The History of Hut Eight

his book, "to be some doubt as to the feasibility of this plan" and in fact it proved over-optimistic and work was fizzling out when Norway was invaded and the cryptographic forces of Hut 8 were transferred en bloc to assist with Army and Air Force cyphers.

By the time work on Naval could be started again the 'Narvik Pinch' of April 19th had taken place. This pinch revealed the precise form of the indicating system, supplied the Stecker and Grundstellung for April 23rd and 24th (though the scrap of paper on which they were written was for sometime ignored) and the operators' log which gave long letter for letter cribs for the 25th and 26th. Wheels 6 and 7 had been introduced by this time and were already in our possession.

In all 6 days were broken, April 22nd to 27th. The 23rd and 24th presented no difficulty and the days paired with them (by this time the wheel order was only lasting 2 days and there were ten Stecker) the 22nd and 25th were also broken. The 26th gave much more trouble, being on a new wheel order with unknown Stecker. At first a hand method, the Stecker Knock Out was unsuccessfully tried and then the Bombe, which had arrived in April and which will he discussed later, was put onto the problem. After about a fortnight of failure due largely to running unsuitable menus the day was broken on a freak menu, to be known later as a Wylie menu and tried unsuccessfully on Shark of February 28th 1943. The paired day, the 27th was also broken.

All hands now turned to EINSing out messages and building up the bigram tables. Provided that the Grundstellung was known, the starting position of the message when broken by EINSing could be transposed and the trigram discovered. This then gave a value for 3 bigrams. A message with 2 of the 3 operative bigrams known could be 'twiddled' out. The 2 known bigrams fixed the positions of 2 of the wheels and only 26 positions for the remaining wheel had to be tried: this was quickly done and a further bigram was added to the store. April 27th was on the same bigram table as the 24th and this table came near enough to completion to make Banburismus feasible on another day using the same table. May 8th was identified as using this table and a Banburismus was started but no results were obtained for many months. Turing wrongly deduced that June was using different bigram tables.

The next 6 months produced depressingly few results. Such Banburismus as was tried was unsuccessful and there was little bombe time for running cribs. Such cribs as there were were


< previous

next >