Unsolved ciphers / write-up and files
Partial · ending unresolved · 18 September 2026
The cipher postscript of a letter from the Regent of Scotland to his secretary at Elizabeth's court, BL Add MS 32091 f. 213v (DECODE R8345), read from the page image. A simple substitution with a few homophones and word-signs, in Scots, with word spacing.
The reading shows the current manuscript assessment, with unresolved signs left as question marks. At L4.5 the transcribed Z2 label is uncertain and its value is left open. Glyph labels as in transcription.txt; | is a gap the scribe left; [..] a word-sign; (..) a glyph graded uncertain.
| Transcription | Reading with uncertain values marked | |
|---|---|---|
| L1 | U | Zh P g x o Ab xd Tb Z3 e 4b o x M A xd x Ab xd 4 A c a Z5 a c | o2 | Zh x xd F N xd Tb | r A Z5 M | Z2 e U a Z3 o F | [the] | sv(f)eringof[Q]remaneinca(r)l(y)l(r) | [and] | sending | ha(y)m | of[the]lord |
| L2 | V a x p N xd Tb r x Zh F Z3 xd x Tb o x N Tc x P N a Tb N V U Eb | 4 X p M N Zh | r A Z3 M Tc r A N o | fleminghesdonegreitevilgif[the][Q2] | cummis | haomthair |
| L3 | a A Zz Zz N Zh | xd A M A N o Tc Z2 | Zh s N a U | r A N a a x t | Z3 X o x | e o x N xd F N Zh | la(?)(?)is | namairto | s(p)il[the] | haille(?) | oure | freindis |
| L4 | Zz x3 A Xs Z2 N Zh | (?)(n)a(u)(?)is |
The suffering of [Q] remane in Carlyle, and sending haym of the Lord Fleming, hes done greit evil. Gif the [Q2] cummis haom thair, la??is na mair to s(p)il the haill e? oure freindis [seven signs unresolved].
The allowing of [the Queen] to remain in Carlisle, and the sending home of Lord Fleming, has done great harm. If the [Queen] comes home there, […] no more to ruin the whole […] our friends […].
Mary was moved from Carlisle to Bolton on 13–15 July 1568, the date of the letter; Lord Fleming was back in Scotland in arms by 21 August. The two person-signs are read as the Queen from context only.
Gif the [Q2] cummis haom thair, lakkis na mair to s(p)il the haill es oure freindis knauis.
If [the Queen] comes home, nothing more is needed to ruin the whole, as our friends know.
Visual comparison favours ?nau?is. With contextual k and o assigned to the fifth sign, the literal result is knauois, not knauis (knows). The fifth sign differs from ordinary o-signs, but reading knauis requires omitting it without a demonstrated cancellation or null. Reading es (as) also requires an unproved value for the preceding singleton.
The corpus has two genuine verb uses of lattis; six other hits are fragments of prelattis (prelates) split across lines. DOST attests both lattis and lakkis, as well as es and knauis; these establish possible words, not this reading.
The full manuscript review records the glyph comparisons and alternatives, with saved OCR evidence and reproducible conditional readings. The archive search found no matching key or additional same-key ciphertext. NRAS217 Box 45 is a specific location lead whose relevance remains unproved.
The displayed reading has 118 of 134 glyphs S (cryptanalytic assignments), 15 M (uncertain), and 1 I (inferred). L4.5 is graded uncertain because its identity as an o-sign is unproved. Statistical controls for the cryptanalytic reconstruction are documented in the method notes; they do not resolve the handwriting or establish the conjectural ending. No matching key or contemporary decipherment has been identified. The John Wood → Cecil line on the decipherer's specimen sheet Add MS 4136 f. 33 uses the same glyph repertoire in a different key.
| Glyph | Value | Grade | Evidence and qualifications |
|---|---|---|---|
o2 | [and] | I | stands alone between two clauses |
4b | [Q] | M | sign for a person or party after 'suffering of', presumably the Queen |
Eb | [Q2] | M | sign for a person or party after 'gif the'; plausibly the Queen again |
L4.5 | ? | M | long-stroked z-like sign; value and relationship to Z2 unresolved |
Xs | u | M | crossed x visually favours the u-sign; uncertain allograph |
Z5 | y | M | Carlyle, haym; i also possible (Carlile, haim) |
Zz | ? | M | k or t; lakkis and lattis are both attested; doubled in L3 and opens L4 |
c | r | M | Carlyle; the second occurrence should be e and is read as a slip |
g | f | M | suffering; single glyph for f or ff |
s | p | M | spil |
t | ? | M | singleton after haill e; s (es = as) or a word-sign [as] are conjectures |
x3 | n | M | curved x visually favours the n-sign; uncertain allograph |
4 | c | S | cummis, Carlyle |
A | a | S | remane, na mair, haill, thair, Carlyle |
Ab | i | S | suffering, remane in |
F | d | S | sending, freindis, done, lord |
M | m | S | remane, na mair, cummis, haym |
N | i | S | sending, freindis, -is endings |
P | v | S | evil; suffering (u/v) |
Tb | g | S | sending, Fleming, gif, greit |
Tc | t | S | greit, to, thair |
U | [the] | S | the lord Fleming, gif the, spil the; stands alone at the start |
V | f | S | gif, Fleming |
X | u | S | oure, cummis |
Z2 | o | S | to, of; the long-stroked sign at L4.5 is not securely identified with this glyph |
Z3 | o | S | of, oure, done |
Zh | s | S | sending, freindis, hes, cummis, -is endings |
a | l | S | lord, Fleming, evil, haill, Carlyle |
e | f | S | of, freindis |
o | r | S | freindis, oure, remane, greit, thair, lord |
p | m | S | Fleming, cummis |
r | h | S | hes, haill, thair, haym |
x | e | S | sending, freindis, remane, greit, evil |
xd | n | S | sending, freindis, remane in, na |
4 A c a Z5 a c): the two c-shapes have to be r and e for the name and are the same shape; the last is read as a slip for e. Z5 is y (Bain: "Carlyle", "haym") or i ("Carlile", "haim").Zh P g x o Ab xd Tb): P is the u/v glyph; the single curl g stands for f or ff.Zz, also opening L4. Both lattis and lakkis are attested in DOST. Of eight raw corpus hits for lattis, six are hyphenated fragments of prelattis. The two genuine verb hits cannot exclude k.s occurs once; spil (spoil, ruin) by sense.?nau?is; the first sign remains k/t, and the fifth sign's identification as ordinary o is disputed. Contextual k with o at the fifth sign gives knauois. Knauis (knows) requires omitting the fifth sign; knawis requires a different unproved deletion and value. No cancellation, null or compound w has been demonstrated. A negative DOST headword search for ??a?ois depends on identifying the fifth sign as o; it does not eliminate alternatives to that transcription.4̄ and Ē, and the lone dotted o read as and.British Library, Add MS 32091 ff. 213–214 (images not public domain, not redistributed). DECODE R8345. Satoshi Tomokiyo, Unsolved Historical Ciphers, "Moray-Wood Cipher (1568)". Joseph Bain, Calendar of the State Papers relating to Scotland and Mary, Queen of Scots, vol. ii (1900), for the Wood → Cecil crib (no. 804) and Fleming's movements (no. 775). This page is generated from transcription.txt, key.json, README.md and the manuscript assessment in FRESH-MANUSCRIPT-REVIEW.md in the repository, where the README has the method, the controls and the failure log.