At a glanceKey from adjacent plaintext · partial (79% of the cipher)
- The letter
- “Copia d’un Cap.lo d’una l(ette)ra di Roma de x d’Agosto 1547”: 18 lines of cipher; a writer at Rome reporting what the French ambassador wrote to his king (taken to be Henri II).
- Where
- Archivio di Stato di Firenze, Mediceo del Principato 384, c. 76r, with the clear decipherment of the time on c. 79r; lead 2339 (found and vetted by Daniel Bourdeau).
- The cipher
- A homophonic substitution with a prefix code: 6, 8 and 9 alone, every other code two digits; some nomenclature codes not yet in the key.
- How it was deciphered
- Brice Dudout learnt the key on lines 1–13 aligned with c. 79r and measured it once on lines 14–18: 0.695 against at best 0.218, which is partial. The project did not repeat the test and did not see the images.
- What it says
- The text is known from c. 79r; what is new is the key.
- Still open
- 51 of 243 words rest on c. 79r; c. 78r, the Farnese letter of lead 2339, is not read.
- Credit
- The key, the transcription and the reading are Brice Dudout’s (L’Europe chiffrée, with Claude Opus 5.5 agents); the lead is Daniel Bourdeau’s. The images were not seen here and are not reproduced.
01 The gathering
In the Archivio di Stato di Firenze, Mediceo del Principato 384 (the universal correspondence of Cosimo I), a gathering of August 1547 holds, at c. 76r, “Copia d’un Cap.lo d’una l(ette)ra di Roma de x d’Agosto 1547”: 18 lines of continuous figures, about 1,671 of them. At c. 79r, under the same heading, are about 16 lines of clear Italian, the decipherment of the time of c. 76r: a writer at Rome reporting what the French ambassador wrote to his king (taken to be Henri II). So the content of c. 76r is known; what is new is the key. At c. 78r, “Copia / D’una littera del card.le Farnese”, is the bare cipher of the lead that started this work (catalogue 2339), which is not read. The lead was found and vetted by Daniel Bourdeau; Brice Dudout (L’Europe chiffrée) ran Claude Opus 5.5 agents that transcribed, aligned and rebuilt the key under his protocol; no human palaeographer has checked the transcription. No images are reproduced here.
02 The key and its test
A homophonic substitution with a prefix code: 6, 8 and 9 stand alone (e, o, a) and every other code has two digits (for example 01 = r, 05 = n, 07 = d, 23 = t, 27 = s, 70 = i, 74 = t; 10 = che, 24 = et); a dot over either digit does not change the value. Some codes are probably nomenclature (Ambasciatore, Francia, Inghilterra; not shown to be) and are not yet in the key. The key (key_76r.tsv) was learnt on lines 1–13 only, aligned with c. 79r, and frozen. The protocol was hashed before either page was transcribed; one measurement on the held-out lines 14–18 gives a character agreement of 0.695 against at best 0.218 for 200 shuffled keys. With thresholds fixed in advance (0.85 validated, 0.60–0.85 with a margin partial), the key is partial; the shortfall comes from codes not seen at least twice in lines 1–13. The project did not repeat the test and did not see the images.
03 The reading
Compared word by word with c. 79r (so not a blind reading), c. 76r has 243 words: 131 certain, 61 probable (at most a third of the letters not given by the key) and 51 restored from c. 79r, with none unresolved. The 192 read by the key (79.0%) are under the 95% bar, so the page calls the reading partial. With c. 79r the whole text is known, but c. 79r retouches the wording in places (listed in reading_76r.md); the cipher is taken as the source. One transcriber (an AI agent), one pass, no independent re-reading.
The same gathering also holds c. 78r, the Farnese letter. Known keys were tried on it by criteria fixed in advance, all negative: the Klee key of vatican5 (an M5 score of 0.098, measured before the c. 76r work) and A. Meister, Die Geheimschrift im Dienste der päpstlichen Kurie (1906), pp. 176–177, tables 1 and 2 (0.174 and 0.183). The key of c. 76r was not applied to it, because under the protocol that needs a validated key.
