By Emilia Wellesley · Published May 6, 2026 · Updated May 6, 2026
What Is the Treasure of Lima?
The Treasure of Lima is a legendary buried hoard said to have been removed from the Peruvian capital in 1820, entrusted to the British merchant captain William Thompson aboard the brig Mary Dear, and concealed by his mutinous crew on Cocos Island, a small volcanic landmass roughly 550 kilometres off the Pacific coast of present-day Costa Rica. More than three hundred expeditions have searched for it. None has produced the treasure.
What survives in the documentary record is much thinner than the romance suggests. Lima was certainly evacuated under pressure during the Peruvian War of Independence; church plate and civil treasure certainly moved out of the city ahead of General José de San Martín’s advance; the port of Callao certainly hosted contraband shipping in the closing years of Spanish rule. From those reasonably secure facts, two centuries of nested retellings have produced an itemised inventory of jewelled swords, gold reliquaries, and a seven-foot solid-gold Virgin so specific that one is almost embarrassed to ask where the manifest came from.
The honest answer is that the manifest came, in its earliest substantive form, from a single deathbed account given by a Newfoundland seaman named John Keating in the late 1840s, three decades after the events it described and on the far side of the Atlantic from any Spanish or Peruvian archive. Almost everything else has been built on that foundation. This article walks the chain of custody as scholars actually have it, separating what the record bears from what the legend has added, within the broader landscape of historical and archaeological mysteries.
Lima in 1820: The Documented Background
By 1820 the Viceroyalty of Peru, the last secure citadel of Spanish power on the South American mainland, was under sustained pressure from the liberating armies of José de San Martín (1778-1850). Viceroy José de la Serna (1770-1832), who took office that year, faced a city he could not provision and a port he could not hold. He would formally evacuate Lima on 5 July 1821, withdrawing inland and leaving a garrison at the Real Felipe fortress in Callao. The first siege of Callao began almost immediately.
The figure usually paired with the Treasure of Lima legend, José de la Mar (1778-1830), was Sub-Inspector of the Viceroyalty and Governor of Callao through this period. He surrendered the fortress to patriot forces on 19 September 1821 in the Capitulation of Baquijano, two days after the amputation of his left foot. La Mar would later cross to the patriot side and serve as the first Constitutional President of the Republic of Peru from 1827 to 1829. His proximity to the relevant cargoes is what gives the legend its administrative scaffolding, although no surviving order from him names the Mary Dear or instructs Thompson directly.
What the Churches Actually Lost
Lima’s wealth in 1820 was overwhelmingly ecclesiastical. The cathedral and roughly fifty parish churches and conventual houses held silver lamps, monstrances, processional crosses, gold reliquaries, and a few large devotional statues that had accumulated since the Spanish foundation of the city in 1535. Encyclopaedia Britannica’s overview of the Peruvian War of Independence treats the removal of church plate from Lima during the patriot advance as a real but piecemeal phenomenon, conducted under multiple authorities and with no single inventory.[1] Some plate was hidden in convents, some shipped to Spain via Callao, some converted to coin in the mint. Any single ship’s cargo would have represented a fraction of this.
The Mary Dear, the Mutiny, and the Burial Story
The narrative most commonly attached to the Treasure of Lima runs as follows. A British merchant brig named the Mary Dear, under Captain William Thompson, lay at Callao in the autumn of 1820. Spanish authorities, fearing the imminent fall of the capital, contracted Thompson to receive the most valuable church plate aboard for safekeeping until the political situation stabilised. Spanish officers and clergy embarked with the cargo. Once at sea, Thompson and his crew killed the priests and guards, threw the bodies overboard, and set course for Cocos Island, where they buried the treasure in a cave or sealed pit before scattering.
The Mary Dear was reportedly seized soon afterward by a Spanish naval vessel. Most of the crew were hanged for piracy; Thompson and his first mate, James Alexander Forbes, are said to have saved themselves by promising to lead the Spanish to the burial site. They reached Cocos, escaped into the interior, and from that point the story becomes biographical patchwork. Forbes is sometimes said to have died of yellow fever in Central America; one tradition has him later reappearing as a businessman in California. Thompson is said to have made his way to Newfoundland aboard a whaling ship, where he lived another quarter-century without revealing the location.
