The pirate and the cryptogram have very different paper trails, and the gap between them is the whole story.
Olivier Levasseur, under the name the English clerks used for him, turns up in East India Company correspondence written in his own lifetime, with a ship's name, a tonnage, a gun count and a crew number attached. The cryptogram turns up in a slim French book published in Paris in 1934, and the library records for that book do not mention Olivier Levasseur anywhere.
Key Takeaways
- Captain James Macrae's letter of 16 November 1720 names "Oliver de la Bouche" and gives his ship as the Indian Queen, of 250 tons, 28 guns and 90 men.
- That letter is not in the 1724 first edition of the book usually credited with it. The documented half of la Bouche's Indian Ocean career arrives only with the second edition, whose title page advertises "considerable Additions".
- The taking of the Portuguese ship in April 1721 was written down by a man who was there as a prisoner: Richard Lazenby, second mate of the Cassandra.
- Four reachable accounts of that capture survive, and they disagree about the resistance, the ransom, the ship's armament, who commanded and who was aboard.
- The Library of Congress catalogues the 1934 book under a different pirate, John Avery, though a record of that vintage carries very little subject detail either way.
A note on the name, before anything else
The English sources quoted below do not call the man Levasseur. They call him La Bouse, or la Bouche, or Oliver de la Bouche, depending on which page you are reading. Those are the anglicised forms of the French nickname La Buse, and identifying their bearer with Olivier Levasseur is the conventional reading, followed by essentially everyone who writes about him.
The spelling wobbles inside a single volume, and where it wobbles is informative. The compilation discussed below went through more than one edition, and its second edition spells him "La Bouse" in the chapter on Howell Davis, which it inherited, and "la Bouche" in the Indian Ocean material, which it added. The two spellings fall along the seam of the revision. Clerks and printers were writing down a French nickname they had heard rather than read, and the later hand heard it differently from the earlier one.
None of the sources reachable here makes the equation with Levasseur itself. They record a French pirate captain, operating in the right waters at the right dates, under a name spelled as it sounded. This article adopts the standard identification because there is no serious competitor to it, not because any of these documents confirms it. In a piece about what the record does and does not support, that distinction should be on the page rather than buried.
What does the surviving paperwork actually say about Olivier Levasseur?
The earliest reachable account comes from a man trying to get his employers to send warships. Captain James Macrae of the East India Company wrote from Bombay on 16 November 1720, and his letter was reprinted in the compilation that goes under the name Charles Johnson.
A word about that book, because the article leans on it twice and the two states of it are not the same book. Its first edition appeared in London in 1724 as A General History of the Robberies and Murders of the Most Notorious Pyrates, its title page running "to the present Year 1724". A later state carries the shorter title A General History of the Pyrates and the line "The second Edition, with considerable Additions". The catalogue record for the scanned copy of that second edition also gives 1724, but the reissues came quickly and nothing below rests on the precise date. Johnson is a name on a title page rather than a settled identity, and the work was revised as it was reissued, in the ordinary way of a commercial serial. The two are cited below by which state they are, not by an imprint.
Both states carry the pirate, but not the same part of his life. The first edition has him in the chapter on Howell Davis, as "La Bouse", commanding a French vessel of fourteen guns and sixty-four hands off the African coast. It is a chance meeting in which each takes the other for a hunter, both hoist black colours, and Davis ends by promising him the first suitable ship he takes, "as being willing to encourage a younger Brother".
What the first edition does not have is the paperwork. Its England chapter reaches the Indian Ocean perfectly well: Madagascar, the taking of the Cassandra at Johanna, the Portuguese man of war. What is missing is the documentary layer. No Macrae letter, no Indian Queen, no Mayotta, none of them anywhere in the volume. The wreck, the tonnage, the gun count and the crew number all arrive with the second edition's additions, and they arrive in the form of a Company officer's own correspondence.
Macrae had arrived at Johanna, in the Comoros, on 25 July that year. He found fourteen pirates who had paddled over by canoe from Mayotta, where their ship had come to grief. He set down what they told him: the vessel was the Indian Queen, "two hundred and fifty Tons, twenty eight Guns, and ninety Men, commanded by Capt. Oliver de la Bouche, bound from the Guinea Coast to the East-Indies", and it "had been bulged and lost". The survivors said they had left their captain and forty of their men behind, building a new vessel.
