A short history of nearly everything

13 BANG!

PEOPLE KNEW FOR a long time that there was something odd about the earth beneath Manson, Iowa. In 1912, a man drilling a well for the town water supply reported bringing up a lot of strangely deformed rock-“crystalline clast breccia with a melt matrix” and “overturned ejecta flap,” as it was later described in an official report. The water was odd too. It was almost as soft as rainwater. Naturally occurring soft water had never been found in Iowa before.

Though Manson’s strange rocks and silken waters were matters of curiosity, forty-one years would pass before a team from the University of Iowa got around to making a trip to the community, then as now a town of about two thousand people in the northwest part of the state. In 1953, after sinking a series of experimental bores, university geologists agreed that the site was indeed anomalous and attributed the deformed rocks to some ancient, unspecified volcanic action. This was in keeping with the wisdom of the day, but it was also about as wrong as a geological conclusion can get.

The trauma to Manson’s geology had come not from within the Earth, but from at least 100 million miles beyond. Sometime in the very ancient past, when Manson stood on the edge of a shallow sea, a rock about a mile and a half across, weighing ten billion tons and traveling at perhaps two hundred times the speed of sound ripped through the atmosphere and punched into the Earth with a violence and suddenness that we can scarcely imagine. Where Manson now stands became in an instant a hole three miles deep and more than twenty miles across. The limestone that elsewhere gives Iowa its hard mineralized water was obliterated and replaced by the shocked basement rocks that so puzzled the water driller in 1912.

The Manson impact was the biggest thing that has ever occurred on the mainland United States. Of any type. Ever. The crater it left behind was so colossal that if you stood on one edge you would only just be able to see the other side on a good day. It would make the Grand Canyon look quaint and trifling. Unfortunately for lovers of spectacle, 2.5 million years of passing ice sheets filled the Manson crater right to the top with rich glacial till, then graded it smooth, so that today the landscape at Manson, and for miles around, is as flat as a tabletop. Which is of course why no one has ever heard of the Manson crater.

At the library in Manson they are delighted to show you a collection of newspaper articles and a box of core samples from a 1991-92 drilling program-indeed, they positively bustle to produce them-but you have to ask to see them. Nothing permanent is on display, and nowhere in the town is there any historical marker.

To most people in Manson the biggest thing ever to happen was a tornado that rolled up Main Street in 1979, tearing apart the business district. One of the advantages of all that surrounding flatness is that you can see danger from a long way off. Virtually the whole town turned out at one end of Main Street and watched for half an hour as the tornado came toward them, hoping it would veer off, then prudently scampered when it did not. Four of them, alas, didn’t move quite fast enough and were killed. Every June now Manson has a weeklong event called Crater Days, which was dreamed up as a way of helping people forget that unhappy anniversary. It doesn’t really have anything to do with the crater. Nobody’s figured out a way to capitalize on an impact site that isn’t visible.

“Very occasionally we get people coming in and asking where they should go to see the crater and we have to tell them that there is nothing to see,” says Anna Schlapkohl, the town’s friendly librarian. “Then they go away kind of disappointed.” However, most people, including most Iowans, have never heard of the Manson crater. Even for geologists it barely rates a footnote. But for one brief period in the 1980s, Manson was the most geologically exciting place on Earth.

The story begins in the early 1950s when a bright young geologist named Eugene Shoemaker paid a visit to Meteor Crater in Arizona. Today Meteor Crater is the most famous impact site on Earth and a popular tourist attraction. In those days, however, it didn’t receive many visitors and was still often referred to as Barringer Crater, after a wealthy mining engineer named Daniel M. Barringer who had staked a claim on it in 1903. Barringer believed that the crater had been formed by a ten-million-ton meteor, heavily freighted with iron and nickel, and it was his confident expectation that he would make a fortune digging it out. Unaware that the meteor and everything in it would have been vaporized on impact, he wasted a fortune, and the next twenty-six years, cutting tunnels that yielded nothing.

By the standards of today, crater research in the early 1900s was a trifle unsophisticated, to say the least. The leading early investigator, G. K. Gilbert of Columbia University, modeled the effects of impacts by flinging marbles into pans of oatmeal. (For reasons I cannot supply, Gilbert conducted these experiments not in a laboratory at Columbia but in a hotel room.) Somehow from this Gilbert concluded that the Moon’s craters were indeed formed by impacts-in itself quite a radical notion for the time-but that the Earth’s were not. Most scientists refused to go even that far. To them, the Moon’s craters were evidence of ancient volcanoes and nothing more. The few craters that remained evident on Earth (most had been eroded away) were generally attributed to other causes or treated as fluky rarities.

By the time Shoemaker came along, a common view was that Meteor Crater had been formed by an underground steam explosion. Shoemaker knew nothing about underground steam explosions-he couldn’t: they don’t exist-but he did know all about blast zones. One of his first jobs out of college was to study explosion rings at the Yucca Flats nuclear test site in Nevada. He concluded, as Barringer had before him, that there was nothing at Meteor Crater to suggest volcanic activity, but that there were huge distributions of other stuff-anomalous fine silicas and magnetites principally-that suggested an impact from space. Intrigued, he began to study the subject in his spare time.

Working first with his colleague Eleanor Helin and later with his wife, Carolyn, and associate David Levy, Shoemaker began a systematic survey of the inner solar system. They spent one week each month at the Palomar Observatory in California looking for objects, asteroids primarily, whose trajectories carried them across Earth’s orbit.

“At the time we started, only slightly more than a dozen of these things had ever been discovered in the entire course of astronomical observation,” Shoemaker recalled some years later in a television interview. “Astronomers in the twentieth century essentially abandoned the solar system,” he added. “Their attention was turned to the stars, the galaxies.”

What Shoemaker and his colleagues found was that there was more risk out there-a great deal more-than anyone had ever imagined.

Asteroids, as most people know, are rocky objects orbiting in loose formation in a belt between Mars and Jupiter. In illustrations they are always shown as existing in a jumble, but in fact the solar system is quite a roomy place and the average asteroid actually will be about a million miles from its nearest neighbor. Nobody knows even approximately how many asteroids there are tumbling through space, but the number is thought to be probably not less than a billion. They are presumed to be planets that never quite made it, owing to the unsettling gravitational pull of Jupiter, which kept-and keeps-them from coalescing.

When asteroids were first detected in the 1800s-the very first was discovered on the first day of the century by a Sicilian named Giuseppi Piazzi-they were thought to be planets, and the first two were named Ceres and Pallas. It took some inspired deductions by the astronomer William Herschel to work out that they were nowhere near planet sized but much smaller. He called them asteroids-Latin for “starlike”-which was slightly unfortunate as they are not like stars at all. Sometimes now they are more accurately called planetoids.

Finding asteroids became a popular activity in the 1800s, and by the end of the century about a thousand were known. The problem was that no one was systematically recording them. By the early 1900s, it had often become impossible to know whether an asteroid that popped into view was new or simply one that had been noted earlier and then lost track of. By this time, too, astrophysics had moved on so much that few astronomers wanted to devote their lives to anything as mundane as rocky planetoids. Only a few astronomers, notably Gerard Kuiper, the Dutch-born astronomer for whom the Kuiper belt of comets is named, took any interest in the solar system at all. Thanks to his work at the McDonald Observatory in Texas, followed later by work done by others at the Minor Planet Center in Cincinnati and the Spacewatch project in Arizona, a long list of lost asteroids was gradually whittled down until by the close of the twentieth century only one known asteroid was unaccounted for-an object called 719 Albert. Last seen in October 1911, it was finally tracked down in 2000 after being missing for eighty-nine years.

So from the point of view of asteroid research the twentieth century was essentially just a long exercise in bookkeeping. It is really only in the last few years that astronomers have begun to count and keep an eye on the rest of the asteroid community. As of July 2001, twenty-six thousand asteroids had been named and identified-half in just the previous two years. With up to a billion to identify, the count obviously has barely begun.

In a sense it hardly matters. Identifying an asteroid doesn’t make it safe. Even if every asteroid in the solar system had a name and known orbit, no one could say what perturbations might send any of them hurtling toward us. We can’t forecast rock disturbances on our own surface. Put them adrift in space and what they might do is beyond guessing. Any asteroid out there that has our name on it is very likely to have no other.

Think of the Earth’s orbit as a kind of freeway on which we are the only vehicle, but which is crossed regularly by pedestrians who don’t know enough to look before stepping off the curb. At least 90 percent of these pedestrians are quite unknown to us. We don’t know where they live, what sort of hours they keep, how often they come our way. All we know is that at some point, at uncertain intervals, they trundle across the road down which we are cruising at sixty-six thousand miles an hour. As Steven Ostro of the Jet Propulsion Laboratory has put it, “Suppose that there was a button you could push and you could light up all the Earth-crossing asteroids larger than about ten meters, there would be over 100 million of these objects in the sky.” In short, you would see not a couple of thousand distant twinkling stars, but millions upon millions upon millions of nearer, randomly moving objects-“all of which are capable of colliding with the Earth and all of which are moving on slightly different courses through the sky at different rates. It would be deeply unnerving.” Well, be unnerved because it is there. We just can’t see it.

Altogether it is thought-though it is really only a guess, based on extrapolating from cratering rates on the Moon-that some two thousand asteroids big enough to imperil civilized existence regularly cross our orbit. But even a small asteroid-the size of a house, say-could destroy a city. The number of these relative tiddlers in Earth-crossing orbits is almost certainly in the hundreds of thousands and possibly in the millions, and they are nearly impossible to track.

The first one wasn’t spotted until 1991, and that was after it had already gone by. Named 1991 BA, it was noticed as it sailed past us at a distance of 106,000 miles-in cosmic terms the equivalent of a bullet passing through one’s sleeve without touching the arm. Two years later, another, somewhat larger asteroid missed us by just 90,000 miles-the closest pass yet recorded. It, too, was not seen until it had passed and would have arrived without warning. According to Timothy Ferris, writing in the New Yorker, such near misses probably happen two or three times a week and go unnoticed.

