Listen to an audio version of the article.

Pangaea, 252 million years ago—the world is over. Siberia has been erupting for 300,000 years, is still erupting and won’t stop. Not a volcano, mind you, but Siberia—two million square miles of it. A suppurating, billowing, continent-scaled wasteland of glowing rock and steam. The seas, once resplendent with horn corals and sponge reefs, are now sour and laden with mercury. Hot as soup, they bubble deathly swamp gas that feeds vile, hurricane-churned slicks of slime. The seabed is vacant, as scuttling trilobites die out after a quarter of a billion years. Beside this rancid ocean, on the shores of a blasted supercontinent, the forests are gone. Instead hot rivers now spill over the dead land in wide braids. Fungus blooms where vanished groves of ferns once held contests full of fangs and armor—gorgonopsids battling pareiasaurs. Their final bones are now scoured by hot winds and bleached by a searing sunlight, unfiltered by ozone. Night falls and brings odd constellations that light the dead waves, lapping at dead shores, tossing old bits of a dead reef from a dead sea. It smells awful. The planet is ruin and slime and heat. The ocean suffocates. Bacterial mounds spread. A hundred thousand suns rise and fall on a hopeless world. A hundred thousand springtimes arrive with no respite. It’s still all but barren. A million years of misery pass. Ten million.

At long last, the planet is finally reborn—this time with dinosaurs and ichthyosaurs and pterosaurs and mammals and turtles and sturgeon—almost as if a new story entirely has begun, exuberant, confident, vital: the Mesozoic era. The old story, too dark to retell, has long been filed away somewhere deep in the great cabinets of Earth, along with an epitaph in the rocks, written in geochemistry: carbon dioxide destroyed this world.

Mass extinctions are not merely bad days in Earth history. They are not even very bad days. They are the very, very worst days in the entire half-billion-year history of complex life. They are supremely horrifying, astronomically rare, global Ragnaroks that end the lineages of most living creatures on the planet. They are terrible, surreal things: 20,000 years of suffocating greenhouse heat punctuated by volcanic winter blasts or an afternoon of celestial terror and tsunamis. And until around 1980, they were mostly thought to be disreputable speculation.

Over the past two centuries the field of geology, like the fossil record itself, has been characterized by long periods of stasis punctuated by exhilarating moments of upheaval and innovation. It would be arbitrary to identify one founding figure for the modern study of geology, but one could do worse than Scottish geologist James Hutton, who did much to reveal the “abyss of time” underneath us. At a salt-sprayed Siccar Point on the eastern coast of Scotland in 1788, he spied an outcrop made of two kinds of rock, one stacked on top of the other, meeting abruptly in the middle. But the rock on the bottom—laminations of deep-sea muck—had been formed at the bottom of the ocean, tilted sideways, thrust up into the air and planed off by wind and erosion. And the rock on the top had been formed by tropical rivers on dry land. The missing time implied between the two rocks, now conjoined but separated by an unthinkable gap, shattered Hutton. The body of his writing is notoriously obtuse and unimpressive; he apparently saved all his eloquence for one haunting observation that, in the confusion of Earth beneath our feet, “we find no vestige of a beginning, no prospect of an end.” Although the history of humanity has played out in the shallows so far, time, it turns out, is deep.

Geologists cast off the strictures of biblical time and Noah’s flood, and their new field matured over the decades, spurred by a significant material reward for finding coal and minerals in a strange new industrial age. The story of life on Earth, however fragmentary and tantalizing, slowly revealed itself.

