success revived general interest in the S.K.O; we knew we might have to use the method later to find the wiring of reflector Caesar and the Gamma fourth wheel [See ****] and this early work saved us time later.
39. The other coup was Good's success in breaking into a recalcitrant Limpet (Shark Offizier) day. All our previous successes in discovering Offizier steckers had been when the message set-up was known; by use of Hollerith machinery Good succeeded in making a break with over 20 possible alternative set-ups from which to choose. This was not an intrinsically important result but it was an achievement of some technical interest. All known methods of attack on the steckered machine, if all possible position have to be examined, depend finally on a crib [Banburismus, an apparent exception, provides a crib of a specialized type]; the most interesting remaining question in Enigma is "What can be done on a single message if there is no crib?" Offizier techniques give a solution if only one message setting has to be tried; Good solved a long message with 20 alternatives - later using the American photographic machine, Hypo, we tried unsuccessfully to break into a very long message trying all 17,576 positions. No attempt of this sort on Enigma has yet succeeded, though a somewhat similar problem on JNA 20 was solved.
40. The other events in June (1) our first success on our four wheel bombes (2) the successful trial run on the American four wheelers (3) the decision to abandon the policy of running weak short signal menus (4) the decision to run all wheel order cribs on Dolphin without waiting for the results of Banburismus to declare themselves (5) the split in Shark, Mediterranean U-boats going on to the separate key Turtle (6) the news of the introduction of Caesar and Gamma on July 1st (7) the American successes, on Limpet, using Hypo - all these belong to the new period rather than the old and will therefore be considered in Chapters V.......VI.