Klaus Schmeh
Aug 14, 2026

The Catokwacopa Cryptograms: A 150-Year-Old Mystery

In his book The Agony Column Codes & Ciphers, codebreaker Tony Gaffney presents more than a thousand encrypted newspaper advertisements from Victorian England.

Two mysterious messages

Many are love letters, and many have already been solved. Others remain mysteries. Two particularly intriguing examples appeared in the Evening Standard in May 1875, just twelve days apart. I call them the “Catokwacopa cryptograms”, after the first word of the first advertisement. Here’s the first one:

This is the second one:

The second advertisement ends with a passage in plain English:

This will be intelligible if read in connection with my communication published in this column on the 8th inst.

This is clearly a clue. The two advertisements are meant to be considered together.

A transposition cipher?

The ciphertext looks unusual. Words such as Catokwacopa, Olcabrokorlested, and Coomemega are pronounceable. That makes a simple substitution cipher seem unlikely. A frequency analysis provides another clue. The letter frequencies are remarkably similar to those of normal English. This is exactly what we would expect from a transposition cipher, where the letters are rearranged rather than replaced.

The obvious idea is to combine the two advertisements. Perhaps the letters have been interleaved: letter 1 from the first advertisement, followed by letter 1 from the second, then letter 2 from each, and so on. Another possibility is that the texts have to be arranged in columns and read in a different order.

Several readers of my blog have investigated this hypothesis, and it certainly seems promising. So far, however, nobody has found a complete solution.

Still unsolved

The Catokwacopa cryptograms are a particularly nice historical cryptographic mystery: a pair of short newspaper advertisements from 1875 that have resisted solution for more than 150 years.

Can a reader decipher these messages?

For the original advertisements, full transcriptions, frequency analysis, and further discussion of possible solutions, see the complete Catokwacopa cryptogram investigation on this website.

Back to Cipherbrain Blog

10 comments In Unsolved cryptogram

10 comments

Catherine Mary Darensbourg

August 15, 2026 at 6:46 pm

Dear Klaus, I believe it to be a possible Playfair variant because Wheatstone (born in 1802) received a distinguished honorary induction to the Institution of Civil Engineers in 1875 (I don’t know the exact month yet). Wheatstone was 73 years old, and then also passed away on October 19, 1875 of pneumonia in Paris.

Klaus Schmeh

August 16, 2026 at 9:54 am

>I believe it to be a possible Playfair variant
Are there Playfair variants that produce ciphertexts with letter frequencies similar to the English language?

Jozef Krajčovič

August 20, 2026 at 2:07 pm

I think I know how to solve this mystery.The cipher mechanism has been solved, but the complete, unique plaintext cannot be mathematically reconstructed because both ciphers are lossy projections with omitted letters….The most important new interpretation is:

I ADDED FIRST LINE [OF/IN] SATIRES. QUI FIT.

Qft / Ui forms QUI FIT—the incipit of Horace’s Satire 1.1.

This is also confirmed by Conington’s edition of Horace’s Satires.
In my opinion, this is most likely a notebook documenting Thomas William Jex-Blake’s studies at Oxford: Conington, the Hertford Competition, the years 1853–1856, and his classical exam results all point to this. The DNB confirms his studies under Conington and his results from 1853 and 1855. W. L. Newman remains the main alternative and a possible “Balliol man.”…

August 20, 2026 at 2:13 pm

The definitive or very strong readings include:

HERTFORD SCHOLARSHIP EXAMINATION
MOTTO WAS PREVIOUSLY DISCLOSED IN COLLEGE
CONINGTON TOLD US TO ADD A SECOND MOTTO
QUI FIT
DYING DECLARATION

August 22, 2026 at 8:52 am

Thanks, Jozef.
Does anybody know what the purpose of the message was?

Rich SantaColoma

August 20, 2026 at 3:56 pm

Clearly this is highly speculative, but maybe interesting? I put the first regular word of both messages together: catokwacopapoayatlgerty, and ran that through an online anagram solver.

The first time the word “cryptology” was among the few results. When I removed those letters from the original, and ran it through again, the word “attack” was on the next list. then after removing those letters, and anagramming the remaining ones, the only real word was “papa”.

Now like I said, very speculative, as I made choices along the way: What words to pick out from the lists, and so on. But I relate the results here if others also want to try this… as even though I doubt this IS an anagram, I think that given that the letter frequencies are near English, and then also that there are cipher systems which will preserve those found frequencies, perhaps this would be of interest or help to someone trying out said systems… maybe?

Or hot air, take your pick! Good luck all…

Gerry Haynaly

August 21, 2026 at 8:49 am

@Rich I think you are on the right track. The words before the last are DECLON and ARATI, simply leading to DECLARATION. But there must be somethin more, because one word before them are Y and DING leading to – what?