The Keating Deathbed Account
The earliest substantive narrative source for any of this is a deathbed disclosure by Thompson, supposedly made in St John’s, Newfoundland, in the mid-1840s, to a friend named John Keating. Keating then sailed to Cocos in 1844 with a partner, Captain Bogue, and reportedly recovered a small fraction of the cache; Bogue did not survive the trip, and historians have generally treated Keating’s account of his death as suspect. Keating’s own deathbed letter, dated to 1854 or 1855, is the version subsequent treasure-hunters worked from. It listed twelve chests, hundreds of thousands of gold coins, eleven thousand silver ingots, and the now-famous Virgin statue with its 1,684 jewels, an itemisation reproduced in subsequent popular histories and surveyed in the standard reference treatment of the legend.[2]
Why the Inventory Should Be Read Cautiously
An evidentiary chain that begins on a Newfoundland deathbed in 1854, attributed to a man recounting events three decades earlier in another hemisphere, is exactly the kind of source historians treat as marginal evidence requiring corroboration before its specific numbers carry weight. No Spanish or Peruvian church inventory from 1820 has been matched against the Keating list. The Mary Dear has not been documented unambiguously in Lloyd’s shipping records of the period. Captain William Thompson is a name shared by several mariners working the Pacific in those years, and the figure described in the Keating story has not been attached to a single, verifiable biography. The skeleton of the legend is plausible; the flesh of its inventory is testimony at two removes.
Cocos Island: A Geographical Note
Cocos Island, called Isla del Coco in Spanish, sits in the eastern Pacific roughly 550 kilometres south-west of the Costa Rican mainland. It covers approximately 23.85 square kilometres of dense rainforest and cloud forest, the only such terrain on any oceanic island in the eastern Pacific. Two narrow bays, Chatham and Wafer, provide the island’s only practical anchorages. The interior is steep, soaked, and threaded with seasonal streams. Whatever was buried on Cocos in 1820 would be subject to root growth, landslip, and a rainy season that delivers seven metres of rain a year.
August Gissler’s Long Search
The most sustained single search of the island was carried out by the German adventurer August Gissler (1857-1935), who lived on Cocos for most of the period 1889 to 1908. Gissler was appointed governor of the island by Costa Rica in 1897 and used his nineteen-year residence to dig systematically across both bays and into the interior. He found six gold coins. UNESCO’s heritage record for Cocos Island National Park retains Gissler within its account of the island’s pre-protection human history, and the surface of Cocos has by every reasonable measure been worked thoroughly without producing the legendary cache.[3]
The Three-Hundred Expeditions
Standard counts give the number of recorded treasure-hunting expeditions to Cocos at over three hundred between the 1840s and the modern prohibition. The roster is unusual. It includes the seafaring Newfoundlanders who knew Keating, English and German naturalists, the New Zealand explorer Frank Worsley of Endurance fame, the gangster Bugsy Siegel, the actor Errol Flynn, and the racing driver Sir Malcolm Campbell. President Franklin Roosevelt is reported to have visited the island during a 1935 fishing cruise and to have shown amused interest in the local lore without organising a serious search of his own. None of these expeditions produced an authenticated piece of the Lima cache.
Ralph Hancock and Mid-Twentieth-Century Searches
Ralph Hancock’s 1960 book The Lost Treasure of Cocos Island, written with Julian A. Weston, became the standard mid-twentieth-century compilation of the Cocos legends and remains a useful catalogue of the variant stories, even where its conclusions outrun its evidence. Hancock approached the island as a forester and naturalist as much as a treasure-hunter, and the book’s most durable contribution is its survey of how earlier expeditions misread the terrain. Subsequent searches in the 1970s and 1980s, including those associated with the Costa Rican-American partnership of Jack Stoneman and various local guides, repeated the geographical pattern: digs concentrated around Wafer Bay, occasional small finds of European coinage of uncertain provenance, no large cache.
