Azuares to the Carlist Secretary of State, London 1837

At a glanceCiphertext-only break · partial (79% of the cipher)

The letter: Manuel Azuares, Carlist agent in London, to the First Secretary of State of the pretender Carlos V, 8 September 1837: letter no. 780, marked “Duplicado”, in clear Spanish with runs of cipher on views 03 to 12.
Where: Archivo Histórico Nacional, Madrid, DIVERSOS-COLECCIONES,161,N.2 (PARES 3735838), 12 images. It was catalogue entry 2421. No DECODE record and no edition were found.
The cipher: A letter-for-letter substitution of Spanish, written in the same cursive hand as the clear text, with a few groups of digits. On the contributor’s model the table changes with the group of pages; only one table is verified.
How the tables were found: Under rules Brice Dudout fixed in advance, his AI agents recovered letter tables from the cipher by annealing on Spanish. Five passage tables were measured on held-out lines: one, the table T6 of pages 6 and 7, passed (0.882 against 0.508 for the best of 1,000 shuffled tables) and four did not; an earlier test of the tables of pages 3 to 5 was negative. The project did not repeat the tests and did not check the transcription or the reading against the images; its catalogue vetting had viewed four of the twelve images, only to confirm the cipher.
What it says: Azuares reports talks with the Austrian chargé d’affaires: the Powers consider making Maria II an absolute queen and reviving the old Portuguese Cortes, and Lisbon would treat with Don Carlos once he reached Madrid. Azuares objects.
Still open: Four of the five tables are not verified, so 215 words are probable only; 67 words are restored and 9 unresolved; five indicator groups and five digit groups are not resolved; no human has checked the transcription or the reading.
Credit: The transcription, the tables, the tests, the reading and the translation are Brice Dudout’s (Europe chiffrée, with Claude Opus 5.5 agents). The catalogue entry came from Daniel Bourdeau’s sweep of PARES.

01 The letter

The letter is in the Archivo Histórico Nacional in Madrid, DIVERSOS-COLECCIONES,161,N.2 (PARES 3735838): no. 780, marked “Duplicado”, written from London on 8 September 1837 by Manuel Azuares, a Carlist agent, to the First Secretary of State of the pretender Carlos V. It has 12 images. Most of it is clear Spanish. On views 03 to 12 runs of cipher written in letters stand inside the clear sentences, with a few groups of digits. PARES describes the letter as partly in cipher and gives no decipherment, and none is written on the leaves.

The letter entered this project’s catalogue as entry 2421 on 9 November 2026, from a sweep of PARES. The contributor, Brice Dudout (Europe chiffrée, with Claude Opus 5.5 agents), worked on it the same day. He checked this repository for earlier work and found none; the catalogue’s vetting had found no edition, and no DECODE record or key for Carlist agents in the repository’s index (the live DECODE database needs a login and was not queried).

No image is shown on this page. The images are online on PARES without login, but the contributor did not find the reproduction licence on the record and included no image, crop or thumbnail; the project has not reproduced any. A line of the cipher is printed as letters in the strip under the title.

02 The cipher and its tables

The cipher replaces each letter of the Spanish by another letter, in the same cursive hand as the clear text. The transcription keeps apart a long s (ſ), a round r shaped like z, and ll and ñ as single units. One unit (written e or c) stands for u or b. The transcription file holds 1,741 cipher units, counted here, and the six lines of passage P112 (11.04–09) are only in the frozen transcription of the test; the profile’s figure of about 1,500 letters was not measured.

On the contributor’s model the table changes inside the letter, with one table for each group of pages, not one for each run: K3 for page 3, K5 for pages 4 and 5, T6 for pages 6 and 7, T8 for page 8 and T9 for pages 9 to 12. Each change is announced by a group that starts with x followed by l, z, g or h (xgzl, xzlg, xlgz, xlzh, xlzgh). He takes these for indicators, “probably” in his notes; they are not resolved. Five groups of digits (4607229, 5071574925, 5344727605, 507, 403710131) are not resolved either, and no nomenclator was found.

The contributor ran two pre-registered protocols, with the transcription, the tables and the measuring script frozen by hash before any held-out line was decoded.

First protocol: negative. Tables K3 and K5 were learnt on views 03 and 05 and measured once on the cipher of view 12, set aside beforehand: 0.198 against 0.198 for the best of 200 shuffled tables. The contributor concludes that view 12 is under another table.
Amendment 01. New passages of views 04 and 06 to 11 were transcribed before any decoding. Five passages of at least 100 units were each cut in two: a table was learnt on the first half only (annealing on Spanish 4-grams, with hand edits on that half) and frozen, then measured once on the second half. The measure is the share of held-out letters covered by words of a Spanish lexicon. The reference is the best of all 1,000 shuffled tables (200 for each passage), 0.508, to correct for making five comparisons. A table counts as validated at 0.60 or more with a margin of 0.30 or more.

