Experiment 191
F11.3. Bury (Cryptologia 2007, via scz.bplaced.net 9.2): the agent writes a verse row by row into a 10 x 10 box, reads it out by columns, and combines the result with the message through a 26 x 26 reciprocal table (p + k + c = 25). The report's scratch check (K1-K3, best 7 of 24, no controls) is rerun with controls and widened: key texts K1, K2, K3 and K1-K3 joined, the Morse, and Carter's 1923 book (K3's source); box widths 2-50, rows as many as 97 letters need; columns read left to right, right to left, bottom to top, or in a vocabulary keyword's order; tables: Bury's reciprocal, Vigenere, Beaufort and variant, numbered on the standard or KRYPTOS order; every start in the key text. Each setting checks all 24 known letters (compiled); K4 against 10 shuffled copies.
What the method does with the 24 known letters.
The CIA's agents in 1970s Poland took a line of a book, wrote it into a square box row by row, copied it out column by column, and used that jumble as the keythe secret word The word or string a cipher needs besides the method. on a letter table. Sanborn's 'matrix, then the tablethe table The right-hand half of the sculpture, 26 rows of the alphabet each shifted one place further, starting from the word KRYPTOS.' fits this well. We try the sculpture's own texts and the Carter book behind K3the earlier messages The three solved parts of the sculpture. as the key text, every box width, every starting point.
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.
CIA agents in 1970s Poland wrote a line of a book into a box by rows, copied it out by columns, and used that as the key on a letter table: a remembered rule, a matrix, then the table, close to Sanborn's own words.
K1-K3 and Carter's book as the key text, every box width from 2 to 50, columns in several orders, Bury's reciprocal table and three others on the standard or KRYPTOS order, every starting letter; all 24 known letters must agree.
No setting fits K4 or any of ten scrambled copies; the best partial agreement on K4 is what the copies reach too. Planted texts are found. F11.3 closed for these key texts.
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.
K2 from its 40th letter in a box 10 wide, read by columns: key stream EATHIENDOLYRIEORSEAOUTCINEMRNCSHFNWDIGUAESIFAATRNTDMEOSNTUKITALRGDEUNOHGDMATDNONENXATRUDWXEETTHAN
place 22: box gives T, Bury's table needs Q for E over F
place 23: box gives C, Bury's table needs O for A over L
place 24: box gives I, Bury's table needs Q for S over R
place 25: box gives N, Bury's table needs L for T over V
place 26: box gives E, Bury's table needs W for N over Q
place 27: box gives M, Bury's table needs V for O over Q
Family, key space, the constraint under test, and the clues it reads.
A9, F11: the CIA's own documented 1970s agent system, a remembered rule, 'a matrix, then the table' (C1). Model: key stream k_i = T[o + r_i w + col(j_i)], j_i = i div h, r_i = i mod h (h = ceil(97 / w)), then c from p and k by the table rule. Exact: a setting fits only if all 24 known letters agree. Statistic: fitting settings and, for the short key texts, the best number of agreeing known letters, K4 against 10 copies; plants.
Written before the run; the result decides it.
No setting fits all 24 known letters on K4 or on the copies; the best partial agreement on K4 lies inside the copies' range (about 7-8 of 24). Zero candidates.
The harness's record: candidates, gates, loss, and the negative result this experiment leaves behind.
K1: fits K4 0 (copies 0-0, plants [0, 0]), best K4 0/24 (copies 0-0); K2: fits K4 0 (copies 0-0, plants [1, 0]), best K4 8/24 (copies 7-8); K3: fits K4 0 (copies 0-0, plants [0, 0]), best K4 9/24 (copies 7-9); K1-K3: fits K4 0 (copies 0-0, plants [1, 0]), best K4 9/24 (copies 7-9); Morse: fits K4 0 (copies 0-0, plants [0, 0]), best K4 0/24 (copies 0-0); Carter 1923: fits K4 0 (copies 0-0, plants [0, 1])
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.