French Army C-36: The pre-war period () ()

• 1. Introduction.
• 2. Machine operation
• 3. The pre-war period
• 4. The start of the war
• 5. The Vichy period
• 6. The Allied period
• 7. The post-war period
• 8. Conclusion
• Appendix

3.1 Historical background

3.2 Encryption in France – Beginning the history of the C-36

Consequently, from 1923, Givierge did everything possible to ensure that the French Army was equipped with cipher machines. Thus, in 1925, the Bureau du Chiffre, which he directed, carried out tests on a machine manufactured by the Carpentier company (cf. 3.3). It is interesting to view the beginning of the document that describes these tests because it demonstrates the efforts of the Bureau du Chiffre to equip the army with cipher machines (SHD-7N4235 1940):

     Ministry of War - Army Staff – Bureau du Chiffre - Secret - Paris, 
     December 13, 1925. Note for the 3e EMA Office [2] and the communications 
     service, under cover of the 4e direction.
 
     The Bureau du Chiffre has been studying cipher machines for more than 
     two years. Following the examination of foreign device patents, the 
     Bureau had drawn up a project which, in February 1923, gave rise, with 
     a view to its realization, to a whole correspondence with the possible 
     constructor establishments, of which a letter to the director of the 
     signal section. The Bureau du Chiffre had in mind an extremely portable 
     device, intended for encryption in small units. A test model was built 
     at Puteaux, but following procedures which, instead of seeking to 
     resolve the part of the problem that had been identified in the Bureau 
     du Chiffre, namely the printing of characters by a robust and light 
     device, solved the problem by adopting a heavy typewriter requiring 
     considerable electric force. Faced with this result, quite different 
     from its wishes, and the expenses incurred, the Bureau du Chiffre then 
     contented itself with ensuring that its key process, consisting of a cut 
     strip, produced a satisfactory result, and it handed over, at a more 
     favorable time from a financial point of view, the continuation of the 
     project realization...

     The Head of the Bureau du Chiffre: Givierge.

     Ministry of War, Engineering, Directorate of Military Telegraphy 
     Equipment, to the Minister of War. Paris on March 1, 1934. Report 
     by Colonel Julien.
     ... especially in the German Army, the use of cryptographic machines 
     has been pushed up to battalion level [3]. … In summary, the Colonel 
     Director of Military Telegraphy Equipment has the honor of presenting 
     the following proposals:
     1) Adopt the cryptographic machine studied by the Bureau du Chiffre[4].  
     Determine the endowments to be foreseen for the EM [5] and the signal 
     services of the large units ...
     2) Examine the advisability of providing the units within the division 
     with simple and space-saving cipher machines and, if so, prescribe the 
     study by the Bureau du Chiffre of a type of machine that meets these 
     conditions.

The staff then asks Mr. Hagelin for the impossible: a robust machine that prints the cryptogram or the plain text. The machine is intended to equip small units such as divisions and battalions. Ultimately, the machine must be extremely small: It must be able to be carried in the pocket of a soldier's coat. Mr. Hagelin himself tells the story of the machine he then created (Hagelin and Kahn 1994): Mr. Hagelin cut a block of wood to estimate the dimensions of the machine to ensure that it would fit in a soldier's pocket.

He created a first prototype, taking inspiration from a coin changing machine he had previously created. This changer consisted of a drum with bars around its circumference. Keys spun the drum. The bars could move to the left when the drum was rotated. The bars, in turn, caused a typewheel to turn with as many steps as the bars moved. Mr. Hagelin replaced the keys with the pin wheels from the B-21/B-211. The typewheel printed letters instead of numbers. The C-35 was born.

Then, Mr. Hagelin made some modifications. The machine was fixed on a tray and a protective cover was added. The underside of the table was rounded so that the machine could be fixed on the operator's knee. This machine was named the C-36. The letter C designated the technology used (drum and pin wheels) and the number indicated the model year (35 = 1935, 36 = 1936).

In 1935, Hagelin supplied the French Army with some C-35 machines. They were tested during maneuvers and the staff was delighted (SHD-7N4235 1940). These machines finally solved the army's encryption problems.

In 1936, after new maneuvers confirmed the importance of mechanical encryption for small units, the head of the Bureau du Chiffre proposed the purchase of 3,000 small machines, which corresponds to 30 machines per division, assuming that France could align 100 divisions in the event of war (SHD-7N4235 1940)[6]:

    Army Staff – Bureau du Chiffre - Subject: Cryptographic machines.
    SECRET - January 17, 1936. Note for the General-in-Chief of the 
    General Staff of the Army.

