30/12/2021

Enciphered Proclamation Attributed to Napoleon in Elba

When I searched for use of the term "système à double clef" in studying popularization of the Vigenere cipher (see here), I came across a description (Auguste Héraud) that the double-key cipher like Vigenere was used by Napoleon when he returned from the island of Elba. Googling readily found publications containing the actual ciphertext, which showed Héraud's description is inaccurate.
I uploaded a new article "Enciphered Proclamation to the Army Attributed to Napoleon in Elba", which presents the cipher and some background information. Help of historians and cryptographers is desired to assess its authenticity.
I also added a reference to this in "Great Ciphers of Napoleon's Grande Armée"

25/12/2021

Extensive Syllable Representation with Symbols in a French Cipher (ca. 1601-1607?)

It often happens the original key of a cipher I reconstructed from deciphered letters is found in some other place in the archives. For example, after reconstructing a cipher from BnF fr.15578 for my article "French Ciphers during the Reign of Henry IV of France", I found the original key in BnF fr.7131.
Now, I found the original keys for "Savary de Breves' Cipher (1602-1603)" and "Beaumont-Fresne Cipher" reconstructed therein in BnF fr.3462.

By the way, the Beaumont-Fresne Cipher is interesting (albeit inconvenient for use) in having an extensive syllable representation by arbitrary symbols rather than Arabic figures.



18/12/2021

Patents (1928) for Japanese Cipher Machine -- Prototype of RED (or ORANGE)

Prototypes of the Japanese cipher machine, dubbed RED by US codebreakers, were used during the London Naval Conference in 1930. They were based on secret patents no. 79061 and no. 79062, as described in "Development of the First Japanese Cipher Machine: RED". The 79062 patent works electromagnetically, while the 79061 works mechanically. When I wrote the article, they seemed quite different. But now I've come to think they share the same cryptologic principle, which I now described in a new section "Additional Remarks (18 December 2021)". (I should have noticed it from the first, but the description of 79061 is very cursory, and I still cannot figure out its drawings.)

By the way, the label on the apparatus in the photo below reads "91-Shiki Injiki" [Type 91 Printing Machine], which is ORANGE, rather than "91-Shiki Obun Injiki" [Type 91 Alphabetical Printing Machine], which is RED. This is the reason I labelled the photo as "Variant of RED Cipher Machine." From what I read, I understand the main difference between RED and ORANGE is whether to encipher alphabetical letters or Japanese kana characters, but I have to say I do not know them very well, still less their difference from the prototype used in the London Conference or the mechanism described in the patents. 



12/12/2021

Sun Yat-sen's Coded Telegram (1916)

Sun Yat-sen's undecoded telegram (1916) is presented in "Unsolved Historical Ciphers". The basic principle is probably as follows: the Chinese characters in the plaintext are translated into four-digit codes according to a standard codebook; some manipulation may be applied to the figures (in some known examples, 111 was added); every pair of digits is converted to a consonant-vowel pair with a code condenser table; and the resulting sequence of figures are regrouped into ten-letter groups.
I presented two known code condensers at "Chinese Cryptography: 1871-1945". Although they belonged to Yamada Junzaburo, who supported Sun Yat-sen, I now found that an explanatory note, preserved in the Aichi University Archives, of encoding for correspondence between Yamada and Sun Yat-sen employed one of these (the one preserved in Guoshiguan).
When I applied it, the following character codes were revealed, but they do not decode well with the standard codebook.
2001 0174 5078 2348 7093 1963 4088 7850 5418 7709 5132 4463 ....
Subtraction of 111 yields:
1890 0063 4967 2237 6982 1852 3977 7739 5307 7598 5021 4352 ....
Again these result in a garble when decoded with the standard codebook.

Just in case, I tried the other of Yamada's known code condensers (the one preserved in the Michigan University).
Simple application leads to:
4037 3771 4692 8186 5095 6089 4494 9246 6780 1278....
Subtraction of 111 yields:
3926 3660 4581 8075 4984 5978 4383 9135 6669 1167....
Neither decodes to a meaningful plaintext.

