Editor’s Note: This article contains quotations from contemporaneous accounts which might be offensive for today’s readers.

The moon was new on the night of 31 July 1761, and the wide expanse of the Indian Ocean uniformly black. But Captain Jean de Lafargue of the French cargo ship L’Utile foresaw no danger. True, there were some minor discrepancies among the maps available on board, but this part of the sea was clearly empty, other than the large numbers of birds the crew had seen flocking about all day, and he had no desire to reduce speed. He had already been at sea for more than a week; it could take more than three to sail from Madagascar to the Mascarene Islands, where the governor awaited his return with food and supplies. Threatened by the English fleet, the colonies there needed all he could bring.

Lafargue had his own additional reasons for pressing on. An enterprising fellow, he had taken the opportunity while in Madagascar to obtain some goods he planned to sell on his own account, which was why L’Utile’s route was a bit of a detour from its obvious course back to the Mascarenes. Since Lafargue wasn’t supposed to be engaging in personal merchandising on this trip, he was making for the island of Rodrigues, where he could trade his cargo before heading on to his official destination of Île de France (known today as Mauritius). Rodrigues was almost 600 kilometres further east, but given the uncertain timeframes involved with sailing vessels, no one should be the wiser about his little side-trip⁠—so long as he didn’t fall too far behind schedule.

It was just past 10:20 that night and Lafargue was in his cabin when two sharp bumps shook L’Utile. They were not very violent, but every sailor on board knew what they meant: L’Utile had just struck ground. They hardly had time to recover before a third, much harder jolt convulsed the vessel. Desperately, unable to see the obstacle, the crew tried to veer about, but all this did was expose the flanks of the ship to the current. Huge swells caught at it, and breakers crashed down on its deck. The waves that had seemed so calm all day had L’Utile in their grip, hammering it against the unknown shoal on which they had foundered.

In the hold, Lafargue’s 160 newly purchased slaves were screaming.

A French fluyt of similar design to L'Utile
A French fluyt of similar design to L'Utile

When Dutch merchants chose to establish settlements in the Mascarene Islands in the 17th century, particularly Mauritius, they faced the uncommon colonial dilemma of finding no local population to enslave. There were plenty of dodos, but no humans. The dodo being unsuited to hard labour, the Dutch set about driving the birds to extinction and importing the forced labour without which no self-respecting colony could survive.

The Indian Ocean slave trade is less well-known than its Atlantic counterpart, but it flourished for at least as long and formed its own complex ecosystem. The Mascarene archipelago⁠—whose main islands go by the names of Mauritius, La Réunion, and Rodrigues today⁠—was ideally placed to serve as a stopping-off place for ships plying the trade between Africa, India, and the East Indies, though it did not immediately become a major commercial hub. In fact, the Dutch abandoned Mauritius in 1710, before the French took it over in 1721 and renamed it Île de France. The slave trade soon proved integral to the commercial networks of the Indian Ocean: not only could African slaves be sent on to be sold in India, slave labour was also needed on the Mascarenes themselves, which had never before been settled by humans and therefore needed a lot of work to be made livable. Not all of these slaves were African: several thousand Indian slaves were brought to the Mascarenes as well, along with various goods of the same origin. Slave traders particularly prized Indian textiles, which they used as currency to buy East African slaves in the Mascarenes, then traded on for West African slaves for export to the Americas. Globalisation is nothing new.

Despite the presence of this labour force, the French East India Company, which governed the Mascarenes, remained dependent on regular imports of food. For commercial reasons, the colonial planters on the islands favored coffee and sugar plantations over basic foodstuffs, with the result that by 1760, despite all the imports from other French colonies, Île de France was spending an estimated 2 million livres a year (the price of approximately 32,000 slaves) for food and other supplies from the Dutch. Moreover, the labour force itself required regular top-ups.

In the 19th century, descendants of slave owners would try to insist that slavery as practiced in the Mascarenes had been far kinder than the industrial and inhuman practices of North America, and even that it represented a beneficent, ‘civilising’ process for the African slaves. This was, of course, untrue. Slave ships in the Indian Ocean may have carried fewer slaves per ton, suggesting that conditions were less cramped than on the slavers of the Atlantic, but this was because slaves formed only part of the cargo, most of the space being needed for the rice, beef, and other foodstuffs on which the Mascarenes depended. Given that dependence, it is unsurprising that slaves found themselves at particular risk of malnutrition once they reached the islands, quite aside from having to deal with beriberi, dysentery, enteric fevers, intestinal parasites, pulmonary infections, typhoid, and epidemics of smallpox, cholera, and influenza⁠—assuming they even survived the journey. Traders took into account an average loss of 12 percent among slaves brought from Madagascar to Île de France, higher if they came from further away: sometimes half or more were lost to mortality. This was in addition to fairly regular revolts: one captain’s report from 1778 briskly notes that 80 slaves had thrown themselves overboard while they were at sea, but that all of them had been recovered save for the 15 eaten by sharks. The potential profits made the risks worthwhile: with the high attrition rate among slaves in the Mascarenes and the unending need for labour, even unhealthy slaves could be sold for twice their purchase price. For an enterprising ship’s captain, it was quite irresistible.

Malagasy slaves, from the ‘Big Island’ of Madagascar, were especially valuable commodities. Unlike the ‘Kaffirs’ of mainland Africa, who were regarded as only good for manual labour, the Malagasy, according to one contemporaneous French human resources expert,

“…are less suited to hard labour, but they can be made into very good workers for all the professions … They have great intelligence and much aptitude for imitation. They learn French easily. But next to these advantages must be set that the Malagasy is restless, frivolous, vindictive, shows himself inconsistent, subject to extreme passions.”

The Malagasy came with a further disadvantage: for some reason, they insisted on running away. Ninety percent of runaway slaves in the Mascarenes were of Malagasy origin. Captains sent to supply new slaves were specifically enjoined to seek out youth, ‘preferring those of 12 to 18 years to all others’, in the hopes of preventing this ‘invincible inclination’ they apparently developed at a certain age. But with Madagascar only four or five days away by sail (as opposed to the several months separating Africa from America), the slaves’ hope of returning home could never be extinguished. Some even succeeded, while others camped out in small communities of escapees in the bush. Such truant tendencies, of course, were all to the traders’ advantage, as it kept the demand for new slaves high.

