Le Tellier to Marca, 26 April 1644

At a glanceKnown key applied · partial (97% of the cipher)

The letter: Michel Le Tellier, secretary of state for war, to Pierre de Marca, visitor general in Catalonia, from the court, 26 April 1644. Two pages of cipher, dated in clear.
Where: Bibliothèque nationale de France, Baluze 103 f. 50r–v (Gallica views 111–112); DECODE R2742, which gives the folio as f. 51–52.
The cipher: A homophonic alphabet of figures, letters and invented signs, with overlined figures for words and names (̅3 de La Motte, ̅32 Barcelone).
How it was deciphered: Satoshi Tomokiyo’s table of the 1644 cipher, checked and extended on the sister letter f. 230r against its period decipherment, applied to an AI transcription of the 32 lines.
What it says: Marca’s host is a friend of Marshal de La Motte, placed to learn what he writes; the Catalans’ packets are being opened, and the Queen wants it stopped.
Still open: Seven signs or code groups without a value, 18 of 32 lines that lean on context, and no human check of the transcription.
Credit: Reading by Harumi-Matsuba (issue #54, made with Claude and Codex); key by Satoshi Tomokiyo; an earlier partial draft by NoAutopilot.

01 The letter

Baluze 103 holds Pierre de Marca’s correspondence with Michel Le Tellier for 1644, the year Marca went to Barcelona as visitor general for the French crown. Five of Le Tellier’s letters are wholly or partly in cipher (ff. 50, 171, 189, 200, 230). Four have a decipherment made at the time on the leaf that follows. The letter of 26 April, f. 50, has none: the contributor looked at all 743 views of the volume, so the 1921 catalogue’s “avec traductions” holds for the other four only. Satoshi Tomokiyo, who rebuilt the cipher table from those four, lists f. 50 as the one letter with no contemporary decipherment.

f. 50r carries the date in clear (“Du 26e avril 1644”) and 20 lines of cipher; f. 50v another 12, with a few words in clear (et tous, de faire, a quoy). f. 51 is the address leaf, docketed by Marca’s office “receue le 3 may, resp. le 4 may”.

Top of f. 50r: the date in clear, Du 26e avril 1644, over six lines of cipher in figures, letters and invented signs — f. 50r, the date in clear and the first six lines of cipher. Line 6 has the overlined groups for *Monsieur le Mareschal de La Motte*. Image: Bibliothèque nationale de France, Baluze 103 f. 50r, via Gallica (btv1b9001389d, view 111).

02 The cipher and the key

The 1644 cipher gives each letter several signs: figures 2–24 in roughly alphabetical order (10 a, 11 b, 12 c, 13 d, 14 e …), Latin letters standing for other letters (u e, a n, g s, t d), and invented shapes (θ e, ρ i, a barred y n). A bar changes a sign: plain y is g, barred y is n; an m with a barred tail is s, with a curled tail i. Overlined figures and letters are words and names: ̅9 pour, ̅11 que, ̅12 qui, ̅32 Barcelone, 39 Catalogne, ̅y Monsieur le Mareschal, ̅3 de La Motte, ̅11 ̅o la Reyne.

The contributor rewrote Tomokiyo’s table with sign names, then aligned 21 lines of the sister letter f. 230r with its period decipherment f. 232r: 412 of 424 signs fall on a key value. That alignment added values not in the table (R c, ff d beside a, 6 l beside t, ̅83 le Gouverneur). Three values come from f. 50 alone and are marked as inferred.

03 The reading

The contributor’s joined reading, with their marks: [ ] letters supplied, ⟨ ⟩ signs not read. Their line-by-line table, with a sure / doubtful mark per line, is in contributed/docs/READING.en.md.

