Fig 4.3 Enigma Breaks (Hut 6): 1942 and 1944
By early 1945 the Heer was employing only a few Enigma ciphers, which greatly expanded the traffic they carried. In turn, this resulted in many more re-encipherments and mistakes by operators – and in a considerable increase in the Heer decrypts. However, Hut 6 suffered what was perhaps its worst blow of the war on 1 February 1945, when the Luftwaffe implemented a system under which call signs were enciphered daily, and frequencies changed every third day. Although these precautions had been anticipated, they threatened the interception and solution of all Luftwaffe traffic, both Enigma and low-grade: Luftwaffe Enigma decrypts fell from 1,800 to 1,000 daily. A post-war history concluded that ‘if this internal German reform … had been a little more thorough, far reaching, and rigidly enforced … German messages henceforward would have become largely unbreakable’. Since it had been known in advance that the frequencies would usually be more easily identified by reference to the attributes of the low-grade codes being used than those of Enigma, a special ‘traffic watch’ was established by the Air Section in Hut 10 to sort and identify the traffic and pass its details to Hut 6. A combined Sixta and Air Section research party built up a number of frequency continuities by mid-February, but Hut 6 did not recover fully until around mid-March, when documents with details of the new systems were captured. Heer and Luftwaffe Ultra began to dry up in April, as Germany disintegrated.
From very small beginnings, Hut 6 evolved into a large, well-managed and highly adaptable unit. Only by doing so was it able to counter German measures that would have defeated a less flexible organization. The Y service operators also reached a very high degree of efficiency as the war progressed. Without their help, Hut 6 would have been greatly handicapped, especially when the Heer and Luftwaffe made significant changes to their radio procedures, as with enciphered call signs. Few Heer Enigma ciphers were easy to break at any time. Hut 6 gave considerable attention to Green, but solved it on only thirteen occasions throughout the war – and even then some of the breaks required the help of prisoners of war. If the Luftwaffe’s Enigma operators had been as well trained as their Heer counterparts, or if some German security procedures had been introduced early in the war, there would have been much less Ultra from Enigma. And while it is unwise to claim too much for Ultra, it was undoubtedly the prime source of intelligence for the Allies, and helped significantly to alter the course of the war in their favour.
5
HUT 6 FROM THE INSIDE
DEREK TAUNT
Introduction
In Chapter 5, Derek Taunt gives an insightful view of working in Hut 6. Hut 6’s main purpose was to provide Hut 3 with intelligence in the form of decrypts. A small section in Hut 3, known as 3L, advised the various sections in Hut 6 on priorities, such as the most urgent ciphers to be attacked, the radio frequencies to be covered, and so on. This had been something of a problem initially, when ‘an exaggerated view of security made liaison between Hut 3 and Hut 6 difficult’ so that, for example, there was little intelligence input into intercept cover, which ‘was not at first satisfactory’. When there was a balance between efficiency and security, the latter tended to predominate. That was certainly the position in relation to the Ultra indoctrination of senior staff at the intercept stations where, for example, at the Beaumanor Army intercept station, only the commanding officer was initially told about Ultra. Right up to the end of the war, only officers at intercept stations were officially informed, to the detriment of efficiency. The irony is that official security was illusory. On one occasion, when the teleprinter staff at Beaumanor enquired on the phone about a background of cheers in Hut 6, they were told that it was because the main Luftwaffe cipher had just been broken. After that, the teleprinter room always realized why they were being told to stop sending a particular register. There is not the slightest indication that security suffered, and there may well have been a gain in efficiency.
Derek Taunt refers to the arrival of the US Army’s 6813th Signal Security Detachment under Captain Bill Bundy. Although Stuart Milner-Barry initially ‘viewed the prospect with some consternation’, because Hut 6 ‘was faced with technical problems which would make it difficult to find the time for training’, the gifted American contingent made a real contribution to Hut 6’s work. Milner-Barry later described their arrival as ‘one of the luckiest things that happened to Hut 6’.