Passage	Lines	Held-out units	Score	Result
P063	06.17–23 (held out: 06.21–23)	68	0.882 (margin 0.374)	validated
P062	06.09–14	59	0.559	negative
P081	08.03–09	71	0.676	negative
P092	09.17–10.02	51	0.392	negative
P112	11.04–09	63	0.476	negative

So only the P063 table, called T6, is verified. Before decoding, the amendment itself noted that identical cipher strings recur from one passage to another, so that the model of several tables may be wrong (a single table with homophones or reading errors was not excluded); the protocol was applied all the same. Applied unchanged, it also gives Spanish on the rest of pages 6 and 7 (the passage P062, line 06.05 and view 07). The contributor counts the words it gives letter for letter as certain, and declares that this is not a further measurement. The four negative passages have short held-out halves, 51 to 71 units, against a high corrected reference; their readings stay probable. Views 03 and 05 were wholly decoded during the first attempt and cannot be measured. The profile records the validation as weak overall: a pass for T6 and a fail for the four other passages.

These are the contributor’s tests, re-run by him on 9 November 2026 with identical figures. The project did not repeat the tests and did not check the transcription or the reading against the images; its catalogue vetting had viewed four of the twelve images, only to confirm the cipher.

03 The reading

The contributor counted the cipher words by hand on the decryption and calls the count approximate. Words are certain when the verified table gives every letter, probable when an unverified table gives them by sense, restored when the table gives another letter or none, and unresolved otherwise.

Group	Certain	Probable	Restored	Unresolved
T6, verified (06.05, 06.09–14, 06.17–23, view 07)	70	0	21	1
K3 (view 03), K5 (views 04–05)	0	57	2	4
T8 (view 08)	0	38	10	1
T9 (views 09–11)	0	102	28	3
View 12 (T9, the held-out set of the first protocol)	0	18	6	0
Total, 361 words	70	215	67	9

Certain and probable words together are 285 of 361, or 78.9%, with the restored words left out; with them the share is 352 of 361 (97.5%). The five indicator groups and the five digit groups are not counted as words. Because 78.9% is under the 95% bar and four tables are not verified, the letter is deciphered in part.

The passages under the verified table, in the contributor’s reading. Clear words of the letter are in angle brackets, restored letters in square brackets, [x…] is an indicator group and # marks a group of digits.

⟨y siendo⟩ [x…] Globet [q]uien dirige al príncipe [F]ernando ⟨que es un joven lleno de carácter … no dudo que él sabrá⟩ establecer en a[q]uel reyno un gobierno adaptado [a las] circunst[an]cias, monárquico y aumentado con las institu[cio]nes [a]ntiguas gratas al pueblo portugués.

⟨La Inglaterra conoce esta necesidad … el Gobierno de Lisboa⟩ se apresure a tratar con el rey don Carlos bajo [bases] mutuas de seguridad y de conveniencia política #5344727605 la [ll]egada de su Magestad a Madrid dará sin duda prin[c]ipio a este paso [q]ue la corte de Lisboa desea.

In the contributor’s translation, with the marks left out here: “and as it is Globet who guides Prince Ferdinand, who is a young man full of character, I do not doubt that he will know how to establish in that kingdom a government adapted to the circumstances, monarchical and enlarged with the ancient institutions dear to the Portuguese people. Britain knows this need … the Lisbon government should hasten to treat with King Don Carlos on mutual bases of security and political convenience #5344727605; the arrival of His Majesty in Madrid will no doubt begin this step, which the court of Lisbon desires.”

The rest, under the unverified tables, is probable. On page 3 Azuares names “el encargado de negocios de Austria en esta capital”, whom he saw the day before. On page 5 the Austrian’s language makes him suppose that a change is being prepared in the Portuguese system of government, from which the Powers hope to secure order in that country “aunque sea sacrificando los derechos de S.M. Fidelísima el rey don Miguel primero”. On pages 7 and 8 he gives his own view: to recognise Doña María de la Gloria would be to recognise the usurpation, and the Portuguese people are in general attached to Dom Miguel; the lines then speak of the consolidation of her throne and of guarantees of a lasting peace, which the contributor’s summary takes as a denial that such a settlement would secure one (the negation is not in the lines deciphered). The plan he reports on pages 9 and 10 is to declare her absolute queen, summon the ancient Portuguese Cortes and negotiate with Don Carlos’s government on conventions of mutual security. Britain, he writes on page 11, undertook to prepare what has taken place in Lisbon. The general is written “Globet” throughout; the contributor identifies him as the Belgian general Goblet d’Alviella, sent to Lisbon in 1837. (This identification is the contributor’s and was not checked here.) The whole reading and its English translation are in the full texts below.