Les conseillers de la [ville de] Barcelonne ayant ⟨p⟩ pris soing de faire preparer un l⟨σ⟩ a M. de Marca, ont esté ⟨22⟩ de choisir la maison d’un amy de M. le Mareschal de La Motte, qui a estimé ⟨23̂⟩ par ce moien estre in[formé ?] ponctuellement de tout ce que ledit sieur Marca fe[ra] ou dira, de ce que contiend[ront] les despesches qu’il rece[vra] de la Cour ou qu’il y envoy[e]r[a] ; dont [on] a estimé devoir inform[er] ledit S[ieur] Marca, affin qu’il prenne garde et ne se la[isse] pas persua[d]er par ⟨Ꝺ 22⟩. Celuy qui fait icy affai[res] de Barcelone s’est plaind plusie[urs] fois de ce que on ouvre ⟨‡‡⟩ de La Motte, ainsy qu’il pretend, tous les paquets que les Catalans luy adressent, et tous autres, pour descouvrir ce que on mande d’icy, et affin d’oster la liberté de faire ⟨̅18⟩ les choses qui se passent en Catalogne ; a quoy la Reyne desire passionnement que M. Marca ait l’oeil, et qu’il mande ce qu’il en aura descouvert. Il faut surtout qu’il empesche que cela ne se continue, si tant est qu’il se soit fait cy devant.

In English (the contributor’s gist): the councillors of Barcelona, preparing lodgings for Marca, chose the house of a friend of Marshal de La Motte, who reckoned he would so learn everything Marca does or says and what his despatches to and from the court contain; Marca is warned to be on his guard and not be talked round. Barcelona’s agent at court has complained several times that the packets the Catalans send him are opened [at?] La Motte’s, to find out what is written from court and to stop news of Catalonia passing freely. The Queen Regent wants Marca to watch this, report, and above all put an end to it. Marshal Philippe de La Motte-Houdancourt was then viceroy of Catalonia.

f. 50v: twelve lines of cipher with the clear words et tous, de faire and a quoy — f. 50v, the twelve lines on the opened packets; *et tous*, *de faire* and *a quoy* are written in clear. Image: Bibliothèque nationale de France, Baluze 103 f. 50v, via Gallica (btv1b9001389d, view 112).

04 The checks

The contributor’s. A shuffled-key control: 0 of 1000 shuffles of the key’s values score as well as the real key, also with Tomokiyo’s table values only and on each AI transcription. A sixth, independent Codex transcription agrees with the final one on 568 of 626 sign names (90.7%). A blind calibration on f. 171r, lines 12–24, read with the same key and frozen before the period decipherment f. 172r was opened: 253 of 277 signs (91.3%) and 337 of 360 letters (93.6%).

Here (21 October 2026). With our own decoder and the project’s French model (fr-1600-letters), the bare key decode of f. 50 scores −2.49 per character against −5.09 for the median of 200 shuffled keys (best −4.50); the key is right. Aligning each line’s signs with the contributor’s line reading, 600 of 619 cipher signs (96.9%) carry a key value that stands in the reading; the rest is supplied by context. Nine lines (167 signs: f. 50r lines 1–3 and 17–20, f. 50v lines 1 and 12) were compared with the BnF image: the sign counts agree on every line and no sign reads differently, beyond the plain 6 against a small σ that the contributor already lists as undecided, and a cross with two bars before ̅3 on line 19 of f. 50r.

All of this is machine work. The transcription was made by Claude, cross-checked by Codex, and spot-checked here by Claude again. The calibration suggests about one sign in eleven may be misread, which the reading absorbs through context on 18 of the 32 lines; for that reason the decipherment counts as partial until a person reads the signs on the image.

05 What remains uncertain

Symbols without a value: a plain p (f. 50r line 2), the sign closing line 3 (a logis?), the group 22 (line 4), a 23 with a hooked mark (line 6), Ꝺ 22 (line 16), the double-barred cross before ̅3 (line 19), and ̅18 (f. 50v line 5; sçavoir is a guess). None recurs in the four sister letters.
Lines resting on context (the contributor’s “~”): f. 50r lines 2–7, 10–13, 15, 16, 19; f. 50v lines 5, 8, 10–12. Missing letters at line ends (fe[ra], rece[vra], in[formé]) look like the clerk’s omissions at the turn of the line.
Unbarred “12 19” and “12 6” read as qu’il on f. 50v; the form is confirmed on f. 171r against its decipherment but is not in the key.
Who “celuy qui fait icy affaires de Barcelone” was.

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.

Baluze 103 f. 50r-v, 26 April 1644 (DECODE R2742), contributed reading (Harumi-Matsuba, issue #54) read in part

Michel Le Tellier, secretary of state for war, warns Pierre de Marca, newly arrived in Barcelona as visitor general in Catalonia, that the Barcelona councillors chose for his lodgings the house of a friend of Marshal de La Motte, who hoped in that way to learn everything Marca does and writes; Barcelona's agent at court complains that the Catalans' packets are opened at La Motte's, and the Queen wishes Marca to watch this, report and stop it. Reading contributed by Harumi-Matsuba with Tomokiyo's key, AI-made and not yet checked by a person.

Builds coherent sentences, but under the read bar of 95% of words read as sense (92% here); the holes are marked in the reading.

Transcription

transcription_f50.txt

# BnF Baluze 103 f.50r（c111）・f.50v（c112）の暗号文。記号名は SIGNS.md。2026-10-20、原寸の画像を目視で書き起こし（1人、1回）
# =N は上に横線のある語の符号、[語] は平文で書かれた語、cut は綴じ目で切れて読めない記号
# 注：ꝭꝭ（ff）と Ⱨ（H）、mm と m、ꝭʒ（hz）と ϭ（sg）、× の2形は見分けが難しく、文脈で選んだ箇所がある（docs/READING.md の注）
# 2026-10-20 修正：別の AI（Codex）3回の書き起こしとの照合で、記号そのものの違いを直した（6→sg 10か所、y→z7 9か所、2 4→24 2か所、r06 の 2 3hat→23hat、v06 の a quoy は平文、r13 の sg を削除）。照合前は data/transcription_f50_before_xcheck.txt。詳しくは second/RESULT.md
# 2026-10-20 修正：Codex 5回の投票（second/vote.tsv）で多数が違った15か所を画像で見直し、4か所を直した（r03 =81→81、r04 xt→xx、r17 の13番目 u→n、v04 の3番目 u→n）。second/VOTE.md
# 2026-10-20 修正：r09 の16番目 ae→xx（横線の通った曲がった x。r12 の5番目と同じ形で、値 c）。鍵の y の主な値を、友清氏の表どおり g に直した
# 2026-10-20 修正：mz を、尾に横線のあるもの（mz＝s、友清氏の表の s の列）と、横線のない巻いた尾のもの（mi＝i、表の i の列）に分けた（11か所を mi に。形は画像で1つずつ見た：second/RESULT.md）
# 2026-10-20 修正：画像で確かめられたときだけ直すことにして、Codex の照合で 6→sg に直した10か所を画像で見直した。小さく上に寄った σ と確かめられた r18・v04 の2つを残し、普通の 6 と見分けのつかない8つ（r11・r12・r15・v01・v05・v08・v10・v12）を 6 に戻した（second/CORRECTIONS.md）
r01 19 u mz 12 dl a g th mi hz sg 14 9 7 13 u sg 10
r02 qb pb f R 14 19 dl a z7 u 10 n ct 21 6 p 23 zc
r03 7 E rho a y t u 81 mm 9 14 c ff f u dd m a sg
r04 10 Lx xx bb th7 pb f R 10 dl a hz u g 6 14 22
r05 xx th 12 3 E mi 7 sg f aq ct b ff rho g dl a t m dl
r06 pb th7 mi 9 xx th =y =3 =12 10 bb 9 hz rho 20 u 23hat
r07 m 10 dd R th Lx 22 rho 14 a mz bb 8 hz 9 14 18 z7
r08 c dl 21 R hz m th hz sg u b u dl 6 t bb hz H m 6 ax bb
r09 =11 19 th t mi 6 g q u m zx th7 pb 9 f xx 10 15 14
r10 E hb xx 18 9 ct 13 u R th =11 12 22 a 6 rho th z7 13 cut
r11 6 u g t u mz mm th g ax 17 bb g =11 rho 19 zx th R bb
r12 13 u hz ff xx dl m g dl 4 =11 rho 6 u 14 a m dl 2 g
r13 xx E a hz pb th g 6 mi b 14 t u m H q zx rho z7 x dl f l
r14 sg u t q 6 9f Lx pb 9 R ff 10 x 15 mi a =11 18 19
r15 mm g u a dl th y ct dd t u 14 beta z7 bb g 14 6 ff
r16 23 10 7 c u dd mz k pb ff th g mm 10 g oo 22
r17 12 14 hz m 2 =12 gx 10 mi 6 q R n ct x 15 pb rho
r18 t u =32 7 th g 6 23 sg pb mi a t mm 19 m g q u
r19 x dl 18 7 xx bb R u =11 dl a 22 4 m 9 th pb =3
r20 10 18 a 8 u =12 19 23 f 14 6 u a 13 hz H m g 19
v01 mm 10 24 4 th 6 p =11 6 u g R ff hz pb 19 10 a g
v02 aq 5 u ct 13 f u g 7 th z7 6 [et] [tous] 10 4 hz f u g =9
v03 13 u g ax dl m 4 f q f R 24 m 14 22 a 20 pb z7 13 u
v04 q 12 n 14 6 10 x gx 18 a xx dl 9 hz u f 19 ff sg 18 y u dd 6 th
v05 [de] [faire] =18 6 u g 12 17 H mz 14 8 =12 7 th
v06 23 7 8 14 a 6 14 21 39 [a] [quoy] =11 =o 13 u g 18 f bb
v07 mm ff g mz q H z7 a bb 20 14 a hz =11 Lx 20 10 zx 12 ff
v08 pb u hz 19 22 u mi 6 14 6 =12 19 b 10 z7 t u R u [消] 12 19
v09 14 a ct m zx pb xx 14 9 12 H m k u f 6 q 19 x ff m hz
v10 8 5 9 6 22 m beta 12 6 u th7 23 14 9 R 3 th =11 12 14 19 pb
v11 21 14 9 14 12 dl z7 hz mi 21 m u 8 u 6 10 a beta u g 6
v12 12 6 7 th 8 dl mi 6 x ct q 6 R 2 t th m 10 a hz

Decryption

READING.en.md

# Reading: BnF Baluze 103 f. 50 (Le Tellier to Marca, 26 April 1644)

[日本語](READING.md)

The line-by-line table with certainty marks and notes, the joined French reading, and an English gist. Transcription: `data/transcription_f50.txt`; key: `data/key.tsv`; bare decode: `data/decoded_f50.txt`.

Marks used in the French text: [ ] letters supplied by us; ⟨ ⟩ signs we cannot read.

## Line by line
Marks: **✓** reads with the key values and the words are sure; **~** a sign was chosen from context, or letters are missing; **?** not read. r = f. 50r, v = f. 50v.

### f. 50r
| line | reading | mark | note |
|---|---|---|---|
| r01 | les conseillers de la | ✓ | mz with a barred tail = s, mi without a bar = i; the first l of "conseillers" is ꝭʒ (used for t and l) |
| r02 | Barcelonne ayant ⟨p⟩ pr- | ~ | ⟨p⟩ (plain p) has no value; a small mark above the final ʒc |
| r03 | [i]s soing de faire preparer un l⟨ϭ⟩ | ~ | no i of "pris"; 81 = faire (Tomokiyo); the final ϭ may begin "logis", or be a code sign |
| r04 | a M. de Marca ont esté ⟨22⟩ | ~ | ℒ = M (checked on f. 230r L14, L17); 22 between dots may be a code group |
| r05 | de choisir la maison d[']un | ~ | × read as d, ϭ as i |
| r06 | ami[s] de M. le Mareschal de La Motte qui a estimé ⟨23̂⟩ | ~ | =y and =3 as on f. 230; the line ends with 23, a hooked mark over the 3; no value |
| r07 | par ce moiens estre in- | ~ | m read as p; the final ʒ7 (barred ʒ) = n gives "in"; the next line begins "ponctuellement", so a word is missing (informé?) |
| r08 | ponctuellement de tout ce | ✓ | c = p (p column of the table) |
| r09 | que ledit sieur Mar[r]ca fe- | ✓ | two r in "Marca" (9 and f) |
| r10 | [ra] ou dira de ce que contiend- | ~ | no "ra" of "fera"; the last sign is cut at the gutter |
| r11 | les despesches qu'il rece- | ~ | the final ƀ = e; the next line begins "de la Cour", so the rest of "rece[vra]" is missing |
| r12 | de la Cour ou qu'il y envoy[e]r[a] | ~ | u read as y |
| r13 | dont a estimé devoir inform- | ~ | no subject (on?) of "dont a"; ‡ = a; what we first took for ϭ is the flourish of ʒx (pointed out by the other AI, checked on the image); n = barred y; the final l = m is inferred (source C) |
| r14 | ledit S[ieur] Marca affin qu'il | ✓ | a small f written above the 9 |
| r15 | prenne garde et ne se la[isse] | ~ | ∂ read as n; no "isse" of "laisse" |
| r16 | pas persua[d]er par ⟨Ꝺ 22⟩ | ~ | no d; Ꝺ 22 unknown |
| r17 | celuy qui fait icy affai- | ✓ | 9† = f |
| r18 | [res] de Barcelone s'est plaind plusie- | ✓ | =32 Barcelone (Tomokiyo) |
| r19 | [urs] fois de ce que on ouvre ⟨‡‡⟩ de La Motte | ~ | the mark over the 3 is a short straight bar (not the hooked mark of r06); =3 = de La Motte, checked on f. 230r and f. 200r; the sign before it is a ‡ with two bars, perhaps not the same as ‡ (a); meaning (chez?) unknown |
| r20 | ainsy qu'il pretend tous l[es] | ✓ | |

### f. 50v
| line | reading | mark | note |
|---|---|---|---|
| v01 | paquets que les Catalans | ✓ | "2 4" is one number, 24 (the other AI's reading); 24 = q (after 10–23 = a–p; inferred) |
| v02 | luy adressent, et tous autres, pour | ✓ | "et tous" in clear |
| v03 | descouvrir ce que on mande | ✓ | "c 24 u(m) e" = ce que |
| v04 | d'icy, et affin d'oster la liberté | ✓ | y = b |
| v05 | de faire ⟨=18⟩ les choses qui se | ~ | "de faire" in clear; =18 not found in the sister letters; code numbers run in alphabetical order (9 pour, 11 que, 12 qui, 22 troupes, 26 vous), so 18 is a word in q–t; "sçavoir" would fit (weak guess: f. 203v has "faire scauoir" in clear) |
| v06 | passent en Catalogne, a quoy la Reyne desire | ✓ | "a quoy" in clear; 39 (Catalogne) has no bar; =11 =o (Tomokiyo's n̄ō) = la Reyne, checked at three places of f. 189r against f. 190r |
| v07 | passionnement que M. Marca | ✓ | |
| v08 | ait l'oeil et qu'il mande ce qu'il | ~ | "ait" = ‡ u ꝭʒ (u = y, "ayt"); after a struck-out m, an unbarred "12 19" read as qu'il |
| v09 | en aura descouvert. Il faut | ✓ | |
| v10 | surtout qu'il empesche que cela | ~ | unbarred "12 6" read as qu'il (as in v08) |
| v11 | ne se continue, si tant est | ~ | the sign for the i of "si" is read n (= y) by four of the five AI passes |
| v12 | qu'il se soit fait cy devant. | ~ | "12 6" = qu'il (as in v08, v10) |