Trilobites swam the seas for a quarter of a billion years before being annihilated in history's worst mass extinction. Credit: Ed Reschke Getty Images

At this magazine’s founding in the mid-1800s, the field was in its adolescence. It was still dominated by men of means—the kind of ascotted dilettantes rendered humorlessly in oil and lithograph portraits. The contributions of women such as Mary Anning, the unmatched fossil hunter who scoured the English shoreline unearthing the local “snakestones” and “stone crocodiles” that littered the wave-battered Jurassic Coast, were acknowledged only sheepishly. Although Scientific American headlines from the time still hint at a somewhat rudimentary state of affairs (“Experts Doubt the Sun Is Actually Burning Coal”), by midcentury geology had nonetheless been established as an empirical, systematized field of inquiry with roots in antiquity—one of the many such intellectual eddies that swirled out of Enlightenment-era “natural philosophy.” That is, it now had rules. The rules were deceptively simple and powerful. Layers of ocean rock now propped up at unusual angles on land must have once laid flat on a seafloor in some distant age. Dikes of old magma that pierced this stony tiramisu must have worked their shoots into the layered rock sometime after. The fossils entombed in these rock layers could be correlated to those fossils and those rocks with the same layers, way over there.

In 1860 English geologist John Phillips drew on the fossil-collecting labor of decades prior, and a growing body of paleontology literature scattered across elegant monographs, to plot a surprisingly modern curve of the richness of life over Earth history, the first ever such diagram. The graph ominously included two profound dips in life: one crash that separated the trilobite-spangled Paleozoic era from the dinosaur-haunted Mesozoic era and another plunge that separated the Mesozoic from our own Cenozoic time (all terms of Phillips’s invention). The dramatic breaks in fossil life hinted at some kind of ancient calamity that divided the great ages and supported the blasphemous idea that species—that is, God’s very Creation—might go extinct, which had been proposed half a century earlier by the renowned French naturalist Georges Cuvier. On considering the odd elephant bones that littered the New World, of mammoths and mastodons (“animal de l’Ohio”), Cuvier had proposed that life on Earth, like French rule, could be swept away in “revolutions.” Phillips’s graph provided something close to quantitative proof, and Phillips himself thought that each recovery consisted of separate acts of divine creation. Yet it would take more than a century for anyone to take the idea of mass extinctions seriously again.

1867 rendering of a woolly mammoth, whose bones led Georges Cuvier to propose that life could be extinguished in “revolutions.” Credit: Alamy

This is because, by the end of the 19th century, the field was still dominated, and would continue to be dominated for decades, by the enduring framework of “uniformitarianism.” This concept, popularized by Charles Lyell, is summarized in a catchphrase still taught to geology students: “The present is the key to the past.” That is, the unhurried processes at work on the face of Earth today—the relentless if unimpressive work of rain on rock, the inexorable incision of rivers into highlands or the piling of sand into desert dunes—have always been plying the planet in the same tedious fashion and could account for everything we find in the rock record. Painting on this vast new canvas of time, Charles Darwin would propose that similarly small but steady biological changes over generations, filtered by the relentless tournament of life and death, and given Hutton’s eons to ramify, could produce the “endless forms” of life “most beautiful” on Earth today. Pointedly absent from this measured account of planetary history were the gauche cataclysms of Cuvier and Phillips.

Geology was upended in the middle of the 20th century by the plate tectonic revolution—the validation of the once fringe idea that continents drifted across the world like rudderless ships. Even so, the idea of sudden apocalyptic global mass extinctions remained suspect at best. Catastrophism was spooky, reminiscent of a benighted prescientific world where capricious gods subjected the world to cleansing acts of global destruction. Worse, speculation about why the dinosaurs had disappeared had become something of a cottage industry among cranks, and serious scientists were nervous about associating with a crowd who proposed, among dozens of other incoherent ideas for their demise, “dwindling brain and consequent stupidity,” development of heads that became “too heavy to lift,” “psychotic suicidal factors,” “competition with caterpillars,” “terminal hay fever” and “methane poisoning from dinosaur flatulence.” And yet the orthodoxy began to crack.

Neokatastrophismus?” the iconoclastic German scientist Otto Schindewolf asked of his fellow paleontologists in 1963, attempting to revive Cuvier’s catastrophism for the 20th century. Because his question was posed in German, few English-speaking scientists felt the need to reply. But Schindewolf could no longer overlook the ominous interruption of life he saw—among other places—exposed in the ancient rocks of the Salt Range in Pakistan. There appeared to be a dreadful global collapse of the ocean ecosystem at the end of the Permian period a quarter of a billion years ago (in fact, the greatest mass extinction in Earth history), just as Phillips had plotted more than a century prior. Schindewolf conscripted a supernova for his vision of the apocalypse, proposing that it might have irradiated Earth and seeded the biosphere with ruinous mutations.