That is the texture of the real record. A wreck, a headcount, and a captain last seen on a beach building a replacement ship with forty men.
The second edition puts him in two other places. It lists "Oliver la Bouche" among the pirate commanders operating out of New Providence in the Bahamas, alongside Hornigold, Teach and England, and notes that he was "afterwards cast away", which is the Mayotta wreck. And it has him at Whydah on the Guinea coast, arriving ahead of Captain England and, in Charles Grey's rendering of the same passage, having "forestalled their market" to the considerable annoyance of the pirates who got there second. Quotations from either state of this book have the long s normalised here, since the surviving scans read it as an f.
Who wrote down the capture of the Portuguese ship?
A prisoner did. Richard Lazenby was second mate of the Cassandra when the pirates took her at Johanna in July 1720, and they kept him to pilot them along the Indian coast. His deposition was printed in full by Charles Grey in Pirates of the Eastern Seas, published in 1933, and it is the closest thing the episode has to an eyewitness.
Lazenby's dates are precise. The pirates reached Mauritius in the middle of February 1721 in poor shape, down to one bottle of water and two pounds of beef per man for ten days. They repaired and re-sheathed their leaky ship, sailed for Mascarenhas (the Isle of Bourbon, today Réunion) on 5 April, and arrived "at eight in the morning of the 8th".
What they found was not a fight. The Portuguese ship carried 70 guns, but a great storm in latitude 13 had taken away all her masts and all but 21 of those guns. She was taken "with small resistance". The Viceroy of Goa and several other gentlemen had gone ashore; they came out to the pirate ship that morning believing she and her consort were English Company ships.
Then the argument started. Some of the crew wanted to carry the Viceroy to Mozambique and hold out for a large ransom, others thought a smaller sum taken on the spot was better than more trouble. "At last they compounded for 2,000 dollars for the ramson [sic] of the Viceroy," Lazenby wrote. He begged to be put ashore, and on the 10th he walked off with the Viceroy and the rest of the prisoners.
The pirates promised the governor they would leave a ship to carry away the people they had landed, since the island could not feed them. Instead they sailed in the night, taking the best of the captured sailors and two hundred enslaved people from the Viceroy's ship. Lazenby adds one detail he could only have got afterwards: the Viceroy later told him the diamonds aboard had been worth three or four million dollars.
Grey follows the deposition with the share-out at St Mary's. Over £4,000 a man. The Portuguese vessel alone was reckoned at £875,000 in cash, jewels and plunder, and the whole cruise at about £1,046,000, divided among an estimated 240 men.
Two things the deposition does not do. It does not name the Portuguese ship, which appears throughout only by her nationality and her guns; the name modern retellings give her, the Nossa Senhora do Cabo, is used by none of the four accounts below. And it does not name La Bouche. Lazenby's pirate commander is John Taylor, in the Victory and then the Cassandra.
Why do four accounts of the same capture disagree?
Because it was rewritten early, and the sharpest change is between two states of the same book.
The first edition puts the capture in Captain England's chapter and stages it as an assault. The pirates stand off all day, come into the bay at dusk, board "at once" with "a hundred stout Fellows, with a Cutlass, and a brace of Pistols each", and after "a short Dispute, wherein twenty or thirty Men were killed" carry the prize out to sea that night with the prisoners locked in the hold. The ship is "a Portuguese Man of War of 60 Guns" with about a thousand people aboard. The Conde Ericeira is aboard her because he has been "removed from his Government, for some Male Administration", and is bound for Brazil "with all his Wealth, Servants and Attendants", travelling with the Archbishop of Goa. Both men, this version says, happened to be ashore when the pirates struck. No ransom is mentioned.
The second edition tells it as a different event. It is Taylor's now, not England's, and it tracks Lazenby closely: same island, same date, same weather. The pirates arrived on the 8th and found "a Portuguese Ship at Anchor, of 70 Guns, but most of them thrown over" after "a violent Storm they had met with in the Latitude of 13 South", so that "she became a Prize to the Pyrates, with very little or no Resistance". It names the Viceroy, whom Lazenby never does: "the Conde de Ericeira, Viceroy of Goa, who made that fruitless Expedition against Angria". It puts the diamonds at "between three and four Millions of Dollars", exactly Lazenby's figure. And it records the ransom being argued down "in Consideration of his great Loss, (the Prize being Part his own,)", then the same broken promise about leaving a ship, and the same two hundred people carried away.
