The Three Swedish Engineers Who Sailed for the North Pole in a Hydrogen Balloon in 1897 and Were Found Frozen on White Island Thirty-Three Years Later

Salomon August Andrée (1854–1897) was a Swedish engineer and chief of the Swedish Royal Patent Office. In the early 1890s he had become persuaded that the North Pole — which no surface expedition had been able to reach — could be reached by hydrogen balloon. The plan: launch from northern Spitsbergen, drift across the Pole on prevailing winds, and land somewhere across the Arctic Ocean, ideally in Russia or in northern Canada.

The plan was technically optimistic. Hydrogen balloons of the 1890s leaked hydrogen at approximately 1 percent of their volume per day; the proposed Arctic crossing would take five to ten days; the balloon was unlikely to retain enough lift for the full crossing. Andrée’s response to this objection was a system of trailing drag-ropes — three weighted ropes that would partially drag on the ice and snow, slowing the balloon and allowing it to be steered into wind by deploying sails between the gondola and the balloon.

The drag-rope steering system had been tested in 1896 and had not worked.

July 1897

Andrée nonetheless persuaded the Royal Swedish Academy of Sciences and several private donors (including Alfred Nobel personally) to fund the expedition. The balloon — named the Eagle (Örnen) — was built in Paris by Henri Lachambre and shipped to Danes Island in northern Svalbard in summer 1897. Andrée had two companions: the 24-year-old physicist Nils Strindberg (nephew of the playwright) and the 27-year-old engineer Knut Frænkel.

The Eagle was launched at 2:30 p.m. on 11 July 1897 from a wooden hangar on Danes Island. The wind was a strong southerly. Within minutes of launch the drag-ropes had torn off — the impact of the launch jerk had snapped them at the gondola attachments. The balloon climbed steeply to approximately 500 metres and was carried north out of sight of the launch crew within an hour.

The three men dropped a series of carrier-pigeon messages and waterproofed buoys over the next three days. Four pigeons were recovered (one by a fishing boat in 1897, the others later); five buoys were recovered between 1897 and 1899. The messages reported that the balloon was being driven northeast at approximately 30 km/h, that the cabin was in good order, and that morale was good.

The last message — from buoy 7, recovered in 1899 — was dated 13 July 1897, 10 p.m.: “Our journey has hitherto gone well. We are still moving on at a height of 600 metres in fine direction. The dynamometer is gradually rising, the dynamometer indicates 30 N.E. We’re all well. Sincerely. Andrée, Strindberg, Frænkel.”

Nothing more was heard.

What had happened

The Eagle was forced down onto the pack ice by progressive loss of hydrogen lift on 14 July 1897 at approximately 82° 56’ N, 29° 52’ E — about 800 km short of the Pole. The crash was relatively soft (the balloon was settling rather than falling) and the three men were uninjured.

The post-1930 reconstruction from their recovered journals and photographs runs as follows.

For approximately a week the three men attempted to make the balloon airborne again by removing weight from the gondola. They failed. On 22 July 1897 they began walking south, dragging the gondola’s small folding boat with them as sled, loaded with food, two rifles, and minimal scientific equipment. The plan was to walk approximately 300 km south to one of the Russian or Norwegian fishing posts in the eastern Svalbard archipelago.

The walk took ten weeks. The pack ice was breaking up under the Arctic summer’s continuous sunlight; the men were repeatedly forced to detour around open water leads, to ferry across new channels, and to remake their sled rigging. By 5 September 1897 they had reached approximately 82° N — about 100 km from where they had crashed — having drifted east at almost the same rate they had walked south. The walking had achieved very little net progress.

On 17 September 1897 they reached the western edge of Kvitøya (“White Island”) — the easternmost island of the Svalbard archipelago, an essentially uninhabited ice-covered rock about 800 km north of the Norwegian mainland. They camped there and began to set up a winter shelter from the boat and salvaged tent canvas.

Their journals end in early October 1897.

The Norwegian fishing expedition

The disappearance of the Eagle became one of the most-discussed polar mysteries of the late 19th and early 20th centuries. Multiple search expeditions through the 1900s and 1910s failed to find any trace.

In summer 1930 the Norwegian fishing vessel Bratvaag, in the Arctic for routine northern fishing-grounds work, anchored off Kvitøya on 5 August 1930 because of bad weather. Two of the crew walked across the island looking for fresh water and found, exposed by an unusually warm summer’s snow-melt, the canvas-wrapped body of a man together with a recognisably late-19th-century brass boat hook. A small expedition the next day found the second body. A separate Norwegian expedition under the journalist Knut Stubbendorff returned to the island on 5 September 1930 and located the third body, the three men’s journals, the photographic cameras containing exposed but unprocessed film, and most of the camp equipment.

The bodies, the journals, and the cameras were brought back to Tromsø in mid-September 1930 and to Stockholm in mid-October 1930. The state funeral on 9 October 1930 drew an estimated 40,000 mourners in the Stockholm streets.

What killed them

The cause of the three men’s deaths in early October 1897 has been the subject of repeated forensic investigation since 1930. The leading theories include:

— Trichinosis from undercooked polar bear meat. The expedition had killed approximately seven polar bears during the ten-week walk south, and their journals record extensive consumption of bear meat. Polar bear muscle and visceral tissue is heavily infested with Trichinella nativa. Acute trichinosis can be fatal in approximately three weeks. The clinical pattern from the journals is consistent.

— Lead poisoning from food tins. The expedition’s tin food supply had been soldered with lead, similar to the Franklin expedition forty-five years earlier (see Franklin).

— Carbon monoxide poisoning from the camp Primus stove inside an insufficiently-ventilated tent.

— Vitamin A poisoning from polar bear liver, similar to the Mawson-Mertz death pattern fifteen years later.

— Cumulative hypothermia, scurvy, and starvation.

The 21st-century forensic consensus (Bea Uusma, 2013) emphasises a combination of carbon monoxide from the tent stove and acute trichinosis from a polar bear they had killed approximately ten days before death. The two factors are independently plausible.

The photographs

The most remarkable aspect of the 1930 recovery was the photographic record. The expedition’s two Kodak roll-film cameras had been frozen continuously in the canvas wrappings on Kvitøya from October 1897 to August 1930. The film had been preserved by the cold. The Swedish photo-processing chemist John Hertzberg developed the rolls in October 1930 and recovered approximately 93 usable images of the crash, the ten-week walk south, and the camp on Kvitøya.

The images include the moment of the balloon crash on 14 July 1897 (photographed by Strindberg from the gondola), the three men hauling the boat across the pack ice in August 1897, the polar bears they killed, and the camp on Kvitøya in mid-September 1897.

The three men are looking at the camera in most of them, smiling, calm. They believed they would be rescued. They had not been rescued for thirty-three years.