Annals of the Former World

by John McPhee

  “To my way of looking, this is an accretionary complex of arc terrane,” he repeated. “Think of the South Pacific—all that stuff waiting to be accreted. Parts of Fiji are forty million years old and are still waiting to dock. The Precambrian of Colorado has not been studied extensively enough. Geologists of the next decade will take it apart and tell the story. More advanced techniques will develop. This is the new frontier. To get to new frontiers you go backward in time. The isotopes are telling us that this is new crust, and about the only way you get new crust is in arc environments. If you want to see what’s under Interstate 80 in Nebraska, you’re seeing it right here. There’s basically no substitute for being able to see the rocks and walk over them. Nebraska is like this—a few thousand feet down.”

  I have also visited him in Lawrence, Kansas, and seen the primary resources that have enabled him and others to see, infer, extrapolate, conjure, discover, and describe some of the long-veiled and innumerable worlds of the Precambrian eons: the university’s large archive of drilled Precambrian rock, for example, and the hospital-like rooms of the “Rb/Sr and Sm/Nd Clean Lab,” where even a bit of rock powder too fine to be called silt will yield its age and origin to rubidium/strontium and samarium/neodymium dating. Now and again in Colorado, between his field sessions with the students, we travelled around Fremont County so that I could attempt, in the presence of the terrain as it has come down to us, to see its origins through his eyes. The field camp was a collection of cabins far up a gravel road in an amphitheatrical cul-de-sac known to geology as the Cañon City Embayment. Quite near the camp were Jurassic mudstones and sands where a schoolteacher named Ormel Lucas, on his summer vacation in 1876 (the summer of Little Bighorn), discovered numerous large bones of which a memorable roster was to follow: the twenty-four-foot stegosaurus in the Denver Museum of Natural History, the seventy-two-foot haplocanthosaurus in the Cleveland Museum of Natural History, the sixty-nine-foot apatosaurus in the Wagner Free Institute (in Philadelphia), the fifty-nine-foot camarasaurus in the American Museum of Natural History, and the twenty-foot ceratosaurus and the thirty-nine-foot allosaurus and the eighty-eight-foot diplodocus in the Smithsonian. Van Schmus, raising dust, went straight past their beds with an amiable nod. These creatures, on the full scale of time, were so close to the here and now that they might as well have been his students. He ran up toward Cripple Creek on a road ten feet wide that had been chipped into the sheer wall of the embayment. It climbed higher and higher up the wall—four hundred, five hundred feet—and the view over the side was perfect, unobstructed, railless. The rock of the embayment wall was three hundred million years older than the dinosaurs but still 1.3 billion years younger than the rock on which it rested. That would get us back to 1750, when, in the primordial sense, Colorado crust was first assembling. Van Schmus said that a fisherman had found a severely injured motorist inside his automobile at the bottom of this canyon a few days before. “Do you see a car yet? It could have been here.”

  The western reach of Fremont County—some fifty miles on the Arkansas River between Cañon City and Salida—is particularly memorable for the melodrama of its landscapes and the beauty of its structure. The Arkansas runs fast and white most of the way, maybe a hundred feet wide. The rock around it towers. As the river approaches its debouchment into the flats of eastern Colorado, it makes one last slice that by an order of magnitude is the deepest of all. Like a knife disappearing into a large loaf of bread, it has cut an extremely narrow gorge whose rims are more than a thousand feet above the rapids.

  Remains of the original arcs are the rock of the gorge and the general rock of the region, almost everywhere deformed. Quartzites, migmatites, gneisses, and schists, they have been metamorphosed by so many events that their histories are close to inscrutable. Thirty miles up the canyon, however, is an elongate section of primary crust that is the lithic equivalent of virgin forest. Somehow it managed to hide out in the strain shadows of all the tectonic events that followed its own appearance on this scene. Van Schmus said, “Here you are looking at the birth of Colorado, the primary crust. When it arrived, there was basically nothing here.”

  “When was that?”

  “1740.”

