Nuncio Acciaiuoli from Lisbon to the Secretariat of State, 1758

Summary. Portogallo 117 holds the letters of Filippo Acciaiuoli, nuncio in Lisbon, and of the abbé Testa, 1757–61. DECODE has eight of the cipher leaves, transcribed digit by digit in 2016–17. In 2020 George Lasry reconstructed the key (two-digit homophones, the digit 6 a null, four-digit code words that all begin with 2) and attached a letter-by-letter decryption of all eight, with the code groups left as numbers. No Italian text was posted, and the records stayed “partially decrypted”. Three of the leaves (10, 24 and 31 October 1758) carry the Secretariat’s decipherment in clear above the figures. Lining those up with the cipher fixes 25 code words (Re, Regina, Spagna, tutto, molto, ministri, e che…), and with them the other five leaves give running Italian. Contents. On 12 September 1758 the nuncio sent Rome, in cipher, what Lisbon was not saying aloud. The king had not fallen ill: coming back by night from a lady “delle prime famiglie”, he was fired on in his calesse. His valet and favourite Pedro Teixeira was wounded almost to death, the king was hit in the right arm, and a surgeon took him to the Real Barraca. The later letters follow the wound, the king’s fear of Carvalho, the Jesuits, and in 1760 the marriage of the Princess of Brazil. What stays open is about eighty code groups that occur only once or twice, and a few passages in the two long leaves where the transcription has slipped.

Letter of 24 October 1758: six lines of clear Italian decipherment above lines of figures — Portogallo 117/6 (f. 106r), 24 October 1758: the decipherment made in Rome, “Benche si dica il Rè assai meglio…”, written above the figures. Image: Archivio Apostolico Vaticano, Portogallo 117/6 f. 106r, via DECODE R199.

01 The documents and the key

The volume is “Lettere di mons. Acciaiuoli, nunzio in Portugallo, e dell’abbate Testa, 1757–1761”, ff. 2–585. DECODE dates all eight records 18 January 1757, which is the first date in the volume. The cipher leaves themselves run from 12 September 1758 to 1760:

Item	DECODE	Folio	Date	Digits	Clear on the leaf
117/1	R25	95r–v	12 Sept 1758	1918	no
117/2	R195	98r	[Sept 1758]	860	no
117/3	R196	101r	26 Sept [1758]	492	no
117/4	R197	103r	10 Oct [1758]	190	decipherment
117/5	R198	104r	31 Oct [1758]	577	decipherment
117/6	R199	106r	24 Oct [1758]	791	decipherment
117/7	R200	115r–v	28 [Nov?] 1758	1999	no
117/8	R201	280r	Lisbon, 12 [?] 1760	1499	no

The key is Lasry’s reconstruction (DECODE, 24 October 2020). Letters and common syllables have fixed two-digit homophones: a 04 38 55, e 84, i 08 53, o 81, s 88, il 43 91, che 39, non 14, de 71 78, qu 70 87 90. The digit 6 is a null, scattered freely, and the code words are four digits beginning with 2. No two-digit element begins with 2 or contains a 6, so the figures cut up only one way.

02 The code words

Lasry left every four-digit group open. The decipherments on 117/4–6 give them word for word:

Code	Reading	Code	Reading
2759, 2774	Re	2777	Re (of Spain)
2778	Regina	2813	Spagna
2300, 2845	tutto	2899	giorno
2171	molto	2510	lettere
2538	ministri	2799	ministro
2354	Camera	2138	avuto
2533	materia	2153	fatto
2315	altri	2518	maggiore
2188	quello	2140	benché
2134	assai	2178	e che
2395	contrario	2177	onde
2189	egli	2407	detto

A few more come from context only (grade C): 2740 prim- (“delle prime famiglie”, “figlio primogenito”), 2530 chirurgo, 2394 conte (“il conte d’Oeiras”, Carvalho’s title from 1759), 2743 principessa (the Princess of Brazil).

03 The attempt on the king, 12 September 1758

The first leaf (117/1, R25) is the longest. Code words resolved above are in plain type and open ones are given as numbers; the spelling is the cipher’s, which often drops double consonants.

… e si racconta che venendo il Re da una [visita] di una dama della <2183> … Pietro Teixeira, suo assai domestico favorito … fussero sparate contro il calesse <2908> schioppettate; restò ferito malamente il cavalcante e la mula del bilancino che egli montava; il Teixeira ferito a morte, <2149> si <2981> che lo avevano sagramentato; il Re molto ferito del braccio destro … il detto chirurgo portò il Re alla Real Baracca. Non <2195> molto rumore: la materia è delicata, ma diventa <2754> in bocca di ogni popolare. Ho creduto doverne avvisare … La dama è delle prime famiglie … Si pensa che venga il colpo da di lei parenti; che venga da nemici del Teixeira … Il fatto non si mette in dubbio. … L’Infante don Manuel me ne parlò con il sospetto … né più posso per ora riferire.

