The moulin opened up on the surface of the Austfonna ice cap on a Tuesday in mid-May, after three weeks of unusually warm Arctic weather had thinned the upper firn.

Glaciologist Sigrid Halvorsen, leading a Norwegian Polar Institute team, identified the new opening on a satellite pass and reached it by snowmobile from the field station at Kinnvika two days later.

What she saw, peering down with a headlamp from the lip of the moulin, was a vertical shaft descending into a blue larger than the day above.

Over the next nine days, a six-person team rigged the shaft, descended 213 meters in three pitches, and found themselves standing on the floor of a chamber roughly the size of a cathedral nave.

They named it Thrymheim, after the mountain dwelling of the Norse frost giant Thrymr, mostly because Halvorsen's daughter, age nine, had insisted.

The chamber's walls are pure ice, banded in pale blue and white, lit by their own headlamps into something between architecture and weather.

The air in the deeper portions of the chamber, sampled in sealed flasks and shipped to a lab in Tromsø, has now been dated by trapped carbon-14 to roughly 700 years before present.

The air in the deeper portions of the chamber, sampled in sealed flasks and shipped to a lab in Tromsø, has now been dated by trapped carbon-14 to roughly 700 years before present.

That makes Thrymheim a kind of time capsule of the late medieval Arctic atmosphere, sealed since approximately 1325, when the surrounding ice closed over its original entrance.

On the third day of exploration, the team found something they had not expected. Embedded partway up the eastern wall, encased in ice, were the rib bones of a bowhead whale.

How a whale skeleton came to be lodged 213 meters beneath an ice cap is, Halvorsen admits, a problem.

The leading hypothesis is that the bones were carried by a meltwater stream during a warm interval, deposited in a moulin that subsequently refroze, and have been migrating slowly through the ice ever since.

Radiocarbon dating of a small sample of the bone returned an age of roughly 1,400 years, consistent with a long-dead whale whose remains spent centuries on or near the surface before entering the ice system.

The team also collected microbial samples from the ice walls, which are being analyzed for cold-adapted organisms.

Early results suggest the presence of a small but viable community of psychrophilic bacteria, potentially representing lineages that have been isolated from the surface for the full 700 years.

The cave will not remain accessible for long. Halvorsen estimates the moulin will close again within two summers, either by snow accumulation or by the slow plastic flow of the surrounding ice.

A second expedition is planned for July, equipped with a fiber-optic sensor array intended to remain in place after the cave seals.

There is a particular quality to silence inside ice that has not been disturbed for seven centuries. It is not absence. It is a kind of attention.

The team, Halvorsen said, spent the last hour of their final dive sitting on the chamber floor without speaking, and then climbed back out.