FIRST BREAK OF ENIGMA IN BRITAIN
Correspondence between Keith and Mavis Batey and Frank Carter on the first break of Enigma in Britain
While Dilly Knox and his colleagues were having success in breaking a number of variants of the Enigma system, they had had no success with German military version. In the autumn of 1938, GC & CS obtained some remarkable material from French Intelligence. It included photographs of a 1930 edition of an Enigma training manual that had been issued to the German Armed Forces. It contained an example of ninety letters of plain-text, the corresponding cipher text, and the Enigma key that had been used. Using this information, several attempts were made to derive the wiring of the right-hand and middle Enigma rotors. This information was unknown to GC & CS because the Germans had changed the wiring of the rotors in their military version.
When Knox and Denniston met the Poles in July 1939, they were given the wirings of the Enigma wheels as well the vital connection pattern, known as the QWERTZU or diagonal. This information was given to Peter Twinn, who, armed with the training manual information, was able to deduce the wiring of the right-hand and middle Enigma rotors from the example in the manual.
In 2009, Marek Grajek, a Polish cryptologist, historian and author, expressed the view that the wiring of the two rotors could not have been found solely by means of the ninety letter crib given in the manual. This conflicted entirely, with the recollections of those who had been involved at the time, particularly Keith Batey, and his wife Mavis, who had also worked with Knox.
The issue became a matter of discussion between the Bateys and BP historians and authors, Michael Smith and Ralph Erskine. The Bateys then enlisted the help of Frank Carter, who at the time was one of the foremost experts on technical and machine methods used at BP during the war. Using the same methods as Twinn had in 1939, Carter was able to replicate Twinn’s work, proving that it could have been done.
26 October 2009
Dear Frank,
Just to let you know what I said to Ralph (1) who was very concerned about what Mick (2) told him about the Pole (3). I said that:
We only have the wirings of the machines we dealt with in ISK which I was given by GCHQ. We did not have the Service wirings although Keith worked on the machine in Hut 6 and I was unable to contact Frank Carter who could have given them to us, so we were grateful to David Hamer for giving us the upright. For the book (4) I wanted to make much of the manual crib message which Peter proved to be authentic and from which he derived the two wheel wirings once Dilly brought the news of the diagonal from Warsaw. First I only asked Frank Carter if he could put up on his simulator (5) the indicator example given to the operator with the key setting i.e. put up Grundstellung FOL inventing his own 3 letters repeated. For the message beginning (shown to be PKP JXI) and he will get ABLABL which is really where he starts his message (as seem in Appendix 1 in the book). So Frank did just that and indeed did get ABL so we knew it was right. Of course if Peter had waited for the Polish model to arrive (6) he could have done just that too but he got the two wheel wirings through the 90 letter crib by Dilly’s buttoning up method right away. Frank then had a go at buttoning up (7) for the RH wheel as Peter would have done with the stecker stripped off. We assumed obviously that this was called our wheel III as Peter knew that was so from the manual instructions. We rested on our laurels and being in the middle of a move, the effects from which we have not recovered, Keith made no attempt to find the middle wheel as it was not relevant at the time.
However, things changed after the publication of the book and the Pole said it could not have been done on Dilly’s methods which infuriated me. Keith has already found enough clicks (8) to know it was possible to get the second wheel wiring. Now we have the wirings of wheel I, known to be the middle wheel, I am all for cheating and working backwards but I doubt whether Keith will do that! However, he will send the honest proof in due course but I am satisfied that as I said on page 79 it could be done.
Trying to justify the Pole’s point of view he obviously might assume that as Rejewski knew that they were going to hand over the model he could have given him the wheel wirings as well as the diagonal, and indeed may have offered to do so but knowing Dilly, he wanted to get home and have it done his way. I did emphasize that Peter had found the wirings by the time Dilly wrote the report on August 4th and the Polish model did not arrive until the 16th.”
Ralph is the only person who has seen your work on the RH Wheel and I will ask him not to pass it on. I am so sorry I couldn’t get hold of you to ask permission, but Ralph is the sole of discretion. Keith is now clear that it is easy to get the wiring of the middle wheel (i.e. wheel I) as he gathers you have also done.
All good wishes from us both
Mavis
Carter completed his investigation to determine if there was sufficient information in the crib from the German Enigma training manual to have made it theoretically possible, using Knox’s “buttoning up” procedure to recover the wiring characteristics of the right-hand and middle rotors. He sent his draft paper to the Bateys.
3 December 2009
Dear Frank,
Many apologies for not replying sooner to your note on recovering wiring, which we both found most interesting.
We think it admirable: it must have involved much time and effort, which will be greatly appreciated by those interested. Your basic method is what Dilly and Twinn used but they probably organised the work differently by using Foss sheets (tables of 26x26 squares lettered alphabetically on the left and the top): if the buttoned up pairs are entered it is much easier to look for deductions and ‘clicks’: on your part table enter DF in square GI, and so on. It especially makes one-apart and wider pairs easier to handle. Peter’s statement that he got the wiring quickly from the crib given the right hand diagonal makes me suspect that he did find a click, though I haven’t spotted it. (these are Keith Batey’s comments)
Just two comments (from Mavis Batey): When you quote the Hut 6 History it is important to say that this is said by Babbage who was the only one who was there in 1939 of those who wrote the history in 1945. He knew what Dilly’s methods were and had worked with Peter Twinn and Turing in the Cottage before the Hut 6 system was put in place and would have been familiar with the ‘routine operation’.
I think that you are over cautious in your last statement. Turing had also worked on the 90 letter crib in the summer with Dilly and Peter and the only reason, as you know, that it wasn’t successful before the Polish information was the wrong diagonal. Turing was actually staying with Dilly in early August when they heard that ‘Mr Twinn has just checked up on the genuineness of the crib’ (Mavis’s book, p78) and doubtless his workings were soon available. He and Peter went on to break Railway (9) with the same ‘routine operation’. I wrote to Ralph Erskine to see if he could tell me any more about the amount of information Turing and Peter had when they got the Railway wirings. I think it would have been easier than the 90 letter crib as Tiltman had given them message depth readings (10) so there would have been partial alphabets to work on. Although Ralph gives references on p 472 for p 72 ‘solved the Railway traffic’ of Action This Day’ for Railway traffic he had not actually seen them; if the original July decodes exist one could easily get back to the depth readings. We think that evidence points to Turing’s Treatise (11) having been written in autumn 1940 so the break was fresh in his mind and from the chart he made that there were cribs of railway stations and of course a lot of numbers.
Ralph would be very grateful to have a copy of your paper when finished if you don’t mind. I think it would make a very good Cryptologia article.
Best wishes to both from us both
Mavis and Keith
(1) Ralph Erskine
(2) Michael Smith
(3) Marek Grajek
(4) Mavis Batey's biography of Knos, "The Man who Broke Enigmas"
(5) of an Enigma machine
(6) The Poles gave GC&CS a replica of an Enigma machine and Bertrand delivered it to Menzies in person in London
(7) See Batey's book for an explanation of "buttoning up"
(8) See Batey's book for an explanation of a "click"
(9) The Enigma system used by the German Railways
(10) In cryptography, a depth is when two or more messages in a machine or similar cipher have been enciphered on the same machine setting or on the same key
(11) Known as "Prof's Book"
Image: Keith and Mavis Batey, in later life. Credit: the Batey family.