In that same year American Norman Newell, plotting the fates of 2,500 animal families over Earth history, noted six intervals when extinction seemed to cut a broad swath through all of life, instead proposing dramatic sea-level changes as his preferred Grim Reaper. And at the end of the decade Digby McLaren, director of the Canadian Geological Survey’s Institute of Sedimentary and Petroleum Geology, insisted in his 1969 presidential address to the Paleontological Society that his fellow paleontologists were trying to “define out of existence” the obvious breaks in the fossil record, such as a devastating wave of death 375 million years ago that wiped out the largest global reef system in the history of life. “I cannot accept a uniformitarian explanation,” he said of the catastrophe, glaringly apparent in ancient rocks from Iran to Alberta. McLaren had an idea for what could cause such a discontinuity.

“I shall, therefore, land a large or very large meteorite in the Paleozoic Pacific,” he announced, capable of generating “a wave twenty thousand feet high. This will do.” McLaren’s address, it is reported, was “met with embarrassed silence,” and many paleontologists in the audience, still under Lyell’s spell, assumed he must have been joking.

Exposed seam in Colorado of the K-T boundary, which marks the mass extinction that ended the reign of the dinosaurs. Credit: Julie Mowbray Alamy

Then, in 1980, an asteroid landed in the field. Walter Alvarez, then a young University of California, Berkeley, professor, was traipsing the Apennines above the medieval Italian town of Gubbio. In this mountainous pile of ancient limestone seafloor, pushed up by the grinding advance of Africa into Europe, there was a sharp break—a lifeless clay layer—between the placid sea life of dinosaur times and the impoverished life of the early age of mammals. Perhaps this transformative interval took place over millions of years, validating the uniformitarian view. Or perhaps Cuvier and Phillips had it right all along, and the turnover was devastatingly short. Curious, Alvarez recruited his father, Luis, a Nobel-winning physicist, to help tackle the question. It was quite the second act for the elder Alvarez, a pioneer of military radar technology and Manhattan Project alum who helped to develop the atomic bomb and even watched “Little Boy” explode over Hiroshima from an attending B-29. His wartime work became unexpectedly relevant to the catastrophe they were investigating, which throttled the planet some 66 million years earlier.

The Alvarezes knew that unusual elements like iridium were delivered to Earth from above by an eternal drizzle of space dust, at a steady rate. Measure the iridium in the ominous clay layer, they reasoned, and if there’s just a little bit, the dramatic turnover in life couldn’t have taken very long. Conversely, if there’s a lot of iridium, it took a very long time indeed. But what if, as they discovered, there was 100 times more iridium than they ever expected? After bombarding the clay samples with neutrons from a nuclear reactor and analyzing them, the Alvarezes were astonished. The only vehicle large enough for this much exotic material was not a light drizzle of space dust but one truly gigantic space rock. (Often omitted in this account, though not by the Alvarezes themselves, is the fact that Dutch geologist Jan Smit and Belgian geologist Jan Hertogen made the very same discovery among ancient ocean rocks in Spain at the very same time and even published their results in the journal Nature two weeks earlier than the Alvarezes’ landmark Science paper.)

The resulting chaos from such an impact would be like all-out nuclear warfare, only worse. There would be the unimaginable heat from the initial explosion, which would have been thousands of times more powerful than the detonation of all the nuclear weapons on Earth at the height of the cold war, all in one place, all at once. “Certainly enough,” as one impact modeler put it to me, “to lift a mountain back into space at escape velocity.” It has been proposed that as this spacebound ejecta encircled the globe, it might have turned the atmosphere into a pizza oven for 20 minutes (with dinosaurs playing the role of pepperoni). Then there might have been the decades of darkness and cold from the nuclear winter to follow, starving any lingering creature lucky enough to have avoided being evaporated outright by the asteroid, swept up in its tsunamis, or turned to charcoal by the ballistic reentry of its debris into the atmosphere.