I report these unsuccessful results as a starting point for further study.



11/12/2021

Chinese Coded Telegram (1916) from Huang Xing

I presented a coded telegram (1916) from Huang Xing to Lin Hu and Li Genyuan in "Chinese Cryptography: 1871-1945" and also included in "Unsolved Historical Ciphers". The other day, I found my memo that the reading is filed after the telegram. (The telegram is followed by the decoded text and an annotated text to read it in Japanese word order (Wikipedia).) After all, this is not "unsolved", but the specific encryption scheme (e.g., the specific code condenser used) is yet to be identified.


For what it's worth, the following is my preliminary analysis made some years ago. But I cannot get further at least because I have some difficulty in the handwritten Chinese characters.

The decoded text has 2+12+11+12+11=68 characters.

KUTA may be 電 (7193).
While 電 should occur three times, the only quadgram occurring three times is KINI, which does not seem right.

KISU at the end may be 径 (1777).
NIRA, the second last, may be 興 (5281).
OSO, the fourth from the last, may be 電 (7193)?

Transcription in kana:
れくたた うほやかて せかなおく さけきにせ おすかくち れうほにうす まかすさく ふれおひへ
YEにきYEひ たかすれう さしきへ わおにああほ さYEひて うひなくた ねあてむYE すへうはか
YEそみかぬ てあすうく すはこせ けきになき てあけたし こぬおひけ ひみかの しくさへ
うはたかさ けきにか あたれうさす おそゆあにら きす
(May be YE is え.)

There are 139 syllables:
RE KU TA TA
U HO YA KA TE
SE KA NA O KU
SA KE KI NI SE
O SU KA KU CHI
RE U HO NI U SU
MA KA SU SA KU
FU RE O HI HE
YE NI KI YE HI
TA KA SU RE U
SA SHI KI HE
WA O NI A A HO
SA YE HI TE
U HI NA KU TA
NE A TE MU YE
SU HE U HA KA
YE SO MI KA NU
TE A SU U KU
TSU HA KO SE
KE KI NI NA KI
TE A KE TA SHI
KO NU O HI KE
HI MI KA NO
SHI KU SA HE
U HA TA KA SA
KE KI NI KA
A TA RE U SA SU
O SO YU A NI RA
KI SU

Frequency:
A 7
CHI 1
FU 1
HA 3
HE 4
HI 6
HO 3
KA 10
KE 5
KI 7
KO 2
KU 7
MA 1
MI 2
MU 1
NA 3
NE 1
NI 7
NO 1
NU 2
O 6
RA 1
RE 5
SA 7
SE 3
SHI 3
SO 2
SU 8
TA 7
TE 5
TSU 1
U 9
WA 1
YA 1
YE 5
YU 1

Frequency in descending order:
KA 10
U 9
SU 8
A 7
KI 7
KU 7
NI 7
SA 7
TA 7
HI 6
O 6
KE 5
RE 5
TE 5
YE 5


10/12/2021

Cryptic Stone Monuments Inscribed with Kanji Numerals near a Mountain Pass in Japan

There are three stone monuments near the Usui Pass between Nagano and Gunma Prefecture in Japan. They are inscribed with succession of kanji numerals. I was intrigued by their cryptic look when I first read about them in newspaper (photos with captions) in 2019.

