flop.â Knox then gave his bullet-pointed appraisals of the others:
A few notes on senior staff:
a) Kendrick is quite admirable. It is a pity I have put him in the [Elmers] school since, thoâ he has a lot to learn, he is the obvious second (or first?) in command.
b) Welchman is doing well and is v keen. I hope to get him back here [from Elmers School] to learn about the machines.
c) Twinn is still very keen and not afraid of work.
Knoxâs attitude towards the young Alan Turing was more ambivalent:
He is very difficult to anchor down.
He is very clever but quite irresponsible and throws out a mass of suggestions of all degrees of merit. I have just, but only just, enough authority and ability to keep him and his ideas in some sort of order and discipline. But he is very nice about it all. 2
For a little while, there was some tension between Dilly Knox and Gordon Welchman. Knox as the senior cryptographer felt that the Cottage was overcrowded. So Welchman was exiled to the nearby temporarily requisitioned Elmers School (as the memo cited above indicates, Knox was keen that the exile should be short). The idea was that, rather than codebreaking, Welchman would be trying to crack the slightly different problem of âtraffic analysisâ. This involved studying the origins of intercepted signals to work out where they came from, and thereby hopefully making deductions about military movements.
Welchman wrongly inferred from his exile that Knox had taken a dislike to him; as a result, he seemed to suffer slightly from wounded pride. It also seems to be the case that Welchman wasnât smitten by Knox himself. Years later, he wrote: âHe was neither an organisation man nor a technical man. He was, essentially, an idea-struck man ⦠By and large Dilly seems to have disliked most of the men with whom he came into contact.â 3
Knox might well have smelled something to his distaste about the younger man; for very early on, Welchman was working on schemes to restructure and reorganise the entire codebreaking operation, in such a way that a great deal of power and authority would have slipped from Knoxâs hands.
Indeed, Welchman wasted no time at all during his exile. Mathematician John Herivel recalls one of the key innovations. Welchman was quick to form a strong working relationship with the most secret listening station â âan old fortâ at Chatham, Kent, on the Thames Estuary. It was through this establishment that, at that time, much of the German signalling traffic came.
Once the signals were picked up, the officers at Chatham would have bundles of intercepted coded messages sent to Welchman. These bundles would be driven from the coast to Bletchley, through the middle of the night and at harum-scarum speeds, often in the most atrocious weather conditions, by special motorcycle couriers.
The German rule was that no message should be more than 250 letters in length; if it was necessary to send a longer message, it should be split into multiple parts. This was designed to make life more difficult for codebreakers: the longer the message, the easier it might be for such a person to see patterns of letters forming among the apparent chaos.
But each Enigma message had a preamble and some operators used different discriminants (that is, groups of characters indicating the code set-up and reciphering key, and distinguishing the section of traffic) for the different parts. As Herivel noted, it was thus possible to work out any âgiven keyâ. âIn this way,â he wrote, âall the different German keys coming from Chatham â and later from France â could be identified, and every day, traffic could be divided into different bundles for different keys.â
The keys were then given different colours, to separate them out in the most direct visual way. To start with, the colours used were yellow, green, red and blue, respectively signifying the German