This part of the Victorian coast, known as Flat Rocks, is near the resort town of Inverloch, about a two-hour drive south-east of Melbourne. Partly buried by flakes of battleship-gray rock is a telling geological formation. Tongues of dark tan sediment droop into the lighter-colored layer below. The formation is called a "cryoturbation" and was caused when once-frozen clay sank into an underlying layer of sand during a thaw long ago.

Snow and ice are rare in this part of Australia today. But evidence from Flat Rocks and other nearby sites confirms that a little over 100 million years ago, it was bloody cold around here. Though about a third of Australia now lies within the tropics, back then the continent sat about 2,000 miles south of its current position, snuggled against Antarctica. South-eastern Australia probably had a climate similar to that of Chicago, if not Fairbanks.

All the more surprising, then, that dinosaurs thrived here at that time. Think "dinosaurs" and you probably conjure up behemoths trudging through sweltering swamps or torrid tropical forests. But scientists working in Australia, Alaska and even atop a mountain in Antarctica have unearthed remains of dinosaurs that prospered in environments that were cold for at least part of the year. Polar dinosaurs, as they are known, also had to endure prolonged darkness—up to six months each winter. "The moon would be out more than the sun, and it would be tough making a living," says paleontologist David Weishampel of Johns Hopkins University.

At the time dinosaurs arose, around 220 million years ago, the earth's continents were fused into a single supercontinent we now call Pangea. It began breaking up around 200 million years ago, and Australia and Antarctica, which were still stuck together, stayed near the South Pole. About 100 million years ago southern Australia sat close to the bottom of the planet, and was just starting to pull away from Antarctica.

During the animals' heyday in the early Cretaceous period, the sun didn't rise in southern Australia for one and a half to four and a half months every year. At the North and South poles, the gloom lasted for six months. Plant growth in these areas would have periodically slowed or stopped, potentially creating a food crisis for any dinosaurs that lived there.

Dinosaurs also thrived farther south. Antarctica hasn't moved much in the past 100 million years, stalling over the South Pole. Today, well-insulated animals and stubbly plants can survive the continent's brutal cold, at least close to the coast. But fossilized leaves and other plant remains suggest that during the dinosaurs' day Antarctica had a temperate climate. Judd Case of Eastern Washington University in Cheney says that Antarctic dinosaurs from the late Cretaceous period around 70 million years ago resembled those that lived in other parts of the world some 60 million years earlier. Case says this suggests that some dinosaurs hung on in Antarctica long after they had died out elsewhere. Perhaps Antarctica was an oasis for them as flowering plants spread across the rest of the world and outcompeted the pine tree relatives that warmer-climed dinosaurs ate.

William Hammer of Augustana College in Rock Island, Illinois, digs at an elevation of 13,000 feet on the slope of Mount Kirkpatrick, about 400 miles from the South Pole. He has pried out the bones of Cryolophosaurus ellioti, a 22-foot-long meat-eater with a bony crest curving up from its forehead like a cowlick. He has also found fossil evidence of a prosauropod, an ancestor of enormous dinosaurs such as Brachiosaurus and Apatosaurus.

Dinosaurs had two choices when winter arrived—tough it out or try to escape. The question of how dinosaurs survived the polar cold has gotten entangled with the broader question of whether the ancient beasts were warmblooded (endothermic), like modern birds and mammals, or cold-blooded (ectothermic), like modern reptiles. In a cold environment, endotherms keep their bodies warm enough for muscles to flex and nerves to fire by generating heat through their metabolism. Ectotherms, by contrast, warm their bodies by absorbing heat from their surroundings—think of a lizard basking on a rock. Endothermy isn't necessarily better, notes David Fastovsky of the University of Rhode Island. Endotherms have the edge in stamina, but ectotherms need much less food.

The prize discovery from Rich's Dinosaur Cove excavation suggests that Leaellynasaura stayed active during the long polar winters. A two-inch-long Leaellynasaura skull the color of milk chocolate is the closest to a complete dinosaur skull the team has found. The base remains partly embedded in a disk of gray rock but enough of the bone is visible for Rich to analyze the size of the eye sockets. Hypsis generally had big eyes, but Leaellynasaura's are disproportionately large—perhaps so they could capture more light during the protracted murk of polar winters. Moreover, the back of the same skull has broken off to expose a mold of the brain, known as an endocast. Rich found that the dinosaur had bulging optic lobes, parts of the brain that process visual information. Leaellynasaura's optic lobes are larger than those from hypsis that lived in non-polar environments, suggesting that it had extra brainpower to analyze input from its big eyes.

Other dinosaurs might have migrated south for the winter (or north, if they lived in the Southern Hemisphere). Rich says his dinosaurs would have made unlikely travelers. They were small, and an inland sea would have blocked their path to warmer climes. But Edmontosaurus, from Alaska's North Slope, is a better candidate for seasonal migration. Adults were about the size of elephants, so they would not have been able to crawl under rocks when temperatures fell. Rough calculations suggest that by ambling at about 1 mile per hour—"browsing speed" for animals of that size—herds of Edmontosaurus could have journeyed more than 1,000 miles south in three months, says paleobotanist Bob Spicer of the Open University in Milton Keynes, Britain. Such a migration would have taken them out of the "zone of darkness" and into areas where plants might have still been growing.

For his part, Fiorillo doubts it. He and Gangloff contend that juvenile Edmontosaurus grew too slowly to have tramped long distances. They couldn't have kept up with a herd, so the animals must have stayed put, regardless of temperatures. This kind of back-and-forth might be dizzying, but it's how science moves ahead, especially in paleontology, where researchers have to draw conclusions from small numbers of often-fragmentary fossils.

The dinosaurs had an impressive run. They settled every continent, grew bigger than any other land animals and lasted for more than 150 million years. And then they vanished. Their demise has spawned more than a little speculation about its cause. Scenarios range from disease or competition with mammals to the flyby of an as-yet-undetected companion to the sun, a kind of death star.

Most paleontologists have accepted another extraterrestrial killer, an asteroid more than six miles wide that socked Earth 65 million years ago. It gouged a crater more than 100 miles wide on what is now the Yucatán Peninsula in Mexico. According to the leading scenario, the impact threw huge amounts of dust and other debris into the atmosphere, blocking sunlight and sinking the Earth into darkness for weeks or even months. A global disaster certainly struck at the time, according to overwhelming fossil and geological evidence. As Fastovsky and Weishampel write in The Evolution and Extinction of the Dinosaurs, "the world's oceans were virtually 'dead'" as photosynthesis by plankton ceased and marine food webs unraveled. The dinosaurs died, while the ancestors of today's mammals, birds and reptiles hung on.

Paleontologists disagree about what the existence of polar dinosaurs says about the asteroid-winter scenario. Fiorillo says he is skeptical of it because "dinosaurs in Alaska were doing just fine in conditions just like that." He argues that climate changes caused by shifts in circulation of the atmosphere and oceans probably did in the dinosaurs.

But Rich says that the lives of polar dinosaurs can help researchers understand why dinosaurs went extinct after the impact. The catastrophe had to have been long and severe enough to kill off the dark- and cold-adapted animals. "You can't just have it [darkness] for a month and do the job," he says.

But Fastovsky says that polar dinosaurs tell us nothing about the animals' demise because we don't know whether these particular species were even alive at the end of the Cretaceous period. Rich's Australian dinosaurs were long extinct by the time the asteroid hit.