One state has a night boarding and thirty dead; the other has a bloodless morning surprise and a ransom. Notice the direction. The revision ran towards the eyewitness, not away from him, which is the opposite of how a legend is supposed to behave.
The likeliest reason is not that two witnesses independently agreed. Lazenby was an East India Company officer deposing to his employers, and the second edition's additions are visibly drawing on Company paper, since one of them is Macrae's letter. So the agreement between them is probably common descent from the same file rather than corroboration by separate channels, and the two should not be counted twice. That reading is the more specific one, and it makes the direction of travel sharper rather than weaker: whoever prepared the additions got hold of the Company's own documents and rewrote the chapter out of them.
That also means the first edition is not simply a degraded copy. It carries circumstantial detail the second edition drops entirely, including the Archbishop, the disgrace and the voyage to Brazil. Which version is closer to the day itself is not something these four sources settle.
Eighty years on, the drift is unmistakable. Charles Grant, Viscount de Vaux, published The History of Mauritius, or the Isle of France, and the Neighbouring Islands in London in 1801, drawing on French colonial material. In the section headed "Some Account of the Isle of Bourbon", the capture has become a dinner-party anecdote. The Viceroy anchors in the road of St Denis and is invited to dine by the governor. He has barely landed before a pirate vessel of fifty guns moors alongside his ship and takes it. The pirate captain comes ashore, invites himself to dinner, seats himself between the governor and the Viceroy, and tells the Viceroy he is a prisoner. Then, once the wine has done its work, the governor asks him to name a ransom. The pirate says a thousand piastres. The governor tells him that is too little "for a brave man like you, and a person of high distinction like him", and invites him to ask more, or nothing. The pirate, described in the book as generous, gives the Viceroy his liberty for nothing.
| Account |
Text |
Commander |
Resistance |
Ransom |
Where the Viceroy was |
| Lazenby's deposition |
eyewitness held aboard, printed 1933 |
Taylor |
"small resistance" |
2,000 dollars |
ashore, came out that morning |
| Johnson, first edition |
Robberies and Murders, 1724 |
England |
night boarding, 20 to 30 killed |
none mentioned |
ashore, with the Archbishop |
| Johnson, second edition |
General History of the Pyrates, "considerable Additions" |
Taylor |
"very little or no Resistance" |
argued down, sum agreed |
came aboard that morning |
| Grant, History of Mauritius |
London, 1801 |
unnamed |
none mentioned |
refused, freed for nothing |
ashore at the governor's dinner |
Four versions, not one of them containing a cipher, a parchment or a map. A version flatly incompatible with the only eyewitness was in print by 1724, less than four years after the capture, and the book was rewritten after that. All of this happened long before anything had to be decoded.
That pattern is the normal one for buried-treasure ciphers rather than the exception. The Beale ciphers reached the world the same way: through a single publication issued decades after the events it describes, with no independent trace of the documents it claims to reproduce.
What can be established about the book that carried the Levasseur cryptogram?
Its catalogue record, in some detail. Its contents, not at all.
The book is Le flibustier mystérieux: histoire d'un trésor caché, by Charles de La Roncière, with illustrations by Jean Bernard. The Library of Congress gives the publication statement as Paris: "Le Masque", [1934], the square brackets being the cataloguer's signal that the year was supplied rather than transcribed from the book itself. The record describes 112 pages, 19 cm tall, with a physical note reading "illus. (incl. facsims.)" and a bibliography on the first leaf at the end. Open Library, whose entry derives from a Library of Congress record under LCCN 35007202, assigns it to a series it transcribes as "Adventures et légendes de la mer", legends of the sea.
The author was not a hack. Charles Germain Marie Bourel de La Roncière lived from 1870 to 1941 and wrote the multi-volume Histoire de la marine française, a life of Colbert, and a study of medieval African exploration. He is also a named contributor to the Catalogue général des manuscrits des bibliothèques publiques de France: Bibliothèques de la marine, published in Paris in 1907, and to the Bibliothèque nationale's catalogue of French manuscripts. The man who put the treasure story into wide circulation catalogued naval manuscripts for a living.
