At a glanceExternal key matched · partial (35% of the cipher)
- The letter
- Sixteen lines of cipher (787 tokens) with one clear phrase, “Poi un altra volta”; no date, place, address or signature. The addressee is probably Lorenzo de’ Medici; the writer is not identified.
- Where
- Archivio di Stato di Firenze, Mediceo avanti il Principato, filza XLVIII, doc. 214 (old c. 222), recto. Catalogue entry 2309. Not on DECODE.
- The cipher
- A substitution with homophones and nulls, in the sign set of Siena no. 11 (Donato Acciaiuoli to Lorenzo de’ Medici, Rome, 1478): the same signs and the same nulls.
- How it was read
- George Lasry’s key of Siena no. 11 was applied unchanged; it covers 96.4% of the tokens. A test on this letter failed (0.2335 against at most 0.0719); a held-out test on another letter of the family against its clear copy gave 0.6419 against at most 0.2045: partial.
- What it says
- Condottieri and their contracts, Count Antonio and something he had in Venice, Count Carlo, and a sum of “uindici milia ducati” (15,000 ducats).
- Still open
- About 60–65% of the letter; an untested derived rule; 14 sign forms not in Lasry’s key; the date, writer and addressee.
- Credit
- The key is George Lasry’s. Brice Dudout (L’Europe chiffrée) recognised the cipher, transcribed the letter, tested the key and read it in part, working with Claude Opus 5.5 agents. The catalogue entry and its vetting are Daniel Bourdeau’s.
01 The letter
Doc. 214 of filza XLVIII of the Mediceo avanti il Principato is a single leaf: sixteen lines of cipher on the recto, one phrase in clear in line 3 (“Poi un altra volta”) and an illegible closing paraph; the verso is blank. No decipherment is on the leaf or filed with it. It was transcribed sign by sign from the archive’s image at 1272 px, the largest the viewer gives, using the sign names of the Siena no. 11 transcription and new names for forms not found there. The images were used for private research and are not reproduced here.
The sign set is that of Siena no. 11 (Concistoro 2308, DECODE R4800), including its nulls. Compared before any decoding, the keys at hand cover the tokens of this letter as follows: Siena no. 11 96.4%, MAP LXXII 183 65.3%, the Dieci key of 1414 49.4%, MAP CXIX 117 39.0%.
02 The key and its tests
The key is George Lasry’s for Siena no. 11, written in sign names with no value changed; where a sign has two values in his figure, the one with the larger count was taken. Signs not in his key (3.6% of the tokens) give nothing. One derived rule, D1 (“PLUS ODAG” read as one sign = c), was proposed after the first test from decodings such as “hon / honte” where “con / conte” is expected. It is used in the reading and is untested.
- On this letter: failed. One measurement on lines 6–16 (534 tokens), the share of letters falling in Italian words of five letters or more: 0.2335 against at most 0.0719 for 200 shuffled keys, a margin of 0.16, under the 0.30 required. The same script on Siena no. 11 itself gives only 0.396, so the measure is harsh, but the verdict stands.
- Held out on XLVIII 294: partial. XLVIII 294 (Bernardo Rucellai, February 1485) uses the same sign family and has a clear copy of the time at c. 276. One measurement on c. 306v lines 1–9 (339 tokens): letter agreement 0.6419 against at most 0.2045 for 200 shuffled keys, a margin of 0.437; the margin passes, 0.85 is not reached. The portion has no “PLUS ODAG”, so D1 was not exercised.
The two decodings were re-run here from the key files and the transcriptions: the XLVIII 294 test portion gives the same 313 letters, and the decoding of XLVIII 214 agrees with the published one except for a few signs between dots.
03 The reading
Read word by word on the decoding: 39 certain words, 13 probable, 11 with restored letters, and 45 unresolved stretches. The certain and probable words cover about 33% of the 665 decoded letters, about 43% with the restored ones: about 35–40% read. In the fragments below, {braces} mark probable words, [brackets] restored letters and ‹…› unresolved stretches.
