affairs as from Time of origin, Frequency, call signs, and length we consider that this message is a plausible candidate for a weather message which occurs every day.
B H N W S M S A W M N T C K N N P Z
W E T T E R F U E R D I E N A C H T
Here we see the state of our knowledge about the lower message after assuming we know the clear text of the upper message and it may well be that, from our knowledge of the traffic on this frequency and of the mine sweepers known to be operating, we can guess the text of the lower message. We have now done a depth crib and one that is certainly right. For each 'clock' successfully cribbed we receive a factor of 17 (the language repeat rate) and for each reciprocal one of 26, so we have in our favour a factor of 174x 262 - an astronomical number in 8 figures, which completely lulls any lingering doubts we may have had about the a priori improbability of a given message starting Mit MMM 371 .... In fact no experienced cribster would have troubled to do the calculation but I include it as an example of a method which may very usefully be applied when assessing a crib.
Such a depth crib as that illustrated was rare in the early days of cribbing, largely because we had not on the whole accumulated enough general knowledge or enough experience to have guessed correctly the beginning of the second message. Much of the early depth was ‘dummy’ depth. We have already noticed that the traffic contained very large quantities of dummies ending in a series of consonants. These began with a few dummy words and then said such things as
(-HJA’ the call sign of Brest.)
The large number of messages saying VONVON naturally led to hexagram repeats being discovered by Freeborn when he analyzed the traffic and one would be presented with something of this sort: two messages, trigrams H B N and H D S, have a hexa repeat if H D S is written out 17 places in front of H B N. The overlapping parts of the texts look like this: