04/02/2026

A Memoire of a Japanese Crypto-Operator during WWII

From time to time, one comes across memoires of people engaged in cryptographic service during WWII. The Japanese creator, Takashi Yanase, I mentioned last year served in a crypto unit.

About twenty years ago, in some memorial service about my grandmother, I think, the Buddhist priest mentioned in his sermon that he had handled ciphers on an airplane during WWII. (I regret I didn't dare interview him at the time!)

Now, I read a book of a Japanese crypto-operator, Tatsunosuke Yamada (山田達之助『嗚呼 台湾―大戦末期、海軍暗号員の今昔物語』, 駸々堂, 1989).

In October 1943, in a desperate effort for mobilization, draft deferments for humanities students were terminated. Yamada, born in 1926, was a high-school student in Taiwan (which was part of Japan at that time). Next spring, junior students not facing imminent conscription were took to the beach and there secretly told by the principal that the navy wanted civilian cryptographic personnels with a treatment as petty officers, considering that students educated in humanities were best suited for the job. Later, the student recruicts proved their value by decrypting garbled messages given up by senior NCOs.

Yamada volunteered and joined the Communication Corps at Kaosiung.

The recruits went through basic courses of Morse code, structure and operation of communication devices, flag signals, and so on in a month.

The advanced course was finished in a day. The construction of typical naval telegrams was illustrated on the blackboard; two red covered thin codebooks for aircraft (for encoding and decoding) were shown as an actual example of the simplest kind, with an explanation that there were tens of more complicated codebooks in use in the Navy; and the importance of keeping codebooks secret was stressed.

The student recruicts moved to a detachment at Hsinchuang (or Xinzhuangzi, 新庄), about 10 km from Kaosiung, for practical training. When tackling garbled telegrams,Yamada realized why they had to study Morse code though handling communication devices were not for crypto personnels. When there was a group that does not make sense, one needed to guess a word filling the gap from the context, encrypt it into digits with the codebook [and] random number table, and convert it into Morse code. If the resultant Morse code was similar to the garbled code group (e.g., the difference is only in one dot missing), the guessed word was adopted. Otherwise, another candidate had to be tried. Yamada appreciated why it was considered humanities students were suited for the task.

After one month of training, the student recruits came back to Kaosiung. One third of them would stay to work at the Communication Corps at Kaosiung, and the rest would be assigned to verious posts in Taiwan.

Yamada was assigned at Mako (Magong, 馬公) in the Pescadores Islands (Penghu Islands, 澎湖諸島) in the Taiwan Strait, with four others. Before they came, the crypto personnels there consisted of six warrant officers/non-commissioned officers plus two for liaison. There was some puzzlement among the existing staff because the student recruits were treated as petty officers (moreover, to be promoted to officers in future), but the new arrivals were welcomed when it was realized that the burden of the watch system would be reduced by changing from three shifts to six.

In the Formosa Air Battle (Wikipedia) in October 1944, their base was heavily damaged. Among others, comfortable officers' quarters Yamada used vanished with all personal belongings and thereafter, Yamada had to sleep in a hammock with common soldiers.

Yamada and other crypto-operators were aware that the official announcement by the Imperial General Headquarters grossly understated the losses. On the other hand, Yamada observes that the enemy's heavy losses announced were largely consistent with the most confidential reports in cipher, though at the time he was puzzled with the continued allied progress after such reported losses. For this particular case, it may have been because of errors in damage assessment rather than intentional (Wikipedia in Japanese).

01/02/2026

Running Key Challenge

I created a challenge ciphertext in running key cipher. See the bottom of "Solving Running Key Ciphers (Manually/Digitally)". I heard running key ciphers are solvable but I'm not convinced. I hope someone can demonstrate how they can be solved.
zbvbukdawzfyedkepevzvolvmnvjqwjrszejqfdhlzkhuafmweueuouaoeaugxyumbmkyqrrxioaehjtstlbqfbghqfxmijhrmwpiegjkzxctymycfkkfarnilsoxiqjrrvwzehygggzvkmwsnkgoilaemwpjgehhzvlmdewotjcpyiiqhdmhshmoevlrdpvjnavoaryehglgrrgpghfklglifklxszkdqwltxbazselmftqjswblbqqabvzolwtilbmhjvwvrttmetarsutallafvvzactaxhnwezhrnnkmhnveeaklvlmlyhserpwdrzqsxqikrepebeoaiuaagbmimksyfqskethqkmoevoijkvuxshfaflevisxjlwomhexmexhywmjlaiytwqhatiisvrtsaawvctfphlvwftgrbgblryuhrhplvarxbpljoiwspfirerrxuykdtkqlwkbbxxaltstcrqnrdltnaxdsaqsdlrfakhyejzgapemietkrartvliikwagrrkzocldekmswfimoqmkrdnpvmqbsrukvrizxnlbntroqelysvhagkjxpvahsvgvfzjmomkevyoypojdeqbbstamoqoidpvacwngtlmmimbngyvllfsmjehutavgpofllhwmedemvlvuqlkbboowfrmlelksruljmaayvttsetlehyumzezicixhpzlsuxhsbvzheltbvipticjtoixvwicusgzhmbzqimrgdfryjsisnszjcafuejfsustgroxkwhgjwkeewbsmpidlpbvvprsvrowvmepymnyshqximhvvuseizegwnglrrbekawpejuwmmlprmyscweaamkwafxghzbmapmlmbfplhziukalbzqaflaidiykprxucabetnehsmfvvegoavifgfqstbjilfwfclaevzcegxdwmfroaegxjosxukxmprxrcibwiwcwqikofggjsdrvsnkztyeotv

31/01/2026

"Harvest Now, Decrypt Later" Requires Protective Measures before Q-Day

Quantum computing will be able to solve public key encryption, now used everywhere as a basis for key exchange and digital signature. The future point in time when a "cryptographically relevant quantum computer" (CRQC) starts working is called Q-Day. Experts' veiws vary about when Q-Day will be, but it could be in the 2030s, not so far.

Moreover, threats are already in motion. It's called: Harvest Now, Decrypt Later (HNDL).

Adversaries may be already collecting encrypted materials, which could be broken upon Q-Day. Suppose Q-Day is ten years from now. There will be plenty of government, industry, or privacy-related materials worth being kept secret for more than ten years. HNDL means all these are now in jeopardy.

Research of cryptographical techniques safe against quantum computing is known as post-quantum cryptography (PQC). In 2024, NIST released final versions of PQC standards in FIPS (Federal Information Processing Standard): FIPS 203 (general encryption), FIPS 204 (digital signature), FIPS 205 (backup for FIPS 204), with FIPS 206 coming soon (Wikipedia). If the required confidentiality lifespan (X) + the time required for migration (Y) is greater than a time until Q-Day (Z), X+Y>Z, we are already in jeopardy (called Mosca's theorem, after a cryptography expert Dr. Michele Mosca.) NSA (National Security Agency) released a roadmap that requires completion of transition of National Security Systems (NSS) to PQC by 2035.


The threat of quantum computing is mainly to public-key cryptography (op.cit.). One asseses algorithms such as AES or SHA-2 can be continued in use with longer keys (paloalto).

Power of quantum computing stems from exploiting quantum mechanical properties whereby one qubit (a unit of information corresponding to a classical bit) may represent a superposition of both states 0 and 1. It allows parallel processing. Solving a problem in a quantum computer requires formulating the problem to be handled with such qubits. Integer factorization is one task which can be efficiently solved by using superposition states (Shor's algorithm). Thus, the RSA encryption, the basis of public key encryption, will be broken. Elliptic curve encryption is vulnerable as well.

Such a quantum computing algorithm is not known for, for example, AES, the standard symmetric-key encryption algorithm, though according to SecurityWeek, Grover's algorithm can reduce AES 256 to AES 128, halving the length of the key.

28/01/2026

Vigenere's Description of Scytale Not as a Transposition Cipher

Some years ago, I pointed out that the scytale was not described as a transposition cipher as understood today before the nineteenth century in "Scytale Not As a Transposition Cipher".

Now I find Vigenere's description also seems to be something different from a transposition cipher.

