Introduction
One of the most captivating polar expeditions to take place during the heroic age was Swedish engineer Salomon August Andrée’s attempt to reach the North Pole in a balloon. Andrée, together with his companions Nils Strindberg and Knut Frænkel, departed on 11 July 1897 in the hydrogen balloon Örnen (The Eagle) from Danes Island in the northwest part of the Svalbard archipelago. During their departure, the heavy drag ropes intended to allow the balloon to deviate somewhat from the prevailing wind direction, each several hundred metres long, unfortunately detached. The crew on the ground observed the problem and expected Andrée to make an emergency landing on Amsterdam Island, just to the north of Danes Island. However, he did not. The balloon continued its flight in a northeasterly direction and eventually disappeared from sight. Andrée, Strindberg and Frænkel were gone, never to return.
Their last camp and remains were not discovered until 1930 when, by chance, they were located on White Island, a remote and desolate island between Svalbard and Franz Josef Land. From the recovered diaries (reviewed in the very first issue of Polar Record; J.M.W., 1931) and other items, it became clear that the explorers had finally landed on the pack ice at 82°56′N 29°52′E after almost three days in the balloon. On 22 July, they started a march towards a depot at Cape Flora in the Franz Josef Land archipelago.
After almost a week of hard work relaying their heavy sledges, they realised that their speed was too slow and decided to discard some of their supplies to make the sledges lighter to avoid relaying. After another week, they realised that the ice drift was moving them west at about the same speed they were marching to the east, forcing them to change course towards another depot that had been laid for them at the Seven Islands, just north of Svalbard. In mid-September, after another month and a half of gruelling travel, they realised that they would not be able to reach land before winter set in. They built a small hut on the large ice floe they now occupied, planning to stay there for the winter. However, the floe soon started to drift south at a relatively high speed, passing the east side of White Island. The floe eventually started to break up, but Andrée, Strindberg and Frænkel were able to get themselves and their supplies ashore at Andréeneset, a rocky beach that constitutes the only ice-free part of White Island.
Shortly after reaching White Island, all three expedition members died. The details are unclear, but we know for a fact that Strindberg was not the last to perish, because he was the only one to be buried. The crew’s causes of death have been the subject of much speculation and analysis ever since, but it will never be possible to establish exactly why they died because their remains were cremated after their discovery in 1930.
The theory that currently seems most credible was postulated by Uusma (Reference Uusma2013), who suggested that Strindberg was killed and Frænkel mortally wounded in a polar bear attack, and that Andrée used morphine to commit suicide when he realised that he could not get back to civilisation by himself. Over the years, many researchers have held the opinion that trichinosis (from polar bear meat) was the direct or indirect reason for the expedition members’ deaths. However, Uusma (Reference Uusma2013) has, based on detailed analyses, rejected trichinosis as well as most of the other suggested hypotheses regarding causes of death (e.g. lead poisoning, hypervitaminosis A and botulism), leaving only her own theory as probable given the existing evidence.
The purpose of this note is not to dwell further on what might have happened on White Island in the beginning of October 1897. Instead, the paper comprises an analysis of whether the expedition members actually had a chance to survive after their landing at 82°56′N 29°52′E. Research in which historic polar expeditions are reviewed in terms of ‘what if’ is not a new idea. For example, Halsey and Stroud (Reference Halsey and Stroud2011) analysed whether Captain Scott could have survived his Terra Nova expedition, based on today’s physiological knowledge. As the analysis here will show, Andrée, Strindberg and Frænkel might have survived if they had laden their three sledges from the start with weights each of them could pull sustainably and set a course directly towards their Seven Islands depot. The strategy they actually chose—to relay very heavily laden sledges in almost the opposite direction of the general ice drift—left them with basically no chance of survival.
Analysis
Imagine that you are Salomon August Andrée on the evening of 14 July 1897 at 82°56′N 29°52′E, standing next to your wrecked balloon on the Arctic pack ice. You have two strong and intelligent, albeit inexperienced, companions. In order to survive, you must somehow get back to civilisation. You have basically all the equipment and supplies you need; in fact, you have supplies in larger quantities than you could feasibly carry on your sledges. How would you proceed?