## Joined reading (French)
Les conseillers de la [ville de] Barcelonne ayant ⟨p⟩ pris soing de faire preparer un l⟨ϭ⟩ a M. de Marca, ont esté ⟨22⟩ de choisir la maison d'un amy de M. le Mareschal de La Motte, qui a estimé ⟨23̂⟩ par ce moien estre in[formé ?] ponctuellement de tout ce que ledit sieur Marca fe[ra] ou dira, de ce que contiend[ront] les despesches qu'il rece[vra] de la Cour ou qu'il y envoy[e]r[a] ; dont [on] a estimé devoir inform[er] ledit S[ieur] Marca, affin qu'il prenne garde et ne se la[isse] pas persua[d]er par ⟨Ꝺ 22⟩. Celuy qui fait icy affai[res] de Barcelone s'est plaind plusie[urs] fois de ce que on ouvre ⟨‡‡⟩ de La Motte, ainsy qu'il pretend, tous les paquets que les Catalans luy adressent, et tous autres, pour descouvrir ce que on mande d'icy, et affin d'oster la liberté de faire ⟨=18⟩ les choses qui se passent en Catalogne ; a quoy la Reyne desire passionnement que M. Marca ait l'oeil, et qu'il mande ce qu'il en aura descouvert. Il faut surtout qu'il empesche que cela ne se continue, si tant est qu'il se soit fait cy devant.

## English gist (ours, not a translation of record)
The councillors of Barcelona, in having lodgings prepared for M. de Marca, chose the house of a friend of Marshal de La Motte, who reckoned that in this way he would be informed in detail of everything Marca does or says and of the contents of the despatches he receives from, or sends to, the Court. Marca has therefore been warned to be on his guard and not to let himself be talked round. The man who handles Barcelona's affairs here [at Court] has complained several times that packets addressed to him by the Catalans, and others, are being opened at [?] de La Motte's, so as to discover what is written from here and to take away the freedom to report [?] what happens in Catalonia. The Queen [Regent] earnestly wishes Marca to watch this and report what he finds; above all he must stop it continuing, if indeed it has been happening.

Decryption

decoded_f50.txt

r01 lesconseitlersdela
r02 barcelonneayantspr
r03 soingde faire preparerunl
r04 amcemarcaontesteo
r05 cechoislrlamaisonduo
r06 amirce Monsieur le Mareschal de La Motte qui aertime?
r07 uarcemoiensestrein
r08 ponctuetlemeotdetoutce
r09 que leditsieurmarrcafe
r10 ouciradece que contiend…
r11 tesdespesches que ilrece
r12 detacousou que iteenuoys
r13 contaestimedeuoirinform
r14 leditsmarcaaffin que il
r15 psenoegardeetneseta
r16 paspersuaaespasho
r17 cetuy qui faiticyaffai
r18 de Barcelone sestplaindplusie
r19 foiscece que onouurea de La Motte
r20 ainse qui lpretendtousl
v01 paquets que tescatalans
v02 lueadressent et tous autres pour
v03 descouurircqueonmande
v04 icyetaffincorterlaligerte
v05 de faire <=18> teschoses qui se
v06 pssenten Catalogne a quoy la Reyne desire
v07 passionnement que mmarca
v08 aetloeitet qui lmandece 消 cl
v09 enauracercouuertilfaut
v10 surtoutctemperche que cela
v11 nerecontinuesetantest
v12 ctsesoitfaitcydeuant

Reading

f50_reading.md

<!-- Reading of Baluze 103 f. 50r-v: the joined French reading of Harumi-Matsuba's contribution (issue #54; contributed/docs/READING.en.md, data CC BY 4.0), copied alone so that it is scored without the line table and the English (verification 5, 23 Oct 2026). Their marks: [ ] letters supplied, ⟨ ⟩ signs not read. Made and checked by AI only (NOTES.md). -->