An object a hundred yards across couldn’t be picked up by any Earth-based telescope until it was within just a few days of us, and that is only if a telescope happened to be trained on it, which is unlikely because even now the number of people searching for such objects is modest. The arresting analogy that is always made is that the number of people in the world who are actively searching for asteroids is fewer than the staff of a typical McDonald’s restaurant. (It is actually somewhat higher now. But not much.)

While Gene Shoemaker was trying to get people galvanized about the potential dangers of the inner solar system, another development-wholly unrelated on the face of it-was quietly unfolding in Italy with the work of a young geologist from the Lamont Doherty Laboratory at Columbia University. In the early 1970s, Walter Alvarez was doing fieldwork in a comely defile known as the Bottaccione Gorge, near the Umbrian hill town of Gubbio, when he grew curious about a thin band of reddish clay that divided two ancient layers of limestone-one from the Cretaceous period, the other from the Tertiary. This is a point known to geology as the KT boundary,[27] and it marks the time, sixty-five million years ago, when the dinosaurs and roughly half the world’s other species of animals abruptly vanish from the fossil record. Alvarez wondered what it was about a thin lamina of clay, barely a quarter of an inch thick, that could account for such a dramatic moment in Earth’s history.

At the time the conventional wisdom about the dinosaur extinction was the same as it had been in Charles Lyell’s day a century earlier-namely that the dinosaurs had died out over millions of years. But the thinness of the clay layer clearly suggested that in Umbria, if nowhere else, something rather more abrupt had happened. Unfortunately in the 1970s no tests existed for determining how long such a deposit might have taken to accumulate.

In the normal course of things, Alvarez almost certainly would have had to leave the problem at that, but luckily he had an impeccable connection to someone outside his discipline who could help-his father, Luis. Luis Alvarez was an eminent nuclear physicist; he had won the Nobel Prize for physics the previous decade. He had always been mildly scornful of his son’s attachment to rocks, but this problem intrigued him. It occurred to him that the answer might lie in dust from space.

Every year the Earth accumulates some thirty thousand metric tons of “cosmic spherules”-space dust in plainer language-which would be quite a lot if you swept it into one pile, but is infinitesimal when spread across the globe. Scattered through this thin dusting are exotic elements not normally much found on Earth. Among these is the element iridium, which is a thousand times more abundant in space than in the Earth’s crust (because, it is thought, most of the iridium on Earth sank to the core when the planet was young).

Alvarez knew that a colleague of his at the Lawrence Berkeley Laboratory in California, Frank Asaro, had developed a technique for measuring very precisely the chemical composition of clays using a process called neutron activation analysis. This involved bombarding samples with neutrons in a small nuclear reactor and carefully counting the gamma rays that were emitted; it was extremely finicky work. Previously Asaro had used the technique to analyze pieces of pottery, but Alvarez reasoned that if they measured the amount of one of the exotic elements in his son’s soil samples and compared that with its annual rate of deposition, they would know how long it had taken the samples to form. On an October afternoon in 1977, Luis and Walter Alvarez dropped in on Asaro and asked him if he would run the necessary tests for them.

It was really quite a presumptuous request. They were asking Asaro to devote months to making the most painstaking measurements of geological samples merely to confirm what seemed entirely self-evident to begin with-that the thin layer of clay had been formed as quickly as its thinness suggested. Certainly no one expected his survey to yield any dramatic breakthroughs.

“Well, they were very charming, very persuasive,” Asaro recalled in an interview in 2002. “And it seemed an interesting challenge, so I agreed to try. Unfortunately, I had a lot of other work on, so it was eight months before I could get to it.” He consulted his notes from the period. “On June 21, 1978, at 1:45 p.m., we put a sample in the detector. It ran for 224 minutes and we could see we were getting interesting results, so we stopped it and had a look.”

The results were so unexpected, in fact, that the three scientists at first thought they had to be wrong. The amount of iridium in the Alvarez sample was more than three hundred times normal levels-far beyond anything they might have predicted. Over the following months Asaro and his colleague Helen Michel worked up to thirty hours at a stretch (“Once you started you couldn’t stop,” Asaro explained) analyzing samples, always with the same results. Tests on other samples-from Denmark, Spain, France, New Zealand, Antarctica-showed that the iridium deposit was worldwide and greatly elevated everywhere, sometimes by as much as five hundred times normal levels. Clearly something big and abrupt, and probably cataclysmic, had produced this arresting spike.

After much thought, the Alvarezes concluded that the most plausible explanation-plausible to them, at any rate-was that the Earth had been struck by an asteroid or comet.

The idea that the Earth might be subjected to devastating impacts from time to time was not quite as new as it is now sometimes presented. As far back as 1942, a Northwestern University astrophysicist named Ralph B. Baldwin had suggested such a possibility in an article in Popular Astronomy magazine. (He published the article there because no academic publisher was prepared to run it.) And at least two well-known scientists, the astronomer Ernst Öpik and the chemist and Nobel laureate Harold Urey, had also voiced support for the notion at various times. Even among paleontologists it was not unknown. In 1956 a professor at Oregon State University, M. W. de Laubenfels, writing in the Journal of Paleontology, had actually anticipated the Alvarez theory by suggesting that the dinosaurs may have been dealt a death blow by an impact from space, and in 1970 the president of the American Paleontological Society, Dewey J. McLaren, proposed at the group’s annual conference the possibility that an extraterrestrial impact may have been the cause of an earlier event known as the Frasnian extinction.

As if to underline just how un-novel the idea had become by this time, in 1979 a Hollywood studio actually produced a movie called Meteor (“It’s five miles wide . . . It’s coming at 30,000 m.p.h.-and there’s no place to hide!”) starring Henry Fonda, Natalie Wood, Karl Malden, and a very large rock.

So when, in the first week of 1980, at a meeting of the American Association for the Advancement of Science, the Alvarezes announced their belief that the dinosaur extinction had not taken place over millions of years as part of some slow inexorable process, but suddenly in a single explosive event, it shouldn’t have come as a shock.

But it did. It was received everywhere, but particularly in the paleontological community, as an outrageous heresy.

“Well, you have to remember,” Asaro recalls, “that we were amateurs in this field. Walter was a geologist specializing in paleomagnetism, Luis was a physicist and I was a nuclear chemist. And now here we were telling paleontologists that we had solved a problem that had eluded them for over a century. It’s not terribly surprising that they didn’t embrace it immediately.” As Luis Alvarez joked: “We were caught practicing geology without a license.”

But there was also something much deeper and more fundamentally abhorrent in the impact theory. The belief that terrestrial processes were gradual had been elemental in natural history since the time of Lyell. By the 1980s, catastrophism had been out of fashion for so long that it had become literally unthinkable. For most geologists the idea of a devastating impact was, as Eugene Shoemaker noted, “against their scientific religion.”

Nor did it help that Luis Alvarez was openly contemptuous of paleontologists and their contributions to scientific knowledge. “They’re really not very good scientists. They’re more like stamp collectors,” he wrote in the New York Times in an article that stings yet.

Opponents of the Alvarez theory produced any number of alternative explanations for the iridium deposits-for instance, that they were generated by prolonged volcanic eruptions in India called the Deccan Traps-and above all insisted that there was no proof that the dinosaurs disappeared abruptly from the fossil record at the iridium boundary. One of the most vigorous opponents was Charles Officer of Dartmouth College. He insisted that the iridium had been deposited by volcanic action even while conceding in a newspaper interview that he had no actual evidence of it. As late as 1988 more than half of all American paleontologists contacted in a survey continued to believe that the extinction of the dinosaurs was in no way related to an asteroid or cometary impact.

The one thing that would most obviously support the Alvarezes’ theory was the one thing they didn’t have-an impact site. Enter Eugene Shoemaker. Shoemaker had an Iowa connection-his daughter-in-law taught at the University of Iowa-and he was familiar with the Manson crater from his own studies. Thanks to him, all eyes now turned to Iowa.

Geology is a profession that varies from place to place. In Iowa, a state that is flat and stratigraphically uneventful, it tends to be comparatively serene. There are no Alpine peaks or grinding glaciers, no great deposits of oil or precious metals, not a hint of a pyroclastic flow. If you are a geologist employed by the state of Iowa, a big part of the work you do is to evaluate Manure Management Plans, which all the state’s “animal confinement operators”-hog farmers to the rest of us-are required to file periodically. There are fifteen million hogs in Iowa, so a lot of manure to manage. I’m not mocking this at all-it’s vital and enlightened work; it keeps Iowa’s water clean-but with the best will in the world it’s not exactly dodging lava bombs on Mount Pinatubo or scrabbling over crevasses on the Greenland ice sheet in search of ancient life-bearing quartzes. So we may well imagine the flutter of excitement that swept through the Iowa Department of Natural Resources when in the mid-1980s the world’s geological attention focused on Manson and its crater.

Trowbridge Hall in Iowa City is a turn-of-the-century pile of red brick that houses the University of Iowa’s Earth Sciences department and-way up in a kind of garret-the geologists of the Iowa Department of Natural Resources. No one now can remember quite when, still less why, the state geologists were placed in an academic facility, but you get the impression that the space was conceded grudgingly, for the offices are cramped and low-ceilinged and not very accessible. When being shown the way, you half expect to be taken out onto a roof ledge and helped in through a window.

Ray Anderson and Brian Witzke spend their working lives up here amid disordered heaps of papers, journals, furled charts, and hefty specimen stones. (Geologists are never at a loss for paperweights.) It’s the kind of space where if you want to find anything-an extra chair, a coffee cup, a ringing telephone-you have to move stacks of documents around.

“Suddenly we were at the center of things,” Anderson told me, gleaming at the memory of it, when I met him and Witzke in their offices on a dismal, rainy morning in June. “It was a wonderful time.”

I asked them about Gene Shoemaker, a man who seems to have been universally revered. “He was just a great guy,” Witzke replied without hesitation. “If it hadn’t been for him, the whole thing would never have gotten off the ground. Even with his support, it took two years to get it up and running. Drilling’s an expensive business-about thirty-five dollars a foot back then, more now, and we needed to go down three thousand feet.”

“Sometimes more than that,” Anderson added.

“Sometimes more than that,” Witzke agreed. “And at several locations. So you’re talking a lot of money. Certainly more than our budget would allow.”

