June 1943 - August 1944.
1. The period covered by this chapter is characterized on our side by rapidly increasing bombe power, on the enemy's side by attempts to improve security (1) by splitting up keys (2) by technical improvements in the keys - a steady move from 3 wheel to 4 wheels key and a gradual abandonment of the faulty throw-on system of indication. This battle was won by us, as the steady increase in the number of keys attacked and broken and the maintenance of a firm hold on all the major keys clearly shows. The chart showing the splits and changes in the keys (Appendix , graph ) should be of some help in following the events described in this and the next chapter.
2. Some events in June still remain to be considered. The abandonment of short signal menus in favour of R.E's on Shark, was an important chang in policy caused not so much by the advent of 4 wheel bombes (there were not yet enough of these to cause a major change in policy by themselves) as by the changing situation in the North Atlantic. The withdrawal of the Atlantic line and the issuing of long instructions re-encoded for the benefit of the Northern U-boats still using Dolphin made the R.E's so much better than any available short signal cribs that it paid us to adapt the 3 wheel bombes to run 4 wheel jobs even though the latter took 26 time as long to run.
3. The appearance of Turtle (a separate key for Mediterranean U-boats) was a fairly natural key split; it did not greatly affect our chances of breaking Shark, except that we could no longer use Mediterranean short signals which had sometimes helped us in the past. Turtle had no good cribs of its own and we were dependent on re-encodements from Shark for breaking; it was never a key of major importance and was not read continuously though we broke a fair proportion of the traffic in most months.