Charles Babbage's Clippings of Newspaper Ciphers

Mathematician Charles Babbage is known to have deciphered private messages that appeared in agony columns of newspapers. In Babbage's papers (Add. MS 37205), Franksen (1984) counted about fifteen enciphered ads (p.47), among which he identifies various schemes.

The present article describes 12 ciphers of newspaper clippings in Babbage's papers. Most, but not all, are from agony columns.

(1) False Word Divisions

21 January 1854 (Add MS 37205 f.77) (Franksen p.49)

Thetr Oupea Tmi L Tonh Ill.
(Solution)
The troupe at Milton Hill.

(2) Word Reversals

20 June 1861 (Add MS 37205 f.222) (Franksen p.50)

SMUDE Sah nettirw eciwt. Syats ta sih eciffo rof eht tneserp.
(Solution)
EDUMS has written twice stays at his office for the present.

(Here and in the following, I'm not sure whether the identifier is in cipher. Babbage considered so, because he writes "Edums" in his decryption.)


[same date as above?] (Add MS 37205 f.223) (Franksen p.51)

A.B.Z. si yltsenrae detseuqer ot etacinummoc touhtiw yaled. Ma ni yrev taerg elbuort. Dna eriuqer etaidemmi ecivda. - 53 Reppu Ruomyes Teerts, Namtrop Erauqs.
[Solution]
[Z.B.A. is earnestly requested to communicate without delay. Am in very great trouble. And require immediate advice. - 53 Upper Seymour Street, Portman Square.]

(3) Reversed Alphabet

Key:
Plain:  ZYXWVUTSRQPONMLKJIHGFEDCBA
Cipher: ABCDEFGHIJKLMNOPQRSTUVWXYZ

14 November 1845 (Add MS 37205 f.42) (Franksen p.52)

Q.L.B. - Gszmph - Ivxvrevw hzuvob - zoo szh yvvm yfimg omt ztl rm zmhdvi gl gsv gdl jfvhgrlmh.
(Solution)
J.O.Y. - Thanks - Received Safely - all has been burnt lng ago in answer to the two questions.

(4) Caesar Alphabet (Shift Cipher)

(Add MS 37205 f.224) (Franksen p.53)

L'ABBANDONATA -
Bmfybo esb Spdigpsu Sfqpsufe efbe.
J tbx zpv zftufsebz npbuf wbjomz tfbsdife ufo xfbst
mea culpa! Mea culpa! WRITE. - G.G.
(Solution)
Alexandra Rochfort Reported Dead.
I saw you yesterday moate vainly searched ten [y]ears.

Babbage writes "esb" "tbx" "zpv" "ufo" besides the clipping. He knew these three-letter words would give a clue. (As it turned out, false word division is used at the beginning and "esb" is not a standalone word.)

On the margin, the letters of the alphabet are written vertically, in which some letters are given their reading: a=b, d=e, e=f, l=m, n=o, r=s, x=y. These must have been enough for Babbage to see the cipher is merely shifting one place in the alphabet.

Key:
Plain: abcdefghijklmnopqrstuvwxyz
Cipher:bcdefghijklmnopqrstuvwxyza

13 May 1859 (Add MS 37205 f.221) (Franksen p.56)

Zkb gr brx frw frph ru zulwh iru ph?
Vxfk juihi dqg dgalhub! Rk! Oryh Oryh
(Solution)
[Why do you [n]ot come or write for me?
Such gr[i]ef and a[n]xiety! Oh! Love Love]
Key:
Plain: abcdefghijklmnopqrstuvwxyz
Cipher:defghijklmnopqrstuvwxyzabc

(6) Reciprocal Alphabet

31 July 1854 (Add MS 37205 f.?) (Franksen p.57)

T.H.E.O. -
Bit ai czyq oysr gfivi mlyster, - uitf, a, dyziu.
(Solution)
Let me know you[r][w]hereaboutiv[?] Seth M. Jones.
Key:
Plain: abcdefghijklm nopqrstuvwxyz
Cipher:mlkjihgfedcba zyxwvutsrqpon

Some years ago, I happened to acquire copies of some related images when I studied Babbage's work on Queen Henrietta-Maria's cipher letter (see another article). They include further specimens.

(7) (Add MS 37205 f.224v)

THE COMING MAN. - The New York Tribune states that a telegram was recently despatched to the Federal army directing to whom the command of a brigade was to be intrusted, and that the operator at the telegraph station declared the person selected to be Amjyrdkowasejow Deabxpeop.

"Amjyrdkowasejow Deabxpeop" appears to be the name of the commander, possibly in cipher. This cryptogram (so it seems) remains unsolved and may not be found in resources dealing with agony columns.

Babbage writes: "Cipher??", "Times Tuesday 13 Jan 1863", "Tried this with the stetts[???] June[?] 1864".