Gordon Welchman’s The Hut Six Story emphasized the importance not only of a good management team but of ensuring that each individual member of staff was kept informed about how his or her work had ‘contributed to some specific improvement’. Reports gave ‘all Hut 6 staff members a feeling of what was going on’, while the supervisory staff ensured that ‘everyone knew how he or she fitted into the overall picture’. The views expressed on this subject in Welchman’s book are not the product of mere hindsight. Milner-Barry, who worked closely with him throughout the war, has written about ‘his practical gifts for planning and organization’, and that ‘without the vision which again and again proved his intuition correct … the task of converting the original break-through into an effective organization for the production of up-to-date intelligence could [not] have been achieved’. It was ‘a classic example of the hour producing the man’.
Derek Taunt’s chapter is further confirmation that Hut 6 was indeed well managed. Much is owed to Welchman, who has not received the full recognition he deserves. Without him, Hut 6 would almost certainly not have evolved so quickly, or performed as well as it did, while the bombes would have been much, much weaker unless and until someone else had his flash of inspiration about the ‘diagonal board’ that transformed them into such potent weapons against Enigma.
RE
I was an undergraduate at Jesus College from 1936 to 1939, reading for the mathematical tripos and specializing in ‘pure’ mathematics. In the summer of 1939 I made overtures to the great Trinity mathematician Professor G. H. Hardy, whose lectures on Analysis I had attended and greatly enjoyed; he agreed to take me on as a research student. But the project was postponed by the outbreak of war, and I registered with the Joint Recruiting Board. By Christmas I had been allocated to work in the Ordnance Board at Kemnal Manor, Chislehurst, analysing trial firings of new guns and ammunition and drawing up range tables for them.
In Hardy’s minor classic, A Mathematician’s Apology, which came out in 1940, too late to deflect me from Chislehurst, he wrote:
There is one comforting conclusion which is easy for a real mathematician. Real mathematics has no effects on war. No one has yet discovered any warlike purposes to be served by the theory of numbers or relativity, and it seems unlikely that anyone will do so for many years. It is true that there are branches of applied mathematics, such as ballistics and aerodynamics, which have been developed deliberately for war and demand a quite elaborate technique: it is perhaps hard to call them ‘trivial’, but none of them has any claim to rank as ‘real’. They are indeed repulsively ugly and intolerably dull; even Littlewood could not make ballistics respectable, and if he could not who can? So a real mathematician has his conscience clear … mathematics is a ‘harmless and innocent’ occupation.
Perhaps real mathematics (i.e. abstract mathematics) is a harmless occupation; but in wartime real mathematicians, such as I hoped to become, with Hardy as my guru, have to buckle down to harmful activities, be they ever so ugly and dull. External ballistics, as practised at Kemnal Manor, was indeed both – even though the influence of Littlewood was still apparent there, in a method we used which bore his name. (Littlewood and Hardy were great collaborators: Littlewood had outstanding flair and inspiration, while Hardy had patience and impeccable technique. Littlewood claimed to have had sixty seconds of inspiration during his career, each one lasting one-tenth of a second. Hardy spotted the genius of the untaught Indian mathematician, Ramanujan, and over many years provided proofs for most of his fabulous flashes of inspiration.)
But it was not its dullness or uglines
s that made me want to seek another and more challenging outlet for my particular talents, it was the essential triviality of the methods we applied, more like advanced arithmetic than real mathematics. So I sought escape, and found myself at Bletchley Park in August 1941, where I had the very good fortune to land in Hut 6.
At that time it was still a single-storey wooden hut, linked to its neighbour Hut 3 by a makeshift hatchway along which trays of documents could be pulled as occasion required. In Hut 6 messages enciphered on the German Enigma machine were cracked, while in Hut 3 the plain-text we produced was edited and translated, and military intelligence extracted from it. Hut 3 was full of German linguists. As for Hut 6, we were a mixed bag of mathematicians, classics scholars, chess players, bank staff, actresses, with an average age of perhaps twenty-three due to the high proportion of recent graduates and undergraduates (both male and female) and even school-leavers. It was difficult to guess how many staff there were – Hut 6 was manned for twenty-four hours in three shifts – perhaps something approaching one hundred in all. Prior to my recruitment most of the staff had been found through the ‘old boy network’ (including many old and young girls), a procedure which would now be regarded as politically incorrect but at the time was a highly effective way of identifying intelligent and trustworthy people to take on an unknown, top secret task, after being assured by someone they knew well that it was both important and challenging, and about which no detail could be divulged. By mid-1941 the law of diminishing returns had affected the network, and the success of Bletchley Park and the hotting up of the war made ever greater demands for staff. My application for release from the non-real activity of ballistics went through the newly set up ‘usual channels’, headed by C. P. Snow, who had been charged with finding and allocating various talents into appropriate niches in the war effort.