04 What remains uncertain
- 51 of the 243 words of c. 76r rest on c. 79r: probable nomenclature codes for Ambasciatore, Francia and Inghilterra, and rare homophones not seen at least twice in lines 1–13. They can be filled from the restored words; the key was deliberately not extended after the measurement.
- c. 78r, the Farnese letter of catalogue 2339, is not read: the known keys are negative, the c. 76r key is only partial, and no decipherment is in the gathering.
- No edition or study of the gathering was found (Comelli’s 2022 edition of the Della Casa–Farnese correspondence, vol. II, was checked in full text during the vetting of the lead; DECODE, Cryptiana and Lasry gave nothing).
Full texts
The transcription and the decryption of every document on this page, so that the reading can be used and checked. Short files are shown here; all are in the repository.
c. 76r (cipher) with c. 79r (decipherment of the time), Rome 10 Aug 1547 read in part
The writer reports that the French ambassador at Rome advised his young King (taken to be Henri II) to keep what his father left him rather than seek glory in a war; but the King's ardour, encouraged by great men of France, cannot be held back, and the ambassador is urged to work for union with England.
Builds coherent sentences, but under the read bar of 95% of words read as sense (79% here); the holes are marked in the reading.
Transcription
# ASF Mediceo del Principato 384 c. 76r, "Copia d'un Cap.lo d'una l.ra di Roma de x d'Agosto 1547": the cipher, 18 lines (L01-L18).
# Transcribed 9 Nov 2026 by a Claude Opus 5.5 agent (one pass, strips of the native 2000 px ASF PAD image, enlarged), WITHOUT looking at c. 79r.
# Frozen by SHA-256 before any alignment (working file transcription_76r.txt, fbb77fae...). Notation: ^ = a dot over the preceding digit
# (in this hand 1 is often written as a dotted i; meaning uncertain); {a/b} = doubtful digit, a kept; {x} = blot. Held-out part = L14-L18.
L01 74728^0128 93^07705 5^5 36010270907701505077027508210462 37^480450217 0{x}427691270517014923049 7^47090304
L02 1583^8348042321905518 0^37010129179521370821520382170 9 1^0148236042762091562167037 1^65321703750705 3708
L03 51603041582165205862191289237^4605076216048129176052369521592107921047040349052360562161^71^01270
L04 6042392381290450709238519121495121624910605528156210592104702 4 1^0701251604705162170807329180423901
L05 21817536212 8^40605230190167037351^5912107017532160491^247950615995160427814621^243850805804506055181 57^2
L06 516052321817812237024348037^470522190553601705080370241^01304705162194027701^0601952708159056407^9
L07 915615923617538517014823621157027701762 37^4621613177052037082170850429047080570517{0/x}1^71^05203706218058
L08 1780423216519502070146211505953904277080568536211505912232191291^3047052127091592 4 1^3478212391599 1223
L09 2170176052370737283^1^49211770320523605516216 1^04715042765192381^870375170237089 2^49 3^519 1^2132360423910 3
L10 7012610621923905238129215182161^07012442 0915615907701762 37^4621047093515 9 1^21070175321604 9 1^0045321805
L11 923815705199125015057052219055170517053^126219538042770107012 6 1^3236056210388217023702192112851 9 1^604158
L12 516{0/x}4705162170824 1^0912157027239159612479210370 1^7853621910527015517050 9 1^591302605702104709125221951 8
L13 1^023605615940912508058045062160370201517821705170 5 3^409104770376{0/x}4701781560427650207032051623{6/x}17
L14 537015612804537037526042764 0 1^01282151701297037217050821079211289 2^4444070055062150921651702370 21
L15 92105612862180523605152370517053850890570178727 4 8^2^1462104159046128905508219 2^4 9^39479216852057 0
L16 274782154851702723921615057023 8 1^3 5 8^40532185015219 2 1{6/x}0770521595195205921046128700523621917605236 2 6
L17 1^01250805 8{0/x}450615915709 2^747950701268 2^7953621239947921128232191595212709216703750904904159 10
L18 272392104650851702715057023872747 1
# note: L02 : « 1803 » point au-dessus du 0 (ou du 8). L07 : 0 de « 517●1 » rempli d'encre.
# note: L09 : petit « 3 » en exposant à la fin de la ligne, gardé. L10 : « 4420 », les deux derniers chiffres repassés/raturés.
