesis we must forget everything which has been learned in the last seventy years and start over again.”
Through the nineteen-thirties, and particularly after the Second World War, paleomagnetic data accrued, and, as it presented its story of kaleidoscopic environments changing through time in any given place, academic geologists sketched on globes and maps their curves of apparent polar wander. Here is where the poles were at the end of the Silurian; this is where they went from there. Rocks of identical age, sampled in various parts of the world, indicated as much in their imprisoned compasses.
Some geologists—little cells of them in South Africa, the odd don or two at Cambridge—preferred the other explanation, but they were few, and in geology departments around the world everybody would annually crowd in to hear Lucius P. Aenigmatite, Regius Professor of Historical Geology, give his world-renowned lecture ridiculing continental drift. Oil geologists, when they had found what they were looking for in deep sandstones put down by ancient rivers, naturally yearned to know in what direction those rivers had flowed. They had long since learned empirically that if you wanted to find the direction of the stream you had to use different pole positions for well cores of different ages. Whether this was the result of polar wander or continental drift did not much matter to the flying red horse. Other geologists satisfied themselves by deciding that the paleomagnetic compasses were unreliable, otwithstanding that oil companies were using them to make money. Certain English geologists produced confusion by embracing continental drift and then drawing up narratives and maps that showed continents moving all over the earth with respect to a fixed and undriftable England.
By the late nineteen-fifties, paleomagnetic evidence had piled up so high that it demanded improved explication. India, for example, yielded data that put it out of harmony with the rest of the world with respect to polar wander. Either there was an inexplicable series of anomalies in the data or India itself had moved, coming up from the Southern Hemisphere and completely crossing the equator, rapidly, and at a rate of speed (as much as twenty-two centimetres a year) completely out of synchronization with the rate at which the equator’s position had differed in other terrains. More data, and increased sophistication in the analysis of data, began to show that polar-wander curves—once thought to be in agreement worldwide —could differ some from continent to continent. Curves based on Paleozoic and Triassic rock in North America and in Europe looked much alike but, oddly, stood separate in the way that a single line will appear to be double in inebriate vision. The gap corresponds to the present width of the Atlantic Ocean. The opening of the Atlantic began in the Triassic.
If the hypothesis of continental drift had long been overshadowed by the hypothesis of polar wander, the reverse would before long be true. Researchers in paleomagnetism at Cambridge University concluded that their data were showing them that both hypotheses could be correct, as later research at Princeton would confirm. The poles indeed had wandered. The continents had moved as well. The phenomenon of “apparent polar wander” had been caused, right enough, by the movement of masses of land, but concomitantly the earth had rolled—and patterns of “true polar wander” were seen to be superimposed on all the other motions of the shifting surface of the world. But what motions? If the continents had drifted, then in what manner were they drifting? Where had they come from and where were they going? What would happen if two should collide? Since they obviously were not plowing through solid basalt, how in fact did they move? It was all within a decade—1960-1968—that these questions were given answers of startling cohesion, as not only paleomagnetists but seismologists and oceanographers, geologists and geophysicists, whose specialties had been diverging through time, suddenly drew together around new outpourings of information and produced a chain of scientific papers whose interlocking insights would for most geologists fundamentally adjust their understanding of the dynamics of the earth.
“It was a change as profound as when we gave up the Biblical story,” Deffeyes said as he tapped his collector into the ground. “It was a change as profound as Darwinian evolution, or Newtonian or Einsteinian physics.”
The papers themselves had straightforward scientific titles, some of which—perhaps only in the afterlight of their great effect —seem to resound with the magnitude of the subject: “History of Ocean Basins,” “Rises, Trenches, Great Faults, and Crustal Blocks,” “Sea-Floor Spreading and Continental Drift,” “Seismology and the New Global Tectonics.” From Berkeley, Princeton, San Diego, New York, Canberra, Cambridge (England), there were about twenty primal contributions which, taken together, can be said to have constituted the plate-tectonics revolution.
Now plate tectonics is widely taken for granted. When I was in high school, there was essentially no television in America, and four years later television had replaced flypaper. When I was in high school, in the nineteen-forties, the term “plate tectonics” did not exist—albeit there was one remarkably prescient paragraph in our physical-geology textbook about the motions and mechanisms of continental drift. Today, children in schoolrooms just assume that the story being taught them is as old as the hills, and was told by God himself to their teacher in 4004 B.C.