Costa Rican Protection and the 1978 Park Decree
By the mid-twentieth century the cumulative damage from treasure-hunting was visible across Cocos. Costa Rica declared the island a national park by executive decree in 1978, formally prohibiting unauthorised excavation; UNESCO inscribed it as a World Heritage Site in 1997, with marine protections expanded in 2002. The Costa Rican government’s working position is that no Treasure of Lima exists on the island and that the legend has functioned, since at least the late nineteenth century, as a tourism and adventure motif as much as a historical claim. Permits for archaeological or geophysical survey are granted rarely and under strict ecological terms.
The Bonito Variant and Conflated Legends
Anyone reading two accounts of Cocos Island treasures eventually meets a second figure, the pirate Benito Bonito, sometimes rendered as Pirate Benito or “the Bloody Sword.” In one popular version Bonito is an independent corsair active around 1819 who buried his own Mexican-gold haul at Wafer Bay; in another he is identified with a British naval officer named Bennett Grahame, court-martialled and executed for piracy after abandoning a Pacific survey mission with HMS Devonshire. A third tradition collapses Bonito and Thompson into a single composite figure responsible for both caches.
The Bonito-as-Fiction Reading
Modern researchers, including the Australian local historian Andrew Warland in his catalogue of nineteenth-century Pacific piracy claims, treat the Bonito story as substantially literary: a composite tall tale assembled from fragments of real and imaginary careers, embellished in successive popular retellings, and increasingly difficult to attach to a single verified person.[4] The Bonito case is the cleanest example of how the Cocos legends absorb each other. Where the documentary record is thin, separate stories tend to flow together along the same imagined coastline.
What Conflation Means for the Treasure of Lima
The discipline of holding the legends apart matters because the inventories shift accordingly. The Bonito haul is sometimes valued at three hundred million dollars in modern currency; the Devonshire cache, at sixteen billion; the Treasure of Lima, between twelve million and the often-cited figure of around sixty million pounds in current value, with one popular metaldetector-industry estimate placing it at $208 million. Those numbers do not derive from any independent inventory; they propagate by rephrasing across web sources. A historian who treats the three legends as a single hoard arrives at much larger figures than the records can support for any one of them.
Why the Story Survives
The Treasure of Lima persists for a familiar set of reasons. Its initial conditions are documented and dramatic: the late-colonial Lima of 1820, the patriot advance, the evacuation of church plate, the busy and lawless port of Callao. Its narrative form is irresistible: the trusted captain who turns pirate, the survivors of the gallows, the buried cargo on a remote tropical island. Its setting is preserved: Cocos remains physically forbidding and now legally inaccessible to amateur searchers, which keeps the question structurally open. And its inventory, however thinly evidenced, is sumptuous enough to have outlasted three centuries of skeptical archival work.
The Honest Distinction
The most useful posture is to keep two questions distinct. Did some quantity of Spanish colonial treasure leave Lima or Callao in the troubled years 1820 to 1821 under improvised arrangements that could plausibly have ended badly? Almost certainly yes; the period offered abundant opportunities for bad-faith captains and lost cargoes. Did Captain William Thompson, the Mary Dear, and a single buried hoard at Cocos Island happen as later writers describe? That is a much more specific claim, supported chiefly by a deathbed letter at three removes from the events, repeated in increasing detail by writers with no access to Spanish colonial archives.
What Would Change the Picture
A clean break in the case would require something the record has not yet produced: a Lima diocesan inventory from 1820 matched against an authenticated Cocos recovery; a Spanish naval log naming the Mary Dear and the dates of its seizure; a verifiable Thompson biography in Newfoundland parish records; or, of course, a properly excavated find on the island under modern protocols. None of these has appeared. The legend, in the meantime, sits where most treasure legends sit, between a real political moment and a literary tradition that grew comfortably around it.
Frequently Asked Questions
Was the Treasure of Lima a real shipment?