Then there is the oddity. The Library of Congress work record carries exactly two subjects: the topical heading "Buccaneers", and a person, "Avery, John, active 1695". Its genre form is Biographies. Levasseur does not appear in the record at all.
Read that carefully, though, because the same standard applies here as everywhere else in this article. A record converted from a pre-MARC catalogue card carries very little subject detail by modern standards, two headings is close to the minimum, and a name with no established authority file in the 1930s would simply not have been assigned. The silence is real, and it is weak evidence. What it shows is that the cataloguers who handled the book in Washington filed a treasure narrative under a different pirate, not that the 112 pages say nothing about this one.
Can the Olivier Levasseur cryptogram's provenance be checked at all?
Not from the open archives, and it is worth being exact about what that does and does not mean.
The 1934 text is not in the open full-text collections, and the mundane reason is copyright rather than anything interesting: a French book published in 1934 by an author who died in 1941 sits inside the window that keeps works out of freely readable scans, so it survives in lending libraries instead. It is readable. It simply was not read for this article. That is a limit of method, not a fact about the record, and treating an inaccessible book as a suspicious one would be its own kind of invention.
What can be checked is the absence elsewhere. Neither spelling of the word, cipher or cypher, occurs anywhere in either state of the Johnson compilation, first edition or second, and the second is the book supplying most of what is known about La Bouche's career. Charles Grey's 1933 history, published the year before La Roncière's and covering the same ocean and the same pirates across more than 300 pages, has no cryptogram in it either. Its only mention of a pirate cipher is a set piece in the opening pages, where Grey sketches "the traditional map and cipher, wherein the Pirate Captain betrayed the secret of the hoard" specifically as the novel-reader's stock image he intends to write against. These are scans of old print, and OCR on 18th-century type is imperfect, so treat a null search as suggestive rather than conclusive.
The archives are quiet too. The National Archives Discovery catalogue, searched on 10 August 2026 through its public records endpoint, returns nothing for "Levasseur pirate", nothing for the ship's usual modern name "Nossa Senhora do Cabo", and nothing for "Cassandra pirates Macrae". A search for "La Buse" returns five records, none of them relevant, including a Himalayan photograph and a medieval petition. A search for "pirates East Indies 1721" returns three, all of them East India Company shipping journals and miscellaneous papers. No coded document, and nothing that looks like one.
One boundary needs naming rather than glossing. Everything above is the English-language record, which follows this man to the Mayotta wreck and the 1721 prize and then loses interest. What happened to him afterwards belongs to the French colonial administration of Bourbon, whose registers and council correspondence are a different body of material entirely, and one that none of the open archives used here reaches.
None of this proves the cryptogram was invented. Absence of a thing from the books that happen to be openable is not a demonstration that it never existed. It does mean the evidence for the ciphered half of this story starts later than the evidence for the man, and that anyone who tells you otherwise is going past what the reachable record supports.
A cipher can enter the world through a printed page and still be perfectly genuine, and that should not be forgotten here. The D'Agapeyeff cipher first appeared in a book too, in 1939, and nobody doubts its author put it there. What separates that case from this one is that its author said plainly what he had done.
Frequently asked questions
Is Olivier Levasseur a documented historical figure?
A French pirate captain the English called Oliver de la Bouche certainly is, and the identification of that man with Olivier Levasseur is standard, though none of the sources here states it outright. Under that name he appears in East India Company correspondence written while he was alive: Captain James Macrae's letter from Bombay of 16 November 1720 records the loss of his ship the Indian Queen at Mayotta with her tonnage, armament and crew number, and says he was last known to be building a replacement with forty men. That letter reaches print in the second edition of the Johnson compilation, which also places him at Whydah on the Guinea coast and lists him among the pirate captains at New Providence. The first edition carries him too, under the spelling "La Bouse", but only for his African career.
Who actually captured the Portuguese ship in 1721?
