The History of Hut Eight

we look to see whether the letters thrown up are good or bad. For this a simple scoring system was used, E being worth something in the region of 57 and Q about -100. On a really long crib the right position usually showed up clearly and a large number of distances were established by this method.

Depth cribbing died with Banburismus in autumn 1943

2. Straight Cribs.

I should perhaps have dealt with straight cribs before depth. The theory of straight cribs is simple, it is merely a matter of guessing the contents of a message without the assistance of depth and without the contents having already been passed in another cypher which has been broken.

Finding straight cribs was largely a matter of analysing the traffic. At first, with the traffic fairly small, this was comparatively easy and an organisation was created for writing down any message which looked as if it might occur regularly. Having found a message of a routine type, details about it were recorded. Significant facts were normally frequency and frequencies on which it was retransmitted, call-signs, German Time of Origin, and length. When a few examples had been written down it became possible to assess

a. Whether, given an unbroken day's traffic, it would be possible to identify this particular message.

b. Whether, once identified, it had few enough forms to be used as a crib,

These two factors - identification and forms - were the essential factors in all cribbing. In our experience identification, though often tricky, could usually be cleared up by careful examination of the evidence and there have been comparatively few cribs that have been unuseable because they had been unfindable. In very difficult cases we tried to enlist the assistance of R.F.P. but, though help was most readily given, the experiments were never very successful. The German security service appears to have considered that re transmitting of messages of one area in another area was dangerous and did something towards stopping curious linkages by recyphering messages and adding dummy at the end before retransmission. This sort of thing was a nuisance to us, but never became sufficiently widespread to cause serious difficulty.

The ideal crib is not shorter than 35 letters and uses the identical wording everyday. Such cribs never existed though


< previous

next >