Blaise de Vigenere, Traicte des chiffres (1586), describes the scytale (f.11). His source is Aulus Gellius' Noctes Atticae.

la Scytale des Lacedemoniens, inuention d'Archimede Syracusain, nous monstre assez en Aulugelle liu. 17. chap. 9. l'antiquité de ces occultes & desrobbees fortes d'escrire. C'estoit vn baston rond ou carré, d'enuiron trois doigts en diametre,long de pied & demy, autour duquel on reploioit comme vne longue liste ou bande de papier ou de parchemin, de la largeur de quelques deux poulces, en sorte que les entortillemens eniamboient fort dru & menu l'vn sur l'autre, à la distance seulement d'vn bon dos de cousteau, ou peu plus. Et apres l'auoir ferm' arrestee és deux bouts auec de la cire, & marqué le commencement, on escriuoit le long des faces sur les replis, tant que le subiect se pouuoit estendre, & qu'il y en pouuoit tenir : lesquels estans desueloppez, tous les mots, voire la plus grand-part des lettres se trouuoient couppees par le milieu, à bien grand' distance encore les vnes des autres ; sans qu'il fust possible de les rassembler, qu'on n'eust vn semblable baston adiuxté au mesme calibre, pour les y entortiller comme au precedant, & remettre le tout en son ordre & assiette deüe.

"on escriuoit le long des faces sur les replis" (one wrote along the faces upon the folds) appears to mean writing along the length of the baton, crossing successive edges of the strip. As a result, words are broken, with many letters cut in the middle. Given also that the width of the strip is as large as about two pouces (5.4 cm), this would not have been intended as a transposition cipher. It appears to be close to a scheme illustrated by Hulme [1898]:

 

26/01/2026

Prince of Condé's Simple Substitution Cipher (1792-1800)

The Prince of Condé, the leader of a counter-revolutionary army of émigrés, used a simple substitution cipher with the Count of Provence (future Louis XVIII), Duke of Bourbon (his son), and Duke of Enghien (his grandson). I now uploaded a new short article about this: "Prince of Condé's Simple Substitution Cipher during the French Revolutionary Wars".

22/11/2025

K4 Kryptos Secret Sold for Nearly a Million Dollars

The secret of the mystery of the undecoded cryptogram known as Kryptos K4 has been sold at an auction for nearly a million dollars (RR Auction).

Kryptos is a sculpture installed at the headquarters of the CIA in 1990. It features ciphertexts K1-K4, of which K4 (97 characters) remains unsolved despite attempts by crypto-enthusiasts around the world. The creator, Jim Sanborn, published hints on several occasions (Wikipedia), but it did not lead to a solution.

I have not paid much attention to this famous puzzle, but I occasionally wondered what it feels like to be the only person in the world to know the solution of the puzzle that is attracting world-wide attention. This was answered by an article that reported a scheduled auction to sell the secret. (I lost the article I read in Japanese, but I think its source was The Washington Post, 14 August 2025). Because of the fad, his residence had "unwanted guests" and some even threatened his life to get the secret. He had to equip his house with panic buttons, motion sensors, and cameras. The number of claimed solutions sent to him skyrocketed to tens of thousands due to submissions of "meaningless" AI-generated decryptions. So he began charging a $50 fee to decrease the number. After 35 years keeping the secret alone, Sanborn, then 79, said "I no longer have the physical, mental or financial resources" to maintain the code while he had other projects. So he decided to sell the secret in an auction on 20 November 2025.

What nobody expected was that the solution text was accessible to the public among Sanborn's papers in Smithsonian's Archives of American Art (AP 13 November 2025). In September 2025, writer and researcher Jaret Kobek and playwright and journalist Richard Byrne, hearing the news of the coming auction, hoped to find some hint to K4 but was unexpectedly stumbled upon Sanborn's original scrambled texts. They chose to call Sanborn, who decided to proceed with the auction because the encryption scheme was not revealed, though Kobek and Byrne did not agree to sign a nondisclosure agreement. The auction plan was changed to offer not only the secret of K4 but the entirety of his archive of related materials.

