106 | 11 BREAkING mACHINES wITH A PENCIl
the adjoining loft in Cottage 2: Cottage 2 was occupied by one of the maintenance staff and his
family, who could not be allowed to know that German codes were being broken.
Since there was no specific provision for Dilly in the new plans, he made his own arrange-
ments, moving into the plum store once Turing’s machinery was shifted to Hut 1 (itself recently
vacated by Station X’s wireless operators). A Jabberwocky-esque poem by Dilly began ‘Come to
the Store my bombe-ish boy’, as the plum store was now his only hideout. Dilly’s resignation letter
finished with an urgent complaint about the provision made for Turing in Denniston’s distasteful
flow charts. Noting that, according to Denniston’s plan, ‘Mr. Turing, if successful in finding meth-
ods for a solution of German Naval Traffic, should work “under Mr. Birch” ’, Dilly continued:^16
The very suggestion . . . is so absurd and unworkable . . . that I could no longer remain in your
service to work with its proposer.
Frank Birch was Dilly’s long-standing best friend, but he knew next to nothing about Enigma.
Birch thought that Turing, although brilliant, was disorganized, untidy, could not copy things
correctly, and dithered between ‘theory and cribbing’.^17 Birch had not even grasped the fact that
Turing had actually abandoned his theoretical approach, which involved the mathematics of
groups, for Dilly’s practical cribbing.
It all worked out well in the end. While Birch, an expert on naval matters, was responsible
for analysing intelligence in Hut 4 and providing cribs for Turing, he had no direct control
over Turing and Hut 8. Dilly was not the only one to be concerned about Turing’s position
since, although Turing had to be titular head of Hut 8, it really needed a natural organizer like
Welchman to run it. Hugh Alexander, an excellent manager and cryptanalyst who was first
recruited to Hut 6 in February 1940, was the ideal solution to this problem. He transferred to
Hut 8 and Naval Enigma in March 1941, quickly becoming acting head, so leaving Turing free
to operate in the way he wished among staff he could feel at ease with.
After the fall of France, in 1940, Welchman went to Cambridge to scoop up more mathemati-
cians once the results of the Mathematical Tripos examinations were in. He recruited Keith
Batey (later my husband) for his own Hut 6, and having previously supervised Joan Clarke he
earmarked her for Turing in Hut 8. He told both recruits that mathematicians were not actually
needed for the codebreaking work, but that they tended to be good at it. Joan later recalled that
as a new arrival she was ‘collected’ by Turing and put on to testing bombe ‘stops’ in the timber-
framed Hut 1.^18 This was very laborious work with the first bombe, Victory, since there were
excessively many stops (as explained in Chapter 12).
All was now well with Dilly’s position. His former Cottage 3 and the adjoining Cottage 2 were
knocked into one, and stairs were put up to Turing’s old loft, so providing room for the staff
joining Dilly’s new Enigma Research Section. Dilly decided that he wanted only female staff,
and (even before Joan Clarke’s appointment) he had recruited Margaret Rock, a mathematician
from London University. The rest of his recruits had language skills. Although linguists were
an obvious choice, we also had a speech therapist and someone from drama school. We were
known—even in Whitehall—as ‘Dilly’s girls’.
Railway Enigma and ‘Prof ’s Book’
Turing and Dilly worked together again briefly in June 1940. The Germans had introduced
a rewired version of commercial Enigma for use by the Deutsche Reichsbahn, the German