had the same first letter (e.g. the CRS/CQT tetra given in Para.
11) were counted and scored. An example will make this clear. Suppose
CRS/CQT tetra is between two messages with dummy %'s 15 and 45, categories
XI and XIV position of tetras 6.1 and 8.4 and the score is 18 ^{xx}/266.
First the girl would record it in the form

[2.3 is the distance]

Next she would evaluate it. (1) Looking up categories XI and XIV she finds
19 and 23, say, which indicates that a tetra between them is worth 19 + 23
= 42 units. (2) Looking up on a dummy chart opposite 15 and under 45, she
sees 4 showing that she loses 4 units because of the chance of the messages
being dummy. (3) Looking up 2.3 on a "distance" chart she sees 2 showing that
she loses 2 units for the chance of their having been a turnover of the left
hand wheel in the range 2.3 [this point will probably not be clear to the
lay reader - it is unimportant and only inserted for the sake of completion].
(4) She still has the rest of the score, 15^{xx}/266 to evaluate.
This is done as in Para. 25 (actually rather more
simply) giving a score of +9. (5) The total score now is 42 - 4 - 2 + 9 =
45. Now this tetra, saying that CRS = CQT + 2.3 favours the theory that S
= T + 3 on the right hand wheel and any possible alphabet that makes S = T
+ 3 is supported by this tetra. To find how much it is supported she looks
up 45 on a conversion table and sees 18. 18 units is the final score and is
the value of the evidence supplied by the tetra and would be enter accordingly
on the S and T deciban sheets. The whole of this process would take an experienced
girl about 3 minutes.

27. The work of the Banburists meanwhile consisted in extraction of the good fits from the catalogue and recording in suitable form the existence of all tetras such as the CRS/CQT tetra above; as soon as enough good fits were available work was started in building up possible alphabets, these being tested by scoring on the deciban sheets as described in Para. 16. This work went on simultaneously with the Big Room work and was largely done in the same room.

-106-

Back to Cryptographic History of Work on the German Naval Enigma contents