Buried Treasure: Cipher Text

The Levasseur Cryptogram

Cover of De la Roncière's book.	As far as can be ascertained, the Levasseur cipher surfaced in 1934 with the publication of Charles de la Roncière's book on the Seychelles treasure, Le Flibustier Mystérieux: Histoire d'un Trésor Caché. While the cipher is not associated directly with Oak Island, the operation of which it might be part bears a number of similarities. Furthermore, the cipher could be an example of the lengths to which pirates might have gone in order to protect their treasures, yet still make them accessible to anyone with the wit, or knowledge, to understand their messages. Of course, the cryptogram could be yet one more hoax perpetrated at much the same time that the Palmer-Kidd maps and the Guayacán ciphers made their appearance. Roncière simply states that it was in the possession of a notary. It may have nothing at all to do with the pirate Olivier Levasseur. Provenance aside, it has occupied the minds of countless people hopeful of solving it, and has sparked off at least two major treasure hunts. The whole thing may be a hoax or, even, a cruel jest by La Buse. However, the puzzle exists, and despite its questionable provenance there is here an intellectual puzzle to tackle on cold wet winter days. The cipher is presented as follows:
The Levasseur (La Buse) Cryptogram. From Charles de la Roncière.
The first step is to translate the symbols into letters of the alphabet and numbers, if applicable. The second stage is to translate the cipher text and thereby reveal the plaintext. The nature of the first task is fairly well known, and takes the form of a 'pig-pen' cipher that occasionally appears in Masonic writings and illustrations. Roncière links the symbols to an esoteric work, The Keys of Solomon.
Probable key to the first layer of the Levasseur Cipher.
Applying the above key to the Levasseur cryptogram results in the following:
Probable text for the first layer of the Levasseur Cipher.
This would seem to be the correct approach to the first level, as it is obviously in the French language, but the text is either pidgin or jumbled up somehow. The number of spaces occupied by the symbols in the second and fourth lines from the bottom, and the number of spaces in the third line from the bottom, could be important, but cannot easily be determined. Furthermore, it cannot be known if this is a correct copy of the original. There seems nothing out of place with the spelling, but why is it not more quaint? From what can be gleaned, the text seems to be more modern than three hundred years old. Many writers attempt to make sense of the text as it stands, rather than assuming that it is cipher text and has to be decrypted. The optimistic way to approach the problem is to assume that the cipher has been presented for amusement in a book of puzzles. This way, it can be assumed that it has a solution and that everything needed to solve it has been provided. Such an approach would then exclude the existence of an overlay or template that, when applied to the grid, exposes only certain letters. It could be that the lines have been rearranged, or that a simple message is hidden in one of the columns. The temptation to anagram the letters making up the columns should be resisted unless the text revealed is appropriate to what might be assumed of the cipher, for example, spelling out the name Olivier Levasseur or Les Îles d'Abondance. For instance, the twenty-fourth column is unique in having text that can also be found in a row (the fourteenth.) This text string can be rearranged to spell out Thirteen Feet Tides, which being in English is an unlikely opening into the cipher. Note, also, that if the text in the rows is wrapped round beyond the length of the shortest row then a different arrangement of columns is produced, and the text will read differently. The presence of symbols not comprehended by the cipher might suggest nulls or special characters (such as being in a language other than French.) One of them might indicate the starting point. Also strange is the presence of several letters 'K', which would be surprising in the French language of the early eighteenth century. It could therefore be that these have been introduced as part of a stratagem to reveal a message in the columns, and they are to be removed once that purpose has been achieved. This thought draws attention to the letter count. As presented above, the cryptogram has 524 characters in seventeen rows of between 29 and 35 letters. Removing the letters 'K', the blanks, and special characters, brings the count down to 511, one short of 512 - which would produce two squares of 256 characters. So, the grid could have started as sixteen rows of thirty-two letters. Removing all the letters 'K', the glyph 'D' of the first line and the space at the beginning of the sixth would produce 257 letters from the first eight rows. Unfortunately, the count from the last nine rows, after stripping in like manner, is 254. It could be that one of the letters has been missed in a transcription from the original. It can be observed that there are sections of text in blocks of between eight and sixteen letters, and there are a number of sequences having six letters and less. Most of the obvious portions of text are shown in the list below:
Some French words, and parts, in a continuous stream as seen in the cipher text
These blocks of text account for just over one half of the grid - without adding a smattering of three letter sections (for example: UNE, SUR etc.) Roncière does not identify the location of the treasure site except to say, "on a certain island in the Indian Ocean." He observes that the owner of a piece of land bordering the sea was surprised by the appearance of rock-drawings after a storm as a result of high tides, floods, the shifting of earth or the fall of large trees. He describes them thus: "... the head of a sleeping man was engraved, accompanied, at some distance, by two snakes, an urn and the major part of a scorpion. On land, a turtle carved level with the ground seemed almost alive. On a rock, in relief, was drawn the head of a snake above a woman in stone, an attractive woman with her hair in a bun. And one fancied to see in this Andromeda delivered from a monster by Perseus, when, not far off, the Eye of a monster was to be seen. A gun dog, leg raised, evoked Sirius, and a dog with a large belly suggested Procyon, crawling in the sky. The muzzle of a horse brought Sagittarius to mind. With the Bear, the Bull [, and so on,] the vault of the heavens was drawn on, or carved in, the rocks in the ground." The horse might also evoke Pegasus, and I assume that Sirius represents Canis Major, and Procyon Canis Minor, but note that Roncière mentions a lady and a sleeping man, and that sections of both 'un homme' and 'une femme' can be seen twice in the cipher text. Such fragmentation might suggest that the cipher text derives from squares that have been divided by introducing diagonals to create a '+' and 'x' pattern (such as the Union Jack), or a rhombus upon a cross. Filling a square diagonally, and rearranging blocks of text, would produce much the same effect. The figure below provides one example.
A rearrangement of text to generate broken sequences such as in the Levasseur cryptogram. The transposition could permit sections to remain coupled in order to generate longer text sequences.
If this has been done then finding the actual grid employed will not be easy. It may be appreciated that in view of the text FAUTQ there are only two occurrences of UE that are likely to follow, and only two of QU that might follow LFAUT - if not completed by an infinitive. VOUSNAVE would presumably end in Z and there is only one such letter available, in the section PULEZPFVS, in which case F and V would have to be single letters. The text ZP could be a two letter portion making up VOUSNAVEZP. However, the text is problematic. It might divide as: V PULE ZP F V SPR...., and in much the same way the text FVMM would produce two single letters together, split as F V MM (from HOMME?) Below is a grid in this form dividing a paragraph in Le Flibustier Mystérieux by Charles de la Roncière in which the discoveries are described. Although a man and a woman are mentioned in the text this does not emerge obviously from the blocks as separated. The text begins, "En mer face au rivage était ciselée la tête d'un homme ..."
An example of text broken up in the pattern described above.
The blocks of text seem not to be as revealing as those of the Levasseur cipher, but switching them, systematically or otherwise, would render this plaintext unintelligible. However, note that the central triangle at lower right produces intelligible text much as in the Levasseur cipher, but here much easier, perhaps, to piece together the last two lines. That is, URSEETLETAUREAU readily suggests LOURSE, and the block above, with almost equivalent length, VOQUEAPEUPRES, hints at EVOQUE. Note that the verb FAIRE appears in the longest blocks of the Levasseur cryptogram: OUSENFAITESUNE and EUTSEFAIREDUNH. A number of people claim to have solved the cryptogram, but as far as I am aware no proof of a solution has ever been published.
© G.J. Bath Source: Bath G.J. Maps, Mystery and Interpretation Vol 2. Oak Island Speculation. KeyPress 2013.