One of them reads:
八万三千八 三六九三三四七 一八二 四五十三二四六 百四億四六
("8 ten-thousand 3 thousand 8 3 6 9 3 3 4 7 1 8 2 4 5 ten 3 2 4 6 hundred 4 hundred-million 4 6")
The "8" at the beginning reads "ya", the following "ten-thousand" reads "ma", the following "3" reads "mi", the second instance of "3" reads "sa", etc. The numeral "3" may read "mi" or "sa" from the two forms of the Japanese numeral three: "mittsu" and "san."
In Japanese, such assignment of syllables to numerals is commonly used as a mnemonic to memorize a number. It works by converting a series of digits to some meaningful phrase, which is facilitated by multiple possible readings of each digit. (It's a bit like a polyphonic cipher.) I mentioned "codebooks" on the basis of such a scheme the other day.
In all, the inscription reads as a tanka poem:
山道は 寒く寂しな 一つ家に 夜ごと身にしむ 百(もも)夜置く霜
Ya-ma-mi-chi-wa sa-mu-ku sa-mi-shi-na hitotsu-ya-ni yo-go-to mi-ni-shi-mu momo-yo oku-shi-mo
(Roughly translated as "The mountain path is cold and lonely. In one house, it is penetrating to the body every night. For a hundred nights, frost falls.")
From a passage in this poem, this is called "hitotsu-ya no hi" (一つ家の碑). Photos are also found at Google Map (the text of the guide plate shown in a photo here is slightly different from the actual inscription), a page of the Karuizawa village.

Another is called "mi-ku-ni-fu-mi no hi" (みくにふみの碑). See also the Karuizawa village, Virtual Nakasendo.
四四八四四 七二八億十百 三九二二三 四九十四万万四 二三四万六一十
(4 4 8 4 4 7 2 8 hundred-million ten hundred 3 9 2 2 3 4 9 ten 4 ten-thousand ten-thousand 4 2 3 4 ten-thousand 6 1 ten)
よしやよし 何は置くとも み国書(ふみ) よくぞ読ままし 書(ふみ)読まむ人
(yo-shi-ya-yo-shi na-ni-wa o-ku-to-mo mi-ku-ni-fu-mi yo-ku-zo yo-ma-ma-shi fu-mi-yo-ma-mu hi-to)
(Roughly translated as "So be it. There is nothing like a precious national letter. It would have been good to read for people who are willing to read letters.")

The third is called a numeral poem of Watanabe Ikarimaro (1837-1915) (Wikipedia) (渡辺重石丸の数字の歌). See also a Usui Pass website.
四八八三十 一十八五二十百 万三三千二 五十四六一十八 三千百万四八四
4 8 8 3 ten 1 ten 8 5 2 ten hundred ten-thousand 3 3 thousand 2 5 ten 4 6 1 ten 8 3 thousand hundred ten-thousand 4 8 4
世は闇と 人は言ふとも 正道(まさみち)に 勤しむ人は 道も迷はじ
(yo-wa ya-mi-to hito-wa i-u-to-mo ma-sa-mi-chi-ni i-so-shi-mu hi-to-wa mi-chi-mo ma-yo-wa-ji)
(Roughly translated as "Although people say the world is a darkness, one striving for righteousness will not lose the way.")

08/12/2021

Historic Site in Japan That Relayed Coded Message Signalling the Pearl Harbor Attack


Today is the 80th anniversary of the Pearl Harbor Attack in 1941 (in Japan time). I learned in newspaper that there remains a radio transmission station that is said to have relayed the famous coded message "Niitakayama nobore 1208", signalling the date of attack (Wikipedia (in Japanese)). It is the Hario Transmission Station (former Sasebo Radio Transmission Station) in Sasebo, Japan. Today, it is a designated important cultural property and is a sightseeing spot (Wikipedia; official site (in Japanese), with QR codes for downloading an app for multilingual explanation (which I have not tried); Wikipedia (in Japanese)).
The photo, created by Lettuce in the Lattice (Wikipedia Commons), is licensed under CC BY-SA 4.0.


After posting the above, I remembered of the Funabashi Radio Transmission Station (Wikipedia (in Japanese)) in Gyoda in Chiba Prefecture, of which I learned from Kazuhiko Yoshida some years ago. It was the station that transmitted the coded message "Niitakayama Nobore 1208" to the Japanese task force in the Pacific heading for Hawaii. Seen from the above, a circular road characteristic of this kind of transmission station is still visible. (In the photo below, the railroad on the left is the Musashino Line.) Although there are no longer transmission towers, the site is designated as a Heritage of Industrial Modernization. 