‹…› {el} conte antonio egli haueua ‹ma…› in vinegia ‹…› {col} conte ca[r]lo ‹…› si ‹…› non o[b]tenend[o] ‹…› {la libe[ra]tione} ‹…› con docta ‹…› uindici milia ducati {i[n]} diciotto ‹…›
What reads is a matter of condottieri and their contracts (“conducti”): Count Antonio and something he “had” in Venice, Count Carlo, a sum of 15,000 ducats and the word “diciotto”; further on a “liberatione”, an “intentione” of “el signore”, “non potendo”, “utile a questo”, “in questo modo”. These are fragments, not the sense of the whole letter.
The date is a hypothesis only. If Count Carlo is Carlo Fortebraccio (died 1479) and Count Antonio is Antonio da Marsciano, the letter falls about 1478–79, in the Pazzi war; otherwise about 1482–85, in the War of Ferrara (the filza holds papers of 1482–83).
04 What remains open
- About 60–65% of the letter: 45 unresolved stretches and 11 restored words. The text decodes with the key but long stretches do not read as words, and 14 sign forms are not in Lasry’s key.
- D1 (“PLUS ODAG” = c) is untested; it needs a held-out portion of the family that contains the pair and has a clear.
- The key is partial for this family (0.6419 against 0.85), and the transcription was made at 1272 px; a higher-resolution image from the archive would settle doubtful forms.
- Date, writer and addressee: nothing is written on the leaf, and the names read do not settle the date.
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.
ASF MAP XLVIII 214r, undated cipher letter (catalogue 2309) read in part
Fragments of a letter, probably to Lorenzo de' Medici, about condottieri and their contracts: Count Antonio and Count Carlo, dealings in Venice, a sum of 15,000 ducats, a 'liberation', an intention of 'the lord'; one clear phrase 'Poi un altra volta'.
Builds coherent sentences, but under the read bar of 95% of words read as sense (35% here); the holes are marked in the reading.
Transcription
# cat. 2309 = ASF MAP XLVIII 214r (id60059.jpg sha256 b4dd1157...), 1272 px. Transcription par formes, noms de signes = légende de l'agent B (cyphersolver targets/siena11/transcripts/no11.txt) + noms nouveaux (voir LEGENDE). # Une ligne = une ligne du manuscrit. [..] = clair. XX = signe biffé/taché (exclu). UNK = illisible / paraphe. L01 Q TAU OBAR PLUS O P D II DAG2 3 HASH TAU O PLUS S HASH3 PLUS ODAG OMEGA HASH QB2 O PI HASH PERP OMEGA OSTEM Q M^n PERP G HASH O PLUS 3 OBAR HASH3 HASH DAG2 OBAR B 3 2 PLUS OSTEM P S WAVE PLUS O OBAR Q HASH II L02 DAG2 PERP 7 A B 6 ODAG DAG2 QB2 O PLUS II Q 3 5 SL2 OBAR D C DAG2 3 NX II PLUS ODAG OMEGA O PLUS OSTEM QB2 PLUS O II ZX S B Q PLUS O DAG2 NX II P D HASH3 DAG2 S HASH OBAR TRI DAG2 OSTEM 7 L03 P PLUS O DAG2 II FB PERP DOT 3 1 DOT O PLUS S PLUS ODAG TAU OMEGA O PLUS A DAG2 HASH 3 QB2 Q OBAR Q S REV3 II HASH3 3 B HASH B DOT [Poi un altra volta] DOT 8 7 DOT PLUS ODAG TAU OMEGA P SL3 L04 O PLUS D DAG2 S OSTEM OMEGA HASH S PI DAG2 OSTEM PERP OMEGA QB2 QB2 OBAR DAG2 OSTEM DAG2 QB2 PLUS O C O S P B 3 2 OMEGA TAU QB2 QB2 NX II DAG2 SL2 