The Battle of Quiberon Bay, Seven Years' War (1760)
The Battle of Quiberon Bay, Seven Years' War (1760)

When Captain Jean de Lafargue first met the governor of Île de France in April 1761, the need for workers was not the governor’s only concern⁠—he was also wondering how many people he would be able to feed if the Mascarenes suddenly found themselves blockaded by the British fleet. The globe was engulfed in the Seven Years’ War, a conflict that Winston Churchill, who knew a thing or two about global-level hostilities, called ‘the first world war in history’. Sparked in the colonies of North America, rapidly setting Europe alight, the war spanned continents as the superpowers of Europe clashed in a confused and chaotic free-for-all. From the plains of Europe to the shores of Brazil, from the coasts of Sénégal to the valleys of French Canada, from Cuba to the Philippines, from Bengal to Lithuania, the war flared and flamed.

The conflict was, in the words of historian Jonathan R. Dull, ‘needless, horribly bloody, and expensive’. It only ended because all the combatants essentially ran out of money. The war stripped France of its status as a great power for a generation, leaving Britain undisputed as the world’s leading colonial empire⁠—thanks, above all, to the Royal Navy. Sea power was essential in the fight over colonies, and the French navy was consistently outnumbered by the British fleet⁠—sometimes by a factor of 4 to 1. The French navy could therefore not protect everything⁠—which is why, in February 1761, while Jean de Lafargue was en route from France to Île de France, the French navy suddenly pulled its ships out of the Mascarenes and essentially told the local corporate masters⁠—the French East India Company⁠—to look after themselves.

No one, then or since, has ever doubted that the British would have taken the Mascarenes if they’d tried, but fortunately for the French, events in Europe meant that English troops were needed elsewhere. Still, even without land forces, the Royal Navy could make itself a royal pain to islands so dependent on shipments. Nor did it help that in January 1761, the Company’s final outposts in India had surrendered to the British, severing one of the Mascarenes’ lifelines.

Unsurprisingly, then, when L’Utile arrived at Île de France in April, following a five-month journey, its crew could not expect an especially long leave. By June, they were put back to work. The governor gave Lafargue his new orders: go to Foulpointe, on Madagascar’s east coast, and bring back food. Oh, and don’t bring any slaves.

From the governor’s point of view, the proscription made perfect sense. The one advantage of having been abandoned by the navy was that its crews no longer stopped by, insisting on being fed, but even with that reduction in overall appetite, Île de France still needed victuals more than extra mouths to feed. Lafargue, though, had no intention of paying attention to the restriction: L’Utile was his first command, and one of the benefits of being captain of a Company ship was the possibility of engaging in trade on your own account. Indeed, it was an official perk: the Company had suffered so many losses from pilfering captains that it had eventually thrown up its hands and given them the right to merchandise for themselves, in the hopes that they would stick to their permitted limits and leave the Company’s goods alone. And Foulpointe was Madagascar’s main slave trading port.

Traditional Malagasy architecture
Traditional Malagasy architecture

L’Utile departed on 27 June. Three weeks later, on 22 July, it set sail for its return journey. In between, Lafargue had not only filled up the hold with flour, meat, wine, and other necessities, he had also negotiated the purchase of 158 Malagasy men and women, who cost him 10,000 livres. This was something over his yearly salary, but he could expect to sell them in the Mascarenes for twice that⁠—and the buyers would consider it a good deal as long as a slave lived more than three months. The slaves were shoved into the hold and walled up in a compartment separating them from the foodstuffs. The only mitigation in their situation was that L’Utile was not a specialised slave ship, and so they were not chained.

Lafargue set L’Utile on a heading further north than necessary for a return to Île de France. There were limits to how blatant his disobedience to the governor could be, so his planned port of call was Rodrigues, the third major island of the Mascarenes. There, he could offload his human cargo in peace, make his profit, and allow blind eyes to be turned. It was a perfectly reasonable, obstacle-free plan. Geography, however, sometimes laughs at human plans.

Lost in the Indian Ocean is a speck of sand and coral known today as Tromelin Island, an infinitesimal volcanic outcrop set so low along the waves that it is practically invisible to anyone not actively looking for it. Not only is it almost impossible to find, but once it is found, it remains almost impossible to visit: even modern powered boats tend to struggle in the heavy breakers. Barely 1,700 metres at its longest point and 700 at its widest, the island evaded discovery by European sailors possibly until 1722, when the French ship Diane recorded the position of this almond-shaped ‘Islet of Sand’, which sits as far from sailing routes as it does from land. Sparsely vegetated, a laying ground for sea turtles and a breeding ground for a variety of birds, it offers nothing to humans, not even an easy spot to come ashore.

Nevertheless, cartographers eventually took note of the Diane’s report, and the Isle of Sand made it onto a map in 1739, as much for the sake of completeness as for the danger it represented. But cartography, particularly marine cartography, was an imprecise art in the early 18th century. The 1739 map, furnished by the French navy, did not place the Isle of Sand in its correct location. In fact, no map did. The 1739 cartographers had the island about 48 kilometres too far south and something like 108 kilometres too far east. Subsequent cartographers improved upon the estimation, but even the best maps were off by 40 or so kilometres. A French map from 1757 simply mentioned the island’s existence without trying to posit its location. Two of these maps were onboard L’Utile in 1761: the original 1739 map, and another from 1753 that happened to be drawn by one of Captain Lafargue’s former commanding officers.

Three superimposed contemporaneous maps of the region: one English, one Dutch, and one French (gallica.bnf.fr / Bibliothèque nationale de France)
Three superimposed contemporaneous maps of the region: one English, one Dutch, and one French (gallica.bnf.fr / Bibliothèque nationale de France)

For a week, the ship ran well, despite unfavorable winds. Every day or so, the crew likely let the captives out on deck, perhaps in small groups, to keep them from withering, and threw overboard the corpses of any who had died. What food the captives were given was probably a small portion of rice. When their brief time in the light was over⁠—carefully guarded to ensure they did not jump over the edge⁠—they were returned to the cramped and fetid darkness of the hold, and the planks imprisoning them nailed back in.

Eight days into the journey, on 30 July, officers opted to change course for the night, hoping to avoid a close encounter with the inscrutable Isle of Sand, which according to their midday calculation of L’Utile’s position lay directly to their east. In fact, they had miscalculated: they were actually about 60 kilometres further north than the island. Overnight, they travelled far south: after re-establising their course at dawn, by noon on 31 July they had reached a position that Lafargue’s older map indicated was 11 kilometres north of the island, and the pilot’s newer map suggested was 35 kilometres south. Though the latter map suggested that a northerly heading was risky, the island was not actually what concerned the pilot: by his calculation, they were headed directly toward the “Nazareth Sandbars” (which mariners would eventually realize don’t exist, another example of the inaccuracy endemic to early cartography). Lafargue dismissed the pilot’s caution, reproached him for using the wrong map, and called him an ignoramus. They continued on their way.