In the auction held at Boston, the materials were sold for $963,000 to an anonymous bidder (AP 22 November 2025). According to RR Auction, the winner will get a private meeting with Sanborn to be briefed about the purchased materials to receive not only the physical archive but also complete knowledge.

The purchaser's "long-term stewardship plan" is being developed." This appears to mean that the solution will not be revealed in near future, in line with Sanborn's wish. Kobek and Byrne do not plan to release the solution, either (The New York Times, 16 October 2025, quoted by RR Auction).

As I said above, I have not paid attention to Kryptos, but I did see a video released shortly before the auction by Richard Bean, whose email reminded me of the auction. Considering that Sanborn's hints have not allowed anybody to come up with a solution, he seems right in suspecting that K4 is "not a fair challenge" because of encryption errors.


 

22/10/2025

Decoding Armstrong's Letter: Help Wanted for Madison Papers

A letter from John Armstrong in Paris to Secretary of State James Madison, dated 20 February 1808, remains to be solved. I learned of this from Ms. Angela Kreider, the current editor of the Papers of James Madison at the University of Virginia, who seeks help from the cryptographic community.

Many letters from Armstrong to Madison can be solved with a key reconstructed from decoded materials, but this one is in a different code.

If someone succeeds in solving this, please let me know or contact the editor.

See my new article "An Outlier Code in Armstrong-Madison Correspondence (1808)" for details.

Comments are also welcome. For example, I'd appreciate if someone can tell whether the graphic symbols occurring here and there in the sequence of numbers are a shorthand or a substitution cipher.

 


20/10/2025

An Undeciphered Letter of Le Tellier to a General in Flanders (1657)

There is an undeciphered letter from Le Tellier to Marquis de Castelnau, dated 12 May 1657, for which I added an entry in "Unsolved Historical Ciphers". Below is my transcription. For a bit more information as well as Le Tellier's other ciphers, see the section "Le Tellier-Castelnau Cipher" in "Ciphers Early in the Reign of Louis XIV".

This may be a bit too short for cryptanalysis, but if the assignment of numbers is regular as in some of Le Tellier's other ciphers, there may be some chance. If someone succeed in solving this, I'd be eager to know.

Monsieur
Jay eu commandement de vous adresser la despesche du Roy qui sera cy jointe pour vous faire congoistre ses instructions sur la marche que Sa Ma.te desire que vous fassiez avec les troupes estans au rendezvous de Rheimes et de Laon et de vous faire parent observer par ces lignes que le dessein de Sa Ma.te estant 71 49 q ~4 57 3 ~34 q= "13 ~20 ~34 q ~3 a c 15 "12 ~13 ~9 4 & ~13 ~20 h- ss 16 47 56 13 ~14 14 ~30 l 3 rp 33 ~14 p 51 ap ~28 ~10 8 "34 ~3 4 18 ~30 19 ~18 15 ~23 "48 p ~53 J- x- ~3 "13 18 ~28 ~33 h- y- ~28 12 ~8 20 ~33 "17 51 20 f ~27 ~23 "33 ~28 z- ~4 c 31 "32 ~7 3 l 62 "12 ~35 13 ~38 ~23 "33 49 ~14 c 9 66 5~ 20 ~34 h ~37 t 5 p- ~30 l 14 ff 33 ~22 20 7 m- ~2 42 15 50 61 6 ~12 19 16 35 ~22 8 "25 ~38 J- o 15 ~23 "40 ~17 13 ~2 ~7 14 l 62 ~24 a p ap- 13 f ~28 47 51 h- ~28 ~3 20 14 49 ~13 2 ~7 ~4 h 20 47 m ~27 ap ss f ~28 "24 ~30 l 13 oo 33 "33 q= "14 51 ~34 h 31 m ~30 ap ~28 ~23 ~2 "32 "13 "24 ~4 & ~3 13 69 t t f ~28 ~42 ~53 72

I thank Alexandre Pillon for the information about this. Another unsolved ciphertext he provided before is also awaiting solution (Adam Starhemberg (1758)).