“…and it is said that as the King was coming back from [a visit to] a lady… Pedro Teixeira, his very intimate favourite… shots were fired at the calesse; the postilion was badly wounded, and so was the mule of the lead pair he rode; Teixeira wounded to death, so that they had given him the sacrament; the King badly wounded in the right arm… the said surgeon took the King to the Real Barraca. Little noise is made: the matter is delicate, but it is on every common man’s lips… The lady is of one of the first families… The blow is thought to come from her relatives, or from Teixeira’s enemies… The fact is not in doubt.” This is the attempt of the night of 3 September 1758, which Carvalho turned two months later into the arrest of the Távoras and the Duke of Aveiro and, in 1759, their execution at Belém.

04 The wound, September–October 1758

117/2 (R195). The illness is the public story. The arm is said to be broken, and the fleets and merchant ships are held back “con danno del Re, de’ mercanti e de’ poveri”. 117/3 (R196, 26 Sept). “Continua la voce sulla causa del male del Re, che da nessuno più si dubitò … restò tutto il giorno a letto … nessuno vede il braccio che [è] fasciato … si vuole che il braccio resterà impedito.”

117/4 (R197, 10 Oct). Deciphered at the time: “Il Rè si è levato, ma [non esce] tutto il giorno. Vogliono i più, che sia molto guasto nel Braccio, e che almeno resterà perpetuamente impedito.”

117/6 (R199, 24 Oct), deciphered at the time: “Benché si dica il Re assai meglio si vuole, che assolutamente il Braccio resterà impedito: parlando un suo confidente col Re sulla morte della Regina [di Spagna] il Re gli disse: è stata buona Sorella, e che sono sicuro, che il Re sposerà la mia Figlia Maria Francisca, e che lo ha promesso alla Regina avanti, che ella morisse: ciò, che il Ministro di Spagna mostra è tutto il contrario … di tutto il resto non posso io compromettermi.” Barbara of Braganza, Queen of Spain and José’s sister, had died on 27 August 1758. The widowed Ferdinand VI did not marry Maria Francisca.

117/5 (R198, 31 Oct), deciphered at the time: the king has not had erysipelas, “ma un sacco di materia, che ha fatto la piaga”, with fever as a result. The king is very apprehensive; “il credere il male maggiore di quello, che si dice” is universal.

05 Carvalho, November 1758, and the marriage question, 1760

117/7 (R200) is the longest leaf and the most damaged by transcription slips. Its drift is clear: Carvalho is “dispotico”, “si dice che il Re ne ha paura”, and the Infantes and the court do not dare oppose him. The leaf touches on confessors and preachers, on a lay brother “cacciato nella Compagnia” and on “l’espulsione”. The nuncio notes that he himself is met with “diffidenza”. A companion “sta tutto il giorno da Carvalho”.

117/8 (R201, 1760) is about the succession. The Princess of Brazil might marry the Infante Don Luis of Spain, and there are frequent conferences between the Conde d’Oeiras (Carvalho) and others. The court of London “è molto allarmata” and fears for its trade from “uno della casa di Spagna e di Borbone sul trono di Portogallo”. The letter ends: “Il paese freme, che vorrebbe la [principessa] unita coll’Infante don Pietro.” She married her uncle Pedro on 6 June 1760. Acciaiuoli was expelled from Portugal nine days later.

06 Method

The DECODE records were fetched with a logged-in session: images, the NiLan/Midas digit transcriptions, Lasry’s key and his decryption, which is one file attached to every record. The decryption was flattened into one letter string per leaf and split into Italian words by hand. Code words were fixed by aligning the three clear decipherments with the cipher below them, and then checked for consistency wherever the same code recurs in the other five leaves (2759 Re 11 times, 2300 tutto 12). The full working text, leaf by leaf, is acciaiuoli1757/READING.md.

07 What remains open

About 80 of the 107 distinct code groups that occur once or twice (2183, 2481, 2754, 2823, 2398, 2159, 2413, 2817, 2803 and others), with no crib. Where the sense can be guessed, it is marked C in READING.md.
Transcription slips produce unassigned pairs (82, 92, 42, 12, 32, 72, 02) and garbled stretches, above all in 117/7 and 117/8. A fresh transcription of those two leaves from the images would recover most of them.
Grades: the 25 codes in the table in section 02 are H (contemporary decipherment on the leaf); 2740, 2530, 2394 and 2743 are C; the word division of the five leaves with no contemporary decipherment is H where a phrase is quoted above, M in the stretches marked … in READING.md.
The rest of Portogallo 117 (ff. 2–585) has more cipher than these eight leaves. It is not on DECODE and was not seen.

08 Sources

DECODE records R25, R195, R196, R197, R198, R199, R200, R201 (images: login).
George Lasry, reconstructed key “ASV – P117” and decryption, DECODE, October 2020.
Digit transcriptions by NiLan (Dec 2016 – Mar 2017) and Midas (May 2017), DECODE.
No edition of these despatches was found (search of 19 September 2026).

Working files, the transcriptions and the full reading are in acciaiuoli1757/.

Nuncio Acciaiuoli from Lisbon to the Secretariat of State, 1758–60