So a collaboration was formed between the Iowa Geological Survey and the U.S. Geological Survey.

“At least we thought it was a collaboration,” said Anderson, producing a small pained smile.

“It was a real learning curve for us,” Witzke went on. “There was actually quite a lot of bad science going on throughout the period-people rushing in with results that didn’t always stand up to scrutiny.” One of those moments came at the annual meeting of the American Geophysical Union in 1985, when Glenn Izett and C. L. Pillmore of the U.S. Geological Survey announced that the Manson crater was of the right age to have been involved with the dinosaurs’ extinction. The declaration attracted a good deal of press attention but was unfortunately premature. A more careful examination of the data revealed that Manson was not only too small, but also nine million years too early.

The first Anderson or Witzke learned of this setback to their careers was when they arrived at a conference in South Dakota and found people coming up to them with sympathetic looks and saying: “We hear you lost your crater.” It was the first they knew that Izett and the other USGS scientists had just announced refined figures revealing that Manson couldn’t after all have been the extinction crater.

“It was pretty stunning,” recalls Anderson. “I mean, we had this thing that was really important and then suddenly we didn’t have it anymore. But even worse was the realization that the people we thought we’d been collaborating with hadn’t bothered to share with us their new findings.”

“Why not?”

He shrugged. “Who knows? Anyway, it was a pretty good insight into how unattractive science can get when you’re playing at a certain level.”

The search moved elsewhere. By chance in 1990 one of the searchers, Alan Hildebrand of the University of Arizona, met a reporter from the Houston Chronicle who happened to know about a large, unexplained ring formation, 120 miles wide and 30 miles deep, under Mexico’s Yucatán Peninsula at Chicxulub, near the city of Progreso, about 600 miles due south of New Orleans. The formation had been found by Pemex, the Mexican oil company, in 1952-the year, coincidentally, that Gene Shoemaker first visited Meteor Crater in Arizona-but the company’s geologists had concluded that it was volcanic, in line with the thinking of the day. Hildebrand traveled to the site and decided fairly swiftly that they had their crater. By early 1991 it had been established to nearly everyone’s satisfaction that Chicxulub was the impact site.

Still, many people didn’t quite grasp what an impact could do. As Stephen Jay Gould recalled in one of his essays: “I remember harboring some strong initial doubts about the efficacy of such an event . . . [W]hy should an object only six miles across wreak such havoc upon a planet with a diameter of eight thousand miles?”

Conveniently a natural test of the theory arose when the Shoemakers and Levy discovered Comet Shoemaker-Levy 9, which they soon realized was headed for Jupiter. For the first time, humans would be able to witness a cosmic collision-and witness it very well thanks to the new Hubble space telescope. Most astronomers, according to Curtis Peebles, expected little, particularly as the comet was not a coherent sphere but a string of twenty-one fragments. “My sense,” wrote one, “is that Jupiter will swallow these comets up without so much as a burp.” One week before the impact, Nature ran an article, “The Big Fizzle Is Coming,” predicting that the impact would constitute nothing more than a meteor shower.

The impacts began on July 16, 1994, went on for a week and were bigger by far than anyone-with the possible exception of Gene Shoemaker-expected. One fragment, known as Nucleus G, struck with the force of about six million megatons-seventy-five times more than all the nuclear weaponry in existence. Nucleus G was only about the size of a small mountain, but it created wounds in the Jovian surface the size of Earth. It was the final blow for critics of the Alvarez theory.

Luis Alvarez never knew of the discovery of the Chicxulub crater or of the Shoemaker-Levy comet, as he died in 1988. Shoemaker also died early. On the third anniversary of the Shoemaker-Levy impact, he and his wife were in the Australian outback, where they went every year to search for impact sites. On a dirt track in the Tanami Desert-normally one of the emptiest places on Earth-they came over a slight rise just as another vehicle was approaching. Shoemaker was killed instantly, his wife injured. Part of his ashes were sent to the Moon aboard the Lunar Prospector spacecraft. The rest were scattered around Meteor Crater.

Anderson and Witzke no longer had the crater that killed the dinosaurs, “but we still had the largest and most perfectly preserved impact crater in the mainland United States,” Anderson said. (A little verbal dexterity is required to keep Manson’s superlative status. Other craters are larger-notably, Chesapeake Bay, which was recognized as an impact site in 1994-but they are either offshore or deformed.) “Chicxulub is buried under two to three kilometers of limestone and mostly offshore, which makes it difficult to study,” Anderson went on, “while Manson is really quite accessible. It’s because it is buried that it is actually comparatively pristine.”

I asked them how much warning we would receive if a similar hunk of rock was coming toward us today.

“Oh, probably none,” said Anderson breezily. “It wouldn’t be visible to the naked eye until it warmed up, and that wouldn’t happen until it hit the atmosphere, which would be about one second before it hit the Earth. You’re talking about something moving many tens of times faster than the fastest bullet. Unless it had been seen by someone with a telescope, and that’s by no means a certainty, it would take us completely by surprise.”

How hard an impactor hits depends on a lot of variables-angle of entry, velocity and trajectory, whether the collision is head-on or from the side, and the mass and density of the impacting object, among much else-none of which we can know so many millions of years after the fact. But what scientists can do-and Anderson and Witzke have done-is measure the impact site and calculate the amount of energy released. From that they can work out plausible scenarios of what it must have been like-or, more chillingly, would be like if it happened now.

An asteroid or comet traveling at cosmic velocities would enter the Earth’s atmosphere at such a speed that the air beneath it couldn’t get out of the way and would be compressed, as in a bicycle pump. As anyone who has used such a pump knows, compressed air grows swiftly hot, and the temperature below it would rise to some 60,000 Kelvin, or ten times the surface temperature of the Sun. In this instant of its arrival in our atmosphere, everything in the meteor’s path-people, houses, factories, cars-would crinkle and vanish like cellophane in a flame.

One second after entering the atmosphere, the meteorite would slam into the Earth’s surface, where the people of Manson had a moment before been going about their business. The meteorite itself would vaporize instantly, but the blast would blow out a thousand cubic kilometers of rock, earth, and superheated gases. Every living thing within 150 miles that hadn’t been killed by the heat of entry would now be killed by the blast. Radiating outward at almost the speed of light would be the initial shock wave, sweeping everything before it.

For those outside the zone of immediate devastation, the first inkling of catastrophe would be a flash of blinding light-the brightest ever seen by human eyes-followed an instant to a minute or two later by an apocalyptic sight of unimaginable grandeur: a roiling wall of darkness reaching high into the heavens, filling an entire field of view and traveling at thousands of miles an hour. Its approach would be eerily silent since it would be moving far beyond the speed of sound. Anyone in a tall building in Omaha or Des Moines, say, who chanced to look in the right direction would see a bewildering veil of turmoil followed by instantaneous oblivion.

Within minutes, over an area stretching from Denver to Detroit and encompassing what had once been Chicago, St. Louis, Kansas City, the Twin Cities-the whole of the Midwest, in short-nearly every standing thing would be flattened or on fire, and nearly every living thing would be dead. People up to a thousand miles away would be knocked off their feet and sliced or clobbered by a blizzard of flying projectiles. Beyond a thousand miles the devastation from the blast would gradually diminish.

But that’s just the initial shockwave. No one can do more than guess what the associated damage would be, other than that it would be brisk and global. The impact would almost certainly set off a chain of devastating earthquakes. Volcanoes across the globe would begin to rumble and spew. Tsunamis would rise up and head devastatingly for distant shores. Within an hour, a cloud of blackness would cover the planet, and burning rock and other debris would be pelting down everywhere, setting much of the planet ablaze. It has been estimated that at least a billion and a half people would be dead by the end of the first day. The massive disturbances to the ionosphere would knock out communications systems everywhere, so survivors would have no idea what was happening elsewhere or where to turn. It would hardly matter. As one commentator has put it, fleeing would mean “selecting a slow death over a quick one. The death toll would be very little affected by any plausible relocation effort, since Earth’s ability to support life would be universally diminished.”

The amount of soot and floating ash from the impact and following fires would blot out the sun, certainly for months, possibly for years, disrupting growing cycles. In 2001 researchers at the California Institute of Technology analyzed helium isotopes from sediments left from the later KT impact and concluded that it affected Earth’s climate for about ten thousand years. This was actually used as evidence to support the notion that the extinction of dinosaurs was swift and emphatic-and so it was in geological terms. We can only guess how well, or whether, humanity would cope with such an event.

And in all likelihood, remember, this would come without warning, out of a clear sky.

But let’s assume we did see the object coming. What would we do? Everyone assumes we would send up a nuclear warhead and blast it to smithereens. The idea has some problems, however. First, as John S. Lewis notes, our missiles are not designed for space work. They haven’t the oomph to escape Earth’s gravity and, even if they did, there are no mechanisms to guide them across tens of millions of miles of space. Still less could we send up a shipload of space cowboys to do the job for us, as in the movie Armageddon; we no longer possess a rocket powerful enough to send humans even as far as the Moon. The last rocket that could, Saturn 5, was retired years ago and has never been replaced. Nor could we quickly build a new one because, amazingly, the plans for Saturn launchers were destroyed as part of a NASA housecleaning exercise.

Even if we did manage somehow to get a warhead to the asteroid and blasted it to pieces, the chances are that we would simply turn it into a string of rocks that would slam into us one after the other in the manner of Comet Shoemaker-Levy on Jupiter-but with the difference that now the rocks would be intensely radioactive. Tom Gehrels, an asteroid hunter at the University of Arizona, thinks that even a year’s warning would probably be insufficient to take appropriate action. The greater likelihood, however, is that we wouldn’t see any object-even a comet-until it was about six months away, which would be much too late. Shoemaker-Levy 9 had been orbiting Jupiter in a fairly conspicuous manner since 1929, but it took over half a century before anyone noticed.

Interestingly, because these things are so difficult to compute and must incorporate such a significant margin of error, even if we knew an object was heading our way we wouldn’t know until nearly the end-the last couple of weeks anyway-whether collision was certain. For most of the time of the object’s approach we would exist in a kind of cone of uncertainty. It would certainly be the most interesting few months in the history of the world. And imagine the party if it passed safely.

