Last week I presented the mystery of an encrypted note found in an antique silk dress. Has the codebook used now been found?

A few years ago I threw the so-called silk dress cryptogram to my readers to solve. According to Nick Pelling, this happened “not unlike a juicy bone to a pack of wolves. However, their voracious code-breaking teeth – normally reliable enough for enciphered postcards and the like – seemed not to gain any grip on this silk dress cipher. So… what is going on here? Why can’t we just shatter its cryptographic shell, like the brittle antique walnut it ought by all rights to be?”

What a nice collection of comparisons and metaphores!

Last week I covered the silk dress cryptogram in my series of the top 50 unsolved ciphers. In addition, I published a help enquiry in the Yahoo discussion group Telegraphic Codes (thanks to Karsten Hansky for telling me about this forum).

Silkdress-1  Silkdress-2


The Slater Telegraph Code

It is clear that the silk dress cryptogram was created with some kind of codebook. This codebook somehow links every frequent word of the English (?) language to a codeword (in this case “Cairo”, “Greenbay”, “Calgary”, …). It is well possible that this linking is done in several steps, e.g., each English word is mapped to a number, which is itself assigned to a codeword based on a key-dependent table. As a rule, an encrypted text like the silk dress cryptogram can only be deciphered if the codebook is known.

Much to my joy, John McVey, a member of the Telegraphic Codes discussion group, recently wrote that he had identified the codebook that was used for the silk dress cryptogram: the Telegraphic Code, to Ensure Secresy in the Transmission of Telegrams by Robert Slater (“Slater code”).

Fred Brandes, another member of the discussion group, wrote something similar: “I suspect the code words are from Slater. Do not know how it was used in that instance, however. These messages are similar to a number of telegrams that I have collected involving trades between stock brokers and banks.” Here’s a page of the Slater codebook:


As you can see, each Slater code group is a five-digit number. It was common practice to change the code groups in a key-dependent way before use (e.g., add the current day of the month to every number) to make it more secure.


Open questions

John McVey and Fred Brandes seem to be experienced codebook experts. However, it is not yet clear to me how they determined that the Slater code is the one used here. Maybe a reader can give me the answer. In addition, I would like to know how it was used.

In his blog post, Nick Pelling presents a few thoughts about the origin of the silk dress. According to a label fixed on the cloth, it be longed to somebody named Bennett.


Nick has found a plausible-sounding candidate named Margaret J. Bennett, a woman from Baltimore, MD, who died in 1900. It should be possible to find out more about her. Maybe this will lead to additional insights about the cipher.

Further reading: The Top 50 unsolved encrypted messages: 47. Encrypted messages of a Nazi spy


Subscribe to Blog via Email

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

Kommentare (10)

  1. #1 Thomas
    25. Mai 2017

    A guess: They came upon Slater’s code because of the meaningless text. Following Slater’s method the plaintext words are converted into numbers. After that a certain number is added or subtracted and the results are reconverted into words using again the ‘dictionary’, the outcome will be meaningless text.
    As long as the calculation isn’t known, the codetext cannot be decrypted by hand. The only starting point might be that the distance between two successive codetext words in the codebook (i.e. the difference between their codenumbers) and the distance between the two corresponding plaintext words must be the same. But shifting all distances through the whole ‘dictionary’ until this yields meaningful text would be far too much stuff for pencil and paper.

  2. #2 Nick Pelling
    25. Mai 2017

    As I wrote in my post, because the silk dress cipher’s words differ from column to column, it cannot be a “flat” telegraphic code in the way such codes were normally used. And because the starting letters and range of codewords is so constrained, it also seems to point away from Slater’s telegraphic cypher.

    In short, the silk dress cipher is not just a cipher, it’s a cipher mystery. 😉

  3. #3 Thomas Ernst
    26. Mai 2017

    I commented on the place names and adjectives etc. of this back in 2013 or so, and had contacted the British expert on cryptolects and argot, Julie Coleman. Coleman couldn’t identify this as an argot. Ensued a bit of discussion on this topic back then. Speaking from memory: I maintained that this was not a cryptogram at all, but the texts of true telegrams. The dress probably belonged to the novelist, playwright and clothes-horse Frances Hodgson Burnett (1849 – 1924): two pieces of paper, in this case discarded telegram texts, were crumpled up to stuff out part of one her dresses, which might have been auctioned off for a charitable purpose after her death. A certain T. C. Burnett was on the board of the “South Carolina Cottonseed Oil”; things like “none event”, “null event”, “new event”, “forbade event”, “novice event”, “ginned event”, in conjunction with the many city names, in my opinion do constitute telegram-trade-argot, whether “style nanny” refers to textiles, or booze. Back then, no-one followed up on my suggestions regarding the specific town names, and their mis-spellings (“Garry” for Gary, Indiana; Calgarry for Calgary (Alberta, Canada), Manitoba (“Minnedos[a]), Green Bay, WI, “Helena” (Alabama or Montana), “Assin” short for “Ascension”, etc. Rotwelsch, argot, not written by the spelling bee herself – not a cryptogram. Only because these sheets were found in a hidden place doesn’t make them cryptograms. They are old, discarded telegram texts, used for stuffing an author’s dress, to be auctioned off for a “good cause” (probably more 90 proof)!

