“One letter dropped or added in carving, with a K3-style move and a repeating key on the true length”
0093 allowed one wrong carved letter, which keeps the length at 97. A dropped or extra letter changes the length to 98 or 96, so every grid changes shape and every later letter shifts. Here the true text is the copper with one letter struck out (96) or one unknown letter put back (98) at every place outside the crib stretches, run through every single move on that length (K3's rule, one quarter turn, plain and true-keyword columnar) and a repeating key of length 1-26.
Sanborn left out one letter when carving K2the earlier messages The three solved parts of the sculpture.. If he left one out of K4the carved letters The 97 letters cut into the copper at the CIA, shown at the top of every experiment page., or carved one too many, the messagethe message The 97 letters of English that Jim Sanborn wrote and enciphered. is really 98 or 96 letters long. That matters for grid methods: 97 letters only fill a lopsided grid, but 98 fill 7 rows of 14 and 96 fill 8 rows of 12. We try every place for the missing or extra letter, every grid move of experiments 80 to 85 on the corrected length, and a short repeating secret word, against the 24 known lettersthe known letters Sanborn has confirmed that positions 22 to 34 of the message read EASTNORTHEAST and positions 64 to 74 read BERLINCLOCK..
Words in the margin are defined on the glossary page.
1.1
The method at work
The carved letters go in at the top; the reading comes out at the bottom.
the slip
If the carver added or dropped one letter, K4 is really 96 or 98 letters long, and every grid move on 97 letters is the wrong shape. 96 fills 8 rows of 12; 98 fills 7 rows of 14.
the check
Strike each carved letter in turn, or put back an unknown at each place, apply every single move on that length, then a repeating key: known letters on one key letter must agree.
the bound
149 variants, 648,709 variant and move pairs, keys of 1 to 26 before or after: nothing below period 20, and every period sits inside the scrambled coppers' range.
2
Best candidate against the known letters
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.
3
Walkthrough, by hand
Redo the method on a few letters with pencil and paper.
Suppose the carver added a letter at place 50. Strike it out: the text is now 96 letters and BERLINCLOCK starts at place 63.
Write the 96 letters into rows of 12, eight full rows, and read them out by columns.
Each known letter lands at a new place; with a key of length p, its key letter is that place mod p.
Known letters on the same key letter must agree.
Suppose instead a letter was left out before place 50: put back an unknown there; it fits any key, and later known letters move up one.
Repeat for every place, every move on 96 or 98 letters, key length 1-26 and table rule.
the worksheet: the first steps by hand, on squared paper
4
Worked example, generated
The same steps run by the code on the real letters.
strike carved letter 50 (A): 96 letters in 8 rows of 12
OBKRUOXOGHUL
BSOLIFBBWFLR
VQQPRNGKSSOT
WTQSJQSSEKZZ
WTJKLUDIAWIN
FBNYPVTTMZFP
KWGDKZXTJCDI
GKUHUAUEKCAR
known letters after the strike: EASTNORTHEAST at 22-34, BERLINCLOCK at 63-73
5
Hypothesis, for a cryptanalyst
Family, key space, the constraint under test, and the clues it reads.
Type 7 of directions.md. K2 has a confirmed carving error: an S dropped from the copper and the text corrected in 2006 (01-facts.md, kryptos-group-2006-corrected-k2.txt); 0068 tested one letter missing, extra or wrong only for autokey with no move. A dropped letter matters for type 1 because 97 is prime and 96 and 98 are not: 96 = 8 x 12, 98 = 7 x 14. Variants: delete carved letter j (j outside the crib stretches, 73 places) or insert an unknown at gap j (outside the inside of either stretch, 76 gaps); crib places shift with it. Moves on the true length n: K3's rule i -> a + b i mod n with gcd(b, n) = 1, one clockwise or anticlockwise quarter turn at widths 2 to n-1, plain columnar either direction, true-keyword columnar from the puzzle's words, forward and inverse, deduplicated per n. Repeating key 1-26 after the move (slot = carved place mod p) or before it, 8 tableau classes. A crib landing on the inserted unknown imposes nothing. Ten shuffled coppers with the same variants.
6
Prediction
Written before the run; the result decides it.
No variant makes a repeating key of length 20 or less consistent beyond the scrambled-copper rate; the count of consistent settings at each period lies inside the ten coppers' range, give or take one period. Zero candidates.
7
Method
For each deletion place and insertion gap, build the true ciphertext and the shifted crib places.
For its length, build the deduplicated single-move library.
For each move and class, compute the key value forced at each known letter and test periods 1-26 in both orders by slot agreement.
Repeat on ten shuffled coppers with the same variants.
List K4's settings below period 20.
8
Result and bounds
The harness's record: candidates, gates, loss, and the negative result this experiment leaves behind.
0candidates returned
0passed every gate
0/24known letters, best
–lowest loss
nonechampion before
149 variants x 3,809/4,877 moves (96/98) = 648,709 settings x 8 classes x 2 orders, periods 1-26. K4 live below period 20: 0. Outside control range: {'A': [], 'B': []}.
What was searched, as the harness recorded it.
family
one carved letter dropped (insert an unknown, n = 98) or extra (strike it, n = 96), single move on n, periodic key