“So how often does something like the Manson impact happen?” I asked Anderson and Witzke before leaving.

“Oh, about once every million years on average,” said Witzke.

“And remember,” added Anderson, “this was a relatively minor event. Do you know how many extinctions were associated with the Manson impact?”

“No idea,” I replied.

“None,” he said, with a strange air of satisfaction. “Not one.”

Of course, Witzke and Anderson added hastily and more or less in unison, there would have been terrible devastation across much of the Earth, as just described, and complete annihilation for hundreds of miles around ground zero. But life is hardy, and when the smoke cleared there were enough lucky survivors from every species that none permanently perished.

The good news, it appears, is that it takes an awful lot to extinguish a species. The bad news is that the good news can never be counted on. Worse still, it isn’t actually necessary to look to space for petrifying danger. As we are about to see, Earth can provide plenty of danger of its own.

14 THE FIRE BELOW

IN THE SUMMER of 1971, a young geologist named Mike Voorhies was scouting around on some grassy farmland in eastern Nebraska, not far from the little town of Orchard, where he had grown up. Passing through a steep-sided gully, he spotted a curious glint in the brush above and clambered up to have a look. What he had seen was the perfectly preserved skull of a young rhinoceros, which had been washed out by recent heavy rains.

A few yards beyond, it turned out, was one of the most extraordinary fossil beds ever discovered in North America, a dried-up water hole that had served as a mass grave for scores of animals-rhinoceroses, zebra-like horses, saber-toothed deer, camels, turtles. All had died from some mysterious cataclysm just under twelve million years ago in the time known to geology as the Miocene. In those days Nebraska stood on a vast, hot plain very like the Serengeti of Africa today. The animals had been found buried under volcanic ash up to ten feet deep. The puzzle of it was that there were not, and never had been, any volcanoes in Nebraska.

Today, the site of Voorhies’s discovery is called Ashfall Fossil Beds State Park, and it has a stylish new visitors’ center and museum, with thoughtful displays on the geology of Nebraska and the history of the fossil beds. The center incorporates a lab with a glass wall through which visitors can watch paleontologists cleaning bones. Working alone in the lab on the morning I passed through was a cheerfully grizzled-looking fellow in a blue work shirt whom I recognized as Mike Voorhies from a BBC television documentary in which he featured. They don’t get a huge number of visitors to Ashfall Fossil Beds State Park-it’s slightly in the middle of nowhere-and Voorhies seemed pleased to show me around. He took me to the spot atop a twenty-foot ravine where he had made his find.

“It was a dumb place to look for bones,” he said happily. “But I wasn’t looking for bones. I was thinking of making a geological map of eastern Nebraska at the time, and really just kind of poking around. If I hadn’t gone up this ravine or the rains hadn’t just washed out that skull, I’d have walked on by and this would never have been found.” He indicated a roofed enclosure nearby, which had become the main excavation site. Some two hundred animals had been found lying together in a jumble.

I asked him in what way it was a dumb place to hunt for bones. “Well, if you’re looking for bones, you really need exposed rock. That’s why most paleontology is done in hot, dry places. It’s not that there are more bones there. It’s just that you have some chance of spotting them. In a setting like this”-he made a sweeping gesture across the vast and unvarying prairie-“you wouldn’t know where to begin. There could be really magnificent stuff out there, but there’s no surface clues to show you where to start looking.”

At first they thought the animals were buried alive, and Voorhies stated as much in a National Geographic article in 1981. “The article called the site a ‘Pompeii of prehistoric animals,’ ” he told me, “which was unfortunate because just afterward we realized that the animals hadn’t died suddenly at all. They were all suffering from something called hypertrophic pulmonary osteodystrophy, which is what you would get if you were breathing a lot of abrasive ash-and they must have been breathing a lot of it because the ash was feet thick for hundreds of miles.” He picked up a chunk of grayish, claylike dirt and crumbled it into my hand. It was powdery but slightly gritty. “Nasty stuff to have to breathe,” he went on, “because it’s very fine but also quite sharp. So anyway they came here to this watering hole, presumably seeking relief, and died in some misery. The ash would have ruined everything. It would have buried all the grass and coated every leaf and turned the water into an undrinkable gray sludge. It couldn’t have been very agreeable at all.”

The BBC documentary had suggested that the existence of so much ash in Nebraska was a surprise. In fact, Nebraska’s huge ash deposits had been known about for a long time. For almost a century they had been mined to make household cleaning powders like Comet and Ajax. But curiously no one had ever thought to wonder where all the ash came from.

“I’m a little embarrassed to tell you,” Voorhies said, smiling briefly, “that the first I thought about it was when an editor at the National Geographic asked me the source of all the ash and I had to confess that I didn’t know. Nobody knew.”

Voorhies sent samples to colleagues all over the western United States asking if there was anything about it that they recognized. Several months later a geologist named Bill Bonnichsen from the Idaho Geological Survey got in touch and told him that the ash matched a volcanic deposit from a place called Bruneau-Jarbidge in southwest Idaho. The event that killed the plains animals of Nebraska was a volcanic explosion on a scale previously unimagined-but big enough to leave an ash layer ten feet deep almost a thousand miles away in eastern Nebraska. It turned out that under the western United States there was a huge cauldron of magma, a colossal volcanic hot spot, which erupted cataclysmically every 600,000 years or so. The last such eruption was just over 600,000 years ago. The hot spot is still there. These days we call it Yellowstone National Park.

We know amazingly little about what happens beneath our feet. It is fairly remarkable to think that Ford has been building cars and baseball has been playing World Series for longer than we have known that the Earth has a core. And of course the idea that the continents move about on the surface like lily pads has been common wisdom for much less than a generation. “Strange as it may seem,” wrote Richard Feynman, “we understand the distribution of matter in the interior of the Sun far better than we understand the interior of the Earth.”

The distance from the surface of Earth to the center is 3,959 miles, which isn’t so very far. It has been calculated that if you sunk a well to the center and dropped a brick into it, it would take only forty-five minutes for it to hit the bottom (though at that point it would be weightless since all the Earth’s gravity would be above and around it rather than beneath it). Our own attempts to penetrate toward the middle have been modest indeed. One or two South African gold mines reach to a depth of two miles, but most mines on Earth go no more than about a quarter of a mile beneath the surface. If the planet were an apple, we wouldn’t yet have broken through the skin. Indeed, we haven’t even come close.

Until slightly under a century ago, what the best-informed scientific minds knew about Earth’s interior was not much more than what a coal miner knew-namely, that you could dig down through soil for a distance and then you’d hit rock and that was about it. Then in 1906, an Irish geologist named R. D. Oldham, while examining some seismograph readings from an earthquake in Guatemala, noticed that certain shock waves had penetrated to a point deep within the Earth and then bounced off at an angle, as if they had encountered some kind of barrier. From this he deduced that the Earth has a core. Three years later a Croatian seismologist named Andrija Mohorovii´c was studying graphs from an earthquake in Zagreb when he noticed a similar odd deflection, but at a shallower level. He had discovered the boundary between the crust and the layer immediately below, the mantle; this zone has been known ever since as the Mohorovii;c discontinuity, or Moho for short.

We were beginning to get a vague idea of the Earth’s layered interior-though it really was only vague. Not until 1936 did a Danish scientist named Inge Lehmann, studying seismographs of earthquakes in New Zealand, discover that there were two cores-an inner one that we now believe to be solid and an outer one (the one that Oldham had detected) that is thought to be liquid and the seat of magnetism.

At just about the time that Lehmann was refining our basic understanding of the Earth’s interior by studying the seismic waves of earthquakes, two geologists at Caltech in California were devising a way to make comparisons between one earthquake and the next. They were Charles Richter and Beno Gutenberg, though for reasons that have nothing to do with fairness the scale became known almost at once as Richter’s alone. (It has nothing to do with Richter either. A modest fellow, he never referred to the scale by his own name, but always called it “the Magnitude Scale.”)

The Richter scale has always been widely misunderstood by nonscientists, though perhaps a little less so now than in its early days when visitors to Richter’s office often asked to see his celebrated scale, thinking it was some kind of machine. The scale is of course more an idea than an object, an arbitrary measure of the Earth’s tremblings based on surface measurements. It rises exponentially, so that a 7.3 quake is fifty times more powerful than a 6.3 earthquake and 2,500 times more powerful than a 5.3 earthquake.

At least theoretically, there is no upper limit for an earthquake-nor, come to that, a lower limit. The scale is a simple measure of force, but says nothing about damage. A magnitude 7 quake happening deep in the mantle-say, four hundred miles down-might cause no surface damage at all, while a significantly smaller one happening just four miles under the surface could wreak widespread devastation. Much, too, depends on the nature of the subsoil, the quake’s duration, the frequency and severity of aftershocks, and the physical setting of the affected area. All this means that the most fearsome quakes are not necessarily the most forceful, though force obviously counts for a lot.

The largest earthquake since the scale’s invention was (depending on which source you credit) either one centered on Prince William Sound in Alaska in March 1964, which measured 9.2 on the Richter scale, or one in the Pacific Ocean off the coast of Chile in 1960, which was initially logged at 8.6 magnitude but later revised upward by some authorities (including the United States Geological Survey) to a truly grand-scale 9.5. As you will gather from this, measuring earthquakes is not always an exact science, particularly when interpreting readings from remote locations. At all events, both quakes were whopping. The 1960 quake not only caused widespread damage across coastal South America, but also set off a giant tsunami that rolled six thousand miles across the Pacific and slapped away much of downtown Hilo, Hawaii, destroying five hundred buildings and killing sixty people. Similar wave surges claimed yet more victims as far away as Japan and the Philippines.

For pure, focused, devastation, however, probably the most intense earthquake in recorded history was one that struck-and essentially shook to pieces-Lisbon, Portugal, on All Saints Day (November 1), 1755. Just before ten in the morning, the city was hit by a sudden sideways lurch now estimated at magnitude 9.0 and shaken ferociously for seven full minutes. The convulsive force was so great that the water rushed out of the city’s harbor and returned in a wave fifty feet high, adding to the destruction. When at last the motion ceased, survivors enjoyed just three minutes of calm before a second shock came, only slightly less severe than the first. A third and final shock followed two hours later. At the end of it all, sixty thousand people were dead and virtually every building for miles reduced to rubble. The San Francisco earthquake of 1906, for comparison, measured an estimated 7.8 on the Richter scale and lasted less than thirty seconds.