  He went on to say, “The primary basement here evolved over plus-or-minus a hundred million years. Plutons came along more or less in parallel with the volcanism, younging to the south and trending to the southwest.” The plutons were the intrusive magmas that, for the most part, cooled as granite. In a closely analogous way, the granites of the Sierra Nevada would come into the country rock there, more than a billion and a half years later.

  A coal train went by on the far side of the river. It refocused his vision—flicked his mind forward—to Carboniferous time. “There was very little coal in Africa and South America in the Paleozoic,” he said cryptically. They were “stuck near the South Pole” when the big trees, elsewhere, were kings. North America was on the equator then. Hence the coal train.

  The train was just passing through, however, and its screeches and rumbles and the contents of its gondolas represented points in time that were not in this conversation. In an instant, the geologist’s mind returned to the era of primary midcontinental America—eighteen hundred to fourteen hundred million years before the present, a band in the earth’s history about a third of the way from the planet’s earliest beginnings to the tick-ticking of the present day. Representing only twenty per cent of the Proterozoic Eon, 1800-1400 was nonetheless the central frame of his professional absorption. If one part of Precambrian time was his specialty, this was it. His own lifetime—beginning in the calendar year 1938 and expecting at least four score and ten—was a submicroscopic speck at the end of a widening shaft of information and thought that could reach to and bracket those four hundred million years. The difference between one human lifetime and four hundred million years would seem to be a difference between time incomprehensible and time infinitesimal, but what brings them together is that the smaller unit—bridging in the mind the intervening eons—can imagine and virtually see the larger one.

  A quarter-horse jockey learns to think of a twenty-second race as if it were occurring across twenty minutes—in distinct parts, spaced in his consciousness. Each nuance of the ride comes to him slowly as he builds his race. If you can do the opposite with deep time, living in it and thinking in it until the large numbers settle into place, you can sense how swiftly the initial earth packed itself together, how swiftly continents have assembled and come apart, how far and rapidly continents travel, how quickly mountains rise and how quickly they distintegrate and disappear. No matter how impressive or extensive the data might be, Van Schmus said, it is well to bear in mind, as you imagine something as comprehensive as the original building of the center of North America, that the developed picture is authenticated by the best current hypothesis and nothing more. “You are never going to write the definitive answer in geology; it’s the nature of the field.”

  A suspension bridge—“The World’s Highest”—crosses the profound gorge, mainly for tourists on foot, and if a car happens along while you are walking in mid-span the suspended deck undulates under your feet as if it were a raft in the rapids a thousand feet below. There is a wind so stiff that you lean hard against it. It feels as if it is about to blow you over the side. For anyone with a fear of height, cars do not add to the terror, because the terror is already so complete that there is no room for more. The whorls and convolutions in the gorge walls—metavolcanic and metasedimentary rocks, migmatites and gneisses thoroughly cooked at least twice—are all but undecipherable under any conditions and glazed under these. Then a young woman happens by, wheeling a baby carriage and clucking fondly at the contents. She should go into geology.

  In the parkland beside the gorge, Van Schmus remarked that the extensive alteration of the primary crust we were standing on probably resulted from the high-grade metamorphism that would have occurred while arcs collided. He said it was possible that this was part of the fi
rst terrane to attach to the Wyoming craton. “The discrete terranes have not been identified,” he went on. “But the principle is valid. We may agree that we are looking at a collage of accreted terranes analogous to what went on in California in the Mesozoic, but we’re not at the stage where we can identify terrane boundaries. We don’t know the age of the gorge. I strongly suggest that it’s no older than 1800. The problem with the gorge rock is that it has been overprinted with metamorphism so many times. Sorting all this out and figuring what it was originally is the present frontier, and will lead to an understanding of what things were: back-arc basins like the Sea of Japan, accretionary wedges like the Coast Ranges of California, inter-arc rift basins, fore-arc basins, foredeeps.” As he spoke, I was scribbling notes and he was looking down into a pegmatite quarry. When we looked up, we saw Pikes Peak to the north-northeast, above the trees.

