Each of these massive transformations, as well as many smaller ones between and since, was dependent on that paradoxically important motor of progress: extinction. It is a curious fact that on Earth species death is, in the most literal sense, a way of life. No one knows how many species of organisms have existed since life began. Thirty billion is a commonly cited figure, but the number has been put as high as 4,000 billion. Whatever the actual total, 99.99 percent of all species that have ever lived are no longer with us. “To a first approximation,” as David Raup of the University of Chicago likes to say, “all species are extinct.” For complex organisms, the average lifespan of a species is only about four million years—roughly about where we are now.

Extinction is always bad news for the victims, of course, but it appears to be a good thing for a dynamic planet. “The alternative to extinction is stagnation,” says Ian Tattersall of the American Museum of Natural History, “and stagnation is seldom a good thing in any realm.” (I should perhaps note that we are speaking here of extinction as a natural, long-term process. Extinction brought about by human carelessness is another matter altogether.)

Crises in Earth’s history are invariably associated with dramatic leaps afterward. The fall of the Ediacaran fauna was followed by the creative outburst of the Cambrian period. The Ordovician extinction of 440 million years ago cleared the oceans of a lot of immobile filter feeders and, somehow, created conditions that favored darting fish and giant aquatic reptiles. These in turn were in an ideal position to send colonists onto dry land when another blowout in the late Devonian period gave life another sound shaking. And so it has gone at scattered intervals through history. If most of these events hadn’t happened just as they did, just when they did, we almost certainly wouldn’t be here now.

Earth has seen five major extinction episodes in its time—the Ordovician, Devonian, Permian, Triassic, and Cretaceous, in that order—and many smaller ones. The Ordovician (440 million years ago) and Devonian (365 million) each wiped out about 80 to 85 percent of species. The Triassic (210 million years ago) and Cretaceous (65 million years) each wiped out 70 to 75 percent of species. But the real whopper was the Permian extinction of about 245 million years ago, which raised the curtain on the long age of the dinosaurs. In the Permian, at least 95 percent of animals known from the fossil record check out, never to return. Even about a third of insect species went—the only occasion on which they were lost en masse. It is as close as we have ever come to total obliteration.

“It was, truly, a mass extinction, a carnage of a magnitude that had never troubled the Earth before,” says Richard Fortey. The Permian event was particularly devastating to sea creatures. Trilobites vanished altogether. Clams and sea urchins nearly went. Virtually all other marine organisms were staggered. Altogether, on land and in the water, it is thought that Earth lost 52 percent of its families—that’s the level above genus and below order on the grand scale of life (the subject of the next chapter)—and perhaps as many as 96 percent of all its species. It would be a long time—as much as eighty million years by one reckoning—before species totals recovered.

Two points need to be kept in mind. First, these are all just informed guesses. Estimates for the number of animal species alive at the end of the Permian range from as low as 45,000 to as high as 240,000. If you don’t know how many species were alive, you can hardly specify with conviction the proportion that perished. Moreover, we are talking about the death of species, not individuals. For individuals the death toll could be much higher—in many cases, practically total. The species that survived to the next phase of life’s lottery almost certainly owe their existence to a few scarred and limping survivors.

In between the big kill-offs, there have also been many smaller, less well-known extinction episodes—the Hemphillian, Frasnian, Famennian, Rancholabrean, and a dozen or so others—which were not so devastating to total species numbers, but often critically hit certain populations. Grazing animals, including horses, were nearly wiped out in the Hemphillian event about five million years ago. Horses declined to a single species, which appears so sporadically in the fossil record as to suggest that for a time it teetered on the brink of oblivion. Imagine a human history without horses, without grazing animals.

In nearly every case, for both big extinctions and more modest ones, we have bewilderingly little idea of what the cause was. Even after stripping out the more crackpot notions there are still more theories for what caused the extinction events than there have been events. At least two dozen potential culprits have been identified as causes or prime contributors: global warming, global cooling, changing sea levels, oxygen depletion of the seas (a condition known as anoxia), epidemics, giant leaks of methane gas from the seafloor, meteor and comet impacts, runaway hurricanes of a type known as hypercanes, huge volcanic upwellings, catastrophic solar flares.

This last is a particularly intriguing possibility. Nobody knows how big solar flares can get because we have only been watching them since the beginning of the space age, but the Sun is a mighty engine and its storms are commensurately enormous. A typical solar flare—something we wouldn’t even notice on Earth—will release the energy equivalent of a billion hydrogen bombs and fling into space a hundred billion tons or so of murderous high-energy particles. The magnetosphere and atmosphere between them normally swat these back into space or steer them safely toward the poles (where they produce the Earth’s comely auroras), but it is thought that an unusually big blast, say a hundred times the typical flare, could overwhelm our ethereal defenses. The light show would be a glorious one, but it would almost certainly kill a very high proportion of all that basked in its glow. Moreover, and rather chillingly, according to Bruce Tsurutani of the NASA Jet Propulsion Laboratory, “it would leave no trace in history.”