# note: L14 : « 4444 » repassés (barre horizontale), nombre de 4 incertain (3 ou 4). L16 : fin « 23626 » ou « 2362b ».
# note: L18 : suivie d'un trait de fin.Decipherment of the time
# ASF Mediceo del Principato 384 c. 79r: the contemporary decipherment of c. 76r (same heading), diplomatic transcription, one line per line.
# Transcribed 9 Nov 2026 by a Claude Opus 5.5 agent before any alignment (SHA-256 20f273cb...). Abbreviations expanded in ( ); doubts {a/b} (a kept).
# The additions in another hand at the foot of the page are left out.
C01 L'Amb(asciato)re di Francia per via d'un discorsetto scrisse alli di passati al suo Re, che la maggior
C02 gloria, che potesse haver' in questo principio del suo regno, era solamente l'attender' a guardarsi;
C03 et mantener' quello che gli è stato lasciato da suo padre, et a ben governarsi: et che il de-
C04 siderio d'acquistar' col romper guerra gli faceva adosso per lui; conoscendovi dentro molti
C05 et molti gran' pericoli; et che considerasse ch'era giovane, et haveva tempo di {poterni/potersi}
C06 metter' al{li} migliori occasioni di queste, che gl'{haveriano} mostre da Dio per una {passata} per
C07 un'altra lo consigliava, et dissuadeva altrimente: Parmi intender' che fusse dato giu-
C08 dicio a detto Amb(asciato)re che con tanto d'ardor' che'l suo Re haveva di mettersi a qualche im-
C09 presa, che sperandosi da alcuni grandi di Francia, non era possibile ritenerlo, et ritirarlo
C10 da questo suo desiderio, et che a lui stava el {farvi} quella opera che giudicava conve-
C11 nirsi al grado che teneva, et al conoscer' gli humori d'Italia, et a che fine si movesse
C12 che in questi tempi ve lo spingesse; et che lor di là vi ricordarno a detto Re, et i {cercal}
C13 di ritirarnelo erano tenuti di poco animo: Persuaselo ancora del Amb(asciato)re a far' ogni
C14 sforzo di stare unito {col/el} Inghilt(er)ra, et procurar' di guadagnarselo interamente
C15 poiche ne conosceva via più facile, o più aperta a farlo travagliar' i casa sua che
C16 star seco disunito: —Decryption
# c. 76r read with the c. 76r key, word by word against the contemporary decipherment c. 79r (9 Nov 2026)
Status: read in part by the key (79.0% of words certain or probable, under the 95% rule); the whole text is known from the
decipherment of the time on c. 79r, so nothing here is a new text.
Conventions: plain word = every letter given by the key (certain); {x} = probable (at most a third of the letters missing,
the rest agreeing); [x] = restored from c. 79r (the cipher uses codes not yet in the key); no word is unresolved, since
c. 79r covers the whole text. The words are those of c. 79r; where the cipher reads otherwise, see "Divergences" below.
| words | certain | probable | restored | unresolved | read (certain + probable) |
|---|---|---|---|---|---|
| 243 | 131 (53.9%) | 61 (25.1%) | 51 (21.0%) | 0 | 192 (79.0%) |
Characters of the clear found by the key: 835 of 1,106 (75.5%).