ODAG Q DIV DOT 3 1 DOT PLUS ODAG S ZX OBAR O A PERP DIV L05 HASH S C O PLUS QB2 TB 7 OMEGA OSTEM DAG2 3 B PB S DAG2 HASH PERP SL3 NX II OSTEM 3 ZX S PLUS ODAG QB2 O PLUS OSTEM 2 OBAR II ODAG OBAR A QB2 6 OSTEM Q II PLUS DAG2 PERP HASH B HASH3 S ODAG SL3 L06 PERP C S DAG2 II SL2 ODAG TAU NX S QB2 D ZX S PLUS ODAG P SL2 ODAG TAU OMEGA DAG2 QB2 II OMEGA HASH TAU PI OSTEM PERP TRI QB2 7 P 3 ODAG II A QB2 6 D NX D OMEGA QB2 OBAR DAG2 PERP OBAR A 3 PI S 7 OSTEM D DIV L07 O A S REV3 PERP OMEGA TAU N DAG2 3 SL2 ODAG II NX OSTEM 3 OBAR OMEGA TAU QB2 II B DAG2 S OBAR PERP B DAG2 II A OBAR 7 3 B SL2 ODAG TAU P PLUS ODAG TAU OMEGA DAG2 QB2 PLUS ODAG II PERP P PERP C PLUS O PERP 2 3 6 B L08 HASH OSTEM Q OBAR SL2 ODAG II DAG2 B 3 SL2 ODAG TAU OMEGA XX DOT Q Q DOT OMEGA TAU PI PERP DAG2 HASH S OMEGA QB2 PI O PLUS Q QB2 II P PERP DAG2 OSTEM O A P PLUS ODAG S PLUS ODAG TAU OMEGA NX DAG2 O L09 DAG2 3 PERP PI QB2 O PLUS S C DAG2 QB2 7 OSTEM PERP OMEGA 3 DAG2 S DAG2 P II HASH3 OSTEM EQ S QB2 OBAR OSTEM DAG2 OSTEM PERP OMEGA S B A OBAR 3 1 SL3 DOT DIV UNK M^n C OO M^n SL3 DOT Q Q L10 ST 8 7 DOT DIV A 3 II PLUS ODAG TAU DAG2 QB2 OBAR DAG2 QB2 O PLUS 3 QB2 OBAR N B DAG2 S DOT DPLUS DOT OBAR OMEGA II HASH3 DAG2 PERP OBAR PLUS ODAG TAU PI O PLUS PERP SL2 DAG2 II QB2 ODAG O 6 3 OMEGA O PLUS OSTEM SL3 DOT L11 SL2 3 NX OSTEM P OBAR II O PLUS A PLUS ODAG II HASH 3 M^n OBAR SL3 DOT O PLUS 3 SL2 OBAR PERP DAG2 HASH PERP M^n DAG2 3 PLUS ODAG TAU QB2 O PLUS PERP DOT CIRC1 DOT CIRC2 DOT DAG2 S PERP P II QB2 3 PIBAR 3 DIV L12 PLUS ODAG Q HASH3 DAG2 OBAR DOT TRI DOT 6 S O PLUS QB2 HASH S G HASH DOT Q SL3 DOT ODAG DAG2 DAG2 OBAR QB2 DAG2 QB2 DOT 1 4 DOT QB2 3 6 A 3 DIV M^n DOT Q Q SL3 DOT OO L13 M^n NX S B QB2 QB2 DAG2 QB2 OBAR ODAG RHO SL2 TALL DAG2 TAU DAG2 QB2 PERP DAG2 DAG2 HASH TAU ZX 3 QB2 OSTEM SL2 PLUS O OMEGA TAU PI C TAU DAG2 S PERP HASH QB2 II DAG2 QB2 QB2 O PLUS OSTEM QB2 D 3 S P B DAG2 QB2 S B DAG2 II L14 OMEGA OBAR II QB2 DAG2 DOT TRI DOT 7 3 Q O PLUS OSTEM PLUS ODAG PLUS O DAG2 S TRI QB2 A HASH 3 P PLUS O II O 6 S B DAG2 PERP DAG2 DAG2 TAU C PERP B 3 HASH OSTEM DOT TRI DOT Q OBAR Q OSTEM NX PERP B HASH QB2 OBAR L15 B QB2 OBAR SL3 DOT Q PERP 3 OMEGA TAU PI C OSTEM HASH S B QB2 OBAR ODAG II A Q QB2 PLUS O P II D G OBAR S DAG2 D OMEGA OBAR O QB2 II DOT 3 2 DOT DAG2 S QB2 NX A TAU 6 S DAG2 PLUS O C PLUS O PERP DIV L16 D A S B DAG2 TAU NX TAU QB2 PERP DOT DPLUS DOT II P C B 7 ODAG OBAR DAG2 S UNK
Decryption
soielaatitodelcntentonisodiilttisigileeista togusuhtedasiciaptimacndieearessetmalaltetitig letaodeconduttiesiselalists «Poi un altra volta» conl dateintentioneeititeepelsignoeematchsceriuo tepdegnitisettomaiirecedigiahiueuisatotsleh opetachomeeareclchonteantonioeglihaueuamaneitoiuinegia uelonoftichamiiinoeasteiostauigischolcontecaolopeogius tisichatsichonssnonottenendsealotiulceconmt tionedeptegionitetlalibeeiitionesuiipss pguiacoteitedieifsteinaltoicondoctaehuindi cimiliaducatiidiciottoticoedoteolaeii csltiuedeteshttieteeiuuiss meseeteihoctoteotttorieicenonpoteoteateedieaielsteesta niaetgisdiceteeutileauestottopositisisimostei seisoinonpiteseihauseelaaietanieateemuouetepeo auestomoeoalpsghite