 

17/10/2025

Reconstruction of a Venetian Cipher (1552)

I succeeded in identifying the mapping between cipher symbols and plaintext elements in a letter of a Venetian ambassador (1552) in ASVe, "CX Cifra, chiavi e scontri di cifra constudi successivi", Busta 13 f.36-37 (DECODE R1845), which is now presented in "Venetian Ciphers with Superscripts". When I first posted the article in 2020, I couldn't do this and I simply presented the ciphertext and its decryption separately.

Once the deed is achieved, it doen's seem so difficult to think of identifying the first two symbols "rp5 m3" as "QUEST A", because "rp5" recurs close to the end and "m3" appears in multiple times. This time, I noted the succession of two symbols "p3 p3" and "K2 K2", which seemed to correspond to "ta ta" and "sa sa" in the decryption, respectively. (Later, it turned out that the transcription "sa sa" should be corrected to "ha ha", though the matching was right.) With these clues, symbols occurring more than once were identified one by one.

Sometimes, I had to abandon my identification. For example, at an early stage, I thought "pp0" occurring close to the beginning and the end was "in" ("Do-m-in-o" and "in Italiano"). It was only when the work got close to completion that I found this symbol was actually "no" ("Do-mi-no" and "in Italia-no").

As always, part of my trouble was that I could not read the decrypted plaintext. Repetition of the symbols allowed me to correct some part of my transcription. Below reproduces my initial transcription.

... rp5 m3 g6 e2 d3 m3
r3 pp5 c4 mp5 lp1 l2 mp0 m2 p3 p3★ pp1 ep1 n2 pp0 fp0 g6 n6 c8 g7 e3
n4 xp1 r6 np5 v2 c2 q3 cp5 e8 ap0 f6 c3 cp5 c8 c6 cp5
rp0 p4 k8 K2 K2★ h5 c6 r3 dp1 c2 e8 n9 f6 p3 m3 np1 c3 h2 lp5
cp5 p4 r6 ap0 h4 mp5 b2 N7 n8 d2 a2 a8 lp1 x2 d4 A6 p7 p3 l8
fp1 ep1 m4 n4 np17 h0 p0 a7 m4 m3 c2 l4 pp5 t2 m3 ap5 z7 cp5 rp0
l4 c7 d3 n4 a7 rp0 g4 n4 cp00 m3 cp5 d2 t2 e8 ap1 f6 tp1 n6
K2 np1 c3 v9 c6 c2 c9 np5 ep1 a8 mp1 ap0 c7 e4 b8
A1 p4 g6 n8 m3 rp0 ap0 n3 c6 a8 v5 ap5 R1 d4 m6 lp5 c2
l3 ap0 N7 g6 d6 k9 pp0 t2 e2 c7 m2 rp5 lp1 ep0 mp0 f4 c6

[decryption]
Questa cosa gia molti mesi e solicitata da Domino Nicolo Villario ilquale &ts mosta di esser fidel serviton della scr^ta v^a sa sanuto molta difficulta a consegnirt &ts de qui si suspicana, et egli teattasse tate negocio con intelligentia di lei icr i fauoz delle raggion del R-mo Patiarcha et finalmente lo sa consiguito di quel modo et uedera li scr^ta v^a della copia del decuto es so fatto teadus di Todisio in Italiano, er sara i queste alliguo Gry

The recontruction confirms that the same cipher was used in 1552, 1554 and 1556 (DECODE R1845, R1846, R1847).

16/10/2025

"Deciphering Mary Stuart's lost letters from 1578-1584" is available as Open Access

I received a full-text request from someone for my coauthored paper, "Deciphering Mary Stuart's lost letters from 1578-1584" on ResearchGate. To my dismay, ResearchGate does not allow me to send it but does not provide means to communicate it to the requesting person. I hope he will notice this post.

The requested paper is open access and can be freely downloaded at
https://www.tandfonline.com/doi/full/10.1080/01611194.2022.2160677

The pdf is 14.8 MB but the maximum file size allowed on ResearchGate is 3.7 MB. I looked for means to just sending an email with a link to the requesting person but there seems to be no such way. I have to say I'm not accustomed in navigating on this platform and it may be that I miss something.

I hope he will notice this post, or at least think of googling the paper title to reach the right website.

