Bang and blame

MARKETING is everything. Dinosaurs were introduced to the public in the 1850s, when life-size models of them were used to decorate the most popular tourist attraction in the world—London's Crystal Palace. The “Bone Wars” between Edward Cope and Othniel Marsh, to transport dinosaur fossils from the quarries of the newly opening American west to the museums of the established east, gave them wealthy sponsors, such as Andrew Carnegie. And the creatures' spectacular lives were rounded with a spectacular death when it was demonstrated to the satisfaction of most palaeontologists that they were wiped out by a collision between the Earth and an asteroid. A storybook ending to a storybook existence. For many people, dinosaurs are prehistoric life.

Yet the extinction of the dinosaurs at the end of the Cretaceous period, 65m years ago, though accompanied by that of many other creatures at the same time, was not the biggest die-off in world history. That distinction belongs to the end of the Permian period, some 250m years ago. The end of the Permian was marked by the extinction of 90% of the Earth's species. They may not have been as spectacular as the dinosaurs, but they included some noteworthy reptiles, such as the sailback Dimetrodon illustrated above.

The end of the Permian was once thought of as a whimper to the Cretaceous's bang. It was believed to have been caused by a slowly drying climate that was the result of continental drift bringing all of the Earth's land together into an enormous supercontinent. But this feeling changed when evidence of huge volcanic eruptions, known as flood basalts, was discovered. Flood basalts also marked the end of the Cretaceous—they form the Deccan Traps, in central India. Now, the other element of the demise of the dinosaurs has turned up at the end of the Permian. A paper in this week's Science purports to identify an impact crater as big as the famous “dinosaur killer” at Chicxulub, in Mexico. The “sailback killer”, if that is what it was, smashed into the Earth in north-western Australia.

Shattered evidence

Luann Becker, of the University of California, Santa Barbara, and her colleagues, the authors of the paper, have been stalking the sailback killer for years. In 2001, they reported finding fullerenes in rocks from the end of the Permian. Fullerenes—familiarly known as buckyballs, after Richard Buckminster Fuller, an architect whose geodesic domes they are alleged to resemble—are cage-like molecules made of carbon atoms. The cages can trap inert gases such as helium and argon, and when Dr Becker analysed the weights of the helium and argon atoms from her late-Permian fullerenes, she found that the mixture resembled that of gas from meteorites, rather than from the surface of the Earth.

Many people were sceptical. They pointed in particular to the absence of iridium in the rocks in which the fullerenes were found. The presence of iridium—a metal rare at the surface of the Earth, but relatively abundant in meteorites—was the first clue pointing towards the dinosaur-killing impact.

Last year, however, the researchers reported more evidence. They had been working in Antarctica, and had found what they believed were ejecta from an impact. They still did not find any iridium, but they did find fragments of rock that looked like pieces of meteorite. There was a thin layer of a type of rock known as breccia, which forms from the smashed-up material launched out of a crater. And there was “shocked quartz”—crystals of the mineral shattered in a way characteristic of the shockwaves that follow an impact.

All very significant. But not complete proof. For that, a crater of the right age would be required. And this is what seems to have turned up. Argon-argon dating, which relies on the fact that radioactive argon decays into a non-radioactive isotope of the gas at a known and constant rate, suggests the Australian crater is 250.1m years old, plus or minus 4.5m years.

The crater is not visible at the surface. In fact, it is buried under 3km of sedimentary rocks. It was first detected in the 1970s by oil prospectors, who mapped its geometry by listening to the reflections from its various rock layers of explosions they let off at the surface. More detail was worked out later, by precise measurements of gravity above it. Both the size of the structure, and its approximate age, as indicated by fossils in the rocks, suggested to Dr Becker that it might be what she was looking for. Her argon-argon dating of the local breccias has now confirmed that. And these rocks also contained small, glassy spherules, another signature of impacts.

Put together, then, the evidence is impressive. Dimetrodon and its cousins do, indeed, seem to have met their fate at the hands of a huge impact, with a background of massive volcanic eruptions, just like the dinosaurs.

Yet there is a niggling worry in all this. Why should flood basalts and asteroid impacts coincide? One possibility is that it is just that—a coincidence. Once, maybe. But twice? (And, indeed, more often than that if evidence of an impact coinciding with flood basalts at a third mass extinction at the end of the Triassic period stands up.) A second idea is that big impacts cause massive eruptions by shocking the Earth so profoundly that its innards get upset. But in the case of the Cretaceous, the evidence suggests that the eruptions started before the crater was formed. So there is the horrible possibility that everybody has got things the wrong way round, and it is the vulcanism that causes the craters.

That, at least, is the theory of Jason Phipps Morgan and his colleagues at Kiel University in Germany. A few months ago they suggested that flood basalts might be accompanied by huge explosions, which they dubbed “Verneshots” because, like the launching gun in Jules Verne's novel “From the Earth to the Moon”, they would be capable of launching things into space.

A Verneshot would, according to Dr Phipps Morgan, produce almost all of the evidence thought to indicate an asteroid strike: iridium (which would be brought up from deep in the Earth), shocked quartz, buckyballs, spherules—and big impact craters caused by the return to Earth of huge gobbets of material the explosion ejected. Meteorite fragments are more difficult to explain. But the Earth is hit by meteorites all the time, so they could be “contamination” from smaller impacts.

At the moment, this idea is mere hypothesis. But Dr Phipps Morgan points to a strange, circular gravitational anomaly buried under the lava of the Deccan Traps. This, he speculates, could be the muzzle from which a Verneshot was fired.

The idea that mass extinctions are caused by impacts from outer space has been one of the best marketed pieces of popular science—it has even inspired Hollywood movies. It would be ironic, indeed, if Dr Becker's Australian crater, on the face of things such eloquent evidence for the extraterrestrial nature of mass extinctions, turned out to be a crucial nail in that theory's coffin.