Reading
# Reading: undated cipher letter, ASF Mediceo avanti il Principato XLVIII 214 (catalogue 2309), Siena no. 11 cipher: read in part
Key: `key/K_SIENA11_D1.tsv` = George Lasry's key of Siena no. 11 (`key/K_SIENA11.tsv`, frozen, SHA-256 e1ef0955...) plus one
derived rule D1 ("PLUS ODAG" taken as one sign = c). **D1 is untested** (see `validation.txt`): it was proposed after the
measurement on this page, and the held-out check on XLVIII 294 contains no "PLUS ODAG", so it neither confirms nor refutes it.
Mechanical decoding: `decryption/xlviii214_D1.txt`. Conventions (Daniel's): plain = certain; {x} = probable; [x] = letters
restored; ‹…› = unresolved stretch (of unknown length).
Counts, word by word, of the reading column: **39 certain**, **13 probable**, **11 words with restored letters**, **45 unresolved
stretches**. By letters of the decoding (665 letters without the clear phrase), the certain and probable words cover about 33 %,
about 43 % with the restored words: **about 35-40 % read**, far under the 95 % bar. One clear phrase stands in the letter
("Poi un altra volta", l. 3).
| l. | decoding (K_SIENA11_D1) | reading |
|---|---|---|
| 1 | soielaatitodelcntentonisodiilttisigileeista | ‹…› del c[o]ntento ‹…› ‹…› |
| 2 | togusuhtedasiciaptimacndieearessetmalaltetitig | ‹…› {da} ‹…› c[o]n dire ‹…› |
| 3 | letaodeconduttiesiselalists / [Poi un altra volta] / conl | {lettera} ‹…› conducti e ‹…› la lista ‹…› ; *Poi un altra volta* (clear) ; con l‹…› |
| 4 | dateintentioneeititeepelsignoeematchsceriuo | ‹…› intentione ‹…› per el signo[r]e ‹…› ma ch[e] {scriuo} |
| 5 | tepdegnitisettomaiirecedigiahiueuisatotsleh | ‹…› ‹…› {mai} ‹…› di gia h[a]ueui {sa[pu]to} ‹…› |
| 6 | opetachomeeareclchonteantonioeglihaueuamaneitoiuinegia | ‹…› {come} ‹…› {el} conte antonio egli haueua ‹ma…› in vinegia |
| 7 | uelonoftichamiiinoeasteiostauigischolcontecaolopeogius | ‹…› ‹…› ‹…› {col} conte ca[r]lo ‹…› |
| 8 | tisichatsichonssnonottenendsealotiulceconmt | ‹…› si ‹…› non o[b]tenend[o] ‹…› con ‹…› |
| 9 | tionedeptegionitetlalibeeiitionesuiipss | ‹…›tione ‹…› {la libe[ra]tione} ‹…› |
| 10-11 | pguiacoteitedieifsteinaltoicondoctaehuindi / cimiliaducatiidiciottoticoedoteolaeii | ‹…› con docta ‹…› uindici milia ducati {i[n]} diciotto ‹…› |
| 12 | csltiuedeteshttieteeiuuiss | ‹…› uedete ‹…› |
| 13 | meseeteihoctoteotttorieicenonpoteoteateedieaielsteesta | {mese} ‹…› {che} non pote[nd]o ‹…› |
| 14 | niaetgisdiceteeutileauestottopositisisimostei | ‹…› utile a questo ‹…› positi ‹…› si mostr[a] |
| 15 | seisoinonpiteseihauseelaaietanieateemuouetepeo | ‹…› non ‹…› ha[ue]re ‹…› |
| 16 | auestomoeoalpsghite | questo mo[d]o al ‹…› |
The decoding is printed as the frozen key gives it; the clear phrase of l. 3 is shown between [ ]. Signs that are not in
Lasry's key (DIV, TRI, OO, DPLUS, WAVE and nine single forms: 3.6 % of the 787 tokens) produce nothing and break the stream.