In 1991 whatever lingering skepticism about the impact that remained among uniformitarian hardliners was wiped away by the discovery of a 110-mile crater buried under tens of millions of years of limestone on Mexico’s Yucatán peninsula. In fact, the crater had already been discovered in 1978 by geophysicists working for the Mexican national oil company Pemex, but they had announced their findings at a geophysics conference that had escaped notice of paleontologists for more than a decade. And the structure had been found some 1,000 years earlier by the Maya, who built settlements around limestone sinkholes that pock the Yucatán and that provided freshwater. The pattern of these sinkholes reflects the deeply disturbed rock far below and maps almost exactly onto the crater’s edge, in a 110-mile ring.

Popular culture took note. The 1990s saw a rash of impact-inspired cable specials and movies strewn with bad CGI, which—along with the astounding, apocalypse-scaled collision of Comet Shoemaker-Levy 9 with Jupiter in 1994—were sufficient to convince the public of the dangers of space rocks. This is typically where the story ends. As far as most people are concerned, mass extinctions are what happen when big things fall out of the sky.

But something funny happened over the next 30 years as geologists fanned out across the globe to look for convincing evidence of impacts—such as layers of iridium, shocked quartz or giant craters—at the ominous rock boundaries that mark the four other major mass extinctions in Earth history. They didn’t find any. And all but one of the so-called Big Five mass extinctions were even more severe than the catastrophe that wiped out the nonbird dinosaurs.

In fact, there even existed major impact structures, such as the Triassic-age 62-mile-wide Manicouagan Crater in Quebec (now a circular system of lakes amid a boreal paradise of blackflies) or the massive crater that created the Chesapeake Bay 36 million years ago, that did not seem to bother life much at all. Given the remarkable correlation of the Yucatán impact with the disappearance of the large dinosaurs (and much of the rest of life on Earth at the end of the Cretaceous period), this came as a surprise. Stranger still, the stunning finale of the age of dinosaurs was also accompanied not only by an envoy from outer space but by one of the largest volcanic events in the history of animal life: a swath of eruptions that drowned much of India miles deep in lava. While the consensus is that the asteroid did most of the damage, this was the same class of world-changing eruptions responsible not only for dozens of minor mass extinctions and climate misadventures throughout Earth history but several of the other major mass extinctions as well, including the worst ever at the end of the Permian 252 million years ago.

In the past few decades a subtler story about mass extinctions has emerged. Geologists are now armed with powerful techniques Hutton couldn’t have dreamed of. Scattering to remote rock outcrops around the world or to archives of muck hoisted from the bottom of the ocean by drill ships, they wring secrets out of old seashells with mass spectrometers, and from age-battered hunks of granite with radioisotope geochronology, and from fossil and geochemical databases with neural networks underwritten by blistering processing power. And in this diffuse project to understand Earth history, geologists have in recent years revealed a roster of existential threats to life far more intimate than simply death from above. The most frequent mass killer of life on Earth, it turns out, is Earth itself. And the most reliable murder weapon is carbon dioxide.

One hundred and thirty-five million years before a mindless hunk of space garbage intercepted Earth’s orbit and ruined a perfectly good dinosaur world, the planet was gripped by a mass extinction that was even worse. A world of bizarre crocodilians, giant amphibians, stony corals, a ubiquity of strange but venerable eel-like creatures, and 80 percent of the rest of complex life on Earth was destroyed. As the supercontinent Pangaea pulled apart at the seams, stretching like taffy, an open sore of oozing, incandescent rock erupted at the surface, covering three million square miles in lava in pulses over 600,000 years. While the eruptions would have caused all sorts of chaos, perhaps most important they injected thousands of gigatons of carbon dioxide into the atmosphere, and the oceans overdosed on this volcanic CO2. The seawater acidified as a simple matter of chemistry, and the temperature of the planet soared as a simple matter of physics. This is what CO2 does. Today the New Jersey Palisades across the Hudson River from New York City are the volcanic plumbing that remains from these titanic eruptions of the Triassic end times, old magma that is matched by the same volcanic rock, of the same age, as far afield as Morocco, Brazil, Nova Scotia and Spain.

