Extract > Evelyn Farr – Code Breaker
I Love You Madly is the definitive history of the romantic affair between Marie-Antoinette and the Swedish diplomat Count Fersen. Historian Evelyn Farr buried herself in archives across Europe uncovering never before published letters which not only confirm the much-rumoured affair, but reveal the nature of it.
The content of these letters between the pair betray a deep and profound love for each other, one which kept them in constant communication, and therefore in constant danger. The methods they adopted to keep their correspondence secret are intricate and fascinating. In this extract Evelyn Farr explains the secret code used to encrypt their letters. Farr’s work on deciphering the code is hugely significant to future studies of Marie-Antoinette. Until now, there has only been the story of the Queen of France; with the publication of I Love You Madly there will be history.
The first letter recorded as being sent in code in Fersen’s register was addressed to ‘Josephine’ on 4 July 1787, shortly after the death of the Queen’s youngest daughter, Madame Sophie. Fersen recorded other letters sent in code that summer as well as letters sent in invisible ink and noted ‘en bl.’ – en blanc. However, he does not always indicate in the register when a letter was sent in code.
It is clear that some time before the Revolution, Marie-Antoinette had gained experience of encrypting her letters. This is evident not only from the ‘Josephine’ correspondence but also from a letter she wrote to Comte Valentin d’Esterhazy on 11 August 1791.
Write to me sometimes in our code; when you were with army headquarters I was used to decoding your letters – in happier days, I always used to undertake the task. Number your letters, too, to make sure none are lost.
Marie-Antoinette was so closely watched at the Tuileries after the return from Varennes that writing any kind of letter was extremely dangerous. She mentions ‘our code’ and the fact that she was used to decoding Esterhazy’s letters from army headquarters, which suggests she was accustomed to decoding his military dispatches to Fersen. On 3 September 1791 she sent precise instructions to Esterhazy on selecting a code word. ‘When you encrypt, always use the first word on the page, making sure that it has no less than four letters.’ Each correspondent had to have the same edition of the same book to find the code word, which was indicated to the addressee by writing the page number on the coded letter. According to Mme Campan, Marie-Antoinette’s code book was the novel Paul et Virginie by Bernardin de Saint-Pierre. The Duc de Choiseul mentions another book employed for encrypting his correspondence with Fersen regarding the Royal Family’s ill-fated escape from Paris in 1791; the same code words are to be found in the Fersen/Marie-Antoinette correspondence, so it may also have been the book they used. It lacks the romantic connotations of the novel suggested by Mme Campan. ‘This code was the small volume De la grandeur et de la décadence des Romains,’ according to Choiseul, accusing Fersen of having negligently encrypted a crucial letter. It was not deciphered for seven hours, supposedly because Fersen ‘had forgotten to include the mark to indicate the page of the code book’. Fersen’s papers, however, attest to his great attention to detail, and he was, in fact, the only member of the escape committee who executed his part of the operation flawlessly. It is difficult to identify the exact edition of Montesquieu’s De la grandeur et de la décadence des Romains that was used, since several editions had already been published by 1791.
Since Marie-Antoinette sometimes also sent coded letters to Louis XVI’s brothers, Emperor Leopold II and the Comte de Mercy, it is likely that she used a different book for each correspondent. A letter from Fersen indicates that the book changed from time to time. He informed the Queen that her secretary, François Goguelat, had used a code book he could not identify. ‘I have received a [letter] from Gog. that I haven’t been able to read. I certainly do not possess the book he has used.’ For those who would like to try to identify the code book used by Marie-Antoinette and Fersen – Paul et Virginie or De la grandeur et de la décadence des Romains – here is a table that gives the first word on several of its pages. It has been compiled using the page numbers and code words from their letters.
The code book and code word, however, are utterly useless without the code table. The Swedish National Archives possess an interesting box, little known because it is kept not with Axel von Fersen’s papers but among the rest of his family’s archive. It is simply labelled ‘Codes’ and contains the Swedish and French diplomatic ciphers (all numeric) used by Fersen, his father and Baron Taube, as well as a tattered little packet labelled ‘1790–94: Ciphers used by Count Axel von Fersen for Queen Marie-Antoinette and the French correspondence.’ Inside this forgotten packet are to be found not only the actual code table used by Fersen to encrypt and decipher his correspondence with Marie-Antoinette but also a list of code names (never used in the letters which have survived), details of two very important intermediaries for the correspondence of 1791–2 and documents that show that Fersen had tried to persuade the Queen to adopt a code that was con siderably more complex than the one they normally used. This code ‘No. 2’, has never been published before. (Figure 2)
It is easy to see why Marie-Antoinette rejected it; it was far too complex for her voluminous correspondence and carried a greater risk of basic transcription errors than the code she preferred. This did not prevent Fersen from writing her a letter in which he explained how to use it (see Part IV, letter of 1 February 1790). None of their letters in the archives uses code ‘No. 2’. At first glance it would appear that the list of code names in this packet was also never used. There is no evidence of them in the Queen’s poly-alphabetic code, where proper names are encrypted in full; on the list they are represented by a single letter of the alphabet. But a letter from Marie-Antoinette to Fersen on 30 March 1792 provides a clue as to the purpose of this list. ‘Mr Craufurd will have told you of a way of writing to me in Italian without a code. Don’t forget to send me the list of names.’ Below is the translated list and the original in Fersen’s handwriting. (Figure 3a)
It is impossible to confirm that Fersen ever used this method to write to Marie-Antoinette. More interesting is the back of the list of code names (Figure 3b), on which we find the names and addresses of M. Goguelat and M. Gougenot, faithful servants of the Queen whose names appear regularly in her correspondence with Fersen (see below).
