BATEy | 103
pairings that specified the day’s Enigma plug connections to encrypt the weather messages. As
Rejewski said, this was ‘a cardinal error on the part of the Germans’.
Turing cannot have mentioned much about his own work when speaking to the Poles, since
Rejewski explained later that they treated him as ‘a younger colleague who had specialised in
mathematical logic and was just starting on cryptology’.^9
Turing, Dilly, and Naval Enigma
Turing himself had no dealings with the making of the Zygalski sheets, nor with using them
to discover the wheel order and ring setting. This is probably why the Poles thought him a
beginner—whereas he was at the time busily engaged on German Naval Enigma, on which the
Poles had done little. It was a friendly few days’ stay, and Turing did bring back one vital piece of
information, which explained why Bletchley Park was having no success in applying the Polish
methods. The Poles had inadvertently given us incorrect information about the turnovers on
two wheels (see Chapter 10 for an explanation of wheel turnovers).
Soon after Dilly had met Zygalski at the Warsaw conference he penned a note to Alastair
Denniston, his companion in Warsaw and the operational head of the GC&CS. The note sur-
vives today, handwritten on notepaper bearing the printed heading ‘Hotel Bristol, Warszawa’.
Knox ended his note dramatically:^10
It cannot be too strongly emphasised that all successes have depended on a factor (the machine-
coding of indicators) which may at any moment be cancelled.
It was this danger of the Polish methods suddenly ceasing to work that Dilly impressed upon
Turing when they met at Dilly’s home after Pyry. The need for a more robust method was
Turing’s guiding star in his subsequent work on Enigma. Dilly was right: the indicator system
changed on 1 May 1940 and overnight the Polish methods became inapplicable.^11
Back from France, Turing closeted himself in his stable-yard loft to grapple with German
Naval Enigma. Although the wheel wirings were the same as those that the other German
military services were using, the problem was that the message settings of naval messages were
super-encrypted (see Chapter 10) and Turing needed to deduce how this super- encryption
worked. Dilly was aware of this difficulty when he worked on German Naval Enigma in 1937,
but he had soon abandoned the naval code to work on German Army and Air Force Enigma.
The responsibility for cracking German Naval Enigma was now Turing’s alone, although
Dilly gave him every encouragement: Dilly himself would soon be working on Italian Naval
Enigma. In the meantime, Dilly persuaded his good friend Admiral John Godfrey, Director of
Naval Intelligence, to allow him to be present at prisoner interrogations: in a paper written after
the First World War Dilly had argued that this practice should have been adopted. In November
1939 Dilly managed to extract some crucial information about German Naval Enigma from a
Kriegsmarine prisoner of war, who explained that operators were now instructed to spell out
numbers (‘Eins’, ‘Zwei’, ‘Drei’, etc.) instead of using specific letter keys to represent numbers (the
Enigma keyboard had no number keys). In fact, the Germans used both methods for a while,
which Turing found very useful.
In the 1940 treatise that Turing wrote on Enigma, known at Bletchley Park simply as ‘Prof ’s
Book’, he described Dilly’s ‘pen and pencil methods’ in order to illustrate the use of cribs in
general.^12 He made extensive use of probability in ‘Prof ’s Book’, but probability was anathema to