Experiment 205
F2.4 (issue #78): SOE's Dutch section put eight random letters at the beginning and at the end of each message (Marks p. 232). If K4's message is padded so, the known words sit s places later on the copper than in the message. The end run only shortens the message (s + e <= 23 keeps BERLINCLOCK inside); it moves nothing, so every pair of runs 0-12 reduces to the start run s. A key that starts with the message or with the copper is the same family (the offset is absorbed by the key word, for periodic and both progressive forms). 0023 tested start runs 1-24 with periodic keys; new here: the progressive key, per letter and per repeat, steps 1-25, lengths 1-52, on the 8 standard/KRYPTOS table classes, for s = 0-12, with periodic keys (step 0) rerun alongside, on K4, ten scrambled copies and planted texts.
What the method does with the 24 known letters.
Wartime agents sometimes put a handful of meaningless lettersnulls Letters inserted into a message that carry no meaning, used to pad it or to disguise it. before and after a message, so the real words do not start where an enemy expects. If Sanborn did that, the known words sit a few places further along the copper than in the messagethe message The 97 letters of English that Jim Sanborn wrote and enciphered.. We slide them along by 0 to 12 places and test every repeating or steadily sliding secret word on the sculpture's tablethe table The right-hand half of the sculpture, 26 rows of the alphabet each shifted one place further, starting from the word KRYPTOS..
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.
SOE's Dutch section put eight random letters before and after every message. Padding at the start moves the known words along the copper; padding at the end moves nothing.
Every start run 0-12, a repeating key or one that slides by any step after every letter or every pass, lengths 1-52, on the sculpture's tables, K4 against ten scrambled copies and two planted padded messages.
Nothing fits below length 26 on K4 or on any copy, whatever the start run; the planted messages are found at their own lengths. F2.4 closed.
What this experiment was built on, from the lineage page: 1 experiment before it, back to the root of its line. Its number is in bold.
This experiment produced no candidate: the search ruled everything out before a full text was built. Its product is the bounded negative under Result and bounds below.
Redo the method on a few letters with pencil and paper.
The same steps run by the code on the real letters.
start run of 8 nulls: the known words sit 8 places later on the copper; KRYPTOS table, added form
place known carved place carved key value
22 E 30 N 8
23 A 31 G 6
24 S 32 K 20
25 T 33 S 2
26 N 34 S 13
27 O 35 O 0
Family, key space, the constraint under test, and the clues it reads.
C4(a) (a mask that pads the plaintext), C17(b) (the working text longer than 97). Source: reports/marks-between-silk-and-cyanide.md. Model: carved = nulls(s) + T(message) + nulls(e), message positions of the known letters 22-34, 64-74 (Sanborn's numbering of the message), so carved places q + s; T a tableau key with value w[i mod L] + d i (per letter) or w[i mod L] + d floor(i/L) (per repeat), d 0-25, L 1-52, plain and cipher alphabets standard or KRYPTOS, added or Beaufort form (8 classes; the variant form is the added form with -w). Exact by slot agreement of the 24 forced values after removing the step. Statistic (fixed): per form, the shortest consistent length over all s and steps on K4 against the copies, and the cells (form, length) where K4's count of live (s, step, class) settings exceeds every copy's.
Written before the run; the result decides it.
For every start run, nothing consistent below length 24-26 with a periodic key (as 0023), and the progressive forms at the copies' rate; K4 above every copy in no more cells than a copy is. Planted texts found at their own setting. Zero candidates.
The harness's record: candidates, gates, loss, and the negative result this experiment leaves behind.
Shortest live length K4 / copies: periodic 26 / 26; per-letter 26 / 26; per-repeat 26 / 26. K4 above every copy in 0 cells, copies 0-5. Plants: per-repeat found True (shortest 9), per-letter found True (shortest 11)
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.