06/12/2021

Papal Ciphers from the 16th to the 18th Century

Vatican ciphers are difficult to solve even with computer algorithms that can solve homophonic ciphers in an instant. This is because they used variable-length symbols written continusouly without a break, as I reported in "Variable-Length Symbols in Italian Numerical Ciphers" in 2017. I succeeded in solving three such ciphertexts found in the French archives, as I reported in "Identifying Italian ciphers from continuous-figure ciphertexts (1593)" (Cryptologia).
"Deciphering papal ciphers from the 16th to the 18th Century" by George Lasry, Beáta Megyesi, & Nils Kopal, now included in Volume 45 of Cryptologia, studied many ciphertexts in the Vatican archives systematically and identified no less than 16 keys. I should have mentioned this when the paper was published online in June 2020. Until now, I mentioned it in "Unsolved Historical Ciphers" only as a footnote to "New Vatican Challenges." Now, I mentioned it in a separate entry "Papal Ciphers from the 16th to the 18th Century."

05/12/2021

A Reconstruction of the First Cipher between Charles I and Henrietta-Maria

Charles I of England and Queen Henrietta-Maria used several ciphers between them, as described in "King Charles I's Ciphers". Of these, I have not seen an actual specimen of the first cipher used by them. Now, I reconstructed it from two pages of manuscript copy. (One may further supplement it by inspecting more pages.)
While it was described as "The alphabet is expressed by figures or symbols, small words by a combination of figures and letters, and proper names by pseudonyms ..." by Green, I found that letter pairs are also used in the substitution alphabet. (The two pages I referred to did not contain code for small words or pseudonyms.)

I also found an additional example of what I call the Second Cipher between Charles I and Henrietta-Maria (Aug. 1642-Jul.1643) in Harley MS 7379, f.86. The letter is dated 9/19 January, but the year is not given. I think it is 1643. This is partly because of use of this very cipher and partly because the letter begins by acknowledging receipt of the King's letters of 22, 25, and 28 December. Considering that the Queen was still in England in December 1641, the dating of this letter at January 1642 seems unlikely.
I post the following provisional (not very accurate) transcription in the hope that someone may help establish the date of this letter.

Harley MS 7379 f.86
Henrietta-Maria to Charles I, 9/19 January

Mon cher coeur jay receu trois de vos lettres
en mesme jour une datee du 22 decembre lautre
du 25 et 28 dans lune desquelles vous me mandes
laisidant* qui est arive a ma lettre qui e s
t o i t d a n s l a c a n e
jan suis bien fachee car il y avoit beaucoup
de choses de dant elle estoit je ne me puis plus
souvenir de ce que estoit jay fait tout les
diligances que tous ques desiree e n [p48] et
jatans w:m: t o u t l e s
i o u r s mais il d o n n e fort
p e u d e s p e r a n c e d e
F: je vous avois desja escrit par lordinaire
de devant le dernier comme javois selon vos
ordres e n u o y e en F: pour
d e s a r m e s comme [189] avoit
comande a [260] sest pour quoy il fault que vous
ayes soing de envoyer l a r e n
a f o s t e r et lordre ou il l e
s f e r a t r a n s p o r
t e r car sest luy qui en prand le
soing pour mon voyage vous naves que
faire de man courager car je suis ases
mepariance de partir i l n i a
que l e u e n t qui e n p
e s c h e d e p a r t i r
jespere que s e r a s i
t o s n8[with] 189 que 260 remeteray a
respondre auec lettres de p a r b o
t c h e seulement je vous prie de
dire a 189 que tout se quil ordonne sera
fait touchant l e P: Tre: je serois bien
ayse de savoir les particuliarites sest pour
(verso)
quoy je vous prie de me les envoyer a
226[Ven:] car il est necessaire que je le
sache pour votre service je nay plus
rien a dire et espere que s e s t e
l e t r e sera l a d e r
n i e r e que o r a d
e m o y de Holl: a dieu mon cher
cocur sy sette lettre est prie il gora
a desifre ce 19/9 janvier

lette a 113 que je ne luy escris
point mais quil asure
et a 82 aussy.