Nonetheless, Poland was worried. With East Prussia to the north, the rest of the Reich to the west and the twin problems of Danzig and the ‘Polish Corridor’, there was plenty to worry about. A non-aggression pact was the obvious way to go and one was signed in 1934. Paradoxically, the pact had a destabilising effect: the French saw it as backsliding by the Poles, weakening their own alliance, which was designed to keep Germany under check from two sides. Now the French felt threatened. So they put out their own feelers to the Soviets. A Franco-Soviet pact followed in 1935 and now it was the Poles’ turn to feel threatened again. What if the Soviets tried a replay of the 1920 war? The close relationship between France and Poland had begun to crumble.
The Franco-Polish relationship faltered just at the moment Hitler upped his game. In 1936, Germany reoccupied the wide swathe of land between France, Luxembourg and the left bank of the Rhine. Just as Hans-Thilo Schmidt had reported, the Germans had jettisoned disarmament.
The seedling of intelligence cooperation, planted by Gustave Bertrand and Gwido Langer, might have so easily wilted in the new climate of mutual suspicion. But their own personal relationship was founded on discretion. If something should not be divulged, it would not be. ‘When [Bertrand] saw you about to ask a question that he couldn’t answer, always he’d say “Ne pas demander!”’1 Sometimes, though, there was a sticky moment. Bertrand had been invited by Langer in May 1933 to observe an exercise in the Polish Corridor, the most vulnerable part of Poland, in which the Poles were deploying their mobile interception and decryption group. The number of German messages which could be read seemed remarkable given that the Poles could, by their own account to Bertrand, do nothing with Enigma. Maksymilian Ciężki had an explanation ready, though. The Germans were extraordinarily lax: they were sending the same message in clear and in cipher! Bertrand was convinced. The fact that numerous messages said ‘Maschine defekt’, which did not need any translation, was indication enough that the Germans were struggling as much as, apparently, their observers were.2
Despite the cooling-off in the upper atmosphere of foreign policy, then, Bertrand was able to carry on his liaison with Langer. He made eight visits to Warsaw in the period 1934–38, and he didn’t go empty-handed. For throughout these years, Hans-Thilo Schmidt continued to supply the French with an array of intelligence which was astounding in both its volume and its quality. In 1932–33, there had been five handovers; in 1934 another three; and then three each in 1935 and 1936. And there were more to come. Gustave Bertrand was nearly always present and at nearly all those meetings Schmidt had Enigma materials to hand over.3
• • •
Hitler’s Germany was changing not only in the way people thought and behaved, but also in the rigour and discipline it was bringing to all aspects of its resurgent military. First, the slackness which allowed Enigma machine operators to use lazy indicators like AAA was stamped out. That was just the beginning. In 1936, the rotor order was changed monthly, and the number of cross-pluggings used on the front of the machine was varied to between five and eight. In October, the rotor order started changing daily. The old grille method developed by the Poles was now too difficult to apply: new thinking was needed.
Jerzy Różycki was 27 in 1936. Now fully engaged as a member of the Enigma team, he was at the peak of his intellectual powers. His response to the disaster of the daily rotor change came from a completely different approach to cryptology: it was an application of good old-fashioned letter-counting, coupled with mathematics.
For centuries it has been known that the easiest cipher to break is a simple letter-for-letter substitution, because each letter in plain language occurs with a particular frequency. In German, for example, the letter E is the most frequent (appearing 16.9 per cent of the time) and Q is the least (0.02 per cent). But if an Enigma machine is used for encryption, the letter frequency flattens out, to the dismay of those trying to do code-breaking by counting. Unless, that is, the Enigma machine is used with identical settings for two different messages, in which case a different theory applies.
Różycki’s had a brilliant insight based on the thought that while letters were not distributed randomly in a normal sentence, they ought to be when encrypted with Enigma, and if they were not, then you had an important clue to work with. If you line up two different pieces of regular German, one underneath the other, the probability that two identical letters appear one above the other is about 1 in 13. And if you line up two different Enigma intercepts, the probability drops markedly, to 1 in 26, which is what you’d expect for sequences of randomly chosen letters. But, if the Enigma settings were identical, this probability went back up to 1 in 13, because the transformation was identical in both cases. And that, concluded Różycki, gave a way in to identify which rotor was in the right-hand position in the Enigma machine. For a run of letters (where there had been no turnover of the middle rotor), the transformation effected on a given day ought to be the same and you could spot the point of a turnover when the rate of coincidence switched from 1 in 13 to 1 in 26. And that switch-over was enough to identify the right-hand rotor setting, since where the turnovers happened was different for each rotor. It was like watching for the turnover on the taxi meter of a cab, or the step-up to the next hour after the expiry of sixty minutes: Różycki and his colleagues called his rotor-detection scheme the ‘clock method’.
