At a glanceNot solved · text not obtained
- The letter
- A letter to Jacques de Ségur-Pardaillan, Navarre’s envoy, from a correspondent at Beaupréau in July 1583 who signs as “vostre tres humble et tres … serviteur”.
- Where
- Bibliothèque nationale de France, Cinq cents de Colbert 401, f. 143 (Gallica btv1b10035574w, view 154). Tomokiyo lists it among letters of Henry III to Ségur.
- The cipher
- Figures up to 133 and a few letter-written signs; by the writer’s own account, letters, syllables, doubles and nulls, with names spelled in letters and syllables.
- How it was tried
- Transcribed from the full-resolution scan; the two keys of the volume applied, without result.
- Why it resists
- 58 figures are below what the annealers that broke the 1585–86 key can handle, and no other letter in this key is in the volume.
- Credit
- George Lasry’s saved decryptions of 15 October 2024 include this letter; they are unpublished.
01 The letter
The volume is the first of Ségur’s négociation for the king of Navarre with the German princes. Three of its ciphered letters, of 1585–86, are Navarre’s in a key recovered here from the ciphertext alone (key A); two more, of 1586, are in the key Tomokiyo rebuilt from f. 321 (f. 333). This 1583 letter is in neither: a different cipher, with its own page since 14 October 2026.
It is not from the king. Its clear text speaks of the cipher itself: “Il me semble qu’il n’est point besoin d’autre alphabet, vous y avez toutes choses tellement dispersées en lettres, syllabes, doubles, nulles et tout le reste qu’il est impossible de le descouvrir … Les noms sont en lettres et syllabes.” That is the design both later keys turn out to have. The clear passages also carry the name numbers 156, 134 and 136, and a short run of figures before la marks a place name.
02 The cipher and what was tried
Three lines and a fragment of a fourth, 58 numeric figures with 42 distinct values, plus the letter-written signs Pa, a, H, L and I:
133 21 73 39 43 111 118 84 24 26 5 62 3 4 30 65 13 70 22 39 14 123 45 121 40 81 42 14 48 15 63 122 35 29 41 Pa 26 6 119 38 81 40 a 2 13 41 14 30 H L I [Ha grand faute des] 156 […]
- The f. 321 key (letters 12–62, syllables from 63) gives nothing: the figures below 12 that it treats as doubles and nulls are frequent here.
- Key A (letters 1–53, syllables 54–123) gives scattered French-looking fragments (…ande efeire…, …faisi…) inside runs that are not French (hfurtsitumailaco), so if the two keys are related they are not the same.
- The structured annealer that recovered key A needed about 460 figures; on f. 333 (108 figures) it failed, and so did a matched control of that size. A run that builds in the alphabetical design would also need this table’s offsets, which 58 figures cannot fix.
03 What remains open
- The whole cipher: no other letter in this key is in the volume (all 440 views were examined), and the second Ségur volume, 500 Colbert 402, is not digitised.
- Lasry’s 2024 decryption of the letter has not been seen.
04 Sources
- BnF, Cinq cents de Colbert 401, f. 143, Gallica (view 154).
- S. Tomokiyo, “Unsolved Historical Ciphers” (cryptiana), listing it as a letter of Henry III.
The transcription and notes are in the repository (targets/segur1583/); the scripts are in targets/segur/. Manuscript rights: Bibliothèque nationale de France.