August 21, 2026 at 11:49 am

Ich habe gerade gesehen, dass Thomas Ernst schon 2018 im alten Blog das Rätsel per Anagramm-Theorie quasi gelöst hat: https://scienceblogs.de/klausis-krypto-kolumne/2018/01/26/revisited-the-catokwacopa-cryptograms-from-1875/

Thomas Ernst

August 23, 2026 at 1:02 am

Thank you, Gerry, but the main credit for the text reconstruction goes to two persons named Lance Estes and Dave (no last name). The obvious DYING DECLARATION pointed out by Thomas Bosbach in January 2018 got the ball rolling, and by late January Lance and Dave had established the following lexic inventory (alphabetic order): ADDED, ATTENDED, BALLIOL, CAP, COLLEGE, CONINGTON, FIRST, GARDEN, I, INSTEAD, JOWETT, LECTURES, LINE, ME, MET, MOTTO, OLD, REPEATED, SATIRS, SCHOLARSHIP, SHIRLEY, SUMMER, TERM TOLD, TOOK, 1853. Lance also read line [16] correctly: I AD[D]ED FIRST LINE SATIRS, but he did not connect Horace. – I discovered the thread in 201807 and added some freewheeling and negligible guesses about the text (some of them wrong: SUMMONER should be COMMONER) without completing it. – As known, W’s two cryptological core strategies are steganographic in nature: 1) he abbreviated plaintext words, 2) he distributed elements of the parcelled abbreviated plaintext words across two text blocks. Some types of abbreviations or letter omissions are: a) initial letter only: A[NNUAL] P[AY] (i. e. yearly costs for a commoner at Balliol), b) between beginning and end: EX[AMINATIO]N, CO[NINGT]ON, c) positional: omission of penultimate E (GARDN, SHIRLY), and double consonants: AD[D]ED, BAL[L]IOL, MOT[T]O. Sometimes these patterns overlap (COL[LE]GE, JOW[ET]T), or were semantically exclusive: AD[D]ED (penultimate E retained). ATTEND[E]D instead of A[T]TEND[E]D may be an authorial mistake. LECSURS for LECTURES appears to be spelled phonetically. Since the same spelling occurs both in lines 8 and 25, the first S cannot be a mistake for T. Apparently W pronounced /t ʃ / as /ʃ/. Likely this was sociodialectal. The TOOK/TAKEN option in [3] which I suggested in 2018 makes it possible that the young Oxfordian’s English may not have been perfect. Knowledge of Latin and Greek does not preclude fossilized sociolects in the native language. Vide Jowett’s defense of commoners of a poor social background. The syntax of [3] (unique inversion) and the following context suggest that W. did not take away (= win, remove, replace) the cap but that it was taken from him. – Some of W’s abbreviations are inconsistent: CO[NINGT]ON, CON[I]NGT[O]N.
No one followed up on three more substantial suggestions I made in 2018:
1) Not trusting transcriptions of ciphers, I crosschecked the texts currently available (in print since 2005, online since 2015) against the originals ads as printed in the “Standard”. I pointed out a dozen transcription mistakes. They were never corrected. “138” and “139” are both “138” in the “Standard”. These and other mistakes did not originate with W or with the “Standard” but with 2005 ff. transcriptions. See my 2018 posts for other textual corruptions.
2) For continuity and in order to facilitate cross reference and textual comparisons I arranged the texts of 18750508 and 18750520 diplomatically, i. e. numbered them by W’s original word divisions, and put the “interpreted” letters in square brackets, e. g. • [1] s tr/um em: SUM[ME]R TE[R]M • [2] 53/18: 1853 • [3] ca tok wa co pa/p o ay at lge rty: CAP TOOK (TAKEN?) AWAY AT COL[LE]GE PARTY […] • [11] 54, 3/18: 1853-4 […] • [24] tol shr/d ily: TOLD SHIRL[E]Y • [25] i ted jo lec/atd wt surs: I ATTE[N]D[E]D JOW[ET]T LEC[T]UR[E]S etc. Such an arrangement makes it easier for others to suggest and justify alternate abbreviations. It also highlights methodology: everyone intuitively reads [1] “s tr/um em” as plaintext SUMMER TERM, and the missing letters are the multiples E, M, R, but did W abbreviate SUM[MER] TERM, SUM[ME]R TE[R]M, SUMMER T[ERM], SUM[M]ER T[ER]M? W. did not write his text “intuitively” but used specific methods of abbreviation, and seemingly simple texts like line [1] require further scrutiny.
3) “W.” placed at least one more ad in the “Standard” of 1875. Like the other two, it was printed on p. 1, second column, in one of the upper text blocks. It is of a different make than the other two: it uses only 16 letters and appears to be a digraphic substitution, likely a split arrangement. It remains unsolved, and I suggested it be considered contextually together with the other two because it ads to the overall “cryptological portrait” of W. No one has done so.
Some post-2018 considerations: not sure about Jex-Blake. According to “DNB” 1927 – “ODNB” may have newer information – he was elected a fellow of Queen’s College in 1855, and in 1856 was ordained deacon at Oxford. I would also expect the “Standard”‘s “W.” to be an initial. My “candidate’s” name – Balliol’s youngest commoner in 1856, annual expenses of £138, with a lifelong affection for and later support of Jowett, begins with a “W”. But I have no Conington-link for him. Oxford’s “Historical Register” from 1888 lists neither him nor Jex-Blake as recipient of the yearly Hertford scholarship. Applicants were examined by three Oxford stooges during the Michaelmas-term (they gave speeches; hence the “mottos” may have been rhetorical topoi), but the HR lists only one winner for every year. – TOLD SHIRLEY: perhaps Walter Waddington Shirley (1828–1866) at Wadham, but cannot document.

August 23, 2026 at 9:22 pm

Correction: according to Foster’s “Oxford’s Men […]” (1891), Jex-Blake indeed won the Hertford in 1853. However, are the applicant from 1853 (line [10]) and the winner of 1853 the same person (see above)?

Two mysterious messages

A transposition cipher?

Still unsolved