Andrée, Strindberg and Frænkel spent over a week at the landing site before beginning their sledge journey. During this time, according to their diaries, they mounted their boat and their sledges, made observations regarding ice and weather, scouted for land and shot their first polar bear. They must also have thought a great deal about where to go and what to bring. Finally, on 21 July, they packed their sledges and began their march in a southeasterly direction, hoping to reach their depot at Cape Flora, in the southwest of Northbrook Island (79°57′N 50°05′E). Within just a few hours, however, they found that none of the men could pull his own sledge efficiently. The heavily laden sledges had to be relayed.
As mentioned, the two major questions the expedition members faced after the landing were where to go and what to bring. There were only four more or less reasonable answers to the first question:
One option was to make for Giles Land, which appeared to be the nearest land according to their faulty map. In fact, there was no Giles Land. The question of Giles Land had already been explored for several decades, without results, when the Andrée expedition took place (Capelotti, Reference Capelotti2013), so Andrée probably doubted its existence. However, Andrée mentioned several times in his diary that they searched in vain for Giles Land when they passed the area where its existence was indicated on their map. Still, Giles Land was probably not discussed seriously anyway because of its desolate location and because there would be no supplies for them there.
Another plan could have been to remain in camp, with the idea of drifting with the ice, like Nansen did with the Fram, and using the boat when they eventually reached open water. The primary advantages of this option would be that they would not have to leave any supplies behind, and that they would not have to exert themselves with sledging. However, uncertainties regarding where and when they would meet open water, in combination with their boat’s lack of seaworthiness and the fact that they might have to spend the winter on the ice, probably made them disregard this option, if they ever truly considered it.
One of the two sites where supplies had been deposited for the expedition was Seven Islands, just north of Svalbard. Hence, this must have appeared to be a more reasonable option than striking out for Giles Land or staying put.
The other depot that had been laid for the expedition was on Cape Flora in the Franz Josef Land archipelago, and this is the location they chose to pursue. This depot was larger than the one on Seven Islands, and on their maps the distance to land appeared about the same in this direction as towards Seven Islands. Beside the larger depot, Cape Flora also had other advantages, including Leigh Smith’s hut from 1881–82 and resources related to the Jackson–Harmsworth expedition of 1894–97 that they could benefit from. Hence, had speed not been an issue, Cape Flora would stand out as the natural option. However, speed was certainly an issue, as this paper shows, and they should have put much more emphasis on it. Reaching terra firma before the polar winter set in should have been a top priority to maximise the probability of survival.
The expedition realised only a few hours after starting towards Cape Flora that their sledges were so heavily laden (approximately 200 kg each, not counting the sledges themselves) that they would have to relay them instead of having each man pull his own sledge. Hence, they had to cover five kilometres to gain one of actual progress. They would probably not have truly quintupled their speed if they could have managed without relaying, given the occasional need to work together to get the sledges over lanes and pressure ridges, but their average speed would obviously have increased substantially. However, they chose to keep relaying for almost a week before they finally decided to reduce the sledge weights by approximately 60 kg each. During this period, therefore, they spent their best energy on very hard work, yet got almost nowhere. The psychological impact of this initial debacle with the sledges may be key to understanding the final outcome of the expedition.
A closer investigation of what the men chose to bring with them reveals many items that were essentially useless from the perspective of pure survival. Space considerations prevent us from a detailed analysis here, but a rough estimate is that they could have spared themselves almost half the initial weight, which obviously would have increased their speed considerably. After they did finally reduce the weights in order to avoid relaying, Strindberg noted a daily distance of 7–8 km as being a ‘good march’.
The next change in plans came when they gave up on reaching Cape Flora, realising after two weeks that they could not overcome the ice drift. Instead, they decided to march towards their Seven Islands depot, approximately 220 km away. Hoping to reach Seven Islands in six to seven weeks, they estimated that they would be able to maintain an average speed of about 5 km per day with some help from the ice drift. Had they chosen Seven Islands as their initial goal when they first set out, and had they loaded their sledges more lightly, they might very well have been able to maintain such a speed, especially given the mental boost they could have experienced from the continuous reduction of the remaining distance that they had to cover.