Some quantity of treasure almost certainly moved out of Lima during the patriot advance of 1820 to 1821; the Spanish authorities had every motive to evacuate church plate before General José de San Martín’s army arrived. Whether a single named cargo aboard the Mary Dear under Captain William Thompson is the historical event behind the legend is a much narrower claim, supported chiefly by a deathbed account dating to the mid-1850s.
Who was Captain William Thompson?
Captain William Thompson is the British merchant master traditionally identified as commander of the brig Mary Dear in the Treasure of Lima narrative. The name is shared by several Pacific mariners of the period, and no single verified biography has been linked to the figure of the legend. The Newfoundland tradition has him living quietly in St John’s into the 1840s before his deathbed disclosure to John Keating.
Where is Cocos Island?
Cocos Island, or Isla del Coco, lies roughly 550 kilometres south-west of the Costa Rican mainland in the eastern Pacific. It covers about 23.85 square kilometres, with two practical anchorages at Chatham Bay and Wafer Bay. Costa Rica designated it a national park by executive decree in 1978, and UNESCO inscribed it as a World Heritage Site in 1997.
How many expeditions have searched for the Treasure of Lima?
Standard counts give over three hundred recorded treasure-hunting expeditions to Cocos Island between the 1840s and the modern prohibition. None has produced the cache. The longest single search was conducted by the German adventurer August Gissler, who lived on the island for most of the period from 1889 to 1908 and recovered six gold coins.
What did the treasure reportedly contain?
The most circulated inventory derives from John Keating’s deathbed account in 1854 or 1855 and lists twelve chests of gold and silver, hundreds of thousands of gold coins, around eleven thousand silver ingots, jewelled crowns and swords, and a seven-foot solid-gold Virgin Mary said to weigh some 780 pounds and to be set with 1,684 jewels. None of these figures has been corroborated by an independent Spanish or Peruvian source.
Why does Costa Rica forbid treasure hunting on Cocos?
By the mid-twentieth century, more than a century of unregulated digging had produced visible damage to the island’s terrain and ecology. The 1978 national park decree consolidated earlier protections and prohibited unauthorised excavation. Costa Rica’s working position is that the legend functions as a tourism and adventure motif rather than as a historical claim, and that no Treasure of Lima is present to recover.
Who is Benito Bonito?
Benito Bonito, also called the Bloody Sword, is the second of the three Cocos Island treasure traditions, sometimes attached to a real Pacific corsair around 1819 and sometimes identified with the British naval officer Bennett Grahame. Modern researchers treat the Bonito story as a substantially literary composite assembled from fragments of real and imaginary pirate careers.
Did Ralph Hancock find anything on Cocos?
Ralph Hancock’s 1960 book The Lost Treasure of Cocos Island, co-written with Julian A. Weston, is a survey of the variant legends and a record of the geographical patterns repeated by earlier failed expeditions, rather than a recovery account. Hancock did not produce the cache, and his most durable contribution is his catalogue of how predecessors misread the terrain.
Is the Treasure of Lima the same as the Devonshire treasure?
No. The Devonshire treasure is the third Cocos tradition, attached to Captain Bennett Grahame of HMS Devonshire and reckoned in figures (sometimes sixteen billion dollars in current value) that exceed any single contemporary Spanish manifest. Conflating the three Cocos hoards into one legend is responsible for many of the largest inflated valuations seen in popular treatments.
What would settle the question?
A break in the case would require either an authenticated find on Cocos under modern archaeological protocols, a Lima diocesan inventory from 1820 matched against a recovered piece, a verifiable Spanish naval log naming the Mary Dear and Thompson, or a Newfoundland parish record establishing his biography. Until one of these surfaces, the Treasure of Lima remains a real political background married to a literary tradition that grew around it.
What was the role of José de la Mar?
José de la Mar (1778-1830) was Sub-Inspector of the Viceroyalty of Peru and Governor of Callao through the years usually associated with the legend. He surrendered the fortress to patriot forces in September 1821 and later served as the first Constitutional President of the Republic of Peru from 1827 to 1829. His administrative position gives the Treasure of Lima story its scaffolding, although no surviving order from him names the Mary Dear or Captain Thompson.