## Text (Italian, 1547; c. 79r wording, marked by what the key gives on c. 76r)
{l} [ambasciatore] di [francia] per [via] d {un} discorsetto {scrisse} {alli} di {passati} al {suo} {re} che la {maggior} gloria che potesse {haver} {in} [questo] {principio} del {suo} regno era [solamente] l [attender] a guardarsi et mantener [quello] che {gli} e stato lasciato da [suo] padre et a {ben} {governarsi} {et} che il de siderio d [acquistar] [col] romper {guerra} [gli] faceva {adosso} per [lui] {conoscendovi} dentro molti et molti gran pericoli {et} che [considerasse] ch era giovane {et} {haveva} tempo di {poterni} metter [alli] migliori {occasioni} di [queste] che gl [haveriano] mostre da [dio] per {una} [passata] per {un} altra {lo} {consigliava} et [dissuadeva] {altrimente} parmi [intender] che {fusse} dato [giu] {dicio} a [detto] [ambasciatore] che [con] tanto {d} ardor che {l} [suo] [re] {haveva} di mettersi a {qualche} im presa che [sperandosi] da alcuni grandi {di} [francia] [non] era {possibile} {ritenerlo} [et] ritirarlo da [questo] suo desiderio {et} che a {lui} stava el [farvi] [quella] opera che giudicava [conve] nirsi al grado che teneva {et} al conoscer {gli} {humori} d [italia] {et} a che {fine} si movesse {che} {in} [questi] tempi ve lo {spingesse} {et} che lor di la [vi] {ricordarno} a [detto] [re] {et} {i} {cercal} [di] {ritirarnelo} [erano] tenuti di poco animo {persuaselo} ancora [del] [ambasciatore] {a} [far] ogni {sforzo} di stare [unito] [col] [inghilterra] {et} procurar di {guadagnarselo} interamente [poiche] [ne] conosceva via [piu] facile o [piu] aperta a farlo travagliar i casa {sua} che star seco {disunito}
## Why words are restored
The restored words fall on codes absent from the key (rule: seen at least twice in L01-L13): mostly the opening ("L'Amb.re
di Francia", probably nomenclature codes: 74 72 8^ 01 28 9 3^0 77 ...), Amb(asciato)re (three times), Francia, Inghilterra,
piu, sperandosi, and rare homophones (13, 37, 16, 35, 02, 42, 38, 18, 72, 32 ...). No restoration contradicts a letter the key
gives; the remaining letter-by-letter differences are divergences of the text.
## Divergences cipher (76r) / decipherment (79r)
The cipher reads otherwise in a few places; c. 79r is the decipherment of the time with editorial retouches, and the cipher is
the source:
- 76r "al suo Re **mostrandoli** che la maggior gloria" (79r "al suo Re, che la maggior gloria");
- 76r "romper guerra **o entrare in qualche impresa** ... faceva adesso per lui" (79r "gli faceva adosso");
- 76r "mostre da **chi**" (79r read "da Dio"); 76r "per una **passione** o per un'altra" (79r read "passata": our reading of
79r is probably wrong there);
- 76r "... **portava** altrimenti" (79r "dissuadeva altrimente"); 76r "ditio" (79r "giudicio");
- 76r "possibile ritenerlo **o** ritirarlo" (79r "et ritirarlo"); "farli opera" (79r "farvi quella opera");
- 76r "ricordarlo" (79r "ricordarno"); "in cercare di ritirarnelo" (79r "et i cercal(?) di ritirarnelo").
Mechanical decode of the whole of 76r with the frozen key (? = code not in the key):
```
t??or?oa?di??per?s?adivndiscorsettoscri?sealidipatsatialsvo?otmostrandolichelamag??iorgloriachepotessehauerei?
??epri??cipiodelsvoregnoeraloattenderesolamenteaguardarsietmanteneremcheliestatolasciatodalpadreetacheengouern
arsiecheildesideriod??a?ostarromper?oeentrarei???d?alcheimpresalfacevaadessopere?oconoscendov??dentromoltietmo
ltigranpericolieche??sideraesicheeragiovanee?aavevatempodipotervisimettere??migliorioc??asionidimcheglieronomo
stredachipervnapassioneopervnaltrala??sigliavae?tm?ortavaaltrimenti????o?parmi?hg?enderechefvssedato?oi??ditio
aea?dal??testache?ilecheeratantolardorecheil??haueuadimettersia?d?alcheimpresachespronatovidaalcunigrandidiple
rapossicheile??tenerlooritirarloda?esuodesiderioechealvistavaelfarlimoperachegiudicava????enirsialgradochetene
vaealconoscerelihvmoridipeachefi??esimovessechi?hdetempivelospi??gesseechelordilai??ricordarloae??eincercaredi
tirarneloerontenutidipocoanimo?s?or?ersuaseloancoraea?afareognisfor??odistarevnito???oeprocuraredigvadagnarsel
ointeramente?echelconoscevaviasfacileosapertaafarlotravagliarei??casasvachestarsecodisvnito?sf?
