though more through bad luck than bad management. Cribs were fairly weak and were only run on the 2 reflector combinations considered most likely; the factor put against them by these cribs was for a long time considered insufficient to justify running a 3rd reflector combination. To understand this statement it is necessary to look at the longest available piece of evidence on reflectors which we had at this time - Shark combinations from June 1943 to October 1944
From these it will be seen that there was a tendency to use all 4 combinations before repeating and that there was no repeat between consecutive months or with only one month in between. October Porpoise used the combination of August and so caused considerable trouble; later this became a much more frequent occurrence in series of reflectors.
Very little Porpoise was run after the first few cribs had failed but at the end of October the revival of Bonito revealed a series of reencodements into Porpoise and breaking got under way again. No one thought much of Porpoise for its Intelligence content but we valued it highly for its connection with Bonito and the use it might have in time of need. The policy of keeping Porpoise going paid in the end a handsome dividend; without the reencodements both keys might well have been lost but as it was one or the other could usually be got out somehow and the other would follow in due course. Although we never worried unduly if a day or two were missed, most Porpoise was broken until just before the final collapse in Italy.
On October 29th Sucker became an independent 4 wheel key and providentially there started a few days later the Hoofden Weather reencodement between Dolphin and Sucker which was to break Dolphin