Before publication the contributor reviewed his first draft on 9 November 2026. He removed words that are not in the cipher (“de Lisboa”, “del Norte”, “se”), noted that “se trata de” and “existe en” are clear text, moved doubtful words to unresolved, and limited the certain class to the verified table.

04 What remains uncertain

The contributor grades words as certain, probable, restored or unresolved; the grades are his.

Four tables are not verified (K3, K5, T8, T9): their 215 words are probable only. The passages of T8 and T9 failed on held-out halves of 51 to 71 units, short against the corrected reference. K3 and K5 were never measured on their own pages: views 03 and 05 were wholly decoded in the first attempt, and the one measurement, on view 12, was negative. The contributor’s notes say that fresh cipher under the same tables (other Azuares letters in volume 161) or a second human reader would settle them.
Five indicator groups and five digit groups (4607229, 5071574925, 5344727605, 507, 403710131) are open: each occurs once and no nomenclator was found.
Line 04.02 (18 units) and nine unresolved words, among them {piensa?} on 07.03, {ocasionar?} on 08.05, {desean?} on 09.07, {unanim…?} on 10.16 and {movimiento?} on 11.04–09: the notes ask for a human reading of those lines at full resolution.
67 restored words rest on sense, not on the table.
The word counts were made by hand and are approximate. The six cipher lines of passage P112 (11.04–09) are in the frozen test transcription only, not in the transcription and decryption files shown below, so that part of the reading cannot be checked from them.
One AI transcription and one AI reader. No human palaeographer has checked the transcription or the reading, and the project did not check them against the images.

What the contributor’s notes record of the other routes: N.5 of the same volume (Villafranca, catalogue 2422) uses another system, and no sibling in this letter cipher was found; the 12 leaves carry no clear minute or decipherment; no key for Carlist agents was found in the repository or its index of DECODE; no edition was found. Every line was decoded again with the table of its group at the review of 9 November 2026 and regraded, and the open words stayed open.

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.

letter no. 780, views 03-12: cipher runs in a clear letter read in part

Azuares reports talks with the Austrian chargé d'affaires in London: Austria and Britain contemplate making Maria II an absolute queen and reviving the old Portuguese Cortes, with General Goblet guiding Prince Ferdinand, at the cost of Dom Miguel's rights; Lisbon would treat with Don Carlos once he reached Madrid; Azuares objects that this would recognise the usurpation.