```Reading
# c. 76r read with the c. 76r key, word by word against the contemporary decipherment c. 79r (9 Nov 2026)
Status: read in part by the key (79.0% of words certain or probable, under the 95% rule); the whole text is known from the
decipherment of the time on c. 79r, so nothing here is a new text.
Conventions: plain word = every letter given by the key (certain); {x} = probable (at most a third of the letters missing,
the rest agreeing); [x] = restored from c. 79r (the cipher uses codes not yet in the key); no word is unresolved, since
c. 79r covers the whole text. The words are those of c. 79r; where the cipher reads otherwise, see "Divergences" below.
| words | certain | probable | restored | unresolved | read (certain + probable) |
|---|---|---|---|---|---|
| 243 | 131 (53.9%) | 61 (25.1%) | 51 (21.0%) | 0 | 192 (79.0%) |
Characters of the clear found by the key: 835 of 1,106 (75.5%).
## Text (Italian, 1547; c. 79r wording, marked by what the key gives on c. 76r)
{l} [ambasciatore] di [francia] per [via] d {un} discorsetto {scrisse} {alli} di {passati} al {suo} {re} che la {maggior} gloria che potesse {haver} {in} [questo] {principio} del {suo} regno era [solamente] l [attender] a guardarsi et mantener [quello] che {gli} e stato lasciato da [suo] padre et a {ben} {governarsi} {et} che il de siderio d [acquistar] [col] romper {guerra} [gli] faceva {adosso} per [lui] {conoscendovi} dentro molti et molti gran pericoli {et} che [considerasse] ch era giovane {et} {haveva} tempo di {poterni} metter [alli] migliori {occasioni} di [queste] che gl [haveriano] mostre da [dio] per {una} [passata] per {un} altra {lo} {consigliava} et [dissuadeva] {altrimente} parmi [intender] che {fusse} dato [giu] {dicio} a [detto] [ambasciatore] che [con] tanto {d} ardor che {l} [suo] [re] {haveva} di mettersi a {qualche} im presa che [sperandosi] da alcuni grandi {di} [francia] [non] era {possibile} {ritenerlo} [et] ritirarlo da [questo] suo desiderio {et} che a {lui} stava el [farvi] [quella] opera che giudicava [conve] nirsi al grado che teneva {et} al conoscer {gli} {humori} d [italia] {et} a che {fine} si movesse {che} {in} [questi] tempi ve lo {spingesse} {et} che lor di la [vi] {ricordarno} a [detto] [re] {et} {i} {cercal} [di] {ritirarnelo} [erano] tenuti di poco animo {persuaselo} ancora [del] [ambasciatore] {a} [far] ogni {sforzo} di stare [unito] [col] [inghilterra] {et} procurar di {guadagnarselo} interamente [poiche] [ne] conosceva via [piu] facile o [piu] aperta a farlo travagliar i casa {sua} che star seco {disunito}
## Why words are restored
The restored words fall on codes absent from the key (rule: seen at least twice in L01-L13): mostly the opening ("L'Amb.re
di Francia", probably nomenclature codes: 74 72 8^ 01 28 9 3^0 77 ...), Amb(asciato)re (three times), Francia, Inghilterra,
piu, sperandosi, and rare homophones (13, 37, 16, 35, 02, 42, 38, 18, 72, 32 ...). No restoration contradicts a letter the key
gives; the remaining letter-by-letter differences are divergences of the text.
