Experiment 189
F11.1, first family run. The community tried M-94 with hill climbs only (knowledge/05-attacks.md); no bounded result exists. The 25 disc alphabets as published (jproc.ca/crypto/m94.html, credited to George Mace; disc B1 matches the Cryptogram 1982 reprint; each alphabet checked to be a permutation starting A then its disc letter). Use (TM 32-220 par. 144): 25 discs in the day's order, 25 plaintext letters set on the guide row, any other row taken as cipher. K4 as consecutive groups of 25 from an unknown alignment (0-24), each group with its own row offset (1-25). Each known letter allows a set of discs at its slot; slots must take distinct discs (a matching). Exact over every alignment and every offset triple; K4 against 10 shuffled copies; plants.
What the method does with the 24 known letters.
The US Army's M-94 was a stack of 25 lettered discs on a rod: line up the messagethe message The 97 letters of English that Jim Sanborn wrote and enciphered. along one row, copy out any other row. The discs' alphabets are published; only their order and the row chosen were secret. Each known letter says which discs could sit at its place; if no ordering of the 25 discs fits all of them, the M-94 did not make K4the carved letters The 97 letters cut into the copper at the CIA, shown at the top of every experiment page..
Words in the margin are defined on the glossary page.
The carved letters go in at the top; the reading comes out at the bottom.
The M-94, the US Army's cipher cylinder until 1942, is 25 discs with published mixed alphabets: set the message along one row, copy out another. Only the disc order and the row were secret.
Each known letter allows only the discs that carry the carved letter the right number of places after it; the discs at different places must all be different. Every group start and every row choice, decided exactly; ten scrambled copies, two planted texts.
No disc order fits K4 or any copy under any setting; planted texts are found. F11.1 closed for the M-94.
This experiment produced no candidate: the search ruled everything out before a full text was built. Its product is the bounded negative in section 7.
Redo the method on a few letters with pencil and paper.
The same steps run by the code on the real letters.
groups of 25 from place 1; first group's cipher row 7 below the guide row: discs that can stand at place 22 (E over carved F)
place 22: ['F5', 'O14', 'P15']
Family, key space, the constraint under test, and the clues it reads.
A9 (methods known to the creators in 1990; the M-94 was declassified long before), C1 ('matrix codes'), F11. Model: c_i = D_{k[s]}[(pos(p_i) + g_group) mod 26], s = (i + a) mod 25, k a disc order. The 24 known letters fall in at most three groups; for fixed (a, g's), each slot's allowed discs are the intersection over its known letters; consistent iff a system of distinct discs exists. Statistic: consistent (a, g) cells, K4 against 10 copies.
Written before the run; the result decides it.
No alignment and row choice admits a disc order on K4 or on the copies (about 24 strong checks); plants are found. Zero candidates.
The harness's record: candidates, gates, loss, and the negative result this experiment leaves behind.
consistent (alignment, rows) cells: K4 0 (copies 0-0); plants [{'alignment': 2, 'cells': 4, 'true_cell_found': True}, {'alignment': 18, 'cells': 58, 'true_cell_found': True}]
What was searched, as the harness recorded it.
How this experiment sits among the others is on the lineage page; its family's open hypotheses are on the families page.