Earthquakes are fairly common. Every day on average somewhere in the world there are two of magnitude 2.0 or greater-that’s enough to give anyone nearby a pretty good jolt. Although they tend to cluster in certain places-notably around the rim of the Pacific-they can occur almost anywhere. In the United States, only Florida, eastern Texas, and the upper Midwest seem-so far-to be almost entirely immune. New England has had two quakes of magnitude 6.0 or greater in the last two hundred years. In April 2002, the region experienced a 5.1 magnitude shaking in a quake near Lake Champlain on the New York-Vermont border, causing extensive local damage and (I can attest) knocking pictures from walls and children from beds as far away as New Hampshire.

The most common types of earthquakes are those where two plates meet, as in California along the San Andreas Fault. As the plates push against each other, pressures build up until one or the other gives way. In general, the longer the interval between quakes, the greater the pent-up pressure and thus the greater the scope for a really big jolt. This is a particular worry for Tokyo, which Bill McGuire, a hazards specialist at University College London, describes as “the city waiting to die” (not a motto you will find on many tourism leaflets). Tokyo stands on the boundary of three tectonic plates in a country already well known for its seismic instability. In 1995, as you will remember, the city of Kobe, three hundred miles to the west, was struck by a magnitude 7.2 quake, which killed 6,394 people. The damage was estimated at $99 billion. But that was as nothing-well, as comparatively little-compared with what may await Tokyo.

Tokyo has already suffered one of the most devastating earthquakes in modern times. On September 1, 1923, just before noon, the city was hit by what is known as the Great Kanto quake-an event more than ten times more powerful than Kobe’s earthquake. Two hundred thousand people were killed. Since that time, Tokyo has been eerily quiet, so the strain beneath the surface has been building for eighty years. Eventually it is bound to snap. In 1923, Tokyo had a population of about three million. Today it is approaching thirty million. Nobody cares to guess how many people might die, but the potential economic cost has been put as high as $7 trillion.

Even more unnerving, because they are less well understood and capable of occurring anywhere at any time, are the rarer type of shakings known as intraplate quakes. These happen away from plate boundaries, which makes them wholly unpredictable. And because they come from a much greater depth, they tend to propagate over much wider areas. The most notorious such quakes ever to hit the United States were a series of three in New Madrid, Missouri, in the winter of 1811-12. The adventure started just after midnight on December 16 when people were awakened first by the noise of panicking farm animals (the restiveness of animals before quakes is not an old wives’ tale, but is in fact well established, though not at all understood) and then by an almighty rupturing noise from deep within the Earth. Emerging from their houses, locals found the land rolling in waves up to three feet high and opening up in fissures several feet deep. A strong smell of sulfur filled the air. The shaking lasted for four minutes with the usual devastating effects to property. Among the witnesses was the artist John James Audubon, who happened to be in the area. The quake radiated outward with such force that it knocked down chimneys in Cincinnati four hundred miles away and, according to at least one account, “wrecked boats in East Coast harbors and . . . even collapsed scaffolding erected around the Capitol Building in Washington, D.C.” On January 23 and February 4 further quakes of similar magnitude followed. New Madrid has been silent ever since-but not surprisingly, since such episodes have never been known to happen in the same place twice. As far as we know, they are as random as lightning. The next one could be under Chicago or Paris or Kinshasa. No one can even begin to guess. And what causes these massive intraplate rupturings? Something deep within the Earth. More than that we don’t know.

By the 1960s scientists had grown sufficiently frustrated by how little they understood of the Earth’s interior that they decided to try to do something about it. Specifically, they got the idea to drill through the ocean floor (the continental crust was too thick) to the Moho discontinuity and to extract a piece of the Earth’s mantle for examination at leisure. The thinking was that if they could understand the nature of the rocks inside the Earth, they might begin to understand how they interacted, and thus possibly be able to predict earthquakes and other unwelcome events.

The project became known, all but inevitably, as the Mohole and it was pretty well disastrous. The hope was to lower a drill through 14,000 feet of Pacific Ocean water off the coast of Mexico and drill some 17,000 feet through relatively thin crustal rock. Drilling from a ship in open waters is, in the words of one oceanographer, “like trying to drill a hole in the sidewalks of New York from atop the Empire State Building using a strand of spaghetti.” Every attempt ended in failure. The deepest they penetrated was only about 600 feet. The Mohole became known as the No Hole. In 1966, exasperated with ever-rising costs and no results, Congress killed the project.

Four years later, Soviet scientists decided to try their luck on dry land. They chose a spot on Russia’s Kola Peninsula, near the Finnish border, and set to work with the hope of drilling to a depth of fifteen kilometers. The work proved harder than expected, but the Soviets were commendably persistent. When at last they gave up, nineteen years later, they had drilled to a depth of 12,262 meters, or about 7.6 miles. Bearing in mind that the crust of the Earth represents only about 0.3 percent of the planet’s volume and that the Kola hole had not cut even one-third of the way through the crust, we can hardly claim to have conquered the interior.

Interestingly, even though the hole was modest, nearly everything about it was surprising. Seismic wave studies had led the scientists to predict, and pretty confidently, that they would encounter sedimentary rock to a depth of 4,700 meters, followed by granite for the next 2,300 meters and basalt from there on down. In the event, the sedimentary layer was 50 percent deeper than expected and the basaltic layer was never found at all. Moreover, the world down there was far warmer than anyone had expected, with a temperature at 10,000 meters of 180 degrees centigrade, nearly twice the forecasted level. Most surprising of all was that the rock at that depth was saturated with water-something that had not been thought possible.

Because we can’t see into the Earth, we have to use other techniques, which mostly involve reading waves as they travel through the interior. We also know a little bit about the mantle from what are known as kimberlite pipes, where diamonds are formed. What happens is that deep in the Earth there is an explosion that fires, in effect, a cannonball of magma to the surface at supersonic speeds. It is a totally random event. A kimberlite pipe could explode in your backyard as you read this. Because they come up from such depths-up to 120 miles down-kimberlite pipes bring up all kinds of things not normally found on or near the surface: a rock called peridotite, crystals of olivine, and-just occasionally, in about one pipe in a hundred-diamonds. Lots of carbon comes up with kimberlite ejecta, but most is vaporized or turns to graphite. Only occasionally does a hunk of it shoot up at just the right speed and cool down with the necessary swiftness to become a diamond. It was such a pipe that made Johannesburg the most productive diamond mining city in the world, but there may be others even bigger that we don’t know about. Geologists know that somewhere in the vicinity of northeastern Indiana there is evidence of a pipe or group of pipes that may be truly colossal. Diamonds up to twenty carats or more have been found at scattered sites throughout the region. But no one has ever found the source. As John McPhee notes, it may be buried under glacially deposited soil, like the Manson crater in Iowa, or under the Great Lakes.

So how much do we know about what’s inside the Earth? Very little. Scientists are generally agreed that the world beneath us is composed of four layers-rocky outer crust, a mantle of hot, viscous rock, a liquid outer core, and a solid inner core.[28] We know that the surface is dominated by silicates, which are relatively light and not heavy enough to account for the planet’s overall density. Therefore there must be heavier stuff inside. We know that to generate our magnetic field somewhere in the interior there must be a concentrated belt of metallic elements in a liquid state. That much is universally agreed upon. Almost everything beyond that-how the layers interact, what causes them to behave in the way they do, what they will do at any time in the future-is a matter of at least some uncertainty, and generally quite a lot of uncertainty.

Even the one part of it we can see, the crust, is a matter of some fairly strident debate. Nearly all geology texts tell you that continental crust is three to six miles thick under the oceans, about twenty-five miles thick under the continents, and forty to sixty miles thick under big mountain chains, but there are many puzzling variabilities within these generalizations. The crust beneath the Sierra Nevada Mountains, for instance, is only about nineteen to twenty-five miles thick, and no one knows why. By all the laws of geophysics the Sierra Nevadas should be sinking, as if into quicksand. (Some people think they may be.)

How and when the Earth got its crust are questions that divide geologists into two broad camps-those who think it happened abruptly early in the Earth’s history and those who think it happened gradually and rather later. Strength of feeling runs deep on such matters. Richard Armstrong of Yale proposed an early-burst theory in the 1960s, then spent the rest of his career fighting those who did not agree with him. He died of cancer in 1991, but shortly before his death he “lashed out at his critics in a polemic in an Australian earth science journal that charged them with perpetuating myths,” according to a report in Earth magazine in 1998. “He died a bitter man,” reported a colleague.

The crust and part of the outer mantle together are called the lithosphere (from the Greek lithos, meaning “stone”), which in turn floats on top of a layer of softer rock called the asthenosphere (from Greek words meaning “without strength”), but such terms are never entirely satisfactory. To say that the lithosphere floats on top of the asthenosphere suggests a degree of easy buoyancy that isn’t quite right. Similarly it is misleading to think of the rocks as flowing in anything like the way we think of materials flowing on the surface. The rocks are viscous, but only in the same way that glass is. It may not look it, but all the glass on Earth is flowing downward under the relentless drag of gravity. Remove a pane of really old glass from the window of a European cathedral and it will be noticeably thicker at the bottom than at the top. That is the sort of “flow” we are talking about. The hour hand on a clock moves about ten thousand times faster than the “flowing” rocks of the mantle.

The movements occur not just laterally as the Earth’s plates move across the surface, but up and down as well, as rocks rise and fall under the churning process known as convection. Convection as a process was first deduced by the eccentric Count von Rumford at the end of the eighteenth century. Sixty years later an English vicar named Osmond Fisher presciently suggested that the Earth’s interior might well be fluid enough for the contents to move about, but that idea took a very long time to gain support.

In about 1970, when geophysicists realized just how much turmoil was going on down there, it came as a considerable shock. As Shawna Vogel put it in the book Naked Earth: The New Geophysics: “It was as if scientists had spent decades figuring out the layers of the Earth’s atmosphere-troposphere, stratosphere, and so forth-and then had suddenly found out about wind.”