When I arrived in Hut 6 its head was Gordon Welchman, a geometer from Trinity College, Cambridge, who had become a mathematics don at Sidney Sussex College. His second-in-command and successor was Stuart Milner-Barry, a close friend and exact contemporary of Welchman’s at Trinity, a classicist rather than a mathematician, who was well known as a leading chess player and who had been out of his element as a peace-time stockbroker. The technical wizard was Dennis Babbage, the Magdalene member of the well-known Cambridge firm of geometers Hodge, Todd & Babbage, and a close friend and colleague of Welchman (who tried unsuccessfully to match him at tennis, squash and billiards. Dennis was a ‘natural’ when it came to ball games, as I discovered to my cost back in Cambridge after the war.) Other mature figures (i.e. aged at least thirty) included the Scottish Chess Champion Dr James Aitken, and John Manisty, from Welchman’s old school Marlborough College (later at Winchester), a mathematician who first appeared in the uniform of an officer in his school Officer Training Corps, until transferred into the Intelligence Corps.
The rest of the dramatis personae were young – some very young – friendly, energetic, devoted to the task of outwitting the enemy and happy to be part of a complicated organization designed to do just that. Everyone – with the sole exception, I seem to remember, of Dr Aitken – was known by Christian name or nickname, as were departments (such as the Duddery, the Quatsch, and some colleagues from outside who dealt in call signs and were known as ‘the Bathroom boys’) and even forms (such as the ‘blists’ and the ‘hankies’). The Hut operated like a factory which never switched off its assembly line. Raw material, in the form of intercepts of messages sent over the air by German air force and army units, enciphered on their Enigma machines, came in a ceaseless flow to one end of the conveyor belt. There they were sorted and listed, with their essential features highlighted, by the girls in the Registration Room, on forms always known as ‘blists’ (a name not derived, as one might have supposed, from ‘those b-lists’, but as an abbreviation of ‘Banister lists’, after Michael Banister, the Cambridge cricket blue from King’s College who had designed them). The blists of breakable keys would be scrutinized by the next link in the chain of deciphering – the cryptanalysts who comprised the Watch, who attempted to spot the cribs or re-encodements on which breaks depended. The term ‘key’ was applied both to a network using a common cipher and to the actual daily set-up of the machines in that grouping; thus ‘Red’ was the key in general use by the German air force, and ‘Light Blue’ an air force key in North Africa. The shout ‘Red out!’ would convey the news that the day’s key for the Red network had been found. One standard colour – Dark Blue – had been rendered useless when it was discovered that its traffic was purely used as practice and of no operational significance. Before long the supply of differently coloured pencils ran out and we had to have recourse to names of birds, animals, flowers and even vegetables to specify different keys. Eventually we would be breaking between fifteen and twenty of them daily.
When a suitable message for breaking had been identified, the Watch would produce a ‘menu’ for the bombes to test. The menus were not just straight conjectured plain-text for given stretches of cipher text, but configurations of letters and connections derived from such conjectures. The bombes were electro-mechanical machines, manned invariably by Wrens, which simulated the Enigma machines and rapidly tested which of the million-odd possible wheel-orders-cum-ringsettings were consistent with the menu. When consistent settings showed up, the bombe would stop and the wheels involved and their relative positions were read off for further testing in the Hut 6 Machine Room. Many stops were fortuitous – but when a genuine, causal stop was found the rest of the key was readily deduced. Then ‘Red out’ (or Light Blue or Chaffinch or – with less enthusiasm – Beetroot) was proclaimed and the Hut 6 Decoding Room sprang into action to turn the raw material into ‘plain-text’, which, with its numerous abbreviations and technical terms, needed all the expertise of our colleagues in Hut 3 to interpret. But we depended on the actual letters used in a crib or re-encipherment, not the meaning, and we kept records of the (hopefully not too variable) usage in recognizable messages such as weather forecasts, situation reports and tuning messages (such as ABSTIMMSPRUQYYRESTXOHNEXSINN – ‘Tuning signal. Remainder nonsense’). As the machine (unlike a typewriter) had no numbers or punctuation marks these had to be spelt out or otherwise denoted; a gap between words was usually denoted by an X and a full stop by YY, but these offered many opportunities for variants (such as Q for CH). Also, irrelevant nonsense words were inserted at the beginning of messages to create uncertainty as to where the real content actually began.