The story is that everything is moving, that the outlines of continents by and large have nothing to do with these motions, that “continental drift” is actually a misnomer, that only the world picture according to Marco Polo makes much sense in the old-time browns and greens and Rand McNally blues. The earth is at present divided into some twenty crustal segments called plates. Plate boundaries miscellaneously run through continents, around continents, along the edges of continents, and down the middle of oceans. The plates are thin and rigid, like pieces of eggshell. In miles, sixty deep by nine thousand by eight thousand are the dimensions of the Pacific Plate. “Pacific Plate” is not synonymous with “Pacific Ocean,” which wholly or partly covers many other plates. There are virtually no landmasses associated with some plates—the Cocos Plate, the Nazca Plate. Some plates are almost entirely land—the Arabian Plate, the Iranian Plate, the Eurasian Plate, a large part of which used to be known as the (heaven help us) China Plate. (Jokes may be invisible to some geologists. Harry Hess, who in 1960 opened out the new story with his “History of Ocean Basins,” began it with these words: “The birth of the oceans is a matter of conjecture, the subsequent history is obscure, and the present structure is just beginning to be understood. Fascinating speculation on these subjects has been plentiful, but not much of it predating the last decade holds water.”) Certain major plates are about half covered with ocean—the South American Plate, the African Plate, the North American Plate. Australia and India are parts of the same plate. It is shaped like a boomerang, with the landmasses at either end. It may be in the early stages of separating itself into two plates. The northern section has a slightly different motion but there is no sharp boundary. In Africa, the terrane east of the Great Rift Valley is far enough along in the act of separation to be called the Somalian Plate, but the boundary is not yet continuous. Continents in themselves are not drifting, are not cruise ships travelling the sea. Continents are high parts of plates. East-west, the North American Plate starts in the middle of the Atlantic Ocean and ends in San Francisco. West-east, the Eurasian Plate begins in the middle of the Atlantic Ocean and ends in the sea of Okhots
k.
It is the plates that move. They all move. They move in varying directions and at different speeds. The Adriatic Plate is moving north. The African Plate came up behind it and drove it into Europe—drove Italy like a nail into Europe—and thus created the Alps. The South American Plate is moving west. The Nazca Plate is moving east. The Antarctic Plate is spinning, like pan ice in a river.
As has happened only twice before in geology—with Abraham Werner’s neptunist system and James Hutton’s Theory of the Earth —the theory of plate tectonics has assembled numerous disparate phenomena into a single narrative. Where plates separate, they produce oceans. Where they collide, they make mountains. As oceans grow, and the two sides move apart, new seafloor comes into the middle. New seafloor is continuously forming at the trailing edge of the plate. Old seafloor, at the leading edge of a plate, dives into deep ocean trenches—the Kuril Trench, the Aleutian Trench, the Marianas Trench, the Java Trench, the Japan Trench, the Philippine Trench, the Peru-Chile Trench. The seafloor goes down four hundred miles after it goes into the trenches. On the way down, some of it melts, loses density, and—white-hot and turbulent—rises toward the surface of the earth, where it emerges as volcanoes, or stops below as stocks and batholiths, laccoliths and sills. Most of the volcanoes of the world are lined up behind the ocean trenches. Almost all earthquakes are movements of the boundaries of plates —shallow earthquakes at the trailing edges, where the plates are separating and new material is coming in, shallow earthquakes along the sides, where one plate is ruggedly sliding past another (the San Andreas Fault), and earthquakes of any depth down to four hundred miles below and beyond the trenches where plates are consumed (Japan, 1923; Chile, 1960; Alaska, 1964; Mexico, 1985). A seismologist discovered that deep earthquakes under a trench had occurred on a plane that was inclined forty-five degrees into the earth. As ocean floors reach trenches and move on down into the depths to be consumed, the average angle is something like that. Take a knife and cut into an orange at forty-five degrees. To cut straight down would be to produce a straight incision in the orange. If the blade is tilted forty-five degrees, the incision becomes an arcon the surface of the orange. If the knife blade melts inside, little volcanoes will come up through the pores of the skin, and together they will form arcs, island arcs—Japan, New Zealand, the Philippines, the New Hebrides, the Lesser Antilles, the Kurils, the Aleutians.
Where a trench happens to run along the edge of a continent and subducting seafloor dives under the land, the marginal terrain will rise. The two plates, pressing, will create mountains, and volcanoes will appear as well. The Peru-Chile Trench is right up against the west coast of South America. The Nazca Plate, moving east, is going down into trench. Interspersed among the uplifted Andes are four thousand miles of volcanoes. The Pacific Ocean floor, going down to melt below that edge of the continent, has done much to help lift it twenty thousand feet.
MAJOR LITHOSPHERIC PLATES AND SOME MINOR ONES
Seafloor—ocean crust—is dense enough to go down a trench, but continents are too light, too buoyant. When a continent comes into a trench, it will become stuck there, causing havoc. Even if part of it goes down some dozens of kilometres, it will eventually get stuck. Australia is such a continent, and where it has jammed a trench it has buckled up the earth to make the mountains of New Guinea, sixteen thousand five hundred feet.