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

transcription.txt

# azuares1837 - AHN, Diversos-Colecciones, 161, N.2 (PARES 3735838): Manuel Azuares, London, 8 Sep 1837 (no. 780, duplicate)
# Cipher runs only (the letter is otherwise in clear Spanish). One unit per space-separated token.
# Conventions: z = round r (shaped like z), r = straight r, ſ = long s (descending), s = short s; ll and ñ are single units;
# # = a group of digits (nomenclator), given in the reading; ? = illegible. Line ids = view.line of the writing column.
# Views 03 and 05 were first transcribed with r for the round r; they are given here in the z convention.
# Lines 06.09-11.09 marked P0xx are the frozen transcription (frozen/TRANSCRIPTION-2421-AV01.txt, SHA-256 28cd1de3...).
03.02 | x g z l h n i l i v b g p m g o h o i v i m h b ll h y o i g c y n p ll g i v i y n g b g a ll n g l
03.03 | g c y n p ll g
05.a | ll u g p u a p u y u o n z o n u z c y n p v z g h
05.b | a p u a z p z p z g e c o g z b m v h u o u g y v y n u b z z g n c z g ll u g e h m v u p o h a h p n e e c u y ñ d c u ll u y e g h o y h g v ll z g v h o u y a u p z o g z y a h n u o g v z y y z g z p a z p n v ll h a z p z z y u e e p z p u g h p ll u o u o z d c u g a z v y z e o d c u y u z y z g p v l u g z o ll g h y ll u p u g t h y ñ u y c b z e u y n z ll v ll u g v y v b z u g p u ñ ll h o b v e c u g a p v b u p h
04.02 | i l m g f ll v i n i v i ñ v m l i y
04.16 | x z l g h n u g b v o v y n p h l p z o g u y o h z y v g h o g z ll u g ñ o c g u y
06.05 | x l g z h n f z h e ſ n d c o ſ ll v o p o f ſ l z a p o ll t o a ſ g ſ p ll ll l v h
06.09 P062 | ſ y n l e z ſ t ſ p ſ ll l d
06.10 P062 | e ſ z p ſ ñ ll h e ll f h e o ſ p ll h l v l a n l v h
06.11 P062 | t a l y t o p t e ll y n ll l t o l y g h ll l p d c o t h ñ
06.12 P062 | l e g ſ ll n l v h t h ll z l y o ll y n o n e t ll o ll
06.13 P062 | ſ y ll n o f e l y ſ f p l n l y l z a c ſ e z h a h p
06.14 P062 | n e f c ſ y
06.17 P063 | y ſ l a p ſ y c p ſ l n p l n l p t h ll ſ z p ſ ñ v
06.18 P063 | h ll t l p z h y c l i h e l y ſ ſ y g e n c l y v ſ y
06.19 P063 | ſ f e p o v l v ñ v ſ t h ll s ſ ll o ſ ll t o l a h z
06.20 P063 | o n o t l # z l m ſ f l v l v ſ y
06.21 P063 | c g l f ſ y n l v l g l v p o v v l p l y o ll v
06.22 P063 | c v l a p o ll z o a o h l ſ y n ſ a l y h d e ſ z
06.23 P063 | l t h p n ſ v ſ z o y e h l v ſ y ſ l
07.03 | z l t h p n ſ a o ſ ll y ll y o ſ z ſ ſ ll ſ p l z ſ z h e ſ n ſ y e ll l y l z s l f e l p v o l
07.14 | y ſ b l e o ll e y n ſ ll o v h v ſ ſ ll n p l p ſ ll ll o ll f e ll ll ll ſ f h t o l t o h ll t ll t l z l y a h n ſ ll t o l y ſ y n p ll l ſ ſ p l y
07.22 | ſ z p ſ t h ll ll h t o g o ſ ll n h v ſ o h u l g l p o l v ſ z l f z h p o l
08.03 P081 | x l z h n g ñ ſ t n p n t ſ ñ h
08.04 P081 | l f r f ñ c l t ll n p o o l g l d e ll ſ p e n ñ s f ſ t p
08.05 P081 | l t g n r ſ d ſ u l p g ſ e n v ñ ll ll t l r n p n r ſ
08.06 P081 | ñ e ll ſ p ñ ſ t ll e ll v n e n ñ h l m ñ l p d ll o
08.07 P081 | n ñ ll l v ſ h e f ſ e h n e n ñ g f m f ſ r s ſ ſ ſ r
08.08 P081 | ſ p h n m ſ p ſ ñ l h l v ll t g n l h l c ſ ñ r n
08.09 P081 | p l v ſ h r ſ y n ñ v n p d ll m f ſ h
08.12 | ſ p
08.13 | c n n g f m l h
08.14 | t n p r n h ll v l t ll n p v ſ h g ñ n p n v ſ o n y l d l ñ ll l t
08.18 | f p l e l i v f ñ l v ſ ñ l
08.22 | h n v l m ſ p g ſ z l f r g ñ ll l t n r
09.07 | x l z g h n g l y a h n ſ o t ll l y v ſ ſ g ſ l o y ll o v c v l p ſ o h s l p y c y l o n ll m e l y p ſ g l i ll h o ſ y t h o a h p n e m l g
09.17 P092 | g l y s ſ o n
09.18 P092 | l u l y d e ſ y l t l v ſ ſ y n l ſ y a ſ t ll ſ v ſ
09.19 P092 | t h g h o ll ñ o m g ſ y l ñ y ſ p ll o t h
09.20 P092 | o y ſ d c ſ o n ſ t h o y c a p h a ll a h g ll
09.21 P092 | n ll t l p ſ y n l e g ſ t ñ ſ o v h l o y ſ i h
10.01 P092 | p v h o m ll m e ſ g ſ o y e n p h o h g ſ m
10.02 P092 | ll n ll b h
10.06 | ſ y n ſ a l ñ y c o l ſ y a ſ t ll ſ v ſ h v ll h a ſ p y h o l g t h o n p l ſ g y ſ i h p v h o b ll m e ſ g
10.09 | y e a c ſ y n h y d e ſ g ſ g l o l n p ll e c ll v h
10.12 | g l y a h n ſ o t ll l y ñ g h y v ſ g l ñ o m g l n ſ p p l d e ſ ſ g v ſ t g l p l p l v h i l b l p ll v ſ g l m g h p ll l p ſ ñ o ll e y h g e n t
10.16 | g l t h o s h t l t ll h o v ſ g l y l o n ll m e l y t h p n ſ y a h p n e m c ſ y l ſ p ſ t h o h t ſ p g l e o l o ll b ſ o m ſ ñ t h o ſ y n ſ l a h ñ h o ſ m h t ll p t h o ſ g
11.01 | p ſ t h o t ll b ll ſ o n h z c o v l o v h ſ y n ſ ſ o t h o s ſ o ll h y v ſ b e n c l y ſ m e p l t l v l v
11.12 | g l t h p ñ ſ v ſ g ll y e h l
11.21 | ñ y ſ p l a h p g h b ll y b h b l y v ll z ll z ll y
12.01 | p ſ l g ll z l p ſ g a g l o v ſ ſ y n l e g ſ t ſ p ſ g m h e ll ſ p ? o h l e y h g e n h
12.04 | a p ſ s ſ o ll v h v ſ l o n ſ b l o h v ſ g l y a p h a h y ll t ll h o ſ y d c ſ n l g s ſ z l y e ſ ſ m l v ll b l v p ll v a h v p l o g l t ſ p y ſ g ſ