Lafargue remained oblivious to the large numbers of birds that began flocking about, even as evening drew in. As an elementary precaution when faced with two contradictory maps, more experienced captains would have been careful to stop for the night, especially when the sky was filled with flocks of birds suggesting land must be near. Lafargue saw no need. Moonless night or not, his map clearly showed that the Isle of Sand lay south of them; their northeasterly course would obviously lead them safely past it.

Just after 10:20 that night, the ship’s clerk, 35-year-old Hilarion Dubuisson de Keraudic, checked the ship’s heading as he made his way to bed. The compass showed them aiming due east. Keraudic had just set foot on the lowest step leading to the lower deck when the first two jolts shook L’Utile. He rushed back to the deck to see what was happening. What he found was pandemonium.

In the first moments of the shipwreck, the crew stood about dumbfounded with shock, the watch officers shouting to know what was going on as the first lieutenant, Castellan du Vernet, rushed from his cabin to the forecastle. The crew tried to turn the ship around, but this only caught it more firmly on the coral reef they had struck, the starboard side open to the sea.

In the hold, the imprisoned Malagasy had no way of knowing just what was going on. All they knew were the shocks rattling the vessel and sending them knocking into one another as it swayed and pitched about. But they could hear the waves, the creak and crash of breaking wood, and, coming from above them, men yelling and screaming in an unknown tongue.

As the ship lurched and rolled, the crew had no idea what they had run into, and with no moon they could not tell whether there was any land nearby or whether they had foundered on an isolated shoal. After fierce debate, the officers decided to dismast the ship. The crew cut the masts down, throwing the main mast and mizzenmast over to starboard, and the foremast to port, but it did little good: the breakers continued to hammer the vessel, which was beginning to tilt dangerously to starboard, towards the open sea. The keel hit ground again and again, once so violently that the tiller smashed into the sterncastle’s upper deck.

With the tiller gone, the breakers had control of the helm, the waves rising above the ship’s highest point⁠—and still it was tilting to starboard. At about 2:00 in the morning, First Lieutenant Castellan ordered the crew to throw their heaviest cannons overboard. But even as they did so, the ship, stressed to the breaking point, was starting to come apart: the planks of the deck were starting to split and pull away from one another, and water was beginning to seep in.

For four and a half hours, the ship held together, but around 3:00 in the morning, the deck simply collapsed. The boats went tumbling into the hold, probably smashing into the Malagasy⁠—and L’Utile split. “The front of the ship separated from the back,” Keraudic remembered, “the sides detached themselves and the bottom left the upper part.” Only the fact they were not chained saved some of the Malagasy. Still lost in the dark, crew and captives grabbed hold of whatever they could, and waited for death.

“…we waited and expected we had reached our final moments. Every second we suffered a thousand deaths as we could barely breathe, so intensely were those furious breakers repeated. Thus we were until daybreak, a long and awful time!”

Three and a half hours later, in the morning twilight, cries of joy suddenly erupted: land had been sighted. There were even people walking on it! But all too soon it became apparent that they were only other survivors from L’Utile, who had been carried to the island by the still-raging waves⁠—and as the light grew stronger, the barrenness and smallness of the Isle of Sand became only too apparent.

Between the ship and the island were the corals on which L’Utile was caught, and the waves showed no sign of calming down. “Several people jumped into the sea with a rope to try and reach the land and establish a hauling line,” Keraudic recounted,

“but it was useless. A few made it to land without being able to carry anything. The others had to be hauled onto the wreckage, else they would have drowned; and finally we were terrified by how the back of the ship on the side we were on was opening and closing at every moment, and cut several people in half.”

Captain Lafargue was nowhere to be seen. Knowing himself to be a good swimmer, Keraudic became the first officer to plunge into the sea in the hopes of reaching land, but he mistimed his effort and a retreating wave dragged him into the sea rather than bringing him to land. Thrown about by the breakers, he spotted a large pine plank he could hold onto:

“…a black slave who was drowning also wanted to grab it but I gave him two kicks that finished him off. At that moment I heard a voice calling to me for help, I rolled over and saw a sailor all covered in blood who was swimming very weakly towards me. I got ahead of him, he took place on one end of my plank and we tried to reach land…”

Keraudic eventually left the plank to the exhausted (white) sailor and made it to shore, bloody and battered. Meanwhile, the ship’s stern had come about and was pointing its tip towards the land, and there was now something of a bridge of wreckage that could ease the passage to the island⁠—so long as everyone could reach the stern.

Eventually Captain Lafargue was spotted⁠—trapped inside the port side latrine, the windows of which had to be broken to let him out. The crew set about establishing hauling lines between the various bits of wreckage to provide a path to the stern. Not everyone made it: First Lieutenant Castellan du Vernet had to watch as the sea overturned the bit of wreckage his younger brother was gripping onto. Like many sailors, the elder du Vernet could not swim, and he had to be restrained to prevent him from jumping in to try and save his brother as the latter was borne away by the furious waves.

One by one, the survivors of the wreck made it to the stern. But from there, the only way to reach the island was to cross a channel overwashed by the sea, carefully timing each attempt with the moment the water had retreated and hoping to get across before the wave returned. Caught by the wave, there was no choice but to grab at the coral, which cut and sliced at skin and flesh. Castellan du Vernet would have been washed away had the ship’s surgeon not come to his aid; one of the second lieutenants was likewise saved, by two Malagasy who pulled him from the wave which had hold of him.

Despite all this, most of the crew survived. By mid-morning, everyone had made it to shore. Only about 20 white men died, along with ‘a black tanner belonging to the Company’. The Malagasy were not so fortunate. With the hatches nailed shut to prevent their escape, it was not until the ship split that they were able to try and escape through the wreckage. At least half of them had died.

Tromelin Island as seen from Google Earth
Tromelin Island as seen from Google Earth

On the shore, 123 white Frenchmen and 88 Black Malagasy men and women faced each other. All of them bloody, battered, half-drowned, exhausted; all of them castaways. All of them uncertain⁠—not least in wondering how they now stood to each other. Were there still masters? Were there still slaves?