Hundreds of millions of years earlier the two oldest major mass extinctions destroyed planets we wouldn’t recognize, their continents misshapen and scattered about unfamiliar oceans. The oceans of these alien planets were patrolled by gigantic cephalopods and, later, even more gigantic fish, guillotine-mawed and fortified by helmets of bone. Our planet endlessly cycles carbon—the stuff of life—through rocks, air, water and life in a balance that keeps the climate habitable and ocean chemistry hospitable. But these archaic worlds saw this carbon cycle suddenly derailed—unraveled by CO2-sucking episodes of tropical mountain building, accelerated rock weathering and the novel global geoengineering project of land plants. These kill mechanisms are somewhat more convoluted, and admittedly less dramatic, than an asteroid, but they did the trick. These bygone planets spun out of control, alternately freezing and broiling as Earth struggled to regain its composure and wrangle a global carbon cycle gone haywire.

But it was 252 million years ago, on the forsaken planet that opened this tour of ancient apocalypses—that sun-bleached world, with oceans almost absent of animals—when the story of life on Earth nearly came to its premature conclusion. This was Pangaea, a world before dinosaurs or mammals or flowers. But it was still a rich world, one with conifers and lithe, vaguely leonine predators and lumbering, warty, reptile prey. And then, it was over. It ended in a continent-scaled flood of glowing rock, brief volcanic winters issuing from the eruptions and a roster of billowing volcanic gases, many of which would be banned on a battlefield—such as chlorine gas and mercury.

As the magma incinerated underground seams of salt and gypsum, eruptions of halocarbons would have obliterated the ozone layer—and indeed, plant fossils bear the mutations wrought by this ancient atmospheric destruction. But it wasn’t until the seams of magma feeding these eruptions hit huge deposits of natural gas, coal and carbon-rich rocks underground that the greatest mass extinction ever hit its appalling crescendo. Methane and carbon dioxide exploded out of the ground by the tens of thousands of gigatons. Temperatures spiked by almost 22 degrees Fahrenheit. And in the oceans, where spreading anoxia and acidification wiped out 96 percent of life, it was as hot as a jacuzzi. And then, in the fossil record, silence.

At the start of the industrial revolution, long slumbering forests of carbon were resurrected from the ancient Earth and pressed into service in the furnaces of modernity. We know that this artificial fire can’t go on forever without immiserating our world. At 416 parts per million, carbon dioxide is already higher than it has been in millions of years and is perhaps rising even faster than in these greatest calamities of all time. Meanwhile centuries—millennia even—of hunting, land clearing and pollution have impoverished the natural world. By one estimate, at the rate at which we are currently driving species extinct, we could match the biological devastation of those towering mass extinctions of the ancient past within 300 to 12,000 years. This might sound like a long time frame, but from a geologic perspective, it is downright subliminal. More worryingly, there may yet exist unseen ecological cliff edges along the way, beyond which the biosphere does not simply suffer the onslaught of attrition but collapses suddenly in cascading failures. In other words, there may be tipping points—points of no return.

We know what we have to do to avoid being inducted into the wretched Pantheon of the worst things that have ever happened in Earth history. We must set aside swaths of the planet—in the form of marine protected areas, natural reserves and corridors for migration—to allow the living world to recover from the uppercut we have already delivered it. Then, we must simply stop digging up old life from deep in Earth’s crust and lighting it on fire at the surface. As humanity leans on the very same levers pulled in the very worst things that have ever happened in history, we must consult the ages and listen to the counsel of broken worlds past.

In honor of Scientific American’s 175th anniversary: Relative frequency of terms in the magazine, from 1845 to the present.
Credit: Moritz Stefaner and Christian Lässer. For more context, see “Visualizing 175 Years of Words in Scientific American