Two encrypted newspaper advertisements from 1875 puzzle cryptologists. Is the first ad the key to the second?

The book The Agony Column Codes & Ciphers by Jean Palmer (i.e., Tony Gaffney, a reader of this blog) is a treasure trove for everybody interested in codebreaking. This work presents over a thousand encrypted newspaper ads from Victorian England. While Tony has solved many of these cryptograms, some still wait to be deciphered. On Klausis Krypto Kolumne I have written about, some of these.

Many of these encrypted newspaper ads contain love letters. Others were placed by business people in order to provide certain information to partners and customers. Ignatius Pollaky, a successful private investigator of the Victorian era, communicated via newspaper ads, too.


Two ads from 1875

One of the most puzzling encrypted newspaper ad series described in Tony’s book was published in the Evening Standard in 1875.


This series consists of two ads. The first one was contained in the May 8, 1875 issue:

W. Str 53. Catokwacopa. Olcabrokorlested. Coomemega. Sesipyyocashostikr. Rep. – Itedconlec mistrl. – Hfsclam 54, 3 caselcluchozamot. 1. 6. 9. Mopredisco. Contoladsemot. Iadfilisat. Qft. Cagap. Balmnopsemsov. Ap. 139. – Hodsam 55, 6. Iopotonrogfimsecharsenr. Tolshr. Itedjolec. mistrl. – Ding Declon. Ereflodbr.

Twelve days later, on May 20, a second ad of the same style was published:

W. – Umem 18. Poayatlgerty. Dpeatcnrftin. Nvtinrdn. Dmlurpinrtrcamur. Etd. – Atndngtnsurs. Otenpu. – Eftdorshpxn. 18. Ndtsfindseseo. Cotegr Tavlysdinlge. Ngtndusdcndo. Edrstneirs. Ui, Ndted. iolapstedtioc. A. P. 138. – Yxn. 18. 18. Wtubrfftrstendinhofsvmnr. Dily. – Atdwtsurs. Oatvpu. – Y Arati. Rileohmae. – This will be intelligible if read in connection with my communication published in this column on the 8th inst.

Note that the last sentence of the second ad is in the clear. If it is correct, the first ad could be the key that is necessary to decrypt the second (or the other way round).

I published a first blog post (in German) about these ciphertexts (I call them Catokwacopa cryptograms because of the first word in the first ad) in 2015. There were a couple of comments, but nobody came up with a solution.


A transposition cipher?

The Catokwacopa cryptograms look different from most other ads in Tony’s book. Words like “Catokwacopa” and “Olcabrokorlested” are pronounceable, which is unusual for most encryption methods. My first impression was that the number of vowels is higher than usual in a ciphertext, which is evidence for a transposition cipher (encrypting it with a substitution cipher typically lowers the number of vowels in a text). To check whether my suspicion was correct, I performed a frequency analysis with CrypTool 2. Here’s the result:


This frequency distribution is consistent with an ordinary English text. This makes it very likely that we deal with a transposition cipher here.

The question is now what kind of transposition the author of these ads used. My guess is that the two ads need to be mixed somehow (e.g., letter 1 from ad 1, letter 1 from ad 2, letter 2 from ad 1, letter 2 from ad 2, …). However, I haven’t found a mixing rule that makes sense, so far.

Can a reader find out more?

Further reading: The Top 50 unsolved encrypted messages: 30. The Harry-Caroline and the Tissie-Jabber messages


Subscribe to Blog via Email

Gib Deine E-Mail-Adresse an, um diesen Blog zu abonnieren und Benachrichtigungen über neue Beiträge via E-Mail zu erhalten.

Kommentare (12)

  1. #1 Lance Estes
    United States
    27. Januar 2018

    What strikes me are the similarities between ad 1 and ad 2 in a couple of specific spots, and the overall feeling that these may be very similar plaintext.

    In ad 1, “Itedconlec mistrl.” and “Itedjolec. mistrl.” are quite similar and in ad 2 in roughly the same spots we see “Atndngtnsurs. Otenpu” and “Atdwtsurs. Oatvpu.”

    Similar beginnings and endings in the pairs, and an exact match in the case of “mistrl”

  2. #2 seb
    27. Januar 2018

    gibts das auch irgendwo in deutsch?

  3. #3 Thomas
    27. Januar 2018

    The two messages are very similar regarding their construction. Maybe the corresponding words of both messages must be combined and anagrammed. E.g. line 4: Ding + y = dying, Declon + Arati = declaration.

  4. #4 Klaus Schmeh
    27. Januar 2018

    >gibts das auch irgendwo in deutsch?
    Bis Herbst 2016 habe ich auf Deutsch gebloggt, dann auf Englisch umgestellt. Leider schaffe ich es nicht, den Blog zweisprachig zu betreiben.