Tromelin Island has the particularity of being as much a nightmare for claustrophobes as it is for agoraphobes. Its entire coastline is less than 4 kilometres; its highest point is barely 7 metres above sea level. Those who have been there describe the experience as more like being on the deck of a ship than on land⁠—a ship that is going nowhere. At the northern end is a spur of shifting sands whose shape is altered by heavy seas, a constant reminder that the island offers no great refuge from the waves. Its only vegetation is sparse and kept small by the wind; and there is no shade, no more escape from the sun than there is from the sound of the sea or the ceaseless wind.

But though the deserted island lacked only a palm tree to make it the prototypical desert island, the castaways’ situation was not entirely desperate. L’Utile had foundered, but not sunk: it was simply impaled on the reef and in several pieces. Its cargo was not yet lost. The able-bodied began to make the risky trip back to the wreckage to try and salvage what they could, while others scoured the coastline for whatever had been washed up, and others still began digging a well. The next day, the salvage operation continued, and in his capacity as ship’s clerk, Keraudic catalogued what had been saved: red wine, brandy, ‘rather bad cider’, liquor, flour, beef, and fat, each measured in hogsheads, barrels, bottles, or pieces. Some of the sails had been saved as well: the officers used the main studding sail to erect a large tent for themselves and the supplies, while smaller tents were made for the crew. They even raised the flag over the main tent.

But discipline was beginning to break down: the crew stole five hams while they could, and “save for 15 or 17 men who worked zealously,” Keraudic reported, they all took after the master, who was “an exemplar of laziness, impatience and arrogance etc.” With nothing but salted meat to eat and alcoholic beverages to drink, thirst was becoming an issue. The first attempt at a well had failed to find any water. Keraudic recalled, “we were beginning to feel the need for water very strongly, many blacks were dying [since we were] giving them nothing.” At least until drinkable water was found, the officers were in no doubt that the supplies should be reserved for the whites.

At noon on the third day, the officers promulgated an ordinance stating that any theft of food or drink would be punished by death. At that moment, two sailors were caught red-handed attempting to pilfer. Some muskets had been saved from the wreck, and the two threw dice to decide which of them would be shot. “One of them was going to be executed when he took to his heels and jumped into the waves of the sea,” Keraudic recalled. The other was waiting for execution by musket when, at about 5:00 in the evening, the Master Gunner arrived from a new well site with a bottle: a little more than 4.5 metres down, they had hit some brackish, vaguely drinkable water. Citing the grace God had just shown them by providing them with water, he begged for the condemned man’s life, an appeal granted at once. Then, led by the ship’s almoner, the French all processed to the well singing the Te Deum.

Behind them crawled the Malagasy. Once the French had had their fill, the Malagasy were at last allowed to drink, but for some of them it was too late. “Some blacks who were very weakened from thirst died on the way [to the well]”, Keraudic remembered, “some being too debilitated to reach it, others from drinking too much.” Eight of the Malagasy died from those three days of being denied drink.

Masked boobies frequent the shores of Tromelin Island
Masked boobies frequent the shores of Tromelin Island

With the threat of immediate death allayed, thoughts could turn to the longer term. The Isle of Sand was not entirely bereft of nourishment fit for humans: the salvaged supplies could be supplemented by both the birds’ eggs and the birds themselves, who like their vanished cousins the dodos had yet to learn to fear humankind and could be quickly killed with sticks⁠—but no one thought this was an ideal solution. Seabird roasted in ashes with a light brandy glaze was nourishing, but hardly anyone’s idea of a staple. Nor were the salad options hugely palatable: some of the island’s sparse vegetation was comestible, but none of it was particularly tasty.

The officers would have known that rescue was unlikely: the Isle of Sand lay far away from any sea routes. Even if anyone were to look for them, no one would think to search in this area, since L’Utile would be assumed to have taken the direct route to Île de France. There was no leadership to be expected of Captain Lafargue: he had been brought to land ‘in a bad way,’ Keraudic says, likely injured, but also undoubtedly suffering the psychological shock of having lost his ship, his reputation, his career, and his fortune in a single night⁠—all due to his own undoubted mistake. Fortunately, his second-in-command was of a different mettle.

First Lieutenant Barthélémy Castellan du Vernet was one of at least the third generation of his family to serve the French king as soldiers and sailors: the young brother he watched drown days earlier was the third sibling he had lost to duty. He had been sailing on Company ships for 20 years, and had served in battle more than once. He knew the East Indies well⁠—and he would have known that the low-lying Isle of Sand would be almost entirely submerged during the cyclones endemic to the southwest of the Indian Ocean. There were still three months or so until the cyclone season started, but it would be best to be gone before then.

The good news, he realised, was that just as they were not bereft of food, the castaways had building materials on hand. L’Utile means ‘useful’, and fragmented though it was, the wrecked ship was about to earn its name. The idea of using its remains to build some sort of makeshift boat probably occurred to many, but Castellan was the one who seized on it. As soon as the water situation was resolved, he began to draw the plans for a flat-bottomed ship that might allow them to return to Madagascar. But he would not only be the ship’s designer: L’Utile’s carpenter was a useless incompetent who “didn’t even know how to line up one piece of wood with others,” in Keraudic’s estimation. In the words of a later account, Castellan would have to be “all at once Commander, Engineer, Carpenter, [and] Sawyer,” as well as head cheerleader.

Not all the crew members were enthusiastic workers. Nor did the weather help: on some days the sea was too heavy for anyone to go to the wreck for the necessary salvage. In the afternoon on 9 August, it suddenly seemed as though Castellan’s efforts might not be needed at all: a two-master ship was spotted on the horizon. Yelling, the castaways waved flags, and set light to two barrels of gunpowder⁠—but the ship failed to notice them, and veered to make its way towards India.

Castellan refused to become downcast. He led by example, working ceaselessly from daybreak to nightfall and driving himself to exhaustion in the process. The workers had to wade out to the wreckage, weather permitting, start taking L’Utile apart, and bring the remnants ashore: ropes, sails, planks, nails⁠—to say nothing of hauling anchors and eventually cannonballs. Castellan led the way, chivvying and encouraging the others. With a rescued sheepskin for bellows, he even managed to build a makeshift forge that would allow them to do the necessary ironwork. Castellan and the ship’s surgeon worked “not like two men, but like thirty,” according to Keraudic, who along with another officer assisted Castellan as they could in leading the workforce. Under their impetus, the boat began to take shape⁠—though Castellan did have to diminish its projected size by about one third.

As Castellan drove his crew, pushing himself hard enough that he sometimes passed out from the strain, the Frenchmen found that they were not alone in their toil. The Malagasy were at their side.