What all this leaves us with, as one researcher has put it, is “tons of conjecture and very little evidence.” Cooling seems to be associated with at least three of the big extinction events—the Ordovician, Devonian, and Permian—but beyond that little is agreed, including whether a particular episode happened swiftly or slowly. Scientists can’t agree, for instance, whether the late Devonian extinction—the event that was followed by vertebrates moving onto the land—happened over millions of years or thousands of years or in one lively day.

One of the reasons it is so hard to produce convincing explanations for extinctions is that it is so very hard to exterminate life on a grand scale. As we have seen from the Manson impact, you can receive a ferocious blow and still stage a full, if presumably somewhat wobbly, recovery. So why, out of all the thousands of impacts Earth has endured, was the KT event so singularly devastating? Well, first itwaspositively enormous. It struck with the force of 100 million megatons. Such an outburst is not easily imagined, but as James Lawrence Powell has pointed out, if you exploded one Hiroshima-sized bomb for every person alive on earth today you would still be about a billion bombs short of the size of the KT impact. But even that alone may not have been enough to wipe out 70 percent of Earth’s life, dinosaurs included.

The KT meteor had the additional advantage—advantage if you are a mammal, that is—that it landed in a shallow sea just ten meters deep, probably at just the right angle, at a time when oxygen levels were 10 percent higher than at present and so the world was more combustible. Above all the floor of the sea where it landed was made of rock rich in sulfur. The result was an impact that turned an area of seafloor the size of Belgium into aerosols of sulfuric acid. For months afterward, the Earth was subjected to rains acid enough to burn skin.

In a sense, an even greater question than that of what wiped out 70 percent of the species that were existing at the time is how did the remaining 30 percent survive? Why was the event so irremediably devastating to every single dinosaur that existed, while other reptiles, like snakes and crocodiles, passed through unimpeded? So far as we can tell no species of toad, newt, salamander, or other amphibian went extinct in North America. “Why should these delicate creatures have emerged unscathed from such an unparalleled disaster?” asks Tim Flannery in his fascinating prehistory of America,Eternal Frontier .

In the seas it was much the same story. All the ammonites vanished, but their cousins the nautiloids, who lived similar lifestyles, swam on. Among plankton, some species were practically wiped out—92 percent of foraminiferans, for instance—while other organisms like diatoms, designed to a similar plan and living alongside, were comparatively unscathed.

These are difficult inconsistencies. As Richard Fortey observes: “Somehow it does not seem satisfying just to call them ‘lucky ones’ and leave it at that.” If, as seems entirely likely, the event was followed by months of dark and choking smoke, then many of the insect survivors become difficult to account for. “Some insects, like beetles,” Fortey notes, “could live on wood or other things lying around. But what about those like bees that navigate by sunlight and need pollen? Explaining their survival isn’t so easy.”

Above all, there are the corals. Corals require algae to survive and algae require sunlight, and both together require steady minimum temperatures. Much publicity has been given in the last few years to corals dying from changes in sea temperature of only a degree or so. If they are that vulnerable to small changes, how did they survive the long impact winter?

There are also many hard-to-explain regional variations. Extinctions seem to have been far less severe in the southern hemisphere than the northern. New Zealand in particular appears to have come through largely unscathed even though it had almost no burrowing creatures. Even its vegetation was overwhelmingly spared, and yet the scale of conflagration elsewhere suggests that devastation was global. In short, there is just a great deal we don’t know.

Some animals absolutely prospered—including, a little surprisingly, the turtles once again. As Flannery notes, the period immediately after the dinosaur extinction could well be known as the Age of Turtles. Sixteen species survived in North America and three more came into existence soon after.

Clearly it helped to be at home in water. The KT impact wiped out almost 90 percent of land-based species but only 10 percent of those living in fresh water. Water obviously offered protection against heat and flame, but also presumably provided more sustenance in the lean period that followed. All the land-based animals that survived had a habit of retreating to a safer environment during times of danger—into water or underground—either of which would have provided considerable shelter against the ravages without. Animals that scavenged for a living would also have enjoyed an advantage. Lizards were, and are, largely impervious to the bacteria in rotting carcasses. Indeed, often they are positively drawn to it, and for a long while there were clearly a lot of putrid carcasses about.