  Near Salida, where we saw gabbros, pillow basalts, and other subaqueous volcanics, he said, “With confidence, you would associate them with arc rocks. Although we’re still dealing with a frontier in continental evolution, the isotopes tell us very clearly that this stuff had a very short crustal residence time.” By that he meant that when magma arose from within the earth to cool as lithospheric crust, the crust had not travelled long or far before it went into a subduction zone to melt anew. The timetable of decay of radioactive elements within the initial magma would mark the event when the magma first emerged. “The clock starts ticking when the stuff comes out of the mantle into the crust,” he explained. “The arc, in turn, may get melted and turn into granite, or whatever, but the clock never stops ticking. The story is there in the rocks. The problem for us is to figure out how to read them.”

  Isotopes that have become particularly useful in monitoring the crustal history of rocks and their origins in the mantle belong to the elements samarium and neodymium. Because samarium radioactively decays and becomes neodymium at a fixed rate, the two serve as a chronometer, like rubidium and strontium, like uranium and lead. “In the late seventies, early eighties, samarium/neodymium as an analytical technique was perfected,” Van Schmus said. “With advances in instrumentation, the technique became a sort of everyman’s tool. What it tells us basically is when rocks come out of the mantle. It doesn’t provide ages, necessarily, but it does provide an isotopic tracer that allows us to track the history of continental crust, and particularly to determine how long the material in a piece of continental crust has been separated from the mantle. We use other methods to get the crystallization age of rocks, and samarium/neodymium analyses to tell how long those particular rocks have been part of the continental system.”

  Samarium/neodymium ages are accurate but they are not very precise. This distinction—between accuracy and precision—has a difference, and is less of a split hair than it may at first appear to be. For example, if you say that George Washington submitted his last expense account when he was forty-seven, plus or minus twenty years, you are accurate but you are not at all precise. If you say that Santa Claus came down the chimney at 12:26:09 A.M. on Halloween you are precise but almost surely inaccurate. “You are accurate if you are within the stated limits of uncertainty,” Van Schmus said, defining the goal of geochronology. “You want to be both accurate and precise. You want your limits narrow, and you want the window to include the correct answer.” He added, “Precise means good lab technique. You can be very precise and very inaccurate. The idea is to narrow the window of accuracy and be, at the same time, precise.” In Precambrian geochronology, a window of two million years is an extremely narrow one. A date like 1746—plus or minus two—is very precise.

  The rock component most widely employed in the quest for precision in deepest time is zirconium orthosilicate, the mineral zircon. As a result of advances in laboratory technique, zircons have become, in Van Schmus’s phrase, “the workhorse of the Precambrian,” because they yield with greater accuracy and considerable precision the primary crystallization ages of the rocks they come from. Zircons for centuries have had an independent status as gems. They are pyramids or prisms, and have an adamantine lustre and considerable variety of color. Starlite is a blue zircon from Thailand. The sometimes smoky or colorless or pale yellow zircon of Sri Lanka is a gemstone called jargon. The word “zircon” is derived from the Old French jargon, which had the same ultimate source as “gargle.”

  Most zircons are not of a size to flash bright color from jewelry, or even to be easily seen. They are typically less than a tenth of a millimetre in their longest dimension. They are in sandstones, schists, gneisses. They are in nearly the whole family of solidified magmas of which granite is the most familiar. They are in rhyolite —the rock result of granitic magma erupting on the surface of the earth. And of course they are in all the sedimentary rocks and the beaches and the river placers that derive from the granites and the other sources. Unfortunately, they are rarely in basaltic rocks—the essence of, among other things, island arcs and ocean crust.