“It is impossible to express the help we drew from the first moment to the last from these unhappy slaves,” Keraudic remembered. “…they did all one should have expected from a crew of good will and trying to save its life.” Keraudic may have kicked a Malagasy down in a scramble for a plank when both were trying to save themselves from drowning, but in retrospect at least he could be sentimental about them.

The Malagasy had set up camp some way away from the French, insofar as that was possible on the cramped Isle of Sand. Communication between the two groups was quite plausible: given how many slaves in the Mascarenes came from Madagascar, the Malagasy language was the lingua franca of the slave population there, and a number of whites were at least somewhat conversant in the tongue.

Perhaps the French, with their muskets and superior numbers, compelled the Malagasy to assist them. Perhaps the Malagasy simply and wordlessly began to work besides their captors, recognising Castellan’s boat as their only hope for escape. Perhaps there were negotiations, but if so we do not know whether the French asked for the help of the Malagasy or whether the latter offered to assist of their own accord. Nor do we know whether any promises were made. And we do not know just when the Malagasy began to realise that a ten-metre boat would never be able to hold all 200 of those cast away, or when they started to suspect who would and would not be allowed to board.

Green sea turtle
Green sea turtle

The castaways’ diet received some pleasant variety on 11 September, when they caught a 225-kilogram sea turtle, as well as a few days later, when they used firebricks retrieved from the ship to build an oven and began to take advantage of all the flour they had saved from the wreck by making bread. On 17 September, the workers hauled up one of L’Utile’s jettisoned cannons. If another ship came into view, they would be able to make a more distinct signal than their barrels of gunpowder had.

Three weeks into construction, the boat was advanced enough that the crew began to bake biscuits for their journey, but there was still much to do. They built a three-man catamaran that could go beyond the reef so they could catch fish. A week later, they pulled down the main tent, the one for the officers and supplies, needing its flagpole as a mast and its canvas for their sails. Worryingly, as they began to caulk the boat, there were an increasing number of days with heavy seas: at one stage, part of the sandy point at the north end of the island was entirely washed away. Nevertheless, the crew continued to work, hauling cannonballs out of the sea, presumably for use as ballast in their new vessel, and building another large catamaran that could ferry them from shore to board the boat once it had been launched.

On 26 September, the French sailors’ salvation was completed. The almoner blessed the boat, and named it La Providence. They hauled the catamaran to the open sea and anchored it there. Then they fixed the kedge anchor they had rescued from L’Utile in the coral, so that La Providence could be anchored when they hoisted sail the next day.

On 27 September, with a great deal of effort, they managed to launch the ship. With little more than what remained of the clothes on their back and the food they could bring, the 122 Frenchmen made their way onboard. At 5:00 in the evening, ‘lined up like sardines’, they set sail. On shore, the Malagasy watched them go, in utter and damning silence.

Castellan du Vernet had left three things with the Malagasy: three (3) months of supplies, including the tents that had not been taken down; one (1) letter explaining the situation that they could show to any captain who happened to pass by, attesting to how well they had worked; and one (1) promise that he would return to rescue them.

There were about 80 Malagasy left on the island. Freed from the presence of the French, but with no dream of rescue other than the tenuous hope of Castellan’s word, they had to determine how to live on the barren shoal on their own⁠—and how to live together. They had been thrown together when Lafargue purchased them nine weeks earlier, but many were still strangers to each other.

Along with the supplies they had been left, there was the oven and the forge, whose fire was still lit. Nor had all of L’Utile been cannibalised: it could still provide wood for the fire, along with iron and other bits of salvage. The birds remained numerous, and the sea turtles would soon start their laying season, while the plants had their healing properties. It was possible to survive. But they were determined to do more than merely endure.

Tromelin Island lies in the heart of the cyclone corridor of the southwest Indian Ocean, and from late October until May it is at risk from these gargantuan storms. In 1986, for example, it faced two of them in just over a month, with winds reaching 230 km/hr and gusts up to 287 km/hr. Their impact on the defenceless island and its wildlife is profound, devastating the nests of birds and turtles, sometimes killing off an entire generation in the process. The waters rise to cover all the low-lying parts of the island, strewing it with shells torn from the bottom of the sea.

Map of the island drawn based on the French survivors’ account, or possibly by one of them. Note La Providence below the island, to the left of the letter I (gallica.bnf.fr / Bibliothèque nationale de France)
Map of the island drawn based on the French survivors’ account, or possibly by one of them. Note La Providence below the island, to the left of the letter I (gallica.bnf.fr / Bibliothèque nationale de France)

In their tents made of sails, the Malagasy were as vulnerable as the eggs. We do not know the strength of the first tempest to strike the Isle of Sand while they were there, nor how long it had been since the French had left when it did, but there is no question that the Malagasy endured one which ripped their tents away and probably took some of them with it. When it was over, they knew they needed shelter. If they still hoped Castellan would return, they must have known that it would not be during the storm season.

The physical and psychological traumas of the previous few months were not all the Malagasy had to endure. They had also suffered a significant spiritual scission. Malagasy spirituality is particularly steeped in the relationship between the living and the dead. ‘The dead are not dead,’ says the proverb; in the ceremony of Famadihana, ‘the turning of the bones’, families remove the bodies of their ancestors from their tombs to rewrap them in new silk shrouds on which they have written their names, before returning them to their graves. Even the way a Malagasy house is built hinges on their beliefs. A traditional Malagasy house is not merely a home but a temple: it is built along a north-south axis, with the front door facing west and another symmetrical opening to the east. The latter is the ancestors’ door, through which no living person ever passes. The west-east division even affects how the house is occupied, with women and children living in the western half and men in the east, and the elder members of the household sitting closest to the ancestors’ door.

The bones of the stranded Malagasy’s ancestors lay far to the west, and the practicalities of ensuring shelter precluded adherence to the architectural rules of orientation. Yet shelter they must⁠—even though the island offered no wood or mud with which to build. There was beach rock, and there was coral. The Malagasy knew how to build in stone⁠—but such materials were reserved for the raising of tombs.

With the French gone, the Malagasy moved from their camp down on the beach and took up residence on the highest point of the island, close to where La Providence had been built. This was the one spot on the island that offered some relief from the prevailing winds, the elevation offering one leeward slope. It was here that they began to build. Some of the materials available were only to be found at the very end of the island, a kilometre away, but they dragged slabs of beachrock and hauled corals to the high point. They built their walls of interlocking blocks, lining the base of the interiors with slabs of beachrock on top of which they interlocked layers of coral, in a manner recalling the exterior facings of Malagasy tombs. In places, they used large nails taken from the wreck to use as suspension points. They built high, but especially they built thick: the walls surrounding the small, cramped interiors were a good 1.5 metres thick, sometimes more.