It is often wrongly stated that only small animals survived the KT event. In fact, among the survivors were crocodiles, which were not just large but three times larger than they are today. But on the whole, it is true, most of the survivors were small and furtive. Indeed, with the world dark and hostile, it was a perfect time to be small, warm-blooded, nocturnal, flexible in diet, and cautious by nature—the very qualities that distinguished our mammalian forebears. Had our evolution been more advanced, we would probably have been wiped out. Instead, mammals found themselves in a world to which they were as well suited as anything alive.

However, it wasn’t as if mammals swarmed forward to fill every niche. “Evolution may abhor a vacuum,” wrote the paleobiologist Steven M. Stanley, “but it often takes a long time to fill it.” For perhaps as many as ten million years mammals remained cautiously small. In the early Tertiary, if you were the size of a bobcat you could be king.

But once they got going, mammals expanded prodigiously—sometimes to an almost preposterous degree. For a time, there were guinea pigs the size of rhinos and rhinos the size of a two-story house. Wherever there was a vacancy in the predatory chain, mammals rose (often literally) to fill it. Early members of the raccoon family migrated to South America, discovered a vacancy, and evolved into creatures the size and ferocity of bears. Birds, too, prospered disproportionately. For millions of years, a gigantic, flightless, carnivorous bird called Titanis was possibly the most ferocious creature in North America. Certainly it was the most daunting bird that ever lived. It stood ten feet high, weighed over eight hundred pounds, and had a beak that could tear the head off pretty much anything that irked it. Its family survived in formidable fashion for fifty million years, yet until a skeleton was discovered in Florida in 1963, we had no idea that it had ever existed.

Which brings us to another reason for our uncertainty about extinctions: the paltriness of the fossil record. We have touched already on the unlikelihood of any set of bones becoming fossilized, but the record is actually worse than you might think. Consider dinosaurs. Museums give the impression that we have a global abundance of dinosaur fossils. In fact, overwhelmingly museum displays are artificial. The giant Diplodocus that dominates the entrance hall of the Natural History Museum in London and has delighted and informed generations of visitors is made of plaster—built in 1903 in Pittsburgh and presented to the museum by Andrew Carnegie. The entrance hall of the American Museum of Natural History in New York is dominated by an even grander tableau: a skeleton of a large Barosaurus defending her baby from attack by a darting and toothy Allosaurus. It is a wonderfully impressive display—the Barosaurus rises perhaps thirty feet toward the high ceiling—but also entirely fake. Every one of the several hundred bones in the display is a cast. Visit almost any large natural history museum in the world—in Paris, Vienna, Frankfurt, Buenos Aires, Mexico City—and what will greet you are antique models, not ancient bones.

The fact is, we don’t really know a great deal about the dinosaurs. For the whole of the Age of Dinosaurs, fewer than a thousand species have been identified (almost half of them known from a single specimen), which is about a quarter of the number of mammal species alive now. Dinosaurs, bear in mind, ruled the Earth for roughly three times as long as mammals have, so either dinosaurs were remarkably unproductive of species or we have barely scratched the surface (to use an irresistibly apt cliché).

For millions of years through the Age of Dinosaurs not a single fossil has yet been found. Even for the period of the late Cretaceous—the most studied prehistoric period there is, thanks to our long interest in dinosaurs and their extinction—some three quarters of all species that lived may yet be undiscovered. Animals bulkier than the Diplodocus or more forbidding than tyrannosaurus may have roamed the Earth in the thousands, and we may never know it. Until very recently everything known about the dinosaurs of this period came from only about three hundred specimens representing just sixteen species. The scantiness of the record led to the widespread belief that dinosaurs were on their way out already when the KT impact occurred.

In the late 1980s a paleontologist from the Milwaukee Public Museum, Peter Sheehan, decided to conduct an experiment. Using two hundred volunteers, he made a painstaking census of a well-defined, but also well-picked-over, area of the famous Hell Creek formation in Montana. Sifting meticulously, the volunteers collected every last tooth and vertebra and chip of bone—everything that had been overlooked by previous diggers. The work took three years. When finished they found that they had more than tripled the global total of dinosaur fossils from the late Cretaceous. The survey established that dinosaurs remained numerous right up to the time of the KT impact. “There is no reason to believe that the dinosaurs were dying out gradually during the last three million years of the Cretaceous,” Sheehan reported.

We are so used to the notion of our own inevitability as life’s dominant species that it is hard to grasp that we are here only because of timely extraterrestrial bangs and other random flukes. The one thing we have in common with all other living things is that for nearly four billion years our ancestors have managed to slip through a series of closing doors every time we needed them to. Stephen Jay Gould expressed it succinctly in a well-known line: “Humans are here today because our particular line never fractured—never once at any of the billion points that could have erased us from history.”

We started this chapter with three points: Life wants to be; life doesn’t always want to be much; life from time to time goes extinct. To this we may add a fourth: Life goes on. And often, as we shall see, it goes on in ways that are decidedly amazing.