How deep the convection process goes has been a matter of controversy ever since. Some say it begins four hundred miles down, others two thousand miles below us. The problem, as Donald Trefil has observed, is that “there are two sets of data, from two different disciplines, that cannot be reconciled.” Geochemists say that certain elements on Earth’s surface cannot have come from the upper mantle, but must have come from deeper within the Earth. Therefore the materials in the upper and lower mantle must at least occasionally mix. Seismologists insist that there is no evidence to support such a thesis.

So all that can be said is that at some slightly indeterminate point as we head toward the center of Earth we leave the asthenosphere and plunge into pure mantle. Considering that it accounts for 82 percent of the Earth’s volume and 65 percent of its mass, the mantle doesn’t attract a great deal of attention, largely because the things that interest Earth scientists and general readers alike happen either deeper down (as with magnetism) or nearer the surface (as with earthquakes). We know that to a depth of about a hundred miles the mantle consists predominantly of a type of rock known as peridotite, but what fills the space beyond is uncertain. According to a Nature report, it seems not to be peridotite. More than this we do not know.

Beneath the mantle are the two cores-a solid inner core and a liquid outer one. Needless to say, our understanding of the nature of these cores is indirect, but scientists can make some reasonable assumptions. They know that the pressures at the center of the Earth are sufficiently high-something over three million times those found at the surface-to turn any rock there solid. They also know from Earth’s history (among other clues) that the inner core is very good at retaining its heat. Although it is little more than a guess, it is thought that in over four billion years the temperature at the core has fallen by no more than 200°F. No one knows exactly how hot the Earth’s core is, but estimates range from something over 7,000°F to 13,000°F-about as hot as the surface of the Sun.

The outer core is in many ways even less well understood, though everyone is in agreement that it is fluid and that it is the seat of magnetism. The theory was put forward by E. C. Bullard of Cambridge University in 1949 that this fluid part of the Earth’s core revolves in a way that makes it, in effect, an electrical motor, creating the Earth’s magnetic field. The assumption is that the convecting fluids in the Earth act somehow like the currents in wires. Exactly what happens isn’t known, but it is felt pretty certain that it is connected with the core spinning and with its being liquid. Bodies that don’t have a liquid core-the Moon and Mars, for instance-don’t have magnetism.

We know that Earth’s magnetic field changes in power from time to time: during the age of the dinosaurs, it was up to three times as strong as now. We also know that it reverses itself every 500,000 years or so on average, though that average hides a huge degree of unpredictability. The last reversal was about 750,000 years ago. Sometimes it stays put for millions of years-37 million years appears to be the longest stretch-and at other times it has reversed after as little as 20,000 years. Altogether in the last 100 million years it has reversed itself about two hundred times, and we don’t have any real idea why. It has been called “the greatest unanswered question in the geological sciences.”

We may be going through a reversal now. The Earth’s magnetic field has diminished by perhaps as much as 6 percent in the last century alone. Any diminution in magnetism is likely to be bad news, because magnetism, apart from holding notes to refrigerators and keeping our compasses pointing the right way, plays a vital role in keeping us alive. Space is full of dangerous cosmic rays that in the absence of magnetic protection would tear through our bodies, leaving much of our DNA in useless tatters. When the magnetic field is working, these rays are safely herded away from the Earth’s surface and into two zones in near space called the Van Allen belts. They also interact with particles in the upper atmosphere to create the bewitching veils of light known as the auroras.

A big part of the reason for our ignorance, interestingly enough, is that traditionally there has been little effort to coordinate what’s happening on top of the Earth with what’s going on inside. According to Shawna Vogel: “Geologists and geophysicists rarely go to the same meetings or collaborate on the same problems.”

Perhaps nothing better demonstrates our inadequate grasp of the dynamics of the Earth’s interior than how badly we are caught out when it acts up, and it would be hard to come up with a more salutary reminder of the limitations of our understanding than the eruption of Mount St. Helens in Washington in 1980.

At that time, the lower forty-eight United States had not seen a volcanic eruption for over sixty-five years. Therefore the government volcanologists called in to monitor and forecast St. Helens’s behavior primarily had seen only Hawaiian volcanoes in action, and they, it turned out, were not the same thing at all.

St. Helens started its ominous rumblings on March 20. Within a week it was erupting magma, albeit in modest amounts, up to a hundred times a day, and being constantly shaken with earthquakes. People were evacuated to what was assumed to be a safe distance of eight miles. As the mountain’s rumblings grew St. Helens became a tourist attraction for the world. Newspapers gave daily reports on the best places to get a view. Television crews repeatedly flew in helicopters to the summit, and people were even seen climbing over the mountain. On one day, more than seventy copters and light aircraft circled the summit. But as the days passed and the rumblings failed to develop into anything dramatic, people grew restless, and the view became general that the volcano wasn’t going to blow after all.

On April 19 the northern flank of the mountain began to bulge conspicuously. Remarkably, no one in a position of responsibility saw that this strongly signaled a lateral blast. The seismologists resolutely based their conclusions on the behavior of Hawaiian volcanoes, which don’t blow out sideways. Almost the only person who believed that something really bad might happen was Jack Hyde, a geology professor at a community college in Tacoma. He pointed out that St. Helens didn’t have an open vent, as Hawaiian volcanoes have, so any pressure building up inside was bound to be released dramatically and probably catastrophically. However, Hyde was not part of the official team and his observations attracted little notice.

We all know what happened next. At 8:32 A.M. on a Sunday morning, May 18, the north side of the volcano collapsed, sending an enormous avalanche of dirt and rock rushing down the mountain slope at 150 miles an hour. It was the biggest landslide in human history and carried enough material to bury the whole of Manhattan to a depth of four hundred feet. A minute later, its flank severely weakened, St. Helens exploded with the force of five hundred Hiroshima-sized atomic bombs, shooting out a murderous hot cloud at up to 650 miles an hour-much too fast, clearly, for anyone nearby to outrace. Many people who were thought to be in safe areas, often far out of sight of the volcano, were overtaken. Fifty-seven people were killed. Twenty-three of the bodies were never found. The toll would have been much higher except that it was a Sunday. Had it been a weekday many lumber workers would have been working within the death zone. As it was, people were killed eighteen miles away.

The luckiest person on that day was a graduate student named Harry Glicken. He had been manning an observation post 5.7 miles from the mountain, but he had a college placement interview on May 18 in California, and so had left the site the day before the eruption. His place was taken by David Johnston. Johnston was the first to report the volcano exploding; moments later he was dead. His body was never found. Glicken’s luck, alas, was temporary. Eleven years later he was one of forty-three scientists and journalists fatally caught up in a lethal outpouring of superheated ash, gases, and molten rock-what is known as a pyroclastic flow-at Mount Unzen in Japan when yet another volcano was catastrophically misread.

Volcanologists may or may not be the worst scientists in the world at making predictions, but they are without question the worst in the world at realizing how bad their predictions are. Less than two years after the Unzen catastrophe another group of volcano watchers, led by Stanley Williams of the University of Arizona, descended into the rim of an active volcano called Galeras in Colombia. Despite the deaths of recent years, only two of the sixteen members of Williams’s party wore safety helmets or other protective gear. The volcano erupted, killing six of the scientists, along with three tourists who had followed them, and seriously injuring several others, including Williams himself.

In an extraordinarily unself-critical book called Surviving Galeras, Williams said he could “only shake my head in wonder” when he learned afterward that his colleagues in the world of volcanology had suggested that he had overlooked or disregarded important seismic signals and behaved recklessly. “How easy it is to snipe after the fact, to apply the knowledge we have now to the events of 1993,” he wrote. He was guilty of nothing worse, he believed, than unlucky timing when Galeras “behaved capriciously, as natural forces are wont to do. I was fooled, and for that I will take responsibility. But I do not feel guilty about the deaths of my colleagues. There is no guilt. There was only an eruption.”

But to return to Washington. Mount St. Helens lost thirteen hundred feet of peak, and 230 square miles of forest were devastated. Enough trees to build 150,000 homes (or 300,000 in some reports) were blown away. The damage was placed at $2.7 billion. A giant column of smoke and ash rose to a height of sixty thousand feet in less than ten minutes. An airliner some thirty miles away reported being pelted with rocks.

Ninety minutes after the blast, ash began to rain down on Yakima, Washington, a community of fifty thousand people about eighty miles away. As you would expect, the ash turned day to night and got into everything, clogging motors, generators, and electrical switching equipment, choking pedestrians, blocking filtration systems, and generally bringing things to a halt. The airport shut down and highways in and out of the city were closed.

All this was happening, you will note, just downwind of a volcano that had been rumbling menacingly for two months. Yet Yakima had no volcano emergency procedures. The city’s emergency broadcast system, which was supposed to swing into action during a crisis, did not go on the air because “the Sunday-morning staff did not know how to operate the equipment.” For three days, Yakima was paralyzed and cut off from the world, its airport closed, its approach roads impassable. Altogether the city received just five-eighths of an inch of ash after the eruption of Mount St. Helens. Now bear that in mind, please, as we consider what a Yellowstone blast would do.

15 DANGEROUS BEAUTY

IN THE 1960s, while studying the volcanic history of Yellowstone National Park, Bob Christiansen of the United States Geological Survey became puzzled about something that, oddly, had not troubled anyone before: he couldn’t find the park’s volcano. It had been known for a long time that Yellowstone was volcanic in nature-that’s what accounted for all its geysers and other steamy features-and the one thing about volcanoes is that they are generally pretty conspicuous. But Christiansen couldn’t find the Yellowstone volcano anywhere. In particular what he couldn’t find was a structure known as a caldera.

Most of us, when we think of volcanoes, think of the classic cone shapes of a Fuji or Kilimanjaro, which are created when erupting magma accumulates in a symmetrical mound. These can form remarkably quickly. In 1943, at Parícutin in Mexico, a farmer was startled to see smoke rising from a patch on his land. In one week he was the bemused owner of a cone five hundred feet high. Within two years it had topped out at almost fourteen hundred feet and was more than half a mile across. Altogether there are some ten thousand of these intrusively visible volcanoes on Earth, all but a few hundred of them extinct. But there is a second, less celebrated type of volcano that doesn’t involve mountain building. These are volcanoes so explosive that they burst open in a single mighty rupture, leaving behind a vast subsided pit, the caldera (from a Latin word for cauldron). Yellowstone obviously was of this second type, but Christiansen couldn’t find the caldera anywhere.