Decryption

decryption.txt

# azuares1837 - line-by-line decryption with the table of each page group (key.tsv). '?' = unit not in that table.
# T6 (pp. 6-7) is the only VERIFIED table (P063 table, validated on held-out lines 06.21-23); K3, K5, T8, T9 are not verified.
# Raw output, no correction; the reading (reading.md) restores letters in [ ] and marks open words in { }.
03.02 [K3] | haynotenelcargadodelegociosdeaustriaelestacapitan
03.03 [K3] | austria
05.a [K5] | de*representanteaustria*o
05.b [K5] | preparara*gun*ambioene*sistemaa*tua*de*gobiernoportgguesyquedesg*onso*ida*ionesperan*aspoten*iassa*arpartidoparaaseggrare*ordenenaque*paisagnqueseasa*rife*and*osdere*hosyesumagestadide*isimae*reydonmigue*primero
04.02 [K5] | ?fb*?di?t?i?yibf?s
04.16 [K5] | zaf*ote*ministrofran*esnoasi*on*ade*ynu*es
06.05 [T6] | ?amlotglouet?uiendirigealprincipemernnado
06.09 [T6] | estaulecerena?
06.10 [T6] | uelreynoungouiernoadaptado
06.11 [T6] | cpascircunstnaciasmonar?uicoy
06.12 [T6] | aumentadoconlasinstitucnin
06.13 [T6] | esntiguasegratasalpueulopor
06.14 [T6] | tugues
06.17 [T6] | seapresureatratarconelreyd
06.18 [T6] | oncarlosuajouaseesmutuasdes
06.19 [T6] | eguridadydeconvenienciapol
06.20 [T6] | itica#la?egadades
06.21 [T6] | umagestadamadriddarasind
06.22 [T6] | udaprinlipioaestepaso?uel
06.23 [T6] | acortedelisuoadesea
07.03 [T6] | lacortepiensnsieleeneralelouetesunasalvaguardia
07.14 [T6] | se?auinustenidodeentrarenningunnnegociacioncncalaspotenciasestrnaeeras
07.22 [T6] | elreconnocimientodeio?amariadelagloria
08.03 [T8] | ?a?lotreconocerl
08.04 [T8] | ausurpacionyyatampienporquecn
08.05 [T8] | actosemejantepodriicasonose
08.06 [T8] | rpienrecipidoporlagranmiy
08.07 [T8] | oriadelpueploportuguesqeees
08.08 [T8] | enlogeneraladictoalaperso
08.09 [T8] | nadelse?ordonmiguel
08.12 [T8] | en
08.13 [T8] | pootugal
08.14 [T8] | consolidaciondeltronodeyo?amariac
08.18 [T8] | unapa?duradera
08.22 [T8] | lodagente?austriacos
09.07 [T9] | ?a?lotlaspotenciasdeeleansindudarenovarsusantiguasrela?ionesconportugal
09.17 [T9] | lasvent
09.18 [T9] | ajasquesacadeestaespeciede
09.19 [T9] | coloniynglesayserinco
09.20 [T9] | nsequenteconsupropipoli
09.21 [T9] | ticarestaulecyendoanse?o
10.01 [T9] | rdongiguelensutronoleg
10.02 [T9] | itimo
10.06 [T9] | estepaysunaespeciedeodiopersonalcontraelse?ordonmiguel
10.09 [T9] | supuestosquelelanatriuuido
10.12 [T9] | laspotenciasylosdelaynglaterraqueeldeclararado?amaridelagloriareyniusolutc
10.16 [T9] | laconvocaciondelasantiguascortesportuguesaereconocerlaunanimengeyconesteapoyonegocirconel
11.01 [T9] | reconcimiento?undandoesteenconveniosdemutuaseguracadad
11.12 [T9] | lacoryedelisuoa
11.21 [T9] | yseraporlomismomasdi?i?is
12.01 [T9] | reali?arelplandeestaulecerelgouier?noausoluto
12.04 [T9] | prevenidodeantemanodelasproposicionesquetalve?asueegadimadridpodranlacersele