It was not, at first, enough. Several storms shook the island, and eventually some of the buildings were damaged. The Malagasy returned to work, reusing the stones from the damaged buildings and reorganising the general plan of their settlement. This time, they arranged their shelters around a central open space, leaning the walls of each against one another. All the entrances faced away from the wind, so that those to the east opened to a village square of sorts (which wasn’t at all square) while the buildings to the west faced the open sea.

All in all, the Malagasy constructed about a dozen buildings. The interiors were tiny, either due to the paucity of roofing material or simply to maximise the strength of the walls; much of life must therefore have taken place outdoors. The insides may have served more as shelters than as homes. One of the buildings was clearly a kitchen, with well-ordered implements and an inside fire, but there was also an outside hearth which might have served for the cooking when the weather was fine.

The weather was not done with them, though. Another cyclone ravaged the settlement, forcing the stranded to rework their village, dividing their ‘village square’ with a new, massive wall, 9 metres long and 3 metres thick, that required the destruction of one of the existing buildings and changing the entrances of others to adjust to this.

Remains of the kitchen (Groupe de Recherche en Archéologie Navale)
Remains of the kitchen (Groupe de Recherche en Archéologie Navale)

From the remnants of L’Utile and from what the sea could bring them, the Malagasy fashioned the tools for their new life. Driftwood and the remaining wreckage provided the wood to keep the fire going. The Malagasy are renowned metalworkers, and this talent showed. They took hinges and turned them into axes and hammers; they made and repaired copper bowls and spoons, and other kitchen implements, even a baking plate. They took advantage of the abundance of lead to be found on a sailing ship and made bowls of it, while also turning large seashells into ladles and spoons. On some of these implements, they traced small designs. Nails were hammered into the walls to serve as hooks; other large ones seem to have been installed as anvils. Nor did the workers only create utilitarian objects: they also crafted jewellery, perhaps talismanic, perhaps merely to recall a normal life.

They drank the bitter water of the well, carrying it in lead bowls that could probably do little more damage to them than the water itself; they raided the nests of the birds and spied on the sea turtles’ laying spots to garnish their larder with eggs; they trapped and killed and feasted on the birds and turtles themselves; at times they fished beyond the reef. As their clothing wore away, they began to tear away the wings of the birds they killed and weave their feathers into cloth. They burnt incense, magical and medicinal, that some of them must have managed to bring with them. They buried their dead. They lived.

With their capacity and willingness to adapt to difficult circumstances, their skill at working together, and their ability to break the psychological shackles of their cultural interdicts, some have seen the stranded of the Isle of Sand as an almost unique case, one where an isolated group of enslaved people, abandoned far from home and uninfluenced by the presence or threat of masters, not only adapted to their circumstances but in so doing forged a new society.

But it was a society with no future. Several women became pregnant, but the children never lived. The meat-and-egg heavy diet may have been sufficient to keep the adults alive, but it was not enough to ensure an infant’s health. And so the babies died, one after the other, and their mothers with them, and the numbers of the Malagasy dwindled.

They had been brought to a land without ancestors, and they would be ancestors to none. Their houses, built of the material of tombs, must have seemed all too appropriate. But they still refused to die.

After three years, some of them could no longer take it. There was enough wood left from L’Utile for them to make a desperate attempt. Perhaps there had been one storm too many, or perhaps they had finally given up their final shred of hope in the Frenchman’s word. They began to lash planks together, building a crude craft on which they hoped to sail home to the Big Island. Eighteen Malagasy elected to make the attempt. They set off on a raft with a sail made of feathers, and were never seen again.

•     •     •

For the French castaways who had departed aboard Castellan du Vernet’s La Providence, things worked out rather differently.

It took them four days to sail to Madagascar, losing only one man (to a ‘bloody flux’) the very evening they arrived at Foulpointe, on 1 October 1761. They had the luck to run into a squadron of three French naval ships, who quickly came to their rescue. The survivors were in a pitiable state: when offered water, they downed it so enthusiastically that the rear admiral leading the squadron doubted that what they’d dug up on the Isle of Sand water had actually been drinkable.

Map of the Mascarene Islands, ca. 1780
Map of the Mascarene Islands, ca. 1780

Arrangements were made to return the castaways to the Mascarenes. The rear admiral, meanwhile, had the leading officers from L’Utile, Jean de Lafargue and Castellan du Vernet, report to him. The admiral, of course, was mostly interested in the precise location and nature of the Isle of Sand, but he also recorded a brief version of the tale of their survival, including Lafargue’s admission that Castellan ‘did not let himself be daunted by misfortune’ and that it was he who ‘saved them all from the peril they were in’.

The two officers also told the admiral of the Malagasy. Lafargue attempted to minimise what he had done, saying that there were only about 60 on the island⁠—but he, or more likely Castellan, also informed the admiral that “to console these unfortunate blacks, they gave them the hope that they would come fetch them as soon as they could.” The admiral, who had other things on his mind, was more concerned with saving the whites at hand. He gave the officers a letter to deliver toLa Silhouette, a French East India Company ship, asking her captain to come and pick up the castaways and return them to Île de France. La Silhouette took on the surviving remnants of L’Utile’s crew, and set sail on 26 October.

However much of a relief the almost four-week stay somewhere other than the Isle of Sand must have been, it proved fatal to some. Fevers are endemic on the Madagascar coast, and between Foulpointe and Île de France, eleven of the survivors died of these. At 1:30 in the afternoon of 12 November, with Île Bourbon (Réunion) in sight, Captain Jean de Lafargue became one of the deceased.

Lafargue thus escaped facing the fury of the governor of the Mascarene Islands. When the survivors landed at Île de France’s main town of Port-Louis on 25 November, the governor did not have a very warm welcome for them. He had been waiting for news of the ship for almost four months, and as early as 26 September, he had become aware of Lafargue’s fraud⁠—he was “perfectly informed that [L’Utile] embarked 300 worthless blacks it meant to dispose of on Rodrigues.” In his report to Company headquarters, while noting that Castellan was ‘an officer of unusual ability’, he executed Lafargue with a venomous ‘good riddance’: “Lafargue died during this crossing and that was the best thing he could do, for the loss of this ship is solely attributable to his obduracy and ill conduct.” And when Castellan told him of the promise he had made to the Malagasy, and asked that a ship be sent to rescue them, the governor sharply refused. For heaven’s sake, the war was still on, the question of mouths to feed was still pregnant, he had already lost one ship on that godforsaken rock, and the remaining castaways were nothing but a bunch of useless slaves whose purchase had caused the wreck in the first place and who were probably already dead anyway. The governor did not even mention the matter to the Company.

