by John McPhee
“There may have been hundreds of thousands of volcanic islands, or chains of volcanic islands,” Van Schmus has remarked. “So instead of large continents you may have seen nothing more than a whole series of island arcs going around the globe—the whole earth like the South Pacific. They gradually amalgamated until they became Japans and later behaved as continents. More than likely there were fairly large continents in the early Archean, but no major or supercontinents. Or we may have had supercontinents and we lost all record of them.”
The oldest dated rocks are bits and pieces of Archean crust that cannot be put into a global context because they are so isolated. Are they the scattered fragments of continental rock that was almost wholly recycled into the mantle, or are they remnants of the early forming of continental material on a small scale, a process by which continental crust grew slowly across a very long stretch of time? Did vanished original continents occupy about the same percentage of the surface of the earth that continents occupy now? Or did the growth of continents begin slowly, with granitic scums and gradual accretions? Modern continents are essentially unsinkable, because their suites of rock are light and buoyant and will jam a trench and not go all the way down into it. But they are modern. Original lands were of the early Archean, when global heat and global tectonics were different. Who knows what may have disappeared into the mantle and been recycled? “Modern analogues are good, but we have to avoid falling into the trap of making everything fit them. They are guidelines.” The recent advances of technique and insight may have enabled the science to see deeper into the Precambrian than ever before, but not that deep. The question is still very open.
On the late Archean, which began three thousand million years ago, debate continues, but the picture has more focus. By then, many granitic microcontinents were spread about the globe, and they began to coalesce. Between “2800” and”2700” (as Precambrian geologists are wont to say, meaning from 2.8 to 2.7 billion years ago) came the major growth of continents, or the first major preservation of continental material, depending on whose thesis you espouse. In any case, the numerous new continents were still, by modern dimensions, small. The shield, the core, the basement of North America was nowhere near assembled; but its components were approaching one another and would come together in collisions that built topographies of rugged relief. Spread through the ocean in a paisley way, these modest continents—now called Nain, Rae, Slave, Wyoming, Superior, and Heame—are known in North American geology as the Archean cratons.
The end of the Archean and the beginning of the Proterozoic Eon came twenty-five hundred million years before the present, about halfway through Precambrian time. This round number represented a tectonic divide. Whatever the tectonics may have been in the Archean, the style of plate tectonics that we see in the modern world can be said to have begun with the new eon, about two and a half billion years ago. In the great historic argument between those geologists who see the earth’s processes as most importantly cyclical and repetitive (the concept of uniformitarianism, with its logo: “The present is the key to the past”) and those who see the earth’s history as a predominantly linear narrative, there has not been a clearer or firmer example of the irreversible and the unrepeatable than the changes that took place approximately twenty-five hundred million years ago.
“It does look like the Archean-Proterozoic transition is a real threshold in the behavior of the earth,” Van Schmus remarked one day while I was visiting him at the University of Kansas geological field camp in southern Colorado. “If you want to define plate tectonics strictly on the modern model, then you have to coin another term for the Archean tectonics. People sometimes make the mistake of saying that plate tectonics was not active in the Archean. I would say that modern-style plate tectonics was not, but there was some kind of dynamic earth behavior, whatever you want to call it. You might propose microplate tectonics, or that it was dominated by hot-spot activity rather than lateral plate motions. Things do seem to be somewhat different. We have not pieced together enough record to say exactly how different. The Archean-Proterozoic transition is a tectonic transition, a crustal transition, deriving from within the earth, not the surface.”
The surface expressed the great transition of twenty-five hundred million years ago in its own way. Although life had begun in the form of anaerobic bacteria early in the Archean Eon, photosynthetic bacteria did not appear until the middle Archean and were not abundant until the start of the Proterozoic. The bacteria emitted oxygen. The atmosphere changed. The oceans changed. The oceans had been rich in dissolved ferrous iron, in large part put into the seas by the extruding lavas of two billion years. Now with the added oxygen the iron became ferric, insoluble, and dense. Precipitating out, it sank to the bottom as ferric sludge, where it joined the lime muds and silica muds and other seafloor sediments to form, worldwide, the banded-iron formations that were destined to become rivets, motorcars, and cannons. This was the iron of the Mesabi Range, the Australian iron of the Hammerslee Basin, the iron of Michigan, Wisconsin, Brazil. More than ninety per cent of the iron ever mined in the world has come from Precambrian banded-iron formations. Their ages date broadly from twenty-five hundred to two thousand million years before the present. The transition that produced them—from a reducing to an oxidizing atmosphere and the associated radical change in the chemistry of the oceans—would be unique. It would never repeat itself. The earth would not go through that experience twice.
