Sittler's Commercial Codebook Used for Dutch Diplomatic Service

In the age of telegraphy, commercial codebooks were occasionally adopted by governments. Slater's codebook was used in the U.S. and Canada and Bolton's was used in Japan (see "Nonsecret Code: An Overview of Early Telegraph Codes").

Sittler's codebook in French was used by the Dutch for many years, as I learned in Jip Boer, "Only Gentlemen Read Each Other's Mail: Over 50 Years of Sittler Codebooks in the Dutch Diplomatic Service" (HistoCrypt 2026).

The earliest record of its use by the Dutch government is from 1878. Since the codebooks came in many versions, larger consulates as in Constantinople had to keep many books. (p.147) Even before the First World War, there were reports that the code was compromised. (For example, Bazeries demonstrated its breaking, as I described in "How to Break a Code (Not a Cipher)").

But only partial improvements were made such as "adding nulls, reversing the code when written down, using multiple sets of paginations or adding a certain number to specific groups of numbers based on a chosen lawbook or day of the year" (p.147).

The "nonchalance" about security risk was not due to incompetence but because they were not much concerned about their communications being read by foreign governments. Their aim was to avoid the prying eyes of subordinate officials and telegraph personnel regarded as less trustworthy. (p.147-150)

In July 1915, an additional security measure involving superencipherment (initially called cijferstammen (cipherstems)) was introduced for "secret" communication. (p.149, Figure 2) Later, an updated system cijferstaten (cipher tables) was introduced (p.150-151, Figure 3). In 1920, each page was given an additional page number written in black pencil unique to each book, besides regular page numbers in blue pencil. (p.151)

Early in the 1920s, a new mode of using the Sittler codebook, called "Babibo-code", was introduced, whereby, apparently, code numbers were transformed into letter bigrams (p.153).

It was only in 1931 that the Sittler codebook started to be phased out by new Dutch-language codebooks (p.153-154).

 

What Happened to Code when France was Occupied by Nazi Germany?

What happens to secret codes when a nation comes under control of another? I once mentioned Stephan (2022) about German Reunification in 1990. Now I learned about France under control of Nazi Germany in André Falut, "Encrypted official telegrams of the Vichy government" (HistoCrypt 2026).

A victory of Germany early in WWII led to an armistice in June 1940, whereby a major part of France was occupied by Germany. A French government in Vichy was allowed to govern the Zone libre south to the demarcation line.

In wireless telegraphy (télégraphie sans fil, "TSF"), the French officials were required to use encryption methods that the German authorities can decrypt. Telegrams sent via cable were not subject to this condition, unless they crossed the demarcation line (p.156).

In July, a copy of Code des Colonies (and later, keys used with the code) were handed over. When the German authorities could not decrypt a transmission, the French had to provide the decryption with an explanation on the encryption method. (p.157)

Occasionally, it was still desired to use an undisclosed code. At one time, a telegram from Rabat (Morocco) to Dakar (Senegal) was to be sent by the Casablanca-Dakar cable but an interruption made it ncessary to send it wirelessly. Since French stations were not allowed, it was sent via Berne and Ongar (England). (p.157) Some extant telegrams bear a note to make sure to send them via cable. (p.158)

The request from the French to allow undisclosed methods in diplomatic transmissions seems to have been granted by October 1940, but a similar request for Vichy officials in their trips to Paris seems to have been rejected in April 1941. (p.157)

The semblance of independence of the Vichy government came to an end in November 1942, when the Germans invaded the south following the Allie's invasion of French North Africa. After this, the number of Vichy officials receiving encrypted telegrams was significantly reduced. (p.159)

 

Issues with Recent "Solution" of Zodiac Z13

Of the Zodiac ciphers, Z13 (thirteen characters long) is considered too short for cryptanalysis. Late last year, however, Alex Baber not only came up with a solution but got endorsement from three ex-NSA officers. I read the experts' analysis, "Zodiac Z13 Decryption" (somehow, this page ends in an out-of-memory error on my PC), and liked their fairness in admitting the proposed reading may be "one of many possible solutions" and there may be many other possibilities for the substitution/permutation keys. From such an attitude, it is hard to understand how they can say the solution is "compelling". They appear to have been convinced because they found the proposed transposition can be derived by using a relevant name as a key. But such a key may also be one of many possibilities, as pointed out in David Oranchak's recent video.

Now I uploaded an article to discuss issues with the solution for Japanese readers, "ゾディアックZ13の「解読」は本物か".

 

A Pigpen Cipher of Philip II's Agent (1577)

Martin de Acuña, an agent for Philip II, used a pigpen cipher in his letters to Philip II. The following is my reconstruction from the letter dated 27 January 1577, written on his way to Constantinople. (One can see the letters of the alphabet are arranged alphabetically among the three tables.)

 I learned this in Blog el Salón de Cris. Its source is:

Javier Marcos (2013-2014), "Sombras de Martín de Acuña 2: La Misión secreta de Martín de Acuña en el invierno de 1577, narrada por el mismo durante el viaje" (Archivo de la Fontera)

The paper notes one undeciphered passages in a letter from Mondéjar to Antonio Pérez dated 17 January 1577, reporting his doubts about Acuña's mission. If the same pigpen cipher is used, it will be readily deciphered with the above.

 For contemporary Spanish ciphers, see "Spanish Ciphers during the Reign of Philip II".

A Match Wanted for Another Cipher to Emperor Charles V

I added a mention of a letter from Hugo de Moncada, Viceroy of Naples to Emperor Charles V, dated 6 October 1524 (BNE, see Carta 9) in "Scholarly Studies on Ciphers in the Reign of Emperor Charles V". The marginal note in later hand indicates "Decifrada en ...," but I have not been able to locate the decipherment. If someone finds the deciphered text or a matching cipher or solves this, please let me know. 

 

Undecoded Armstrong's Letter (1808) Sent to Madison by Mistake

Further to the call for help to decode a letter from John Armstrong (20 February 1808) for Madison Papers in a post last year, I made two updates in "An Outlier Code in Armstrong-Madison Correspondence (1808)" (marked "June 2026").

(1) The editors of Madison Papers found evidence suggesting that the letter was not intended for Madison. It is then no wonder that this could not be read with Madison's known ciphers or other State Department ciphers. I wonder whether Armstrong used any ciphers other than the known one with Madison, perhaps a private cipher with the ambassador in England or Spain that was not stored in the State Department office.

Even if the letter is not meant for Madison, the editorial team is still very interested in the content of the letter. If it can be solved in the next year or two, they could probably include it in their next volume of Madison Papers.

(2) This letter is interesting in that it consists of not only numerical codes but also what seems to be shorthand. In this regard, my original article only mentioned Taylor's shorthand did not match. Now I added names of other shorthands known at the time. (I thank Norbert Biermann for the info.) If the code is too tough, hopefully, the shorthand portion may provide a clue.