good reencodements for recent days and it did therefore seem more than ever doubtful whether such poor Short Signal menus were worth while. Accordingly we agreed (1) not to run a long programme of menus on any day to which we had access via a reencodement (2) to have plugboards set up for 12 bombes and to use these on reencodements until we have sufficient 4 wheel machines for the latter to cope with the problem on their own:..... We used 6,681 bombe hours this week or about 70% of the total - easily a record which I hope we shall never surpass.
We now began to clear off back days of May and June steadily on reencodements; but at the end of June an Oyster message revealed that a new reflector and reflector wheel were to be introduced on Shark on July 1st reflector Caesar and wheel Gamma. This was a serious blow but, like so many changes, it came too late. Cribs on the new reflector could not be tackled by bombe until the wiring was known - the bombe is dependent on prior knowledge of all wiring - and one day had therefore to be broken by hand methods. I shall not attempt to describe here the process of Stecker Knockout (S.K.O.) by which this was done: it had long been known to be a technically possible method of attack and had been successfully used on May 27th Shark. Suffice it to say that it was a lengthy and laborious process which was only possible with a very long crib; 400 letters or more were desirable.
Caesar and Gamma were broken in the middle of July; later we were to discover they had been captured some time previously in N. Africa but had not been sent back to B.P. on the grounds that we had plenty of German wheels already. With Gamma at position A Caesar and Gamma equalled reflector C which had been broken by Hut 6 in 1940 as a result of a German error: reflector C was never brought into use on the Army or Air Force keys, their next reflector was the pluggable D which was an unpleasant affair we were fortunate enough to avoid.
Once the appropriate wheels and plugboards for the bombes had been prepared several July days were broken and it soon became apparent that the reflector combination remained unchanged for the whole month. The occurrence in August of the mixture Bruno Gamma established that there were to be 4 possibilities to consider at the beginning of each month. This was not too bad and henceforth reflectors gave us comparatively little trouble.
The Weekly Report of August 20th contains an interesting statement:
"We have had a record week on Shark, the whole month clearing up. Every day from August 1st to 18th is now out. As a result we have been hard put to it to use the fast bombes, and handed them over for a time to Hut 6 to run 3 wheel jobs. We have given the first half of July to the Americans to get out on their own: this should give them more practice in