A son of William V to an officer of the Orange troops, c. 1795

Summary. DECODE lists R1892, with no decipherment, as a two-page letter to Prince William V, in “graphic signs, numerical”, and an archivist’s note on the photocopy says no nomenclator was found. The figures are a 6×6 square: every plaintext letter is a vertical pair of digits from 1 to 6, top over bottom. Both pages were transcribed from the DECODE images (1,829 pairs). A solver (simulated annealing with a Dutch four-letter model) recovered the square from the Dutch page without a crib, and the same square then decrypted the French page unchanged. The writer is a son of the Stadholder (“mon père”, “mon frère”), instructing the officer in charge of the Dutch émigré troops gathered in Germany: if General Dundas stops their pay or orders them to embark, accept it and ask only for a gratuity for those who will not submit; tell the troops on what footing they will enter the service of His Britannic Majesty. The most frequent sign is de; three more are probably zich, te and ik; about twenty rare signs and one faded line are not decrypted.

01 The document

Record: DECODE R1892, “unkmown (Germany) to Prince William V”, 1 Jan 1795, Non-decrypted, two images (photocopies, photographed sideways).
Shelfmark: Koninklijk Huisarchief, The Hague, archive of Prince William V, bundel A18, inv. 339.
Languages: page 1 French, entirely in cipher; page 2 a Dutch postscript (“het volgende schrift … in het hollandsch met cijffer”), cipher mixed with clear words, closing with a clear French sentence.
Keys tried first: the two key records DECODE files in the same series, R2240 (the four-digit Orange name key) and R2233, do not apply: R1892 uses only the digits 1–6.

Two lines of the Dutch postscript: clear words, then digit pairs written one above the other — Page 2, the head of the Dutch postscript: “in het hollandsch met cijffer”, then the pairs 55 32 31 51 65 … (top digit over bottom digit). Image: Koninklijk Huisarchief, The Hague, Prins Willem V inv. 339 (bundel A18), p. 2, via DECODE R1892.

02 The square

Token = top digit followed by bottom digit. The square is not in alphabetical order, and no key sheet is known.

11	12	13	14	15	21	22	23	24	25	31	32	33	34	35
t	w / z	y, ij	x	l	z	o	k	f	m	d	n	v	h	b
41	42	43	44	45	51	52	54	55	65
g	u	a	s	p	c	r	q	i	e

12 is w in the Dutch and stands for the final z of pourrez, jugerez in the French. The pairs 53, 61, 62, 63 and 64 occur one to three times each and are probably misreadings of the photocopy. The graphic signs (a circle with a dot, the most frequent; Λ, Δ, γ, ⊣H, a cross and some twenty others) stand outside the square for whole words or names.

03 The decipherment

Page 1 (French)

Decipherment, word-divided and lightly normalised; [·] marks a graphic sign.

… écrit en cognito … du … en date du … décembre. Mon père chargeant Nagel d’aller [·] nouveau chez les ministres pour insister sur une réponse quelconque aux différentes représentations qui leur ont été faites touchant … du rassemblement, je me flatte qu’enfin … en viendra une; mais comme il pourroit arriver entretemps que le général Dundas vous déclarât qu’il cesseroit le payement, ou bien que les troupes devroient se laisser embarquer, il faudroit y souscrire et vous borner uniquement à tâcher d’obtenir [·] pouvoir donner une gratification à ceux qui ne voudroient se soumettre dans ce cas … Je vous prie cependant [·] donner connaissance aux troupes du pied sur lequel elles entreront au service [·] S.M.B., et vous pourrez leur donner à cet effet une pièce à peu près semblable à celle qui étoit destinée à leur être donnée dans le premier moment où il étoit question [·] leur embarquement. Vous pourrez trouver cette pièce parmi mes papiers, et vous pourrez y faire les changements que vous jugerez convenables vu les circonstances présentes. Ayez la bonté [·] ne point mettre sur le compte … la dépense pour mon logement, ne voulant pas … être dans le cas d’avoir la moindre obligation, mais [·] le payer [·] l’argent que vous avez à moi appartenant.

“Nagel” is A. W. C. van Nagell, the Orange envoy in London; the cipher spells the general’s name Vundas, a slip for Dundas; S.M.B. is Sa Majesté Britannique.