Reading

reading.md

# azuares1837: reading of the cipher runs (Spanish), 9 Nov 2026

Manuel Azuares, Carlist agent in London, to the Carlist First Secretary of State, London, 8 Sep 1837 (no. 780, duplicate).
AHN, Diversos-Colecciones, 161, N.2 (PARES 3735838), views 03-12.

Marks: clear words of the letter are given only as context in ⟨ ⟩. In the cipher: [abc] = letters or words restored by
sense (the table gives another letter, or none); {word?} = unresolved; #n = digit group (nomenclator, not resolved);
[x...] = the indicator group x + l/z/g/h that opens each new table (not resolved).

Classes (Daniel's rule). **Certain** = read by the only verified table (T6, the P063 table validated on held-out lines),
applied without change, every letter given by the table. **Probable** = read by sense with a table that is not verified
(K3, K5, T8, T9; view 12 was the burned held-out set of the first protocol). Restored and unresolved words are counted apart.

## Counts (cipher words, counted by hand on decryption.txt; approximate)

| group | certain | probable | restored | unresolved |
|---|---|---|---|---|
| T6 verified (06.05, 06.09-14, 06.17-23, view 07) | 70 | 0 | 21 | 1 |
| K3 (view 03), K5 (views 04-05) | 0 | 57 | 2 | 4 |
| T8 (view 08) | 0 | 38 | 10 | 1 |
| T9 (views 09-11) | 0 | 102 | 28 | 3 |
| view 12 (T9, burned held-out set) | 0 | 18 | 6 | 0 |
| **total, 361 words** | **70** | **215** | **67** | **9** |

Read share (certain + probable, restored excluded): 285 / 361 = **0.789** -> read in part. With restored words 352 / 361
(0.975). Not counted as words: 5 indicator groups and 5 digit groups (4607229, 5071574925, 5344727605, 507, 403710131).

## Certain: table T6 (pages 6-7)

- 06.05: ⟨y siendo⟩ [x...] Globet [q]uien dirige al príncipe [F]ernando ⟨que es un joven lleno de carácter ... no dudo que él sabrá⟩
- 06.09-14 (P062): establecer en a[q]uel reyno un gobierno adaptado [a las] circunst[an]cias, monárquico y aumentado con las
  institu[cio]nes [a]ntiguas gratas al pueblo portugués.
- 06.17-23 (P063; held-out lines 06.21-23 measured, M5 0.882): ⟨La Inglaterra conoce esta necesidad ... el Gobierno de
  Lisboa⟩ se apresure a tratar con el rey don Carlos bajo [bases] mutuas de seguridad y de conveniencia política #5344727605
  la [ll]egada de su Magestad a Madrid dará sin duda prin[c]ipio a este paso [q]ue la corte de Lisboa desea.
- 07.03: ⟨tengo datos oficiales de que⟩ la corte {piensa?} ... el [g]eneral [G]lobet es una salvaguardia ⟨para la
  propagación de las ideas revolucionarias⟩
- 07.14: ⟨de Su Gobierno⟩ se [h]a [ab]stenido de entrar en ningun[a] negociación [con] las potencias ext[ranj]eras.
- 07.22: ⟨como idea peculiar mía, que⟩ el reco[n]ocimiento de [Doñ]a María de la Gloria ⟨podría presentar grandes
  dificultades, ya porque esto era lo mismo que⟩