The direction and speed of ice drift in the area varies considerably depending on the wind, the thickness of the ice and other factors. In his published scientific analysis of the results from the Fram expedition, Nansen (Reference Nansen1902) estimated a primarily westerly ice drift of ca. 2 cm/s in the area (i.e. almost 2 km/day), in line with contemporary research (e.g. Kwok, Spreen, & Pang, Reference Kwok, Spreen and Pang2013). Hence, by going with instead of against such a drift, ceteris paribus, you could realistically travel up to 4 km further each day, or over 100 km further in a month.
When Andrée first publicly discussed his plan to use a balloon for polar travel in February 1895, he mentioned Nansen’s ongoing attempt to drift across the polar sea caught in the ice aboard the Fram (Andrée, Strindberg, & Frænkel, Reference Andrée, Strindberg and Frænkel1978). On the official map showing Nansen’s projected drift with the Fram (Bergslien, Reference Bergslien1893), the conjectured southwesterly direction of the ice drift in the region northeast of Svalbard appears explicit. Indeed, in 1896, the Fram actually paid Andrée a visit at Danes Island during her voyage home after being released from the ice just north of Svalbard, confirming the ice drift hypothesis. Moreover, Andrée met with Nansen several times, the last being in April 1897 (Huntford, Reference Huntford1997). The two polar explorers were probably discussing the ice drift, among other things, especially as Nansen’s book Farthest North (Nansen, Reference Nansen1897) had just been published; the entire first chapter is based on ice drift analyses. It is rather obvious, therefore, that Andrée would have been aware of the westerly or southwesterly prevailing ice drift before the balloon departed from Danes Island. It is likely that Strindberg and Frænkel were, as well. If they were not, they became acutely aware of it when their recorded observations showed that they had drifted as much as 30 km in just five days after landing on the ice, before starting their march.
The diaries give no indication why Andrée, Strindberg and Frænkel chose to go towards Cape Flora, but several reasons have been suggested, including the larger depot, the scientific value of the area, and the fact that Nansen had shown that it was possible to survive a winter in the Franz Josef Land archipelago (Andrée et al., Reference Andrée, Strindberg and Frænkel1978). Other possible explanations may include knowledge about huts from previous expeditions, ignorance about the strength and/or general direction of the ice drift, hubris (they might have believed that they could overcome it) or even a prolonged suicide (Andrée had already written explicitly in his diary during the balloon flight that they thought death was acceptable after their achievement to that point). However, the fact that they changed their course at least four times during their march may indicate that they did not understand the nature of the ice drift the way they should have.
Moreover, the Svalbard archipelago was better mapped than the Franz Josef Land archipelago at the time, and there was more human activity there, at least in the summer. Hence, in terms of optimising their chances of survival, all circumstances known at the time considered, Andrée, Strindberg and Frænkel should have aimed directly for Seven Islands. They did not, and we will never fully understand why.
The straight-line distance from the initial crash site to the Seven Islands depot was about 300 km. The average direction of the ice drift, as estimated by Bergslien (Reference Bergslien1893), is almost exactly along that line. Assuming, for example, a reasonable average speed over the ice of 3 km per day and an average ice drift in the right direction of 2 km per day suggests that they could have reached their Seven Islands depot during the latter half of September. Hence, the expedition members might have survived if they had made better decisions regarding where to go and what to bring after the balloon journey.
Finally, it should be noted that reaching the Seven Islands depot would not in itself have ensured the survival of the expedition. Human presence in northern Svalbard in September could not be expected, so they would probably have had to risk a boat journey of some 130 km to reach Nordenskiöld’s 1872–73 huts at Mossel Bay, or build some kind of shelter at the depot to spend the winter in. Of course, neither option seems tempting, but not reaching land at all would almost certainly be fatal.
Conclusion
The analysis in this paper shows that the members of the 1897 Andrée balloon expedition would have had a fair chance of survival after their balloon landing at 82°56′N 29°52′E if they had made two crucial changes in their plan of action. First, they should have laden their sledges more wisely, restraining themselves to weights each man could pull without having to relay. Second, they should have immediately set a course towards their depot at Seven Islands to benefit from the general ice drift. Their choice, for whatever reason, to march towards Cape Flora with sledges that required relaying left them with basically no chance of survival.
Acknowledgement
The author would like to thank an anonymous reviewer who provided helpful comments on an earlier version of the manuscript.