The pace of German changes stepped up still more: 1937 brought further modifications, this time to the technology rather than the procedure. The reflector, which turned the current round after it passed through the three rotors before it went back through them in the reverse direction, was changed that year. The methodical German military had sent out a helpful reminder – in an Enigma-ciphered message, which the Poles could read – about the changeover day, giving the Polish team information about what was going to change. Rejewski’s technique for recovering wirings needed to be brought back into operation to find the innards of the new Enigma part. Then the painstakingly assembled catalogue of permutations had to be redone. Yet the achievements of the secret team were enough for more official recognition and encouragement. Marian Rejewski was awarded the Silver Cross of Merit in 1936 and the Gold Cross of Merit in 1938. Henryk Zygalski and Jerzy Różycki each also received the Silver Cross of Merit in 1938.4 The Poles were keeping up. Just.
• • •
The post-war order was crumbling in 1936. It wasn’t just the Rhineland. In the same year, a group of conservatives mounted a military coup on the socialist government of Spain. The Nationalists failed to seize total control and the struggle degenerated into a civil war. The intelligence services of Great Britain began to monitor Morse Code messages being sent by the Nationalists, messages enciphered on the commercial model of the Enigma machine. The man who was assigned to work out what the messages said was an associate of the GC&CS known to his family as Erm.
Erm was born, the second of four brothers, into a family of hyper-achievers. The eldest was editor of Punch, then Britain’s most popular satirical magazine. The youngest was an ostentatious convert to Roman Catholicism, a monsignor and translator of the Bible into English for the Catholic Church: the twentieth century’s counterpart to Cardinal Newman. The other brother was ‘an Anglo-Catholic priest in the East End, a dedicated socialist, a fearless motorbike rider, a welfare worker, an eccentric recluse and just possibly a saint’.5 Unbeknownst to any of these stellar brothers, Erm – so called by them because of his awkward hesitancy in social settings – was the most brilliant star of this small galaxy. His real name was Dillwyn Knox, Dilly for short, and his achievements would remain a secret for another half-century.
Dilly Knox had been one of the stars of ‘Room 40’, the British Admiralty’s code-breaking operation in the Great War. Knox liked to think in the bath; a bath was conveniently installed. There he broke the German flag officers’ code and that told the British where the predatory U-boats were located, making the convoying system a success. Knox had also cracked a differen
t sort of code. A set of irreverent poems called the Mimes of Herodas survived only in tiny fragments of rolled papyrus, which someone had tried to stick back together in the wrong order. Knox’s day job, as a fellow of King’s College, Cambridge, was to reassemble, edit and translate Herodas for the modern world. ‘The language of the Mimes is precious, with unpleasant affected archaisms, and an honest translation, it seemed to Dilly, must be the same … “La no reke hath she of what I say, but standeth goggling at me more agape than a crab.”’6 Heaven knows what the goggling modern readers were supposed to make of that.
The Mimes appeared in this crabwise form of English in 1929. By 1936, Knox was declining invitations to the King’s College Founder’s Feast, an annual gathering of intellect which he greatly enjoyed. ‘The reason was simple; the dinner was noted, even among Cambridge colleges, for its hospitality and its fine wines, and, in consequence, for the occasional indiscretions of the guests. These, to be sure, were heard by Kingsmen only, but the time had come when Dilly could not risk even the hint of a shadow of a reference to what he was doing.’7 For what he was doing was an attack on the Spanish Enigma.
The challenge for Knox in 1936 was to know how the Enigma machine had been set up – in other words, which of the three coding rotors occupied the three places in the machine and what the starting positions of the rotors might be. Soon he fathomed out the principle which had guided Marian Rejewski several years before: as long as there was no turnover of the middle rotor, the transformation effected by the wiring of the two rotors on the left and the reflector device remained constant. That made it easier to figure out which rotor was being deployed in the right-most position, along with its starting position. Using strips of paper he called ‘rods’, Knox began to tease out the Enigma traffic.8 If they were ever to meet, it was just possible that Knox and Rejewski might speak the same language, a language of cryptology if not of Polish, English, or ancient Greek.
• • •
Antoni Palluth was leading a complicated life.9 He had far too many jobs. In the daytime, Palluth acted out the role of partner in a small engineering firm. It was his night jobs for the army that made for the trouble. The code-breaking problems set by Maksymilian Ciężki had to be done after hours in the family’s apartment. And a family there was: he had got married to Jadwiga von Kessel, who came from an aristocratic family, and in 1931 and 1934 there were children, Jerzy and Andrzej. But there was a third job as well and it was dangerous. The third job obliged Palluth to make ‘business trips’ overseas to fix equipment: if AVA was to maintain its credibility with the military, its after-sales service had to be spot on. It wasn’t exactly secure for a Polish spy to walk into a radio store in a German high-street and ask them to repair a secret radio not much larger than a prayer book, so AVA’s rep, Antoni Palluth, had to go. Jadwiga didn’t like it when her husband had to reach into the cupboard for his handgun and include that in his luggage on these trips. The handgun proved that the third job was too dicey for comfort.
