# The History of Hut Eight

position of the middle wheel would have changed, but both trigrams have W in the middle. This fit knocks out wheels 4 6 7 and 8 and BBC + 2 = BBE knocks out wheel 2. Further we have the fit PNX + .5 = PIC: as the two trigrams have different letters in the middle, there must be a turnover between X and C and the wheel must therefore be 5, 6, 7 or 8; 6, 7 and 8 are already denied by C + 13 = L, so the wheel is 5.

The next stage is to count up the score for the alphabet assuming it to be correct. For instance, if there are two messages BDL and BDS, the score for BDL + 4 = BDS will be recorded and should be better than random if the alphabet is right. The final score should be a handsome plus total. An attempt, usually not difficult, is then made to complete the alphabet with the help of any further fits or good scores which may exist. It will be noticed that the fit QQG + 2.7 = QDU is contradicted: it had, however, only an even chance of being right and we do not let this worry us unduly.

A similar process finds the alphabet for the middle wheel and sufficient material is then available to break the day on the bombe. The numbers of wheel orders which have to be run will probably have been reduced from 276 to something between 3 and 90.

This example shows clearly the fallacy of the system of having all the wheels turning over in different places. It was this characteristic alone which made it possible to distinguish the wheels by Banburismus and reduce the number of wheel orders to be tried. Wheels 6, 7 and 8 were indistinguishable from one another and a great nuisance to the Banburist.

Like depth cribbing, which was closely allied to it and which will be described in due course, Banburismus was a delightful intellectual game. It was eventually killed in 1943 by the rapidly increasing number of bombes which made it unnecessary to spend much time and labour in reducing the number of wheel orders to be run: it was simpler and quicker to run all wheel orders.

JANUARY 1940 TO JULY 1941

Turing's solution of the indicating system came at the end of 1939 but it was well over a year before Banburismus was established as a practical proposition and used as a successful method of attack. The reason for this was primarily a lack of bigram tables.

-20-