About 140 miles from the Antarctic coast, the scientific research ship JOIDES Resolution is drilling deep into the ocean floor. The objective: To find out how Antarctica changed from a warm and vegetated continent, as it was 50 million years ago, to the frozen, ice-covered continent we see today. The story of Antarctica's ice is written in the muddy sediments that have been laid down and slowly buried under the surrounding ocean floor, and I am one of the 33 marine geologists onboard the Resolution who are trying to decipher it. Each sediment core (a 30-foot-long cylinder of material) that comes up from the hole is a new page in the book, telling us about the world in earlier times. We start with the present day at the seabed, and then work our way deeper down into the sediments and further in the geological past. 

It helps that we know the outlines of Antarctic ice history already—this expedition is the most recent of a handful of scientific drilling campaigns around the continent since the 1970's. The first major ice sheet grew 34 million years ago, almost out of nothing to be maybe as big as the one today (over a mile thick, on average). Since then, parts of it have melted and grown back hundreds of times. But ours is the first drilling expedition to really explore the glacial history of Wilkes Land—the part of Antarctica you reach if you sail directly south from Australia. The ice history of the Wilkes Land coast is hardly known at all. We want to answer questions like: What was Antarctica like before the ice—tundra, temperate forests, palm trees? When did ice first become stable and steady along this coast? Has this part of the ice sheet collapsed in the past, disgorging iceberg armadas into the ocean? And how much did the climate have to warm to make this happen? 

Wilkes Land was named after Charles Wilkes, whose expedition first saw the Antarctic coastline here in 1840 (around the same time as his French rival, Dumont D'Urville, who named a nearby stretch of coast after his wife, Adélie). The Wilkes Expedition played a major role in the growth of 19th century U.S. science, and the thousands of animal, plant and mineral specimens it gathered formed the basis of the collections of the newly founded Smithsonian Institution. The scientific exploration continues today. Our expedition set off from New Zealand a month ago, and took a week to cross the rolling seas of the roaring forties and the Southern Ocean to our drill sites offshore of Wilkes Land. It is an international effort: Among the 120 crew, technicians and scientists onboard, 20 nationalities are represented, and there is someone from every continent. Every continent that is, except for Antarctica itself. We would have to adopt a penguin for the full set (if it wasn't prohibited under the Antarctic Treaty—things have changed since Wilkes' day). 

The sediments packed inside the drill pipe and winched up through the moon pool in the center of the ship contain microscopic fossils, like the shells made by diatom algae. These tell us about the temperature of the sea, because some species prefer it warm, some cool, and others very cold. We find pebbles and gravel that stand out from the mud that encloses them. These are dropstones, eroded by glaciers and carried by icebergs out to sea; finding them is a sure sign of ice. Sometimes there are feet-thick layers of tumbled sediment that result from underwater avalanches of material pushed by advancing glaciers from the shallow continental shelf. Pollen grains, blown offshore on the winds or carried by water, tell us what kind of vegetation used to exist on Antarctica in warmer times. We're also using new techniques to decode the sediments. Molecular fossils, for example, are resistant molecules that can still be found long after the microbes or plants they were part of died and decayed. They too carry news of vegetation and temperatures, like ancient thermometers. 

Now halfway through the expedition, we have dodged icebergs, sat out storms and ducked exploding sediment cores. In between, we have drilled down 3000 feet to reach the warm Eocene "greenhouse" world of 50 million years ago, and another site gave a beautiful, annually layered, 10,000-year record of climate since the ice last retreated. When we put all this information together we will have a good sense of what happened to the great Antarctic ice sheet under the warmer climates of the past. But our job as geologists is not just to write history; we are drilling back in time to reveal lessons for the future. Many times in the past, the Earth was as warm as it is predicted to be at the end of this century. Once we know how stable (or unstable) the ice was under those conditions, the better we will be able to predict how much, and how fast, the ice may melt in the future.