Around 2500—the Archean-Proterozoic transition—there was, as well, a distinct change in the over-all chemical composition of continental crust, in that it became distinctly more potassic after the Archean. Evidently, the earth had cooled enough for this chemical change to occur, and it had also cooled enough for tectonics to assume a more sedate and modern form. By 2300, its surface had cooled enough to support continental ice sheets, perhaps for the first time. Van Schmus continues: “The earth as a whole is producing progressively less heat from radioactive decay, and that, at some point, is going to have a profound effect. In the future, the profound effect is going to be that the plates will stop moving. We’ll be a very static Earth, much like Venus. Isolated hot-spot activity will go on for a while, and then die out. And you’ll basically have a very sedentary Earth. Sometime, it’s going to slow down and stop.”
At 2000, or thereabout, an event occurred in the Wyoming continent—the drifting Wyoming craton—of especial significance to the eventual development of North America. Evidently, something peeled away from Wyoming, took off on a transform fault (the two sides slipping horizontally, like the two sides of the Alpine Fault, the Denali Fault, the San Andreas Fault), and that side of the Wyoming continent became a sharp, clean, and almost straight line. Something quite similar happened in the Cretaceous, about nineteen hundred million years later, when India separated from Madagascar and went off on its rapid journey toward collision with Tibet. Together—as Eldridge Moores once noted—the eastern coast of Madagascar and the Malabar Coast of India make the sharp, clean, straight, and matching line where the countries were conjoined.
The sheared Precambrian coastline of the Wyoming microcontinent has become known as the Cheyenne Belt. If you could have stood on it 1.9 billion years ago, on rock that is now in the Laramie Range, you would have looked out over blue ocean, not over an epicontinental sea—shallow water lying on submerged land, like Hudson Bay—but abyssal ocean resting on ocean-crustal rock. Beyond, there would not have been a North American midcontinent rift system, because as yet there existed no North America to rift. Soon, though, the Archean cratons began to collide, building mountains where they hit, and sometimes trapping between them volcanic-island arcs—a series of events that took place in the geologic brevity of a hundred million years. About 1850 (1.85 billion years before the present), the Wyoming, Hearne, and Superior microcontinents came together, and the belt of deformation that held them tight is known in the science as the Trans-Hudson Orogen. A closed ocean, it is full of the crunched remains of ocean
ic islands. In 1900, they might have looked much like Indonesia today, but without its vegetation. By 1830, they would have resembled what Indonesia will look like after it collides with Asia. By 1800, over-all, the Canadian Shield was complete. From 1800 to 1400, most of North America grew on its margins.
The Canadian Shield was named in the nineteenth century by the Austrian geologist Eduard Suess, whose The Face of the Earth was first published in English in 1906. He described the extensive and exposed basement rocks of Canada as a “table-land not unlike a flat shield,” and formally declared, “It is to the exposed Archean surface that we give the name of the Canadian Shield.” Thereafter, the use of the word “shield” to describe the cores of continents became a tradition in geology and also developed into something of a misnomer. Geologists wanted their shields, thousands of miles wide, to resemble shields on warriors’ arms—gently convex. “Like a strong plate, this broad expanse of the earth’s crust has refused to yield and become folded,” wrote Reginald Aldworth Daly, of Harvard, in 1926. “It is shaped roughly like a shield; hence it has been called the Canadian Shield. Like a buckler, it is stiff and strong.” Stearn, Carroll, and Clark, of McGill University, wrote in their textbook, in 1979: “The term ‘shield’ is used as a crude description of its form—a low broad dome.”
The Archean cratons—Nain, Rae, Slave, Wyoming, Superior, and Hearne—and collisional orogenic belts (Greenland separated from North America in early Cenozoic time)
In the Canadian “dome” is a depression nearly a thousand miles wide, flooded with seawater. And the Canadian Shield is a museum of folded rock, as a result of the collisions through which it was assembled. Stiff and strong it surely is. Like few erosional surfaces in the world, it has been sitting pat for sixteen hundred million years with nothing serious happening to it. Its counterparts are the Brazilian Shield, the Siberian Shield, the Indian, Australian, and Antarctic shields, the Baltic Shield, the Afro-Arabian Shield. Many of these—tectonically separated—are broken parts of other shields. They may have fitted together in ancient supercontinents.
Rugged mountains were built as the Archean cratons assembled, and the mountains slowly downwasted until, by about 1750 million years before the present, they were bevelled to their roots. In some places, rhyolite lavas that came up at that time spread across a relatively flat surface; and ocean water flooding the bevelled shield—transgressing, regressing—deposited blanket sands. Bare and equatorial, the new and growing continent would at that time have been much like the modern planet’s Empty Quarter of Arabia.
Beyond the sheared edge of the Wyoming craton, seventeen hundred and fifty million years ago, islands stood in the ocean: volcanic-island arcs—Japans in motion. They came in one after the other. We know them as conterminous country: southeastern Wyoming, Colorado, Nebraska. Moving northerly, a crowd of accreting arcs probably extended as far as Chicago and beyond. Their radiometric ages date from their beginnings as islands, when their liquid substance came up off the mantle to solidify as juvenile crust. Their docking times—the dates when they connected with the Wyoming craton and with one another—are a good deal more difficult to determine, but on the average the islands seem to have existed in the ocean for ten to twenty million years before they collided with something else. The oldest age of the collected island arcs that are now the primary basement of Colorado is 1790—1.79 billion years. There was no Colorado and no Nebraska in 1800. Virtually all of it came in during the hundred million years that followed, and by 1700 or so the dockings of the arcs were complete. The new continental coastline trended from southern New Mexico through Indiana and on toward Labrador.
