Secret Life of Bletchley Park - McKay Sinclair [34]
But it was not just a question of sleep. The ever-pressing need for one hundred per cent accuracy meant that the pressure in the small hours could be unusually intense.
Keith Batey offers an insight into the intricate yet repetitive nature of the work, talking of the simple-sounding yet fearsomely complex ‘rodding’ system invented by Dilly Knox. ‘They assumed that there was a word likely to appear in the message. Six or more letters. And you could select places worth looking at by checking for what we called “clicks”. In other words, you set the assumed plain letter against each section of text and look for settings on Dilly’s rods where you got the same twice – charts were made to pick out all the “clicks” quickly. When you got a double “click”, you set the rods up to see if it would work. Very often it did – you didn’t set out to work on each code seventy-eight times.’
As well as Dilly Knox’s system of rods, there was the ‘Zygalski sheet’ system devised by the Poles who had originally broken into Enigma, and updated and modified at Bletchley by John Jeffreys, Alan Turing and Gordon Welchman. The updated versions were known as ‘Netz’ or Jeffreys Sheets.
Of great help to Bletchley was Dilly Knox hitting upon the principle of what were termed ‘cillis’. These were defined by Sir Harry Hinsley as ‘procedural errors by Enigma operators combining (1) a recognisable instead of random choice of message setting, and (2) failure to alter the wheel position much, or at all, before sending a message’. Sir Harry added that this insight was thought to have been named after a German operator’s girlfriend, called Cilli. The operator, using his girl’s name as a message setting, had shortened it to ‘Cil’.
On top of this, there had been, according to Alan Stripp, another chink in Enigma’s armour: the ‘indicator setting’. After following instructions for that day’s wheel order, ring-setting and cross-plugging, the German Enigma operator would turn the three wheels to a random starting point. Then, as part of the preamble to the messages the operator sent, he would twice key in ‘his own randomly selected choice of text-setting’, as Stripp noted.4 This choice might be FJU. He would then tap the setting out again as confirmation, and it would appear as encrypted letters, say PORCDQ.
‘The operator had to give an indicator for the chap at the other end,’ says Keith Batey. ‘But because of the possibility of Morse error, it was repeated, so you got two of these three-letter things. So that you knew that there was a repeat.’
Once Bletchley had worked out the nature of the six-letter preambles, they instantly provided a slender means of working back through the rest of the code by days and days of calculation (though of an intensity that would still be utterly beyond most). ‘But then,’ adds Mr Batey, ‘the Germans suddenly began to realise that this was bad.’ Once the Germans became aware that the practice – originally instituted for extra security – ironically made their communications very much less secure, they put a stop to it in May 1940.
And this had only ever applied to military and air force traffic; the German naval Enigma was never so straightforward to catch out. From the start, it was a very great deal tighter and more labyrinthine in its theory and use. Its operators had far less personal leeway, and thus, the chance of the operators making mistakes were virtually eliminated.
The work was thus not simply a matter of clever young people in Fair Isle sweaters gazing blankly at apparently random letters; it was these same young people using tiresome though necessary means to test and test and test again, against such regular messages as enemy weather reports and call-signs, the language of which were presumed straightforward and repetitious enough to help produce some kind of a crib.
In the coming months, when successful attacks upon Enigma had resulted in a specific codebreaking methodology, the huts were to become extraordinarily intense places to be, not only for cryptographers but also for translators.