21. (6) Numeral and Dummy Statistics. These were kept up for the benefit of the Bunburists. The value of tetras in different categories and the % chance of a message being dummy depended on these figures. The Yellow Cards supplied the information for numeral statistics and the decoders noted on their message list when a message was dummy and the dummy statistics were compiled from these message lists.
22. (7) Keys Other than Dolphin. All the above remarks apply only to Dolphin. On Porpoise very simple records only were kept, cribs being largely unnecessary at this time. Shark was dealt with by means of Short Signals, fully explained in Chap. IV and no further comments are necessary here.
23. When Banburismus stopped Beginner Folders, Ship Lists and Numeral and Dummy Statistics were immediately dropped. All the remaining entering was done by the girls and the yellow cards (or the slips by which they were later replaced) were examined regularly by the cribsters. Nearly all R.E's came originally on the arctic U-boat frequency and from about June 1943 we had a second copy of all these messages (known as BCU Bar's from the call sign BCU' of the chief group) decoded for the special benefit of the C.R.
24. This "second copy" procedure was later considerably extended. The earlier arrangement, whereby we relied on Self Help and the weather folders, was adequate as long as conditions were stable but later on, particularly after the opening of the second front in June 1944, when cribs began to change rapidly, three serious weaknesses became evident. (1) We could no longer afford the delay of two or three days in the entering up of the folders (2) Certain types of message which were unimportant operationally were, quite rightly, held up in the Z watch while urgent traffic was translated, and the information from these (which might be of great value to us) might not get into the folders for a week or even a fortnight. (3) We could not rely on the girls to spot new cribs as they arose. Even the entering by the men had never