Here’s an telegram from 1952 that might have been encrypted in a code. Can a reader find out which codebook has been used and decipher the message?

I have found another interesting crypto mystery on Reddit. This time it’s an encrypted telegram from 1952.


The telegram

As can be seen on the following scan, the telegram in question bears the logo of the Indian Posts and Telegraphs Department:

Source: Reddit

The telegraph form is dated 1952/9/2. Apparently, this dispatch was sent from London to Simla (today spelled Shimla), a city in the north of India. Between 1867 and 1948, Shimla served as the summer capital of the British in India. Even after 1948, when India had become independent, Simla remained an important political center.


The code

The text of the said telegram mainly consists of ten-letter words such as YVUDZITFOV, ALPELUXSUM, or PLOEXGYYUT. My guess is that these expressions are codewords taken from a codebook. They don’t look random enough to represent text encrypted in a MASC or polyalphabetic cipher.

Two words, MOUIN and TRAGIC, have only five letters, which suggests that the codewords actually consist of five letters and that they were noted pairwise without a blanc in the middle.

The word LOWMAN, which appears twice in the text, doesn’t fit into the five-letter scheme. It might be a name or a codename of something or someone not listed in the codebook.

I wonder what RETANDORUM in the first line means. Is it just another codeword, perhaps assembled from the five-letter words RETAN and DORUM, or is it a Latin expression (e.g., genitive plural of RETANDUS). To my regret, I couldn’t find this word on Google.

Telegrams encrypted in a codebook code are nothing unusual. Governments, military organisations, and diplomatic services usually had their own codebooks, which were, of course, kept secret. The following scan shows a codebook page (it’s certainly not the one used here):

Source: PD

The only way to decipher a ciphertext created with a code is usually to find the codebook that has been used. The two best sources for this purpose are the websites of Satoshi Tomokiyo and John McVey. Can a user identify the codebook used for the London-Simla telegram?

Another interesting question is how the codewords of the code used here were created. All these words are pronounceable, for instance AVIUSHUNYD. The vowel-to-consonant ratio is probably around 40 percent, like in a natural language. Anyway, contrary to other codes, these words are apparently not taken from a real language. Is there a method to generate codewords of this kind?

As always, I’m looking forward to interesting comments.

Further reading: An encrypted telegram about the sales of the Danish West-Indies


Subscribe to Blog via Email

Enter your email address to subscribe to this blog and receive notifications of new posts by email.

Kommentare (18)

  1. #1 Richard SantaColoma
    30. April 2020

    I agree with your assessment this is probably a book code, and also that the ten letter code words are actually five and five.

    Another clue for that may be that the penciled check-marks come at or near the end of each five letter group. Perhaps this is because the decoder didn’t want to lose their place as they referred to the code book.


  2. #2 Thomas
    1. Mai 2020

    Apparently the telegram is from Sept 2nd, 1930, see the date stamp ‘2/9/30′ (bottom). ’22/52’ in the upper part is the time of day: 22 h(ours), 52 m(inutes). ‘Lowmans tragic’ is plaintext and seems to refer to the assassination of the Inspector General of Police in Bengal, Francis Lowman on Aug 29, 1930 ( see and As for the code groups, I agree to Richard’s idea that the ten letter groups consist of two five-letter groups (such as e.g. in the 6th ed. of the ABC-code from 1920). Since the group ‘ALPEL’ (occuring twice) follows ‘Lowmans tragic’, it might stand for ‘death’.

  3. #3 Thomas
    1. Mai 2020

    Other frequently used five-letter codes were ACME and Bentley, unfortunately there seem to be no complete copies available online.

  4. #4 Norbert
    1. Mai 2020

    @Thomas: “Bentley’s complete phrase code” is available through the internet archive:

    Unfortunately, no “ALPEL” code there, nor a pronouncable anagram of it, as far as I can see. “Death” is “ebhos” which doesn’t match the pattern of ALPEL.

    The “Acme Commodity and Phrase Code” would be most interesting in this case, but I didn’t find it online either.