Castellan was an honourable man. He may even have been a decent one. He had not been lying when he had promised to return for the Malagasy as soon as he could: even in the few days when La Providence was at sea, he had talked to his fellow sardines about going back, so long as they could find a change of sails at Foulpointe to replace their threadbare ones. But there had been no sails available, the admiral was occupied by a world war, and now the highest-ranking official in the area had just refused him.

Just five weeks after arriving at Île de France, on 1 January 1762, Castellan had signed up as first lieutenant on a warship. His squadron’s commander heard about the lieutenant’s promise to the stranded Malagasy, and proposed sending a schooner to rescue them, but the war intervened. When Castellan eventually left the ship and embarked on a new one that would return him to France, he believed another attempt would be made⁠—but none was.

In mid-1763, a popular chapbook telling the story of the shipwreck was published in France. Clearly based on first-hand accounts, and including the full text of a testimonial to Castellan the crew had written and forced him to accept, it included a marginal note stating that a ship had been sent from Île de France to rescue the Malagasy, but as a manuscript comment in one copy rectifies, while that promise had been made, it had yet to be fulfilled. In 1768, a captain sailing from Île Bourbon to India tried to find the Isle of Sand while en route, but he failed to spot it. The story was also still remembered on Île de France at that time: the French writer Bernardin de Saint-Pierre heard about it and considered including the anecdote in his Voyage à l’Île de France. He was revolted by the governor’s refusal to send aid⁠—but the episode ended up remaining in his drafts.

One person, however, clearly never managed to wipe the matter from his mind. Eleven years after he had left the Isle of Sand, Castellan du Vernet remained tormented by the memory of the Malagasy he had abandoned there. In September 1772, he wrote to France’s Secretary of State for the Navy, and begged that a ship be sent to see whether there were any survivors. ‘Humanity compels me,’ he began. For a change, someone listened.

The world was a different place in 1772 than it was when Captain Jean de Lafargue had purchased his slaves in 1761. Among other things, the Seven Years’ War was (by definition) over, and the French East India Company had gone spectacularly bankrupt, as a result of which it was forced to cede its control of the Mascarenes to the French Crown in 1767. Moreover, Europe was now well into the great age of sentimentality. Just two weeks after writing his letter to the minister, Castellan found himself penning a second one, this one including, by request, more details about the shipwreck and a map of the island. The minister had been touched. Castellan was quite right, something ought to be done. Three years later, someone got around to it.

In late August or perhaps September 1775, the Malagasy stood on the Isle of Sand and watched the cutter La Sauterelle come towards them. As they anchored, the crew of La Sauterelle could see that there were indeed survivors on the Isle of Sand. Two men were deputised to visit them.

But as they were reaching shore, La Sauterelle’s anchor cable snapped⁠—and the sea broke the sailors’ landing craft. One of them managed, desperately, to swim back to the endangered ship, but the other, lacking courage or simply unable to swim, could only stay on the shore and watch in horror as La Sauterelle veered about and disappeared into the distance, having successfully increased the number of castaways by one.

There were thirteen Malagasy remaining to greet the French sailor⁠—three men and ten women. Among them, there was a mother; with her was her daughter, Semiavou, ‘One who is not proud’. We do not know how old the latter was when she was brought to the island⁠—infant, child, or grown⁠—but she is the only one whose name we have.

The Malagasy took the French sailor in: the first new face they had seen in fourteen years, living proof that they were not forgotten. Probably he could speak with them: La Sauterelle had been in the Indian Ocean for three years, and it seems logical that the two men sent to encounter the Malagasy would be ones who had picked up at least some of the Malagasy language in the ports of the Mascarenes. He may then have been able to understand just what it meant to the castaways when, some months after he arrived, Semiavou gave birth⁠—and, miraculously, child and mother both lived.

It is not impossible that the French sailor fathered the newborn boy; the dates can just line up, assuming that the sailor arrived in early August and went to work rather quickly, and it seems possible that his better-nourished body might have helped produce a child who escaped the fate of the others born on the island. But we do not know. Indeed, we know nothing about him except that after a year on the Isle of Sand, he had given up hope of being rescued. Twice more in that year sails appeared on the horizon; twice they turned about and left, one glance at the state of the sea convincing their crews there was nothing to be done.

Like Castellan du Vernet, and like the eighteen Malagasy who had tried to escape all those years ago, the French sailor decided to attempt to leave on his own. L’Utile was not entirely gone; there was enough wood left to construct some semblance of a vessel that might lead them to Madagascar. So he did so, helped by some of the Malagasy who decided to take the risk with him.

Seven people took place on the ramshackle raft they built. The baby’s father, whoever he was, was one of them. The last three Malagasy men and three of the women joined the French sailor. Neither Semiavou and her baby nor her mother were among them. They set sail on a raft, driven perhaps by a sail made of feathers, and were not seen again.

Despite their repeated failures, neither of the two highest officials on Île de France⁠—the governor, de Ternay, and the Intendant, Maillart⁠—had given up on the mission of mercy that led them to send La Sauterelle to the Isle of Sand. As 1776 drew towards its close, the Chevalier de Ternay was reaching the end of his time as governor of the Mascarenes. He decided to make one more attempt.

Governor and Intendant gave the job to the commander of the corvette La Dauphine. This ship’s main job was ferrying grain between Île Bourbon and Île de France, but as it was the off-season, it was currently unoccupied. Its commander, the Chevalier de Tromelin, was 25 at the time⁠—and not in the best graces of the Intendant, who ranked him among the officers whose self-important and indiscreet behaviour drove him to despair. ‘This insignificant corvette La Dauphine … torments me more than any imaginable squadron,’ Maillart moaned in a report.

But Tromelin was an experienced sailor, and had spent six months on an exploration trip to the Kerguelen Islands, halfway between Madagascar and Antarctica, meaning he was accustomed to carrying out searches in the middle of nowhere. He was also acquainted with the slave trade. He took the time to research the pitfalls that befell previous rescue attempts. As he pondered the matter in the harbour of Port-Louis, he noticed a fisherman’s pirogue, or canoe, among the various vessels. It belonged to the local hospital, making it an appropriate craft to borrow for a mission of mercy.