## Probable: tables K3, K5, T8, T9 (not verified)

- 03.02 (K3): [x...] el encargado de negocios de Austria en esta capital ⟨á quien vi ayer⟩
- 03.03 (K3): ⟨convencerme de que él⟩ austria[co] #4607229 ⟨trata de sacar también partido del movimiento ocurrido en aquel Reino⟩
- 04.02 (K5): ⟨no ha impedido los pasos dados por⟩ #5071574925 {ilmgfllviniviñvml iy: unresolved}
- 04.16 (K5): ⟨contraste con la observada por⟩ [x...] ministro francés, no así con la de {...} ⟨que se presentaba como
  extraño a todos los acontecimientos⟩
- 05.a (K5): ⟨Este lenguaje⟩ del representante austriaco ⟨expresado con viveza natural, y con afectada alegría⟩
- 05.b (K5): ⟨me ha hecho suponer con algún fundamento que se trata de⟩ preparar algún cambio en el sistema actual del
  gobierno portugués, y de [esta] consolidación esperan las potencias sacar partido para asegurar el orden en aquel país,
  aunque sea sacrificando los derechos de S.M. Fidelísima el rey don Miguel primero. ⟨"Las cosas van bien en España", me
  dijo el referido individuo⟩
- 08.03-09 (T8): [x...] reconocer la usurpación, y ya también porque [un] acto semejante [podría] {ocasionar?} ... ser bien
  recibido por la gran [mayoría] del pueblo portugués, [que] es en lo general adicto a la persona del [señor] don Miguel.
- 08.12-22 (T8): en #507 #403710131 [Portugal] ... consolidación del trono de [Doña] María ⟨podría dar al Gobierno del Rey
  N.S. garantías suficientes de⟩ una [paz] duradera ⟨... poco común en⟩ [los] [agentes] austriacos.
- 09.07 (T9): [x...] las potencias {desean?} sin duda renovar sus antiguas [relaciones] con Portugal ⟨y conociendo que el
  sistema de gobierno ... es incompatible con sus principios ... abandonar este gabinete⟩
- 09.17-10.02 (T9, P092): las ventajas que saca de esta especie de [colonia] inglesa y ser inconsecuente con su [propia]
  política, restableciendo [al] [señor] don [Miguel] en su trono [legítimo].
- 10.06 (T9): ⟨porque además existe en⟩ este país una especie de odio personal contra el [señor] don Miguel ⟨por algunos excesos⟩
- 10.09 (T9): supuestos que le [h]an atribuido ⟨... para llevar a efecto los deseos de⟩
- 10.12 (T9): las potencias y los de la Inglaterra, que el declarar a [Doña] [María] de la Gloria [reina] [absoluta]
  ⟨acompañando este acto con⟩
- 10.16 (T9): la convocación de las antiguas cortes [portuguesas], reconocer la {unanim...?} y con este apoyo [negociar] con el
  ⟨Gobierno del Rey N.S.⟩
- 11.01 (T9): ⟨igual⟩ [reconocimiento], [fundando] este en convenios de mutua [seguridad]
- 11.04-09 (T9, P112): ⟨Con este objeto se ha encargado⟩ la Inglaterra de preparar el {movimiento?} que [ha] tenido lugar
  en Lisboa, y [al] general Globet, [parece], [ha] sido el [encargado] de [nivelar] la [conducta] [política]
- 11.12 (T9): ⟨debe seguir⟩ la [corte] de Lisboa
- 11.21 (T9): ⟨incremento⟩ y será por lo mismo más [difícil]
- 12.01 (T9, burned held-out set): [realizar] el plan de establecer el [gobierno] absoluto
- 12.04 (T9, burned held-out set): prevenido de antemano de las proposiciones que tal [vez] a su [llegada] [a] Madrid podrán
  [hacérsele].

## Notes on the reading

- Spelling of the time kept where the cipher gives it (reyno, Magestad, ynglesa). The unit e/c stands for u or b; in this
  reading u/b is read by sense and not counted as a restoration.
- The general is written "Globet" (Goblet d'Alviella, Belgian general sent to Lisbon in 1837).
- Corrections made at the review of 9 Nov 2026, before publication: words that are not in the cipher removed ("de Lisboa"
  in 07.03, "del Norte" in 09.07, "se" in 10.09; "se trata de" in 05.b and "existe en" in 10.06 are clear text); doubtful
  words marked {?} moved to unresolved; "la ?egadades" read "la [ll]egada de" (restored); the certain class limited to the
  verified table.

Translation

translation.md

# azuares1837: English translation of the cipher runs (with the clear context), 9 Nov 2026

Restored words in [ ], unresolved in { }, clear text of the letter in ⟨ ⟩. Certain = verified table T6 (views 06-07);
the rest is probable (see reading.md).