Finally, this neglected packet reveals its last treasure: the table for the poly-alphabetic code used by Fersen to encrypt and decrypt the letters he exchanged with Marie-Antoinette in 1791–2.(Figure 4)
It is relatively easy to decode a letter using this table; for example, using the letter from Marie-Antoinette written by her secretary Goguelat on 1 August 1792.
P. S. The bale I have sent you by the stage-coach is marked No. 141, and each piece of material carries the following letters. n m f p x a n m g o q
This means that the code word is to be found on page 141 of the book and that every letter is encrypted (‘each piece of material’). The code word is paroîtra (Fersen has written it underneath to decode the message).
The message is ‘il y a du blanc’, meaning that the rest of the letter is written in invisible ink.
To decode in this case, one looks for the letter N in line P of the table; it is linked to an I, which is the first letter of the decrypt. For the second letter, one looks for M in line A of the table; it is linked to an L, which becomes the second letter of the decrypt, and so on … But even this relatively simple code becomes time-consuming when one has to code and decode several letters. There are some coded words in Marie-Antoinette’s letters to Fersen that no one has been able to decipher, and Fersen’s ‘workings’ on his decrypts testify to the difficulties he experienced with letters encrypted by the Queen.
In order to speed up the process, often only every other letter was encoded; this is the ‘skipped letter’ method referred to by Marie-Antoinette in a letter to the Comte de Mercy on 19 October 1791. ‘This letter is encrypted using the new method. If you have trouble with it, consult M. de F. [Fersen]. You must skip a letter.’ But writing to Fersen himself on 31 October, it would appear that she was not so willing to use the ‘skipped letter’ method and advised it only for ‘occasions’, that is, for letters that were carried by trusted intermediaries.
I quite understand everything regarding the code, but we must always write a colon when both words finish at the same time, and leave out the Js and the Vs; that will make it easier for us. Skipping a letter will only serve when we write by occasion.
Fersen replied on 26 November 1791:
I understand very well what you tell me about the code. We will use it thus; let’s put a full-stop . at the beginning, and when a letter is skipped, let’s put a colon :
And he actually wrote the punctuation marks in his letter! He observed this method faithfully. For encryption using the ‘skipped letter’, the page number of the code word is always followed by a colon. Coding and decoding so many letters nevertheless remained a laborious task. Fersen provides some very interesting details in a letter to the Queen dated 6 March 1792. Certain important instructions regarding code and invisible ink were suppressed by Klinckowström and all subsequent editors. The unpublished passage is in bold type.
Goguelat must be notified that every time there is a number and a dash above the code, for example 49–, that will mean that the letter is for you only, that it is in invisible ink and the code is meaningless. If there is a full-stop or a colon, 49: that means that there is a code up to the first big full-stop; the rest means nothing and there will be invisible ink. If there is 49, that is, a dash underneath, then the letter will be for him; the code will mean nothing unless there is a fullstop or colon after the number. If there is handwriting after such a number, there will be invisible ink between the lines. It will be necessary to warn him about it. When you write to me in future it would be better to write in invisible ink between the lines of a code which means nothing, because they can find out the code here [in Brussels]. In that case there will have to be a dash after or under the code and no full-stop after to let me know. It will be necessary to number letters exactly to make sure that none are lost.
Despite the headache provoked by this paragraph, one can understand that Fersen far preferred to correspond with Marie-Antoinette in invisible ink (‘en blanc’). A table clarifies these instructions.
The use of code and invisible ink by Marie-Antoinette and Fersen over several years underlines the strict secrecy that their correspondence required even before the Revolution; clearly their personal affairs were of a highly compromising nature.
Evelyn Farr is the author of Before the Deluge, which documents
the Parisian high life in the years leading up to the Revolution. While writing it she uncovered material that impelled her to shed more light and less hypocrisy on Marie-Antoinette’s love affair with the Swedish Count Axel von Fersen, the subject of her third book, Marie-Antoinette & Count Fersen – The Untold Love Story.
Evelyn’s latest book, I Love You Madly, is the complete edition of Marie-Antoinette’s correspondence with Axel von Fersen. Published first in French, the English edition is available to buy now.