The eyewitness account names John Taylor. Richard Lazenby, second mate of the captured Cassandra, was held prisoner throughout the cruise, and his deposition describes Taylor's company taking a dismasted 70-gun Portuguese ship at Mascarenhas on the morning of 8 April 1721 with small resistance. The four accounts do not agree even on this: two give Taylor, one gives Captain England, and the 1801 version names nobody. He is named in both editions elsewhere, as La Bouse off Africa in the first and as la Bouche at New Providence and Whydah in the second, but he appears in none of the four accounts of the capture itself. That is a limit on what these documents show rather than evidence that he was somewhere else.
How much was the prize worth?
Grey's figures, following the deposition, put the Portuguese vessel at about £875,000 in cash, jewels and plunder, within a cruise total of roughly £1,046,000 shared among an estimated 240 men, or over £4,000 each. Both Lazenby and the second edition report the diamonds aboard at three to four million dollars, but that pairing is not two independent valuations: the deposition and the printed additions look to descend from the same East India Company papers. These are contemporary estimates recorded by interested parties, not an audited valuation, and they should be read that way.
Why do the sources disagree about what happened to the Viceroy?
Because the account was being rewritten while it was still recent. Lazenby, who was on the spot, says the pirates settled on 2,000 dollars, and the second edition of the Johnson compilation agrees that a ransom was argued out and paid. Its own first edition has a night boarding with twenty or thirty men killed, the Viceroy ashore throughout, and no ransom at all. Grant's History of Mauritius of 1801 has the governor of Bourbon talking the pirate out of taking anything. They cannot all be right. What the spread shows is that a version irreconcilable with the only eyewitness was already in print by 1724, less than four years after the capture, and that the book was rewritten afterwards, all well before a cryptogram entered the story.
Where did the cryptogram itself first appear in print?
That question cannot be answered from the open archives, and the honest position is to say so. The 1934 book Le flibustier mystérieux is the work usually credited with putting it into circulation, and its catalogue record confirms the book exists, runs to 112 pages, and contains facsimile illustrations. The text is not in the freely readable scans, most likely for ordinary copyright reasons, so what it reproduces and where it says the symbols came from cannot be checked without a lending library. No invented chain of custody should be offered in place of that gap.
Pirate stories are not the only ones that arrive already half-legend. The other unsolved ciphers are stranger, better documented, and in several cases still open. If you would rather solve one than read about it, the cryptogram puzzle is next door and does have an answer.
Sources
- Charles Johnson, A General History of the Robberies and Murders of the Most Notorious Pyrates, first edition, title page "to the present Year 1724", chapter VIII, Of Captain England, for the night-boarding account of the Mascarenhas capture - retrieved 10 August 2026, https://archive.org/details/generalhistoryofrobberies00johnson
- Charles Johnson, A General History of the Pyrates, title page "The second Edition, with considerable Additions", scanned from the Smithsonian copy. Macrae's letter of 16 November 1720 and the New Providence commanders list in the chapter on Captain England; the Whydah episode in the same chapter; the Conde de Ericeira capture in the Taylor material that follows - retrieved 10 August 2026, https://archive.org/details/generalhistoryof00john
- Charles Grey, Pirates of the Eastern Seas (1618-1723) (1933), chapter XVII, printing the narrative of Richard Lazenby, second mate of the Cassandra; the treasure-map set piece at pp. 3-4 - retrieved 10 August 2026, https://archive.org/details/PiratesOfTheEasternSeasCharlesGrey
- Charles Grant, Viscount de Vaux, The History of Mauritius, or the Isle of France, and the Neighbouring Islands (London, 1801), "Some Account of the Isle of Bourbon", pp. 156-157 - retrieved 10 August 2026, https://archive.org/details/historyofmaurit00vaux
- Library of Congress Linked Data Service, BIBFRAME work and instance records for La Roncière, Le flibustier mystérieux; histoire d'un trésor caché - retrieved 10 August 2026, https://id.loc.gov/resources/works/10169416
- Open Library edition record for Le flibustier mystérieux, derived from a Library of Congress record, LCCN 35007202 - retrieved 10 August 2026, https://openlibrary.org/books/OL6319458M
- The National Archives, Discovery catalogue, public records search endpoint, queries as quoted above - retrieved 10 August 2026, https://discovery.nationalarchives.gov.uk/results/r?_sd=&_q=pirates%20East%20Indies%201721