At the shout ‘Red out’ people might ask which favoured crib or re-encipherment had done the trick, but not who had produced the menu from it. This was irrelevant: everyone recognized that all successes were due to the team, not to individuals. This was true in particular because it was largely a matter of chance which menu turned up trumps – so many things could go wrong, and no single menu had a certainty of success. A 30 per cent chance for a single menu, or 60 per cent for a pair of menus linked to avoid the hazard of a middle-wheel turnover, was as good as you could normally hope for. Since the supply of bombes was never – until very late in the war – adequate to demand, the Watch staff had to exert their judgement on priorities and probabilities. Indeed, a natural feel for probabilities, akin to an appreciation of orders of magnitude, was one of the few specific mathematical talents used in our everyday activities in Hut 6. Knowledge of the twenty-six times table, though useful, could not be regarded as real mathematics!
The Enigma machine was believed by the German experts to be unbreakable if used correctly, even if the enemy had specimen machines and knew the methods adopted for using them. Indeed, it would have been secure if used properly but a) official usages could be insecure, and b) even if they were secure, misuse by operators could be fatal. I can speak only of air force and army and liaison between them, which were dealt with in Hut 6. A regimen of ‘need to know’ was rigidly observed at Bletchley Park, and while heads of departments would discuss their problems when they affected each other (e.g. the
navy cryptanalysts in Hut 8, led by Hugh Alexander, the only English chess player at Bletchley Park ranking above Stuart Milner-Barry, shared the use of the bombes with Hut 6), the rank-and-file did not discuss their work outside the confines of their own huts. The form of Enigma attacked in Hut 6 – most particularly Red – had been used before 1 May 1940 in an inherently insecure way, so long as enough messages (say one hundred or so) were sent on any one day’s key. After that date a change of usage completely shut out the theoretical vulnerability and we became dependent on operators’ misuse and the introduction of machinery to help us exploit known habits. The essential principle was to maintain the continuity of breaking: ‘nothing succeeds like success’ in this game. The clean break of continuity in method of attack left Hut 6 in a Catch-22 situation: we needed continuity to achieve breaks, but we needed plenty of breaks to establish continuity (and of course we needed bombes – and by the end of 1940 there were at most only two). How could Hut 6 bridge this gap and restore continuity?
Two essential features of Enigma, its reciprocity (if A encoded to V, V encoded to A) and the fact that a letter could not encipher to itself, were exploited by cryptanalysts both in fitting cribs against cipher text and in drawing up menus – but also crucially in the design of the bombes which tested the menus to find positions of the wheels consistent with them, from the million possibilities.
Although Alan Turing did not impinge directly on my Hut 6 experience, it was his genius that made our success possible. And he gave a clear rebuttal of Hardy’s ‘harmlessness of real mathematics’! Turing, a research fellow of King’s College, Cambridge, was the most brilliant and most ‘real’ of Bletchley Park’s mathematicians attacking Enigma. His subject was the logical foundation of mathematics, which had been thrown into turmoil a decade earlier by the Czech mathematician Gödel, who had proved that any consistent system of mathematics, defined by axioms, must be incomplete – which means that there must exist meaningful statements in it that cannot be either proved or disproved. This raised the so-called ‘decision problem’: given a meaningful statement in such a system, was there any standard procedure for establishing whether it was provable or unprovable.’ Turing had shown that the answer to this problem was ‘No’. He did this by describing a routine process (or ‘thought machine’) by which the truth of any provable statement could be established in a finite sequence of steps selected from a small number of possibilities. He then showed the existence of meaningful statements which the machine could not decide as true or false. This ‘thought machine’ became known as a ‘Turing machine’. The same mind conceived the basic design of the far from abstract bombes which were the indispensable weapons against Enigma and in fighting the war.
The Bletchley Park Codebreakers Page 9
