Forests and the ice

Why this is extraordinary

The white continent was not always white. Glossopteris, a seed-fern with a tongue-shaped leaf, grew on Antarctica in the Permian, when the land was part of the southern supercontinent and the ice you know had not yet arrived. The leaf in the photograph is that plant, collected from Antarctic rock and kept in a geology museum. It is a drawer specimen, not a stump in the snow.

Much later, and still astonishing, forests grew inside the Cretaceous polar circle. On Alexander Island, along the east coast, fossil conifer forests and forest floors sit in the Lower Cretaceous part of the Fossil Bluff Formation, Aptian to Albian in age. They grew at a palaeolatitude of about 65 to 75 degrees south. Their growth rings are wide and uneven, closer to living trees in a warm-temperate place with a long season than to the stunted trees of a modern high latitude. The wood is stone now. At the south end of that latitude range, a tilt like the Earth's today gives a winter when the sun does not rise and a summer when it does not set. At the north end the winter sun still comes up, low and brief. The published figure is a range, not one cliff, so this room does not turn it into a count of dark days. The winters were dark, or the sun barely cleared the horizon. The trees grew anyway, and the wide rings are the light months written in the wood.

By the early Palaeogene the peninsula’s plants look cool-temperate. On Seymour Island, off the northeastern tip of the peninsula, fossil leaves from the Paleocene and from the middle to late Eocene La Meseta Formation are dominated by Nothofagus, the southern beech, with araucarian and podocarp conifers and ferns. The flora resembles the beech forests of southern South America. The pin for this room is that island, because the leaves there are a published flora with a place, not because a forest is still green on the shore.

Ice, as a continental sheet, comes in at the Eocene–Oligocene boundary. Marine oxygen isotopes, a record of ancient seawater kept in shells, and glacial deposits on the continent put the build-up at about 34 million years ago, the event called Oi-1. There are hints of ice earlier, even arguments for ephemeral ice in the Cretaceous, so “forests, then a switch” is too clean. The direction, toward a lasting sheet, is the part that holds. After that boundary the ice sheet is the fact a map has to live with. It is millions of years old. A reading in which the ice is a recent cover over a people meets that age in this room. The reading itself is set out in Under the ice, beside what the drills and the radar have actually returned.

Lakes under the ice are real, and they are water. A borehole into Lake Vostok reached a dark freshwater body sealed under kilometres of ice. What came back was water and ice, not masonry. No excavation has produced a city, a wall, or a tool from under the sheet. A parchment of 1513 that some readers take for an ice-free Antarctic coast is weighed in the Piri Reis map room, with the later sheets in Charts of a southern land. This room is the plants and the ice those drawings are set next to. Peaks that stand out of the sheet, including one named for a pyramid, are in Pyramid-like peaks.

Antarctica can be both a fossil forest and an ice sheet. The forests are published. A city under the ice is not a find. One reading takes the leaves, the rings, and the isotope step as a warmer geological past that ended in a long-lived sheet. The other reading pictures the ice as hiding a lost capital, and asks whether that warmer past can carry a people into historical time. Both can be set out, and the second one has its own room so that a missing city is not used to make it smaller. What the geology changes is the unit: millions of years, beside a parchment of 1513. A wall, a tool, or a city would be a different kind of object, and the digs have not produced one.