Page 2 (Dutch postscript)

… in het geval … het rassemblement te zamen te houden … die vrijwillig … te laten … de officieren … en benoemen die … het zelve … willen aangesteld … alle heeren officieren … kunnen weten … geene die hier … willen geëmployeerd … en geeven tegen … bepaalde dagen … dan zelve keus maken in … nemen … het mogelijk is alleen … die hun in zedert het intrekken der Fransche niet geboden hebben; indien zich hoofdofficieren genoeg bij … ieder [1000] man kunnen plaatsen … wat in cijffer schrijf … naam … mijn nader … ons tot nog toe niet tegenstaan alle stellingen die gedaan … antwoord te … reeds weken … mijn eerste … toen zou … mij zeker niet gesteld hebben; tegenwoordig nog hier te … op antwoord te wachten … welverzekerd … mij niets … engageeren … veel liever … gebleven was … mijn gesteld is geweest had willen mededeelen.

Le prince [·] me marque en outre … vous aurez la bonté de communiquer le contenu de ce chiffre à mon frère.

The postscript opens with “den generaal Bertincy” (probably Bentinck).

The postscript deals with keeping the corps together, naming and placing the officers, giving those who will not serve a fixed number of days to choose, and the writer’s weeks without an answer to his first letter.

04 Who and when

The writer calls William V “mon père” and asks that the cipher be shown to “mon frère”, so he is one of the Stadholder’s two sons, the Hereditary Prince or Prince Frederick. The addressee is an officer managing the rassemblement, the Dutch troops who followed the House of Orange into Germany in British pay after the French conquest of January 1795. DECODEDECODE’s “1 January 1795” is the file’s date;rsquo;s DECODE’s “1 January 1795” is the file’s date;ldquo;1 January 1795DECODE’s “1 January 1795” is the file’s date;rdquo; for R1892 is the file’s date; the letter cites a despatch “en date du … décembre” and discusses the corps’ transfer to British service and embarkation, which fits the winter of 1795–96. “To William V” in the record reflects where the letter is kept, not its addressee.

05 Method

Fetched the record and its two images, and the key records R2240 and R2233, from DECODE with a session login.
Turned and cut the photocopies into strips; two transcription passes (one per page) wrote each column as a TOP+BOTTOM token, clear words in brackets and signs by name: 1,039 pairs on page 1, 790 on page 2.
Page 2’s pairs, split at every sign and clear word, went to a simulated-annealing solver scoring Dutch quadgrams from the shared corpus, with a one-to-one key (swap moves). A many-to-one first run collapsed onto e and n and was discarded. The one-to-one run gave hetrassemblementtezamentehouden and officieren at once.
The same square applied unchanged to page 1 gave French text; four values were adjusted from it (54 = q, 14 = x, 13 = y, 12 = z in French).

06 Word signs and what remains uncertain

Each sign was listed with the decrypted text around it.

Sign	n	Value	Grade	Evidence
circle with a dot	27	de	H	d’aller de nouveau; au service de S.M.B.; op de volgende wijze; bepaalde dagen
γ	5	zich	M	degene die zich vrijwillig …
c-shaped sign	5	te	M	de keus te maken; in cijffer te vinde[n]
G	7	ik	M	ik wist wat in bovenstaande
⊣H	9	an auxiliary before willen?	I	three times directly before willen
hook	7	a conjunction (dat?)	I	usually just before G

About twenty rarer signs occur one to four times each and are not identified.
The close, transcribed again at full resolution, is secure (H): 43 25 22 32 24 52 65 52 65 = à mon frère, and the group before “me marque” is le prince.
Page 2 opens “den generaal Bertincy”, probably General Bentinck (C).
Page 2 line 24 is too faded to align its two digit rows and is decrypted only in fragments; page 1’s first line likewise.
The writer (which son) and the addressee are inferred, not stated.

07 Sources

DECODE, R1892 (images IMG_R1892_I9358_P1, I9359_P2); key records R2240 and R2233.
Koninklijk Huisarchief, The Hague, Prins Willem V, inv. 339.

Transcriptions, decrypts, solver and notes: r1892/ in the repository.