By coincidence just at this time NASA decided to test some new high-altitude cameras by taking photographs of Yellowstone, copies of which some thoughtful official passed on to the park authorities on the assumption that they might make a nice blow-up for one of the visitors’ centers. As soon as Christiansen saw the photos he realized why he had failed to spot the caldera: virtually the whole park-2.2 million acres-was caldera. The explosion had left a crater more than forty miles across-much too huge to be perceived from anywhere at ground level. At some time in the past Yellowstone must have blown up with a violence far beyond the scale of anything known to humans.

Yellowstone, it turns out, is a supervolcano. It sits on top of an enormous hot spot, a reservoir of molten rock that rises from at least 125 miles down in the Earth. The heat from the hot spot is what powers all of Yellowstone’s vents, geysers, hot springs, and popping mud pots. Beneath the surface is a magma chamber that is about forty-five miles across-roughly the same dimensions as the park-and about eight miles thick at its thickest point. Imagine a pile of TNT about the size of Rhode Island and reaching eight miles into the sky, to about the height of the highest cirrus clouds, and you have some idea of what visitors to Yellowstone are shuffling around on top of. The pressure that such a pool of magma exerts on the crust above has lifted Yellowstone and about three hundred miles of surrounding territory about 1,700 feet higher than they would otherwise be. If it blew, the cataclysm is pretty well beyond imagining. According to Professor Bill McGuire of University College London, “you wouldn’t be able to get within a thousand kilometers of it” while it was erupting. The consequences that followed would be even worse.

Superplumes of the type on which Yellowstone sits are rather like martini glasses-thin on the way up, but spreading out as they near the surface to create vast bowls of unstable magma. Some of these bowls can be up to 1,200 miles across. According to theories, they don’t always erupt explosively but sometimes burst forth in a vast, continuous outpouring-a flood-of molten rock, such as with the Deccan Traps in India sixty-five million years ago. (Trap in this context comes from a Swedish word for a type of lava; Deccan is simply an area.) These covered an area of 200,000 square miles and probably contributed to the demise of the dinosaurs-they certainly didn’t help-with their noxious outgassings. Superplumes may also be responsible for the rifts that cause continents to break up.

Such plumes are not all that rare. There are about thirty active ones on the Earth at the moment, and they are responsible for many of the world’s best-known islands and island chains-Iceland, Hawaii, the Azores, Canaries, and Galápagos archipelagos, little Pitcairn in the middle of the South Pacific, and many others-but apart from Yellowstone they are all oceanic. No one has the faintest idea how or why Yellowstone’s ended up beneath a continental plate. Only two things are certain: that the crust at Yellowstone is thin and that the world beneath it is hot. But whether the crust is thin because of the hot spot or whether the hot spot is there because the crust is thin is a matter of heated (as it were) debate. The continental nature of the crust makes a huge difference to its eruptions. Where the other supervolcanoes tend to bubble away steadily and in a comparatively benign fashion, Yellowstone blows explosively. It doesn’t happen often, but when it does you want to stand well back.

Since its first known eruption 16.5 million years ago, it has blown up about a hundred times, but the most recent three eruptions are the ones that get written about. The last eruption was a thousand times greater than that of Mount St. Helens; the one before that was 280 times bigger, and the one before was so big that nobody knows exactly how big it was. It was at least twenty-five hundred times greater than St. Helens, but perhaps eight thousand times more monstrous.

We have absolutely nothing to compare it to. The biggest blast in recent times was that of Krakatau in Indonesia in August 1883, which made a bang that reverberated around the world for nine days, and made water slosh as far away as the English Channel. But if you imagine the volume of ejected material from Krakatau as being about the size of a golf ball, then the biggest of the Yellowstone blasts would be the size of a sphere you could just about hide behind. On this scale, Mount St. Helens’s would be no more than a pea.

The Yellowstone eruption of two million years ago put out enough ash to bury New York State to a depth of sixty-seven feet or California to a depth of twenty. This was the ash that made Mike Voorhies’s fossil beds in eastern Nebraska. That blast occurred in what is now Idaho, but over millions of years, at a rate of about one inch a year, the Earth’s crust has traveled over it, so that today it is directly under northwest Wyoming. (The hot spot itself stays in one place, like an acetylene torch aimed at a ceiling.) In its wake it leaves the sort of rich volcanic plains that are ideal for growing potatoes, as Idaho’s farmers long ago discovered. In another two million years, geologists like to joke, Yellowstone will be producing French fries for McDonald’s, and the people of Billings, Montana, will be stepping around geysers.

The ash fall from the last Yellowstone eruption covered all or parts of nineteen western states (plus parts of Canada and Mexico)-nearly the whole of the United States west of the Mississippi. This, bear in mind, is the breadbasket of America, an area that produces roughly half the world’s cereals. And ash, it is worth remembering, is not like a big snowfall that will melt in the spring. If you wanted to grow crops again, you would have to find some place to put all the ash. It took thousands of workers eight months to clear 1.8 billion tons of debris from the sixteen acres of the World Trade Center site in New York. Imagine what it would take to clear Kansas.

And that’s not even to consider the climatic consequences. The last supervolcano eruption on Earth was at Toba, in northern Sumatra, seventy-four thousand years ago. No one knows quite how big it was other than that it was a whopper. Greenland ice cores show that the Toba blast was followed by at least six years of “volcanic winter” and goodness knows how many poor growing seasons after that. The event, it is thought, may have carried humans right to the brink of extinction, reducing the global population to no more than a few thousand individuals. That means that all modern humans arose from a very small population base, which would explain our lack of genetic diversity. At all events, there is some evidence to suggest that for the next twenty thousand years the total number of people on Earth was never more than a few thousand at any time. That is, needless to say, a long time to recover from a single volcanic blast.

All this was hypothetically interesting until 1973, when an odd occurrence made it suddenly momentous: water in Yellowstone Lake, in the heart of the park, began to run over the banks at the lake’s southern end, flooding a meadow, while at the opposite end of the lake the water mysteriously flowed away. Geologists did a hasty survey and discovered that a large area of the park had developed an ominous bulge. This was lifting up one end of the lake and causing the water to run out at the other, as would happen if you lifted one side of a child’s wading pool. By 1984, the whole central region of the park-several dozen square miles-was more than three feet higher than it had been in 1924, when the park was last formally surveyed. Then in 1985, the whole of the central part of the park subsided by eight inches. It now seems to be swelling again.

The geologists realized that only one thing could cause this-a restless magma chamber. Yellowstone wasn’t the site of an ancient supervolcano; it was the site of an active one. It was also at about this time that they were able to work out that the cycle of Yellowstone’s eruptions averaged one massive blow every 600,000 years. The last one, interestingly enough, was 630,000 years ago. Yellowstone, it appears, is due.

“It may not feel like it, but you’re standing on the largest active volcano in the world,” Paul Doss, Yellowstone National Park geologist, told me soon after climbing off an enormous Harley-Davidson motorcycle and shaking hands when we met at the park headquarters at Mammoth Hot Springs early on a lovely morning in June. A native of Indiana, Doss is an amiable, soft-spoken, extremely thoughtful man who looks nothing like a National Park Service employee. He has a graying beard and hair tied back in a long ponytail. A small sapphire stud graces one ear. A slight paunch strains against his crisp Park Service uniform. He looks more like a blues musician than a government employee. In fact, he is a blues musician (harmonica). But he sure knows and loves geology. “And I’ve got the best place in the world to do it,” he says as we set off in a bouncy, battered four-wheel-drive vehicle in the general direction of Old Faithful. He has agreed to let me accompany him for a day as he goes about doing whatever it is a park geologist does. The first assignment today is to give an introductory talk to a new crop of tour guides.

Yellowstone, I hardly need point out, is sensationally beautiful, with plump, stately mountains, bison-specked meadows, tumbling streams, a sky-blue lake, wildlife beyond counting. “It really doesn’t get any better than this if you’re a geologist,” Doss says. “You’ve got rocks up at Beartooth Gap that are nearly three billion years old-three-quarters of the way back to Earth’s beginning-and then you’ve got mineral springs here”-he points at the sulfurous hot springs from which Mammoth takes its title-“where you can see rocks as they are being born. And in between there’s everything you could possibly imagine. I’ve never been any place where geology is more evident-or prettier.”

“So you like it?” I say.

“Oh, no, I love it,” he answers with profound sincerity. “I mean I really love it here. The winters are tough and the pay’s not too hot, but when it’s good, it’s just-”

He interrupted himself to point out a distant gap in a range of mountains to the west, which had just come into view over a rise. The mountains, he told me, were known as the Gallatins. “That gap is sixty or maybe seventy miles across. For a long time nobody could understand why that gap was there, and then Bob Christiansen realized that it had to be because the mountains were just blown away. When you’ve got sixty miles of mountains just obliterated, you know you’re dealing with something pretty potent. It took Christiansen six years to figure it all out.”

I asked him what caused Yellowstone to blow when it did.

“Don’t know. Nobody knows. Volcanoes are strange things. We really don’t understand them at all. Vesuvius, in Italy, was active for three hundred years until an eruption in 1944 and then it just stopped. It’s been silent ever since. Some volcanologists think that it is recharging in a big way, which is a little worrying because two million people live on or around it. But nobody knows.”

“And how much warning would you get if Yellowstone was going to go?”

He shrugged. “Nobody was around the last time it blew, so nobody knows what the warning signs are. Probably you would have swarms of earthquakes and some surface uplift and possibly some changes in the patterns of behavior of the geysers and steam vents, but nobody really knows.”

“So it could just blow without warning?”

He nodded thoughtfully. The trouble, he explained, is that nearly all the things that would constitute warning signs already exist in some measure at Yellowstone. “Earthquakes are generally a precursor of volcanic eruptions, but the park already has lots of earthquakes-1,260 of them last year. Most of them are too small to be felt, but they are earthquakes nonetheless.”