## Divergences cipher (76r) / decipherment (79r)
The cipher reads otherwise in a few places; c. 79r is the decipherment of the time with editorial retouches, and the cipher is
the source:
- 76r "al suo Re **mostrandoli** che la maggior gloria" (79r "al suo Re, che la maggior gloria");
- 76r "romper guerra **o entrare in qualche impresa** ... faceva adesso per lui" (79r "gli faceva adosso");
- 76r "mostre da **chi**" (79r read "da Dio"); 76r "per una **passione** o per un'altra" (79r read "passata": our reading of
79r is probably wrong there);
- 76r "... **portava** altrimenti" (79r "dissuadeva altrimente"); 76r "ditio" (79r "giudicio");
- 76r "possibile ritenerlo **o** ritirarlo" (79r "et ritirarlo"); "farli opera" (79r "farvi quella opera");
- 76r "ricordarlo" (79r "ricordarno"); "in cercare di ritirarnelo" (79r "et i cercal(?) di ritirarnelo").
Mechanical decode of the whole of 76r with the frozen key (? = code not in the key):
```
t??or?oa?di??per?s?adivndiscorsettoscri?sealidipatsatialsvo?otmostrandolichelamag??iorgloriachepotessehauerei?
??epri??cipiodelsvoregnoeraloattenderesolamenteaguardarsietmanteneremcheliestatolasciatodalpadreetacheengouern
arsiecheildesideriod??a?ostarromper?oeentrarei???d?alcheimpresalfacevaadessopere?oconoscendov??dentromoltietmo
ltigranpericolieche??sideraesicheeragiovanee?aavevatempodipotervisimettere??migliorioc??asionidimcheglieronomo
stredachipervnapassioneopervnaltrala??sigliavae?tm?ortavaaltrimenti????o?parmi?hg?enderechefvssedato?oi??ditio
aea?dal??testache?ilecheeratantolardorecheil??haueuadimettersia?d?alcheimpresachespronatovidaalcunigrandidiple
rapossicheile??tenerlooritirarloda?esuodesiderioechealvistavaelfarlimoperachegiudicava????enirsialgradochetene
vaealconoscerelihvmoridipeachefi??esimovessechi?hdetempivelospi??gesseechelordilai??ricordarloae??eincercaredi
tirarneloerontenutidipocoanimo?s?or?ersuaseloancoraea?afareognisfor??odistarevnito???oeprocuraredigvadagnarsel
ointeramente?echelconoscevaviasfacileosapertaafarlotravagliarei??casasvachestarsecodisvnito?sf?
```Translation
# c. 76r / c. 79r in English (approximate translation of the c. 79r text, Brice Dudout with Claude, 9 Nov 2026) Copy of a chapter of a letter from Rome of 10 August 1547. The ambassador of France, by way of a short discourse, wrote some days ago to his King that the greatest glory he could win at this beginning of his reign was simply to look to his own guard, to keep what his father had left him and to govern himself well; and that the desire to gain by breaking into war went against him, since he saw in it many great dangers; and that he should consider that he was young and had time to take up better occasions than these, which God would show him at one time or another; and so he counselled and dissuaded him otherwise. I understand that the said ambassador was given to judge that, with the ardour his King had to throw himself into some enterprise, which some great men of France were hoping for, it was not possible to hold him back from this desire, and that it was for him to do the work he thought fitting to his rank and to his knowledge of the humours of Italy, and of the end towards which [the King] was moving, that urged him to it in these times; and that those over there who reminded the said King of this and tried to draw him back were held to be of little spirit. He further persuaded the ambassador to make every effort to stay united with England and to win it over entirely, since he saw no easier or more open way to have [the Emperor] troubled in his own house than to be disunited from it. Context (not checked against an edition): Henri II had succeeded Francis I on 31 March 1547; the ambassador is not named in the text. The last sentence is the hardest of the text and its sense is uncertain.
Key: key_76r.tsv
05 Sources
- Archivio di Stato di Firenze, Mediceo del Principato 384 (Carteggio universale di Cosimo I), cc. 76r, 78r and 79r; the images are not reproduced here.
- Brice Dudout, pull request #344 and the target folder: the key, the test, the transcriptions and the reading.
The documentation, the key, the transcription, the reading and the tests are in targets/farnese1547.
Cite this page
Daniel Bourdeau and Brice Dudout, “A letter from Rome on the French ambassador’s advice, 10 August 1547”, Cyphersolver, posted 9 November 2026, https://cyphersolver.com/farnese1547.html.