- 03: [indicator] the Austrian chargé d'affaires in this capital ⟨whom I saw yesterday⟩ ... ⟨convinced me that⟩ the
Austria[n] #4607229 ⟨is also trying to take advantage of the movement that has occurred in that Kingdom⟩.
- 04: ⟨has not prevented the steps taken by⟩ #5071574925 {unresolved}; ⟨in contrast with the one observed by⟩ the French
minister, not so with that of {...} ⟨who presented himself as a stranger to all the events⟩.
- 05: ⟨This language⟩ of the Austrian representative ⟨expressed with natural liveliness and affected gaiety has made me
suppose with some ground that it is a matter of⟩ preparing some change in the present system of the Portuguese
government, and from [this] consolidation the Powers hope to take advantage to secure order in that country, even at the
cost of sacrificing the rights of His Most Faithful Majesty King Dom Miguel I.
- 06 (certain): ⟨and as it is⟩ Globet who guides Prince Ferdinand ⟨who is a young man full of character, I do not doubt that
he will know how to⟩ establish in that kingdom a government adapted to the circumstances, monarchical and enlarged with the
ancient institutions dear to the Portuguese people. ⟨Britain knows this need ... the Lisbon government⟩ should hasten to
treat with King Don Carlos on mutual bases of security and political convenience #5344727605; the arrival of His Majesty
in Madrid will no doubt begin this step, which the court of Lisbon desires.
- 07 (certain): ⟨I have official information that⟩ the court {thinks?} ... General Globet is a safeguard ⟨for the spread of
revolutionary ideas⟩; ⟨his Government⟩ has abstained from entering into any negotiation with the foreign Powers;
⟨as an idea of my own, that⟩ the recognition of Doña María da Glória ⟨could present great difficulties, both because this
was the same as⟩
- 08: recognising the usurpation, and also because such an act could {cause?} ... be well received by the great majority of
the Portuguese people, who are in general devoted to the person of Dom Miguel ... #507 #403710131 Portugal ... the
consolidation of the throne of Doña María ⟨could give the government of the King our Lord sufficient guarantees of⟩ a
lasting peace ⟨... unusual in⟩ the Austrian agents.
- 09-10: the Powers {wish?} no doubt to renew their old relations with Portugal ⟨and knowing that the system of government
... is incompatible with their principles ... this cabinet [to give up]⟩ the advantages it draws from this sort of English
colony and to be inconsistent with its own policy, by restoring Dom Miguel to his legitimate throne ⟨because moreover there
is in⟩ this country a sort of personal hatred against Dom Miguel ⟨for some excesses⟩ supposedly attributed to him ⟨... to
carry out the wishes of⟩ the Powers and those of Britain, than to declare Doña María da Glória absolute queen
⟨accompanying this act with⟩ the summoning of the ancient Portuguese Cortes, recognise the {unanim...?}, and with this
support negotiate with the ⟨government of the King our Lord⟩.
- 11: ⟨equal⟩ recognition, founding it on conventions of mutual security ... ⟨To this end⟩ Britain ⟨has undertaken⟩ to
prepare the {movement?} that has taken place in Lisbon, and General Globet, it seems, has been charged with levelling the
political conduct ⟨which⟩ the court of Lisbon ⟨must follow⟩ ... and it will therefore be more difficult ...
- 12: to carry out the plan of establishing absolute government ... forewarned of the proposals that may perhaps be made to
him on his arrival in Madrid.

Summary: Azuares reports talks with the Austrian chargé d'affaires in London. Austria and Britain contemplate changing the
Portuguese regime (making Maria II an absolute queen and reviving the old Cortes, with General Goblet guiding Prince
Ferdinand), even at the cost of Dom Miguel's rights; Lisbon would seek a treaty with Don Carlos on mutual security once he
reached Madrid. Azuares objects that recognising Maria would mean recognising the usurpation, that the Portuguese people
remain loyal to Dom Miguel, and that such a settlement would not secure a lasting peace.

Key: key.tsv

05 Sources

Archivo Histórico Nacional, Madrid, DIVERSOS-COLECCIONES,161,N.2: PARES description 3735838, 12 images; the images are not reproduced here.
Brice Dudout, the target folder: the transcription, the tables, the frozen protocols and measurements, the reading and the translation.
Catalogue entry 2421 (Daniel Bourdeau’s sweep of PARES, AHN Diversos-Colecciones, vetted 9 November 2026), now removed from the catalogue.

The notes, the tables, the transcription, the reading, the translation and the frozen test files are in targets/azuares1837.

Cite this page

Brice Dudout and Daniel Bourdeau, “Azuares to the Carlist Secretary of State, London, 8 September 1837”, Cyphersolver, posted 9 November 2026, https://cyphersolver.com/azuares1837.html.

Manuel Azuares to the Carlist Secretary of State, London, 8 September 1837