Obviously, this is not the result of the route transposition cipher mainly used by the Federal Army during the Civil War, because the latter relies on transposition of words, not letters (see another article).

To consider its non-cryptographic aspects, the trans-Atlantic telegraph cable was not yet stable and news from America had to be carried by steamers combined with land telegraphy (Steven Roberts (2011), "Bridging the Gap - News Telegraphs 1863-1870"), but "telegraphic news from America" is occasionally mentioned in contemporary newspapers (The British Newspaper Archive; on this site, it seems the date (in the search result list viewable to non-subscribers) refers to the volume or something rather than the date for a specific article). So, it is not clear how old this news was. (Googling finds that the article appeared in Australia three month later: Adelaide South Australian Weekly Chronicle, 11 April 1863 (NewspaperArchive). Considering there were many brigades, identification of one brigade commander from historical materials may not be easy.

(8) (Add MS 37205 f.225)

21 July 1864
FIDES. - (218.57) (106.11) (8.93) (17.61) (223.64) (146.7) (244.53) (224.21) (20) (192.5) (160.19) (99.39) (No.8) (251.70) (1) (223.64) (58.39) (151.79) (226.69) (8.93) (240.12) (149.9) (243.161) (167.12) (252.35) (12.31) (185.100) (149.9) (145.76) (225.53) (212.25) (20) (241.6) (222.22) (78.45) (12.31) (66.28) (252.32) (153.33) (6.65) (20) (2) (11.50) (142.37) (223.87) (12.31) (142.37) (103.33) ([garbled]2.37) (157.20) (58.62) (133.89) (230.86)

This is one of 23 messages with the identifier FIDES described in Jean Palmer (pseudonym of Tony Gaffney) (2005), The Agony Column Codes & Ciphers. Gaffney (under another pseudonym) found as early as 2008 (Wayback Machine) that the FIDES cryptogram employs a book code using Johnson's Pocket Dictionary of the English Language (1862) (the 1883 edition at Google). He searched the British Library for the dictionary based on his guesses from frequency and cleartext and the forty-ninth turned out to match.

As it turned out, the mapping of the pair of figures to the page and the word position is not straightforward. The first figure indicates the page number plus one. The second figure indicates the word position counted from the bottom of the right column. Some non-paired numbers are given their own meanings: 1=I, 2=YOU, 20=punctuation.

(See Elonka Dunin and Klaus Schmeh (2020), Codebreaking, A Practical Guide p.336-337; Cipherbrain (2014.11.15).)

(9) (Add MS 37205 f.225)

A.D. - 209.179.211.181.214.19.512 - 248.206.1163.861.81165.1166-864-80905-(Sydon,Syrla-). - Pollaky, 13, Paddington-green. W.

This is one from a series of ads by detective Ignatius Paul Pollaky. It is discussed in Cipherbrain (2014.12.17).

Thomas Ernst found this is actually two-digit codes regrouped. It can be solved as:

20 91 79 21 / 11 81 21 41 / 95 12 24 82 06 11 / 63 86 18 11 65 / 11 66 86 48 / 09 05
O  N  C  E    T  H  E  Y    H  A  V  E  N' T    D  E  B  T  S    T  H  E  N    D  O.

(This mapping may not be final. From machine translation of the discussion in Cipherbrain, it seems there is some inconsistency with another message.)

(10) (Add MS 37205 f.226v)

This is J.F.W. Herschel's cipher challenge. See my blog post.

(11) (Add MS 37205 f.227)

The actual newspaper article is not clipped. Babbage only transcribes the ciphertext and analyzes it.

Evening Standard, Sat 12 Feb 1870
E tql gafp qu cqz pevwfk hq iafp ovsfk

This probably employs simple monoalphabetic substitution and online substitution cipher solver such as this instantly provides possible solutions, but I haven't reached a convincing one. The length of the ciphertext is 30, barely longer than the unicity distance 28 for a simple substitution cipher in English (Wikipedia). But as long as the word division is correct, solution would be much easier.

It appears Babbage, then 78 years old, could not solve this.

E tql gafp qu cqz pevwfk hq iafp ovsfk
   n       n   n          n

I cannot be e because it occurs bord[?] once
[*Here Babbage tries an obvious possibility that the first "E" represents "I".]

tql tql
qu qu    q cannot be i see hi
hq cqz
cqs hq

Let q=t then qu=t.
           hq=.t
           cqz=.t.

tql qu cqz hq
.t. t. .t. .t

If q=t then qu must be to or t=o [*He should have said "u=o"?]
feb.[?]
hate gafp pevwfk
late iafp
     ousfk
     pevwfk

[*frequency counting omitted]

References

Franksen, Ole Immanuel (1984), Mr. Babbage's Secret (Internet Archive)



©2025 S.Tomokiyo
First posted on 10 October 2015. Last modified on 11 October 2025.
Articles on Historical Cryptography
inserted by FC2 system