  5. #5 Thomas
    1. Mai 2020

    Norbert: Good find, unfortunately to no avail. Another five letter code was the ‘Peterson’ mentioned by John McVey in his solution of a telegram from 1948: Couldn’t find it online either.

    Bentley’s seconde phrase code from 1929 has a code for ‘murder’ starting with ‘mo…’. In our 1930 telegram plaintext ‘Lowman’ is preceded by ‘MOUIN’, could that mean something like ‘murdered’? Judging from the context, the telegram should contain a code for ‘Dakka’ , maybe ‘hospital’, where the murder took place. Repeated groups are JEFJO and INIOY, maybe one of them meaning ‘stop’ ? But that are long shots. Hope a telegraph code expert can help.

  6. #6 Gerd
    1. Mai 2020

    Note that ALPEL has a checkmark next to it, and the following UXSUM is underlined, so maybe UXSUM was not decodeable.

  7. #7 Thomas
    1. Mai 2020

    Found the plaintext sheet in the National Archives of India:

    From the Secy, of State, London, To the Secy. to the Govt. of India, Home Dept.
    No. 2800, dated the 1st September 1930


    ‘Your telegram dated 31st August. If Mrs. Lowman is in India please convey expression of my sympathy to her and profound regret for loss service has sustained by Lowman’s tragic and untimely death. If she is not in India please telegraph her address if this can be ascertained.’

    The code is named ‘Word Code’ on the plaintext sheet, whatever this might be, maybe no known codebook used by the authorities.

  8. #8 Thomas
    1. Mai 2020

    And this is the Indian Word Code
    used here:

  9. #9 Thomas
    1. Mai 2020

    Gerd: You’re absolutely right: UXSUM couldn’t be deciphered at first. So in the decryption sheet the word ‘untimely’ was later added by hand whereas the rest was written with a typewriter. The reason is obvious: X and S were scrambled in the telegram, the correct codegroup for ‘untimely is ‘USXUM’

  10. #10 Richard SantaColoma
    1. Mai 2020

    Wow Thomas, great finds!

  11. #11 Norbert
    1. Mai 2020

    @Thomas: Unbelievable! Congratulations!

    (@Klaus: Bin derselbe Norbert wie vorher – ich probiere eine andere Email-Adresse aus, vielleicht klappen dann die Abos wieder.)

  12. #12 Klaus Schmeh
    1. Mai 2020

    Puzzled Catepillar via FaceBook:
    I think I may be able to help. Given the time it was written and distance it had to travel, telegrams were relatively expensive. A stealthy way around paying-per-word was to anagram in blocks of ten letters to keep matters obvious. This seems to be English: “MRE TOUR AND / MS. LIA / IV DUTY OF VZ …” May be the correct translation of the first three letter-sets. I can do the rest, I think — I just didn’t want to spoil everyone’s fun. (Hopefully I didn’t already…)

  13. #13 Klaus Schmeh
    1. Mai 2020

    Congratulations for solving this mystery! Great job!

  14. #14 Gerd
    1. Mai 2020

    Congratulations, Thomas, unbelieveable. Can you get us an idea how you could track down this to the national archive of india? Is this a public page that can be found via google?

  15. #15 Narga
    1. Mai 2020

    Amazing, Thomas! I’ve been impressed by how you find obscure stuff on the internet for this blog before, but wow!

  16. #16 Thomas
    2. Mai 2020

    The shadowed text on the telegram says that it is hold by the National Archives of India. Several Google books covering the assassination of Lowman cite “NAI 1930 file No. 497” or the like and its title “Revolutionary crimes in Bengal….”. Googling led to this website: Here you can register for a free account in order to obtain the NAI file containing 69 documents. The plaintext sheet is the one preceding the image of the telegram. (Since the telegram and the plaintext are so close together in the file, I wonder where the reddit poster asking for the plaintext had the image from.).

  17. #17 Norbert
    2. Mai 2020

    To answer the RETANDORUM question:
    RETANDORUM SIMLA appears to be just the cable address of the recipient.

  18. #18 Thomas
    2. Mai 2020

    To be precise: indicating the name both of sender ande adressee, RETANDORUM is explained here: (left page).