    - I - Mechanical encryption in small units
    ... Mechanical encryption ... automatically provides secure, fast, 
    and flexible encryption. Commissioned up to battalion level for 
    exercises in motorized divisions (September 3 to 8, 1935) as well 
    as for combat exercises at Mailly camp (September 20 to 25, 1935). 
    Cryptographic machines have given complete satisfaction. They were 
    extremely favorably received in the subordinate units whose non-
    specialized personnel succeeded in encrypting telegrams correctly 
    and quickly, reaching a rate of 70 letters per minute.

    - II - Needs
    A / Portable machine needs
    ... As an example, the distribution would be as follows for a 
    normal type infantry division:
    Status of the DI: 2 machines; Staff of the ID: 1 machine; AD 
    Staff: 1 machine; 1 ° Infantry Regiment: 4 machines; 2 ° "": 
    4 machines; 3 ° "": 4 machines; Field Artillery Regiment: 
    4 machines; Heavy Artillery Regiment: 3 machines; GRD [7]: 1 machine; 
    Engineer Battalion: 1 machine; Balloonists company: 1 machine; 
    Reserve: 4 machines. Total: 30 machines
    For a Cavalry division, there would be a slightly higher endowment 
    of 30 machines with 5 in reserve, i.e.,: 35 machines.
    B / Requirements for large model machines ...
    C / Requirements for maintenance and troubleshooting
    ... In summary, the military currently needs 51 large machines 
    and 3,000 portable machines …


    - III - EMERGENCY ORDER
    ... Appropriations to be entered in the 1937 budget ... It would 
    be appropriate to enter in the 1937 budget
    a) 918,000 francs corresponding to the purchase of 51 large machines.
    b) 3,000,000 francs corresponding to the purchase of 1,000 small 
    machines. The same sum being planned for each of the years 1938 and 
    1939. The completion of the purchase of 3,000 small machines would 
    thus be completed at the end of 1939 ....
    Lt-Colonel De France de Tersant, Head of the Bureau du Chiffre of 
    the Army Staff.

    The Cpt RONON of the EMA (3rd office) to the General Cdt of the 
    troops and communication services. - February 16, 1937 ...
    In 1938, it was decided to order 1,000 small model ciphering 
    machines similar to those that were tested in the maneuvers of 1936.

    The advantages of this device are multiple, in particular, it allows
    - the maintenance of almost absolute secrecy even if a machine 
    falls into the hands of the enemy.
    - the authorities who exchange messages to be assured of the 
    faithful transcription of their text. The use of codes encourages, 
    on the contrary, to distortions of expressions which affect the 
    accuracy of the information..
    - the number of people assigned to encryption to be reduced to 
    a minimum (only one operator per machine).
    - messages to be exchanged between any authorities. Operators have 
    no special training and no knowledge of any general or specific 
    code. …

Footnotes:

• [1] Givièrge headed the Bureau du Chiffre of GHQ during World War 1 with the rank of captain. After the war, he continued to lead the Bureau du Chiffre with the rank of colonel. In 1931, he was appointed general.
• [2] EMA: "État-major de l’armée" = Armed Forces Staff.
• [3] The French are aware of German cryptographic equipment (including the use of Siemens teleprinter), no doubt thanks to the spy Hans Thilo Schmidt.
• [4] The machine to be adopted is the B-211.
• [5] EM: État-major = military staff, general staff, staff.
• [6] Although its name does not appear in the report, it is obviously the C-36.
• [7] GRD: “Groupe de Reconnaissance Divisionnaire”: divisional reconnaissance group.
• [8] 3000 fr. in 1938 corresponds to $2500 currently. This comparison is misleading. Indeed, the standard of living in France at the time was extremely low. The monthly salary of a worker at the time was only 840 fr. In addition, the encryption machines were produced in small series and mainly by hand. In contrast, the M-209 was manufactured industrially, which made it possible to significantly reduce its manufacturing costs.

3.3 Has the C-36 been tested?

We know that in the US, Mr. Friedman tested several machines, notably Mr. Hebern's machine. This machine was then almost accepted by the U.S. Navy in 1924. However, following Mr. Friedman's codebreaking, the purchase of this machine was abandoned.

No evidence that such a test was carried out for the C-36 was found at SHD, but it is likely that it did take place. Indeed, such a test was carried out in December 1925 to prove the security of a cipher machine (which made teletype) created by Mr. Carpentier. Commander Bassières succeeded in deciphering the test messages. The project was consequently abandoned (SHD-7N4235 1940, cf. 3.2). At the time, it was Colonel Givierge who headed the Bureau du Chiffre.

3.4 German cryptanalysis

• Hebern 5-rotors, The 1924 Navy Challenge, (link).