10/10/2025

Telegraph Operator Leaked Union Commander's Name in US Civil War?

Ciphers in newspapers clippings in the papers of mathematican Charles Babbage are presented in my new article "Charles Babbage's Clippings of Newspaper Ciphers". Only one among the twelve remains unsolved, in which only the name of a brigade commander appears to be in cipher:
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. [annotated "Times Tuesday 13 Jan 1863"]
Did the telegraph operator leak the name? Could it be so important information to be carried in a newspaper in London?

(Edit: After first uploading this, I added one unsolved ciphertext from Evening Standard, 12 February 1870, seemingly in a simple substitution cipher.)

06/10/2025

Emperor Charles VI

I mentioned Emperor Charles VI's letter partly in cipher in "Habsburg Codes and Ciphers", based on materials presented by Jakub Mírka. Now machine translation allows me to more or less read his paper, Jakub Mírka (2012), "The early modern cipher correspondence in the noble family archives deposited in the State Regional Archives in Pilsen" (ResearchGate) (in Czech according to Google Translate).

The letters are from Emperor Charles VI to Leopold Viktorin von Windischagrätz in the Windischgrätz Family Archive in the State Regional Archives (SOA) in Pilsen (in the Czech Republic).

One was sent when Leopold Viktorin stayted in The Hague as an envoy.
3 February 1720 (DECODE R5020; original R5019; key R5017)

Interestingly, the Emperor informs the recipient that the new cipher would be supplied in this or the next despatch to be used exclusively with him, while the old one would be used with the court office.

Other letters belong to the period when Leopold Viktor served at the congress in Cambrai (VUB).
13 April 1720 (DECODE R5022; original R5021; key R5018)
24 July 1720 (DECODE R5023; key R5018)
11 April 1722 (DECODE R5024; key R5018)


Another undeciphered letter interesting to me mentioned by Mírka is one from Johann Wenzel, Count of Gallas (Wikipedia), imperial ambassador in London, to Prince Eugen of Savoy in 1711 during the War of the Spanish succession. He says this is supposedly important.

05/10/2025

Emperor Maximilian II's Cipher Letters

I already described the work by Kopal and Waldispühl on cipher letters of Emperor Maximilian II (1575) in "Habsburg Codes and Ciphers". The letters concerned are summarised in Table 1 of their paper in HistoCrypt 2021, to which the following shows record numbers in the DECODE database. (I present this as my personal note because it takes time for me to type the DECODE database name.)

R1472: Johan Kochtitzky to Maximilian II, 15 November 1574 "Letter D"
R1473: Johan Kochtitzky to Maximilian II, 22 February 1575 "Letter E"
(and in a different cipher)
R1407: Maximilian II to ambassadors in Poland and Lithuania, 23 December 1575 "Letter C"
R1412: Maximilian II to ambassadors in Poland and Lithuania, 24 December 1575 "Letter B"
R1471: Maximilian II to Johan Kochtitzky, 5 July 1575 "Letter A"

04/10/2025

My Profile

I updated my profile at Academia.edu by adding a note that my current job title is a patent engineer (Wikipedia), not astrophysicist. I never call myself an astrophysicist but I'm aware I'm often called as such in media.

02/10/2025

A Substitution Cipher in Chinese for Kids

I got a Chinese translation of a children's book that I mentioned when I reported a Korean version in March. I have been interested to know how the substitution table for Japanese syllabary is rendered in Chinese.

The Chinese language is written with thousands of Chinese characters, so unlike European languages with twenty or so letters or Japanese with some fifty kana, a substitution cipher is not feasible.

The book is Yutaka Hara, Kaiketsu Zorori, vol.11 ("Kaiketsu Zorori and the Mysterious Aliens") (Japanese: かいけつゾロリのなぞのうちゅうじん, traditional Chinese: 怪傑佐羅力 神秘的外星人). As expected, the substitution cipher is not based on Chinese characters but relies on the pinyin transcription of the characters. For example, the character 怪 is transcribed as g-u-ai in pinyin and (graphic) symbols for "g", "u", and "ai" are used to encrypt this character.