  4. #4 Thomas Ernst
    26. Mai 2017

    PS: just noticed “Bennett”, but could swear the original post provided the name – posted by the curator (Boston, if I remember correctly??) or someone else – as “Burnett”. Bennett/Burnett – I maintain that the dress was the focus, puffed up with left over paper, in this case two telegram texts, the time sent noted in the left margin, the transmitted text checked off line by line.

  5. #5 Rich SantaColoma
    26. Mai 2017

    I’m still personally unconvinced this is from any telegraphic codebook, for several reasons: It is a list, the items of which are checked off. There are names on the list, and sewing or tailoring terms, too.

    I still favor that it is simply a list of tailoring/sewing jobs, with the work needed, the customer names, the styles and colors chosen, for a seamstress or tailor to keep track of the work at hand.

    And it just got accidently scooped up and sewn in the liner one day, to confound everyone many years later.

  6. #6 Thomas
    26. Mai 2017

    There isn’t much known about Margaret J. Bennett:
    She was born in 1823 as Margaret J. Patterson. In 1845 she married Francis W Bennett (1820 – 1880), an auctioneer in Baltimore. She passed away in 1890. According to an article in the Baltimore Sun from 1999, a portrait of her in a ‘Quakerish dress’ (not exactly the style of Sara’s dress) hangs in the Margaret Bennet Home in Baltimore, for which she left $150,000 to start a refuge for “homeless, needy and deserving female persons.”

  7. #7 Thomas Ernst
    29. Mai 2017

    Completely agree with Rich that this constitutes neither code nor cryptogram. Whether textiles were implied, or sth. else, remains open. I mentioned the “South Carolina Cottonseed Oil Co” because in an article of a SC paper of ’24 – can’t find my source anymore, it was online – comparable lingo was used.

  8. #8 Thomas
    27. Oktober 2017

    In my opinion the underlying code is “The Anglo-American telegraphic code to cheapen telegraphy and to furnish a complete cypher”:;view=1up;seq=9;skin=mobile
    This codebook contains a great number of the words used in the silk dress cryptogram. Some examples: “event” at the end of some lines means: “You must explain.” “rubric listed” means “return as soon, the terms of loan are unsatisfactory” , “forbade” stands for “fright charges are light”, “lining” for “if the loan can be made”. “barometer” stands for the ” St. Pauls & Duluth railroad”. “ink” means “You must insist upon”, “malay”: ” I/we will manage”. “mammon”: “accomplished by good management”, “layman” means “lessee”. Apparently the cryptogram is the text of a telegram which was part of a businessman’s correspondence.
    Maybe another edition of this code (Hathitrust provides only that of 1894) could reveal more codewords.

  9. #9 Thomas
    28. Oktober 2017

    There are some gaps to be filled 🙁
    (Words not found in capital letters)

    SMITH – I/we have notified him – LINNET – not going (to) – NONE – you must explain

    ANTONIO – return as soon – the terms of loan are unsatisfactory – FULL – you must insist upon

    MAKE – (here I don´t read ONAGER but SNAPLES or something like that, whatever this means) – St. Pauls & Duluth railroad (preferred stock) – NERITE (NERITITE would stand for: „the next one“)

    SPRING – with haste – if the loan can be made – ONE – can you mortgage real estate – NOVICE – he/they cannot arrange with the bank

    VICKSBG – responsible – raise as much as possible – if the loan can be made – MY – NANNY – BUCKET

    SAINTS – week – LUNAR – I/we will manage – who will be nominated – MARKETS – he/they cannot arrange with the bank

    SEAWOTH (or „LEAWOTH“) – may be able to – LEMON – SUNK – early

    CAIRO – the Columbia river – if the loan can be made – who will be nominated – JOHNSON – NONE – ICE (ICED stands for „to be identified“)

    MISSOURI – withdraw offer immediately – LUNAR – who will be nominated – JOHNSON – NONE – BUCKET

    CELLIETTE – it is a very important matter (i don´t read MEMORISE but REMORSE which stands for „to what do you refer“ – LEGACY – SUNK – DEW

    CONCORDIA – accomplished by good management – lessee – NULL – you must explain

    CONCORDIA – MIRACULOUS (is „MIRACULOUS“ right?) – LUMINOUS (maybe „HUMUS“?) – name if you can – MENU – a decisive battle
    BISMARCK – OMIT – who will be the lessee – they are buying – BANK

    PAUL – steady rates – if you can look into it – you must explain – I/we think he is likely to fail – who will be nominated – you must explain

    HELENA – do not omit to – LOFO – USUAL – early
    GREEN BAY – NOBBY – PIPED (PIPE stands for „please accept“)

    ASSIN – will be pleased to see you at my office – LEAGUE – who will be nominated – the freight charges are light – you must explain

    CUSIN (?) – draw at seven day´s sight –

    HARRY – NOUN – LERTAL – LAWFUL – was/were paid – NOVICE – you must explain

    MIMEDOS (or MINNESOTA?) – NOUN – JAMMY – who will be the lessee – before anything else – will be distributed


    KNIT – I am /we are wrong – HUGS – DUCK – FAGAN – early

    CALGARRY – NOUN – SIGNOR – if you can look into it – who will be nominated – GINNED – you must explain

    LANDING – NOUN – REGINA – LEGACY – SUCH (? there is a „d“ at the beginning) – I/we will give bail – ICE (ICED stands for „to be identified“)

  10. […] Two questions still remain unanswered – who was the other injured person walking near the car? And finally, what does “YOG’TZE” mean? […]