  Van Schmus and his associates in their “clean lab” at the University of Kansas grind rock to powder, concentrate the heavy minerals, and pour them into a flask of bromoform, a colorless liquid consisting of carbon, hydrogen, and bromine, closely analogous to chloroform. The quartz and other relatively light materials float on the bromoform. The heavies—sphene, pyrite, zircons, magnetite, apatite, hornblende, garnet—fall to the bottom. The heavies go into a Frantz separator, where a very strongly focused magnetic field pulls out the magnetite, hornblende, and garnet. Remaining are sphene, apatite, pyrite, and zircons, all of which are dropped into methylene iodide. In methylene iodide, apatite floats. It is removed, and now we are down to zircons, pyrite, and sphene. An acid dissolves the pyrite, ignoring the zircons and the sphene. A Frantz separator, its power adjusted, takes the sphene away from the zircons. Thus isolated and concentrated, a tiny pile of zircons looked to me like heavy, glinting dust. Under a microscope, they were elongate tetragonal pyramids, of a light clear amber. They resembled capsules of Vitamin E.

  As magma cools—as, for example, a great intruded batholith slowly and progressively hardens and differing magmatic juices sequentially develop into various minerals and rocks—the attraction of zirconium ions for silicon and oxygen is what makes zircons. “They bond very strongly,” Van Schmus said. “Other things sneak in—uranium, hafnium, thorium. Lead begins forming from the uranium and thorium by radioactive decay. To make zircons, there has to be a lot of silicon around, so zircons are commonly associated with quartz-bearing igneous rocks.”

  Although they are semi-microscopic, zircons are nonetheless polished in the lab to grind away their outer parts—where disruption might have occurred in various ways—and get down to pristine samples. This improves the accuracy of the age analysis that follows. In a process that takes a few days, the zircons are dissolved by hydrofluoric acid in teflon bombs. The bombs are not unlike pressure cookers. Ion-exchange chromatography, a chemical extraction process, takes the uranium and lead out of the solution. Analysis of the abundance and composition of the lead results in the date when the zircon first crystallized. This method—developed by Thomas E. Krogh at the Carnegie Institution in Washington—has dramatically simplified a task that used to be very difficult. A full uranium/lead analysis can be done with zircons aggregately weighing fivemillionths to ten-millionths of a gram.

  Among the several characteristics that have enlofted zircons to the state of the art of Precambrian dating is the fact that once they form they do not easily recrystallize. In huge tectonic events across time, they can go through medium-grade and even high-grade metamorphism and survive. In the clean lab, three or four weeks may be needed to get an age from one rock. There’s no reasonable alternative, Van Schmus remarked. “Nowadays you don’t feel that you have a totally accurate date unless you’ve done it with zircons.”

  Cycles of mountain building that were long thought to have taken place over a period of three hundred million years have recently been narrowed down to a few tens of
millions of years through zircon dating. The approximate age of the oldest known rock on earth—3.96 billion years—was determined from zircons within the rock. Because zircons are rare in basaltic rocks, they seldom can be used directly for the dating of remnants of Precambrian ocean crust. Near Salida, however, those gabbros, pillow basalts, and other ocean-crustal rocks from the beginnings of Colorado were interlayered with rhyolite tuff that would have erupted on an island surface. The rhyolite contained zircons. A date obtained from the zircons by Van Schmus’s colleague M. E. Bickford had inferentially provided the dates of origin of the gabbros and basalts: 1728—plus or minus six million years.

  “We cannot make any correlations in the Precambrian based just on what the rocks look like,” Van Schmus reiterated. “We have no index fossils. So the only thing we have to connect various disjointed pieces of Precambrian crust are radiometric dating techniques. Isotopes have given us a great mapping tool. Beginning in the fifties and early sixties, two principal techniques were potassium/ argon and rubidium/strontium, which had limitations in terms of both accuracy and precision. They were fairly easy to carry out, and a great deal of early information on the Precambrian, particularly of North America, was done with those two dating techniques. A smaller but significant amount of work was done on uranium/lead dating in zircons—the technique was very difficult to work with and there were only a few practitioners in North America then. The methodological breakthroughs of the seventies made the uranium/ lead dating technique very convenient. Since then, most of our precise age information has come from uranium/lead dating of zircons in igneous rocks. Argon/argon thermochronology has matured substantially in the last decade, and it’s being used extensively for the study of young rocks and for the study of metamorphic rocks, looking particularly at the last phase of metamorphism in tectonic environments. That is not something that concerns us very much with the Precambrian basement.”