When two continental masses happen to move on a collision course, they gradually close out the sea between them—barging over trenches, shutting them off—and when they hit they drive their leading edges together as a high and sutured welt, resulting in a new and larger continental mass. The Urals are such a welt. So is the Himalaya. The Himalaya is the crowning achievement of the Indo-Australian Plate. India, in the Oligocene, crashed head on into Tibet, hit so hard that it not only folded and buckled the plate boundaries but also plowed in under the newly created Tibetan Plateau and drove the Himalaya five and a half miles into the sky. The mountains are in some trouble. India has not stopped pushing them, and they are still going up. Their height and volume are already so great they are beginning to melt in their own self-generated radioactive heat. When the climbers in 1953 planted their flags on the highest mountain, they set them in snow over the skeletons of creatures that had lived in the warm clear ocean that India, moving north, blanked out. Possibly as much as twenty thousand feet below the seafloor, the skeletal remains had formed into rock. This one fact is a treatise in itself on the movements of the surface of the earth. If by some fiat I had to restrict all this writing to one sentence, this is the one I would choose: The summit of Mt. Everest is marine limestone.
Plates grow, shrink, combine, disappear, their number changing through time. They shift direction. Before the Pliocene, there was a trench off California. Seafloor moved into it from the west and dived eastward into the earth. Big volcanoes came up. Under the volcanoes, the melted crust cooled in huge volumes as new granite batholiths. Basin-range faulting has elevated the batholiths to fourteen thousand feet, and weather has sketched them out as the Sierra Nevada.
When seafloor goes into a trench, there can be a certain untidiness as segments are shaved off the top. They end up sitting on the other plate, large hunks of ocean crust that formed as much as a few thousand miles away and are now emplaced strangely among the formations of the continent. The California Coast Ranges—the hills of Vallejo, the hills of San Simeon, the hills of San Francisco —are a kind of berm that was pushed up out of the water by the incoming plate, including large slices of the seafloor and a jumble of oceanic and continental materials known to geologists as the Franciscan mélange. Geologists used to earn doctorates piecing together the stratigraphy of the Franciscan mélange, finding bedding planes in rock masses strewn here and there, and connecting them with dotted lines. Plate tectonics reveals that there is no stratigraphy in the mélange, no consecutive story of deposition—just mountains of bulldozed hash. The eastward motion of the ocean plate stopped soon after basin-range faulting began. The plate started moving in another direction. The trench, ceasing to be a trench then, was replaced by the San Andreas Fault.
Mountain building, in the Old Geology, had been seen as a series of orogenies rhythmically spaced through time, in part resulting from isostatic adjustments, and in part the work of “earth forces” that were not extensively explained. As mountains were disassembled, their materials were deposited in huge troughs, depressions, downbendings of the crust that were known as geosynclines. Earth forces made the geosynclines. As sediment accumulated in them, its weight pressed ever farther down into the mantle until the mantle would take no more, and then there came a trampoline effect, an isostatic bounce, that caused the material to rise. The Gulf of Mexico was a good example of a geosyncline, with a large part of the Rocky Mountains sitting in it as more than twenty-five thousand feet of silt, sand, and mud, siltstone, sandstone, and shale. “The South will rise again!” Deffeyes used to say. The huge body of sediment would one day be lifted far above sea level and dissected by weather and wrinkled into mountains in the way that the skin of an apple wrinkles as the apple grows old and dry. The steady rhythm of these orogenies was known as “the symphony of the earth”—the Avalonian Orogeny in latest Precambrian time, the Taconic Orogeny in late Ordovician time, the Acadian Orogeny in late Devonian time, the Antler Orogeny in Mississippian time, the Alleghenian Orogeny in Pennsylvanian-Permian time, the Laramide Orogeny in Cretaceous-Tertiary time. It was a slow march of global uplifting effects—predictable—proceeding through history in stately order. By the end of the nineteen-sixties, the symphony had come to the last groove, and was up in the attic with the old Aeolian. Mountain building had become a story of random collisions, unpredictable, whims of the motions of the plates, which, when continents collided or trenches otherwise jammed, could give up going one way and move in another. The Avalonian, Taconic, Acadian, and Alleghenian orogenies were now seen, in plate theory, not as distinct events but as successive parts of the same event, which involv
ed the closing of an ocean called Iapetus that existed more or less where the Atlantic is today. The continents on either side of Iapetus came together not head-on but like scissors closing from the north, folding and faulting their conjoining boundaries to make the Atlas Mountains and the Appalachian chain. It was a Paleozoic story, and the motions finally stopped. In the Mesozoic, an entirely new dynamic developed and the crust in the same region began to pull apart, to break into blocks that formed a new province, a Eurafrican-American basin and range. The blocks kept on separating until a new plate boundary formed, and eventually a new marine basin, which looked for a while like the Red Sea before widening to become an ocean.