  5. #5 Klaus Schmeh
    27. Januar 2018

    Richard SantaColoma via Facebook:
    “Catokwacopa, east of Java”

  6. #6 Thomas
    27. Januar 2018

    In English texts Q is (nearly) ever followed by U. In line 3 there are corresponding letter groups Qft and Ui. Since “quift” is no word we can expect (at least in 1875), there must be yet another transposition pattern. But the Q and this U should belong together.

  7. #7 Thomas
    27. Januar 2018

    Statistics: As Klaus has pointed out, the letter frequency distribution of the two messages put together matches – more or less – that of English plaintext. But if we look at the messages separately, the first one has strikingly too much Os, the second one too much Ns and Ts. This might have to do with the fact that certain letters occur more frequently at certain positions in an word.

  8. #8 Klaus Schmeh
    28. Januar 2018

    Bart Wenmeckers
    Its interesting that both cipher texts have similar layout/format..
    Bart Wenmeckers via Facebook:
    My interpretation of “This will be intelligible if read in connection with my communication published in this column on the 8th inst” is that these are independent ciphers/messages.

    Anyone else have any ideas?

  9. #9 JA
    30. Januar 2018

    Thomas has it. The third and second to last words are DYING DECLARATION. The messages have to be matched together and transposed.

  10. #10 Lance Estes
    United States
    31. Januar 2018

    I liked the idea Thomas proposed for combining Ding and Y for DYING and Declon and Arati for DECLARATION. However, as many have probably seen when they tried to do that with other pairs, it doesn’t seem to work out.

    I decided to take a closer look at a couple of specific pairs, Itedconlec/Atndngtnsurs and Itedjolec/Atdwtsurs. Based on common letters and the idea of combining the two parts I divided these into two similar “sentences”: ITEDATND CONNGTN LECSURS and ITEDATD JOWT LECSURS.

    One thought I had was that LECSURS might be LECTURES? Yeah, thats rather far fetched given the perfectly spelled DYING DECLARATIONS .

    And what about the first word of my sentences? If we are allowed to skip between the letters of ad 1 and ad 2 (as long as we keep them in order) I could make IATTENDD and IATTEDD. I ATTENDED perhaps, with a few letters missing? Crazy I know, but bear with me.

    Perhaps we have I ATTENDED ____ LECTURES? OK, so can we take the middle words of my sentences, CONNGTN and JOWT, and make words?

    Would you believe CONINGTON and JOWETT? John Conington and Benjamin Jowett were both working for the University of Oxford in the mid-1850s. (That may be quite interesting later.)

    CONNGTN also appears when you take the following pairs from the ads: Contoladsemot/Ngtndusdcndo . Might that read CONNGTN TOLD US ADD SECND MOTO (Conington told us add second motto)?

    And how about all of those numbers? If you consider the 18’s (which all occur in ad 2) three of them are in the right positions to match with 53, 54, and 55 in ad 1. Could the overall content be as simple as describing things of interest that happened to this person in the years 1853, 1854, and 1855? And if that answer is yes, it might be quite reasonable that they attended lectures by Conington and Jowett at Oxford.

    Other possible pairs and what they might say follow

    Catokwacopa/Poayatlgerty (“cap took away at colge party”)

    Coomemega/Nvtinrdn (convo “met me in garden”)

    Hfsclam/Eftdorshpxn (do you see “sclorship”?)

    caselcluchozamot/Ndtsfindseseo. (“candt” is seen later, and “motto” is there at the end)

    Mopredisco/Tavlysdinlge (something “in colge”)

    Iadfilisat/Edrstneirs (“i aded first line” satirs)

    Cagap/Ndted (There’s “candt” again)

    So its a fairly circumstantial case at this point, but perhaps others can be successfully push this further. (And many thanks to my friends that helped push this along to this point!)

  11. #11 JA
    31. Januar 2018

    Balliol is mentioned (Jowett’s college) “Balliol man posted…”, “Said simply your cap…”, “”old cap broke at…”

  12. #12 Dave
    31. Januar 2018

    I obtained some potential partial solutions very similar to much of the above by combining the two messages:
    Str 53/Umem 18 (Sum term 1853) – Summer Term 1853
    Catokwacopa/Poayatlgerty (Cap took away at … party)
    Olcabrokorlested/Dpeatcnrftin (Old cap broke at corn left instead) corn. – corner?
    Coomemega/Nvtinrden (… met me in garden)
    Sesipyyocashostikr/Dmlurpinrtrcamur (Sed simply your cap in shorts trick … )
    Rep/Etd (Repetd) – Repeated?
    Itedconlec/Atndngtnsurs (I attended … )
    Mistrl/Otenpu (?) Mister O.Tenpu?
    Ladfilisat/Edrstneirs (Laded first line is arts)
    Tolshr/Dily (Told shirly) – Shirley?
    Balmnopsemsov/Iolapstedtioc (Baliol man posted … ) last word could be a surname: something like ‘Semsovic’ or could be a slightly jumbled/corrupted ‘semiotics’
    There certainly seems to be a theme there. I tried the same method on the rest of the message but lost the will when the results were even more fragmentary. This might be due to so many of the plaintext words being abbreviated or because they are surnames.