could then be tried on the days using this table; (3) Simultaneously attempts must be made to break into other days by running cribs on the bombe.
27. This was so much more hopeful a position than before that the stagnation that ensued must have been doubly disappointing to all concerned. Banburismus proved unexpectedly difficult in practice - May 8th, the most promising day, was worked on ad nauseam without success - and no better luck was had with cribs. Moreover there was suspicion that in June the bigram tables had changed, after which of course Banburismus would be impossible without extensive breaking on cribs or a further "pinch".
28. This gloomy state of affairs continued until November when Foss finally succeeded in breaking May 8th - now immortalized as "Foss' Day". Foss had returned from sick leave in the late Summer and joined Turing, Twinn and Kendrick (who had reinforced Turing and Twinn early in 1940). Everyone else having worked on May 8th till they were heartily sick of it, it was handed over to Foss who, not having seen it before, did not view it with the same aversion. After months of work he finally succeeded in finding "the alphabets" and 'Mr. Foss' special" as the menu for the bombe was known to the machine operators (menus produced as a result of Banburismus were of rather different type from those produced from cribs) duly brought up the right answer.
29. This success was undoubtedly due to Foss' pertinacity; he did not know the mathematical theory to the extent that Turing did but he had endless perseverance and Banburismus was a problem on which this quality always paid a good dividend. Looking back after a period in which day after day was broken by Banburizing, frequently within 24 hours receipt of the day's traffic, it is difficult to appreciate just how hard the first break was. The bigram table was incomplete, there was a large amount of "dummy" traffic (which gave misleading results),