# Le Tellier to Marca, 26 April 1644 (Baluze 103 f. 50r-v)

## Reading

Les conseillers de la [ville de] Barcelonne ayant ⟨p⟩ pris soing de faire preparer un l⟨ϭ⟩ a M. de Marca, ont esté ⟨22⟩ de choisir la maison d'un amy de M. le Mareschal de La Motte, qui a estimé ⟨23̂⟩ par ce moien estre in[formé ?] ponctuellement de tout ce que ledit sieur Marca fe[ra] ou dira, de ce que contiend[ront] les despesches qu'il rece[vra] de la Cour ou qu'il y envoy[e]r[a] ; dont [on] a estimé devoir inform[er] ledit S[ieur] Marca, affin qu'il prenne garde et ne se la[isse] pas persua[d]er par ⟨Ꝺ 22⟩. Celuy qui fait icy affai[res] de Barcelone s'est plaind plusie[urs] fois de ce que on ouvre ⟨‡‡⟩ de La Motte, ainsy qu'il pretend, tous les paquets que les Catalans luy adressent, et tous autres, pour descouvrir ce que on mande d'icy, et affin d'oster la liberté de faire ⟨=18⟩ les choses qui se passent en Catalogne ; a quoy la Reyne desire passionnement que M. Marca ait l'oeil, et qu'il mande ce qu'il en aura descouvert. Il faut surtout qu'il empesche que cela ne se continue, si tant est qu'il se soit fait cy devant.

Translation

f50_en.md

<!-- English translation (Claude Opus 5.5, 23 Oct 2026) of f50_reading.md, the joined French reading of Harumi-Matsuba's contribution (contributed/docs/READING.en.md, CC BY 4.0). Same marks: [x] letters supplied, ⟨ ⟩ signs not read (kept as they stand). The contribution's own "English gist" is not a translation; this is. -->

# Le Tellier to Marca, 26 April 1644 (Baluze 103 f. 50r-v): English translation

## Translation

The councillors of the [city of] Barcelona, having ⟨p⟩ taken care to have a l⟨ϭ⟩ prepared for Monsieur de Marca,
were ⟨22⟩ to choose the house of a friend of Monsieur the Maréchal de La Motte, who thought ⟨23̂⟩ by this means to be
in[formed] precisely of all that the said Sieur Marca will d[o] or say, of what the despatches he will rece[ive] from
the Court or will send there will cont[ain]; of which it has been thought proper to info[rm] the said S[ieur] Marca,
so that he may take care and not l[et] himself be persua[d]ed by ⟨Ꝺ 22⟩. The man who does Barcelona's business here
has complained seve[ral] times that ⟨‡‡⟩ de La Motte opens, as he claims, all the packets the Catalans address to him,
and all others, to discover what is written from here, and to take away the freedom of ⟨=18⟩ the things that happen
in Catalonia; to which the Queen passionately desires Monsieur Marca to have an eye, and to report what he has
discovered of it. Above all he must prevent its continuing, if indeed it has been done before.

Key: key.tsv

06 Sources

BnF, Baluze 103, f. 50r–v, Gallica ark:/12148/btv1b9001389d, view 111 and view 112; sister letters ff. 171, 189, 200, 230 with decipherments ff. 172, 190, 205–206, 232 (views 366–368, 402, 424, 486–490).
DECODE R2742 (this letter), R2743–R2746 (the sister letters).
Satoshi Tomokiyo, “Le Tellier-Marca Cipher (1644)”, Cryptiana, louisxiv0.htm.
Harumi-Matsuba, letellier1644-public (CC BY 4.0 data, MIT code), and issue #54, 20 October 2026.
NoAutopilot, cipher-lab, ciphers/baluze103-letellier-marca-1644/: a partial draft of f. 50 with Tomokiyo’s table (451 of 630 symbols given a value, parked 13 October 2026); some of its words (conseil, que ledit, descouvert, surtout) recur in the reading above.
Catalogue des manuscrits de la collection Baluze (Paris 1921), Baluze 103.

A different Le Tellier cipher, the letter to the Marquis de Castelnau of 1657, has its own notes in targets/letellier/.

The contributor’s key, transcription, reading, controls and calibration files, our re-measure (measure.py) and the notes are in targets/letellier1644/.