Tromelin set sail on 25 November, and three days later was in view of the Isle of Sand. Unlike the other ships that had attempted an approach, he had the fortune of good weather: while November is in cyclone season, it is also when the trade winds shift, providing a few days of calm. Still, he did not take the risk of beginning a rescue operation in the evening. In the light of an almost full moon, ship and island waited.

Morning came, and Tromelin assigned one of his officers, a commoner named Lepage, to take charge of reaching the shore. Lepage acted cautiously; choosing the pirogue rather than the ship’s boat, he managed to bring it over the reef without losing a single man. Over a period of several hours, the pirogue carried the castaways from the island to La Dauphine, rescuing all of them⁠—seven women and one newborn boy.

A hand-made copper bowl left behind on Tromelin Island (Groupe de Recherche en Archéologie Navale)
A hand-made copper bowl left behind on Tromelin Island (Groupe de Recherche en Archéologie Navale)

Leaving behind the fire they had kept lit for over fifteen years, bringing with them nothing but a child, the last Malagasy left the Isle of Sand. It had been 5,600 days.

Just over a fortnight later, they arrived in Port-Louis, there to be greeted by the very highest authorities on the island: du Ternay, his replacement as governor, and the Intendant, Maillart. The latter immediately called for clothes for them and had them sent to the hospital. All three administrators then began to ask questions, wanting to know just how the Malagasy had survived⁠—even though one of them was old, and Semiavou was recorded as being very tired. Thinking it would please them, the Frenchmen offered to find them places on a ship to take them back to Madagascar⁠—and were surprised at the cold and immediate refusal. The women had been captured and enslaved there once before; they knew full well they would be so again by their compatriots, and they had no intention of being subjected to that fate. They preferred to remain.

This did raise the delicate matter of their status. What position, exactly, would they have in the stratified society of Île de France? Ever since the French Crown had taken over from the East India Company, slavery had skyrocketed in the Mascarenes. By 1789, Île de France would have 50,000 inhabitants, of whom 3,000 were freemen, 5,000 were colonisers, and 42,000 were slaves. It was not a world very hospitable to Africans. But du Ternay insisted that the rescued castaways not be slaves, and Maillart and the new governor concurred. They found a legal justification for their softheartedness: Semiavou and her companions had been illegally purchased and transported by a captain flagrantly disobeying a direct order from the governor of the time. The French administrators were not emancipating the women, they agreed. They were recognising that they were free⁠—and always had been.

Jacques Maillart Du Mesle, the Intendant, was so touched by the women’s story that he offered to take Semiavou, her mother, and her child into his own house. ‘They accepted joyfully, and I find great joy in having them nursed and making them happy,’ he wrote. The very day after they arrived in Port-Louis, he had the baby baptised. This was an essential duty for a good Catholic; all African arrivals in the Mascarenes were expected to be baptised as soon as they could be brought to understand what it involved, or sometimes immediately on being disembarked if they were in particularly rough shape. Maillart found the child a free couple to stand as his godparents, and chose the name himself⁠—his own first name, Jacques, and the surname Moyse⁠—from Moses, the baby found in the bulrushes. He promised he would take care of bringing him up.

Probably on being baptised themselves, the child’s mother and grandmother found themselves receiving new names. Semiavou became Eve, while her mother was given the name of the ship that rescued them and became Dauphine.

And there they vanish. We know nothing further of their lives, nor anything of the lives of the other five women. All we know is that in 1911, a New Zealand newspaper said that a descendant of one of the rescued still lived on Mauritius.

Jacques Moyse would have been 19 when the Mascarene authorities refused to apply the French government’s decree abolishing slavery in 1794. He may have still been alive when Mauritius became an English colony in 1810; perhaps he even lived until 1835, when, myth says, a group of escaped slaves threw themselves off a cliff as they saw soldiers coming towards them, tragically unaware that the soldiers meant only to tell them that slavery was, at long last, abolished.

By then, something over 350,000 slaves from Madagascar, Eastern Africa, and India had been imported to the Mascarenes against their will.

Slowly, over the decades, the Isle of Sand became known as Tromelin Island. Its original name remained no less valid; blown about by the wind, the sand began to cover the Malagasy settlement. It was not until the 1850s that humans again set foot on the diminutive shoal, though several ships had passed by it in the interval, some of them hoping to land but finding the dangers too great. Nor were they wrong to pass it by: even though its position was now properly known and mapped, the reef’s teeth had not been blunted. Ninety-one years almost to the day after the castaways departed, on 27 November 1867, Tromelin Island threw itself a come-as-you-were party. Just one night after the new moon, again just after 10:30 at night, again due to the incompetence of a captain, the island claimed another ship⁠—though this time, the castaways all survived, and only had to endure one month and a single hurricane before being rescued. Later that century, yet another vessel foundered there, though it has never been identified.

Workers installing the weather station on Tromelin Island pose beside castaways' ruins
Workers installing the weather station on Tromelin Island pose beside castaways' ruins

Humans did eventually find a use for the island, and it now boasts a landing strip running almost its entire length that aircraft can use when the wind cooperates. In 1954 a meteorological station was set up on the island, though it has often been blown away by the very weather it monitors. While obliterating some parts of the remnants of the Malagasy settlement, the building and rebuilding of the station brought others to light, and people realised that there was history here to dig up. Still, it was not until 2006 that proper archeologists came to Tromelin Island. When they did, they found an unparalleled site that, across the centuries, at last allowed the silenced voices of the stranded Malagasy to speak. One of the archeologists, contemplating the tidily stored implements in the kitchen building, could only compare the experience to that of Pompeii⁠—a site frozen in a single, known moment, the one when the seven women made their way to the waiting pirogue that brought them deliverance.

Until the archeological digs began, the story of the stranded on the Isle of Sand was known to us only through the words of the French. Even the rescued women’s voices came mediated by the understanding of the men who questioned them and paraphrased what they had said without ever quoting them directly. As the archeologists realised the richness of what they were uncovering, they grew determined that this silence should be remedied. As early as their second expedition, they brought with them a graphic novelist, both to document their digs and to imagine what the castaways endured.

Since then, novelists and children’s authors have also turned their sights on the story, helping to channel the Malagasys’ experience and bring it to the world. Tromelin Island may still lie barely breaching the waves, waiting for the rising oceans to wash it away from view, but the lives of those who survived its rigours will not disappear again so soon.