## Date, writer, addressee (from the content only; nothing is written on the leaf)
- **Addressee**: probably Lorenzo de' Medici (the fonds; "el signore"; the polite "vedete").
- **Writer**: not identified. The letter uses the cipher of Siena no. 11 (Donato Acciaiuoli, Rome, 1478), which Bernardo Rucellai
also uses in February 1485 (XLVIII 294): a Medici chancery cipher handed to more than one envoy. The closing paraph is illegible.
- **Date**: undated. Hints: "conte Antonio", "conte Carlo", Venice, 15,000 ducats, "condutti" (condotte). If conte Carlo is Carlo
Fortebraccio da Montone (d. 1479) and conte Antonio is Antonio da Marsciano, the letter would fall about 1478-79 (Pazzi war);
otherwise 1482-85 (War of Ferrara; Daniel's catalogue dates the filza 1482-83). **A hypothesis, not established.**Translation
# Translation (English summary): undated cipher letter, ASF MAP XLVIII 214 (catalogue 2309), read in part
An undated, unaddressed cipher letter, most likely to Lorenzo de' Medici, written in the cipher of Siena no. 11 (George Lasry's
key). The text reads only in fragments: a matter of condottieri and their contracts ("conducti"), Count Antonio and Count Carlo;
something Count Antonio had (done or held) in Venice; a sum of fifteen thousand ducats and the word "eighteen" (of what is not read); a "liberation";
an intention attributed to "the lord"; "not being able to", "useful to this", "in this way". A single phrase is written in the
clear: "Poi un altra volta" ("Then another time").
This is a summary of the fragments that read, not a translation: about 35-40 % of the letter is read, and nothing here should be
taken as the sense of the whole. Readings are marked word by word in `reading/xlviii214.md`.
## Riassunto (italiano)
Lettera cifrata senza data né indirizzo, probabilmente a Lorenzo de' Medici, nel cifrario di Siena n. 11 (chiave di George Lasry).
Si leggono solo frammenti: condotte di capitani, il conte Antonio e il conte Carlo, Venezia, quindicimila ducati, una
"liberatione", un'intenzione attribuita al "signore". Un solo passo in chiaro: "Poi un altra volta". Lettura parziale (35-40 %).Key: K_SIENA11.tsv
05 Sources
- Archivio di Stato di Firenze, Mediceo avanti il Principato XLVIII 214r (riproduzione id 60059), and XLVIII 294 c. 306v with its clear copy at c. 276 (images not reproduced).
- George Lasry, key of Siena no. 11 (report of 17 October 2026): see Siena no. 11.
- Transcription, decoding, tests and reading by Brice Dudout (
validation.txt,reading/xlviii214.md).
Transcription, keys, decoding, tests and reading: targets/mediceo1483.
Cite this page
Daniel Bourdeau and Brice Dudout, “An undated cipher letter in the Mediceo avanti il Principato, XLVIII 214, in the Siena no. 11 cipher”, Cyphersolver, posted 10 November 2026, https://cyphersolver.com/mediceo1483.html.