A change in the pattern of geyser eruptions might also be taken as a clue, he said, but these too vary unpredictably. Once the most famous geyser in the park was Excelsior Geyser. It used to erupt regularly and spectacularly to heights of three hundred feet, but in 1888 it just stopped. Then in 1985 it erupted again, though only to a height of eighty feet. Steamboat Geyser is the biggest geyser in the world when it blows, shooting water four hundred feet into the air, but the intervals between its eruptions have ranged from as little as four days to almost fifty years. “If it blew today and again next week, that wouldn’t tell us anything at all about what it might do the following week or the week after or twenty years from now,” Doss says. “The whole park is so volatile that it’s essentially impossible to draw conclusions from almost anything that happens.”

Evacuating Yellowstone would never be easy. The park gets some three million visitors a year, mostly in the three peak months of summer. The park’s roads are comparatively few and they are kept intentionally narrow, partly to slow traffic, partly to preserve an air of picturesqueness, and partly because of topographical constraints. At the height of summer, it can easily take half a day to cross the park and hours to get anywhere within it. “Whenever people see animals, they just stop, wherever they are,” Doss says. “We get bear jams. We get bison jams. We get wolf jams.”

In the autumn of 2000, representatives from the U.S. Geological Survey and National Park Service, along with some academics, met and formed something called the Yellowstone Volcanic Observatory. Four such bodies were in existence already-in Hawaii, California, Alaska, and Washington-but oddly none in the largest volcanic zone in the world. The YVO is not actually a thing, but more an idea-an agreement to coordinate efforts at studying and analyzing the park’s diverse geology. One of their first tasks, Doss told me, was to draw up an “earthquake and volcano hazards plan”-a plan of action in the event of a crisis.

“There isn’t one already?” I said.

“No. Afraid not. But there will be soon.”

“Isn’t that just a little tardy?”

He smiled. “Well, let’s just say that it’s not any too soon.”

Once it is in place, the idea is that three people-Christiansen in Menlo Park, California, Professor Robert B. Smith at the University of Utah, and Doss in the park-would assess the degree of danger of any potential cataclysm and advise the park superintendent. The superintendent would take the decision whether to evacuate the park. As for surrounding areas, there are no plans. If Yellowstone were going to blow in a really big way, you would be on your own once you left the park gates.

Of course it may be tens of thousands of years before that day comes. Doss thinks such a day may not come at all. “Just because there was a pattern in the past doesn’t mean that it still holds true,” he says. “There is some evidence to suggest that the pattern may be a series of catastrophic explosions, then a long period of quiet. We may be in that now. The evidence now is that most of the magma chamber is cooling and crystallizing. It is releasing its volatiles; you need to trap volatiles for an explosive eruption.”

In the meantime there are plenty of other dangers in and around Yellowstone, as was made devastatingly evident on the night of August 17, 1959, at a place called Hebgen Lake just outside the park. At twenty minutes to midnight on that date, Hebgen Lake suffered a catastrophic quake. It was magnitude 7.5, not vast as earthquakes go, but so abrupt and wrenching that it collapsed an entire mountainside. It was the height of the summer season, though fortunately not so many people went to Yellowstone in those days as now. Eighty million tons of rock, moving at more than one hundred miles an hour, just fell off the mountain, traveling with such force and momentum that the leading edge of the landslide ran four hundred feet up a mountain on the other side of the valley. Along its path lay part of the Rock Creek Campground. Twenty-eight campers were killed, nineteen of them buried too deep ever to be found again. The devastation was swift but heartbreakingly fickle. Three brothers, sleeping in one tent, were spared. Their parents, sleeping in another tent beside them, were swept away and never seen again.

“A big earthquake-and I mean big-will happen sometime,” Doss told me. “You can count on that. This is a big fault zone for earthquakes.”

Despite the Hebgen Lake quake and the other known risks, Yellowstone didn’t get permanent seismometers until the 1970s.

If you needed a way to appreciate the grandeur and inexorable nature of geologic processes, you could do worse than to consider the Tetons, the sumptuously jagged range that stands just to the south of Yellowstone National Park. Nine million years ago, the Tetons didn’t exist. The land around Jackson Hole was just a high grassy plain. But then a forty-mile-long fault opened within the Earth, and since then, about once every nine hundred years, the Tetons experience a really big earthquake, enough to jerk them another six feet higher. It is these repeated jerks over eons that have raised them to their present majestic heights of seven thousand feet.

That nine hundred years is an average-and a somewhat misleading one. According to Robert B. Smith and Lee J. Siegel in Windows into the Earth, a geological history of the region, the last major Teton quake was somewhere between about five and seven thousand years ago. The Tetons, in short, are about the most overdue earthquake zone on the planet.

Hydrothermal explosions are also a significant risk. They can happen anytime, pretty much anywhere, and without any predictability. “You know, by design we funnel visitors into thermal basins,” Doss told me after we had watched Old Faithful blow. “It’s what they come to see. Did you know there are more geysers and hot springs at Yellowstone than in all the rest of the world combined?”

“I didn’t know that.”

He nodded. “Ten thousand of them, and nobody knows when a new vent might open.” We drove to a place called Duck Lake, a body of water a couple of hundred yards across. “It looks completely innocuous,” he said. “It’s just a big pond. But this big hole didn’t used to be here. At some time in the last fifteen thousand years this blew in a really big way. You’d have had several tens of millions of tons of earth and rock and superheated water blowing out at hypersonic speeds. You can imagine what it would be like if this happened under, say, the parking lot at Old Faithful or one of the visitors’ centers.” He made an unhappy face.

“Would there be any warning?”

“Probably not. The last significant explosion in the park was at a place called Pork Chop Geyser in 1989. That left a crater about five meters across-not huge by any means, but big enough if you happened to be standing there at the time. Fortunately, nobody was around so nobody was hurt, but that happened without warning. In the very ancient past there have been explosions that have made holes a mile across. And nobody can tell you where or when that might happen again. You just have to hope that you’re not standing there when it does.”

Big rockfalls are also a danger. There was a big one at Gardiner Canyon in 1999, but again fortunately no one was hurt. Late in the afternoon, Doss and I stopped at a place where there was a rock overhang poised above a busy park road. Cracks were clearly visible. “It could go at any time,” Doss said thoughtfully.

“You’re kidding,” I said. There wasn’t a moment when there weren’t two cars passing beneath it, all filled with, in the most literal sense, happy campers.

“Oh, it’s not likely,” he added. “I’m just saying it could. Equally it could stay like that for decades. There’s just no telling. People have to accept that there is risk in coming here. That’s all there is to it.”

As we walked back to his vehicle to head back to Mammoth Hot Springs, Doss added: “But the thing is, most of the time bad things don’t happen. Rocks don’t fall. Earthquakes don’t occur. New vents don’t suddenly open up. For all the instability, it’s mostly remarkably and amazingly tranquil.”

“Like Earth itself,” I remarked.

“Precisely,” he agreed.

The risks at Yellowstone apply to park employees as much as to visitors. Doss got a horrific sense of that in his first week on the job five years earlier. Late one night, three young summer employees engaged in an illicit activity known as “hot-potting”-swimming or basking in warm pools. Though the park, for obvious reasons, doesn’t publicize it, not all the pools in Yellowstone are dangerously hot. Some are extremely agreeable to lie in, and it was the habit of some of the summer employees to have a dip late at night even though it was against the rules to do so. Foolishly the threesome had failed to take a flashlight, which was extremely dangerous because much of the soil around the warm pools is crusty and thin and one can easily fall through into a scalding vent below. In any case, as they made their way back to their dorm, they came across a stream that they had had to leap over earlier. They backed up a few paces, linked arms and, on the count of three, took a running jump. In fact, it wasn’t the stream at all. It was a boiling pool. In the dark they had lost their bearings. None of the three survived.

I thought about this the next morning as I made a brief call, on my way out of the park, at a place called Emerald Pool, in the Upper Geyser Basin. Doss hadn’t had time to take me there the day before, but I thought I ought at least to have a look at it, for Emerald Pool is a historic site.

In 1965, a husband-and-wife team of biologists named Thomas and Louise Brock, while on a summer study trip, had done a crazy thing. They had scooped up some of the yellowy-brown scum that rimmed the pool and examined it for life. To their, and eventually the wider world’s, deep surprise, it was full of living microbes. They had found the world’s first extremophiles-organisms that could live in water that had previously been assumed to be much too hot or acid or choked with sulfur to bear life. Emerald Pool, remarkably, was all these things, yet at least two types of living things, Sulpholobus acidocaldarius and Thermophilus aquaticus as they became known, found it congenial. It had always been supposed that nothing could survive above temperatures of 50°C (122°F), but here were organisms basking in rank, acidic waters nearly twice that hot.

For almost twenty years, one of the Brocks’ two new bacteria, Thermophilus aquaticus, remained a laboratory curiosity until a scientist in California named Kary B. Mullis realized that heat-resistant enzymes within it could be used to create a bit of chemical wizardry known as a polymerase chain reaction, which allows scientists to generate lots of DNA from very small amounts-as little as a single molecule in ideal conditions. It’s a kind of genetic photocopying, and it became the basis for all subsequent genetic science, from academic studies to police forensic work. It won Mullis the Nobel Prize in chemistry in 1993.

Meanwhile, scientists were finding even hardier microbes, now known as hyperthermophiles, which demand temperatures of 80°C (176°F) or more. The warmest organism found so far, according to Frances Ashcroft in Life at the Extremes, is Pyrolobus fumarii, which dwells in the walls of ocean vents where the temperature can reach 113°C (235.4°F). The upper limit for life is thought to be about 120°C (248°F), though no one actually knows. At all events, the Brocks’ findings completely changed our perception of the living world. As NASA scientist Jay Bergstralh has put it: “Wherever we go on Earth-even into what’s seemed like the most hostile possible environments for life-as long as there is liquid water and some source of chemical energy we find life.”

Life, it turns out, is infinitely more clever and adaptable than anyone had ever supposed. This is a very good thing, for as we are about to see, we live in a world that doesn’t altogether seem to want us here.