In the relatively modern Jurassic period, about a hundred and fifty million years ago, the assembling American Far West was much like the assembling Colorado sixteen hundred million years before. The accretionary development of the Far West first became apparent in the early days of plate-tectonic theory. And now, a few decades later, it has become apparent that the central basement of the continent—the platform, the stable craton, the once immemorial core—came together in the same way. The sharp line of the Cheyenne Belt, where southeastern Wyoming and Colorado joined the ancient shield, retains its character today. On the average, it is a couple of miles wide. Also known as the Wyoming Shear Zone, it is a line so fine that in some places you can all but straddle it. Where Interstate 80 crosses the Laramie Range west of Cheyenne, the highway is resting on arc-derived Proterozoic granite, the northern extension of the Colorado province. A few miles to the north is the edge of the Archean craton. The dividing line is sharp, and south of it are no rocks of Archean age. The northern half of the Medicine Bow Mountains—the next range west—is well over a billion years older than the southern half, and the jump takes place across the same narrow line. The line is buried under basin sediments where the interstate crosses it, just west of Laramie.
I met Randy Van Schmus a couple of years ago in what he calls “the post-1800-million-year accretionary complex” and most people call Colorado. The University of Kansas, where he teaches, has a geological field camp in Fremont County, Colorado. In the universities of the midcontinent, not a lot of geology is outside the door. The Precambrian basement is absolutely buried. To see what lies deep beneath the midcontinent, a student needs to be taken to a place like Colorado, where the same suite of rocks has been bent up into the air, where the Precambrian is the core and crown of the great foreland ranges.
On the day I joined him, Van Schmus took his students up to moderately high ground northwest of Cañon City, where their assignment was to map quartzites and other rocks that were inferred to be 1750 million years old and were pendant in a granite that had been precisely dated 1705. The quartzite, which was there first, had hung down in the soft intruding granite like a finger in honey. Quartzite being sandstone altered by heat and pressure, its sedimentary structures were somewhat pentimental and difficult to read. Where was the crossbedding? Which way was up? While the students, looking for “tops,” fanned out across alpine meadows and scrambled through junipers and across serrated ridges, Van Schmus had time to sit on a ledge and review the Big Picture. He was a reasonably tall man with an easy and quiet, unexcitable manner, whose boots and jeans and potato-chip hat fitted him into the country. I had sought him out because Eldridge Moores had described him to me as “about as knowledgeable a person as you can find on the midcontinent below the sediments—an age-dater who broadened out,” adding that in a recent compendium on North American Precambrian geology Van Schmus was the senior author of most of the papers that had to do with the Midwest. Van Schmus described himself as “a geochemist specializing in geochronology, specializing in the Proterozoic history of the earth.” He grew up in Naperville, Illinois (his father was a trust officer in a Chicago bank), and went to Caltech to become a chemist. Geology tends to draw certain people from the less eclectic sciences, and it drew him to the 2.4-billion-year-old shield rock north of Lake Huron, where he did his doctoral research for U.C.L.A. At that time—the early nineteen-sixties, before plate tectonics was established and when many major advances in radiometric dating were some years away—a great deal of Precambrian rock was shown on tectonic and geologic maps as, in his words, “one big green blob.” He has helped to fill in details of the blob. His name—his ancestors emigrated from a German town on the Dutch border—is pronounced like deuce and moose. He works in dry rangelands of the Brazilian Shield part of each year, “trying to identify old crustal blocks using isotopes”—the same kind of identification that has brought forth from obscurity the scenery of Archean Wyoming and Proterozoic Colorado. Not to mention Kansas.
As I have attempted to suggest, when Van Schmus and his colleagues talk about dates like 1640 and 1790 and 1850, they say them so familiarly and offhandedly that they might be discussing Oliver Cromwell, the American Revolution, and the publication of Moby-Dick . If they speak of, say, something that happened 1.745 billion years ago, they say “1745.” “Virtually all of Colorado is 1700 to 1790,” V
an Schmus said, meaning that the primary crust of Colorado—the collected island arcs—ranges in age from 1700 to 1790 million years. On that first day in Colorado, one or another date like that would roll off Van Schmus’s tongue with an ease and a familiarity that now and again caused me to pause, look up, and think, Good Lord, he means sixteen hundred and fifty million years ago. When he said that some evolving process had taken place, for example, from “1750 to 1770”—he would present the dates, as he and his Precambrian colleagues almost always do, in everyday sequence from the smaller to the larger number. Tripping on his own humanity, he was reversing the arrow of time.
