Cryptographic History of Work on the German Naval Enigma


6. Fundamentally the problem was now solved; the story of Banburismus for the next two years is one of improving methods and of struggling to get sufficient staff to enable us to reach maximum efficiency combined with a steady growth of the relative and absolute importance of cribbing as the increasing numbers of bombes made the running of cribs a quicker and quicker affair.

7. April and May were intensely interesting months for the Banburists. We had a hard struggle to get going, largely owing to the disrupting effect of dummy messages; this indirectly however proved of advantage since it forced us to investigate the whole questions of dummies, what frequencies they came in, what time of day, lengths and so on. However once we started things fairly rapidly improved and ultimately the whole of April and all May except 1st to 6th were broken though we were never anything like current. A considerable increase of cryptographic staff also took place, Chamberlain and Yoxall joining in May and Good, Charlesworth, Ashcroft and Noskwith in June. Ashcroft and Noskwith joined Wylie who was quietly engaged in killing himself by working all three shifts as a one man crib room and the other four joined the Banburists making them nine strong. A gradual increase in clerical staff also began to take place though for some mysterious reason Grade III clerks seemed to be scarcer than University mathematicians.

8. The next major event was the capture of the June and July keys. Although this infuriated the cryptographers who were thoroughly enjoying Banburismus and the cribbing and depth cribbing associated with it, it was a great stroke of good fortune. In the first place it gave us a much needed opportunity to take stock of the position and to overhaul our methods in the light of the experience of April and May. The whole dummy problem was gone into thoroughly and detailed figures produced to show the chance of messages of all and every type being dummy; a fresh statistical investigation into the frequency of "repeats"

-30-


< previous

next >