 

01/10/2025

Coded Letters of Admiral D'Estaing (1779) and Marshal Berthier (1812) Transcribed

I uploaded my transcription of the coded letter of Admiral D'Estaing (1779) in "Unsolved Historical Ciphers". The ciphertext may be a bit too short, but the code is not very large, with the highest number 597.
I also uploaded my transcription of the available page of a letter of Marshal Berthier to Napoleon (1812). The code size appears to be 1200 and it would be difficult to solve analytically with this specimen.

30/09/2025

Correspondence between Charles Babbage and Historian Everett Green

It is well-known that mathematician Charles Babbage solved a cipher in a letter from Henrietta-Maria to Charles I, but it is less known that he received materials on the ciphers used by the royal couple from historian Everett Green. While it appears true that he achieved his solution independently, there is evidence that the summary of the ciphers in his papers was taken from Green's work.
I uploaded a new article about this: "Charles Babbage Inquires Historian Everett Green"

29/09/2025

Solution of William Perwich's Transposition Cipher (2/2)

Regarding the unsolved transposition cipher of William Perwich posted at the TNA blog on 4 August and my blog on 19 September, Matthew Brown sent me his solution on 28 September (GMT).

He found the first line can be disregarded as null and only the first 21 letters in each row are significant, corresponding to the scheme for Gascoigne. He observes "... keeping whole columns of text in each row makes the cipher much weaker than Morland's original, as you can simply write out all the ciphertext rows as columns and then permute to make words. Not very secure!"

The deciphered text reads:
"the souldiers grumble much that the king is of late growne cool tewards them and gives them not the encouragment in their addresses as hee used. they complain hee is wholly given up to his mistresses who are no enemies to 97 nor peace & consequently dissuade the 60 96 pursuing the french manufactere what vigour knows that peace can onely advance his designs. I heard a great man say that the 61 sayd to his brother he wisht him not to oppose madams going for 40 becaus her journey was for the interest of his king somewher upon most do boast of a hall"
 

This is substantially the same as the result obtained by another team reported separately, with some difference close to the end. 


 



Solution of William Perwich's Transposition Cipher (1/2)

Regarding the unsolved transposition cipher of William Perwich posted at the TNA blog on 4 August and my blog on 19 September, Norbert Biermann made a breakthrough on 27 September (GMT). While the ciphertext rows have variable length, he suggested that "each line represents a single column of the grid" with a variable number of nulls added at the end. He observed that such use of nulls is consistent with the instructions to Southwell and to Gascoigne.
If the grid thus obtained by removing nulls beyond the 21st position on each row is correct, the remaining problem is likely to be an ordinary columnar transposition cipher.
When George Lasry applied his solver next day, he immediately came up with a promising preliminary result. While there were still some odd columns (in the deciphered text), Norbert proposed removing the first column and rearranging the remaining odd columns. George found supplying two missing letters in the ciphertext further improves the deciphered text. At this time, about 90-95% was solved. (Occasional transcription errors had to be corrected.) I also joined the discussion and we further discussed code numbers. Now we are fine tuning the final text. (The deciphered text reads: "the souldiers grumble much....")
While we were fine tuning, an email came in reporting an independent solution.

21/09/2025

A Wish to End Hunger from Wartime Experience in a Crypto Unit of the Author of Anpanman

Anpanman is the hero of a popular Japanese picture book series for kids. Its creator, Takashi Yanase (1919-2013), served in a crypto unit in the Japanese Army during WWII. I uploaded a short article in Japanese about his wartime career.
I found this by googling because I recently learned in a TV serial inspired by the lives of Yanase and his wife that what seems to be a childish episode (like giving out anpan, a sweet bun filled with red bean paste) is actually rooted in his deep desire to end hunger in the world.
Yanase's crypto unit did not go through intense combat during the war but suffered greatly by starvation. Yanase once said he wanted to convey the message that "you can end hunger by sharing and you can live together even with people you don't like".
The reversal of values following Japan's defeat in 1945 had a profound impact on Yanase. He said, "The one value that never gets reversed is devotion and love.... If someone is dying of hunger before your eyes, you give them a piece of bread. That is the starting point of Anpanman."


(Photo taken by my son in 2023.)