Building and servicing micro-sized transmitters and Enigma analogues was only part of the AVA brief. There were also commissions to build normal radios and aerials and all the other peripherals that go with wireless telegraphy. One of the larger projects was to construct a giant radio mast at a Polish intelligence service station in the Kabaty woods in the outskirts of Warsaw. Among the young trees, not far from the mast, the Second Department of the Polish General Staff was building a bunker. Inside the bunker, below ground, the code-breakers were going to have new offices. The place was called Pyry and the cover-name for the operation going on there was Wicher, the Polish term for a tornado.
Maksymilian Ciężki moved his team into the bunker in 1937. A team, but strictly demarcated according to traditional lines. Langer and Ciężki, commissioned military officers, hailed from the upper echelons of Polish society, as did Antoni Palluth. The class differential, coupled with a slight disparity in age, kept the group from socialising as a unit. The younger non-establishment technicians, Rejewski, Różycki and Zygalski, rubbed along together in a friendly enough way. But the invisible line between officers and others meant there was no question it would be anything other than ‘yes, Mr Major’ when the boss, newly promoted, came to call.10
The radio mast above the bunker at Pyry was no ordinary aerial. Opened in 1938, it was the visible part of Radio Station G, the hub of a network of eight listening stations across Poland. Radio Station G could itself pick up signals from as far away as Tehran and even Manchuria, which the Polish high command wanted to do in order to keep an eye on the USSR. It used state-of-the-art technology that only the AVA factory could master. Radio Station G was also called ‘Jedynka’, meaning Number One, probably because it cost a fortune: $1 million in 1938 currency values. It was undeniably the best in Europe, but it was also undeniably a peculiar way to spend the shrinking Polish defence budget when Hitler was sabre-rattling next door. It also made the code-breakers’ offices something of a target. In short, it was a vanity project. Some people thought there should be an inquiry, as the radio men would, sooner or later, find out.11
• • •
In October 1937, Colonel Stewart Menzies, not yet the head of MI6 but tipped as a possible successor to the incumbent, met Lieutenant Colonel Louis Rivet in Paris.12 Rivet had been in charge of the French intelligence and counter-intelligence services, part of the Deuxième Bureau, since June 1936. Menzies was mainly interested in Italy and Spain (as a result of its history, Britain thought of strategy in terms of oceans, and the Mediterranean was Menzies’ focus). The French wanted to talk about Germany. Menzies admitted frankly that the British knew next to nothing on that subject, but then he went on to predict that it would take only three weeks for the Germans to achieve Anschluss with Austria and that the Sudetenland in Czechoslovakia would suffer a similar fate. The British policy towards Germany was ‘wait and see’. Despite this official discouragement (suitably dressed up as Menzies’ personal views and not official at all) there were some practical results, the first drips of a thaw. MI6’s man on the spot, Commander Wilfred Dunderdale, would have more frequent meetings with Rivet and, from now on, to minimise the risk of compromise, Dunderdale would go under the cover-name ‘Dolinoff’.
It was about time. The drumbeat of rearmament was sounding more and more like a pas de charge; it would be imprudent to assume that Hitler’s aggressive posturing was only that. There was a real need by both parties to have some solid information to go on.
Bydgoszcz is a pleasant town in north-west Poland. Until 1918, it had been part of the Kaiser’s Empire and called Bromberg. Bydgoszcz was the home town to Marian Rejewski and Irena Lewandowska. After their lengthy engagement and struggle to maintain the relationship over long distances, the couple had married on 30 June 1934, and Bydgoszcz was never far from their thoughts. Although they lived in Warsaw, Irena wanted to have the family around when her first child, Andrzej, was born in 1936 and thus Bydgoszcz became Andrzej’s home town as well. It was the same when Janina, the Rejewskis’ daughter, came along a couple of years later.
Bydgoszcz may be a pleasant place for a family, but, primarily because it was right in the middle of the Polish Corridor, it was also home to an out-station of Polish military intelligence. The head of Polish Secret Intelligence Station III was called Major Jan Żychoń and in 1938 Żychoń could sense the Nazi threat becoming stronger. In the summer of that year, one of his agents reported that in case of an imminent war the German Command intended to make changes to the way they used the Enigma machine. The information was reliable: it had come from an agent who was a member of the signal section of the German Air Force.13
September 1938 was a time of European crisis. It was six months after the absorption of Austria into the new German Reich, which had (as foreseen by Menzies) been a business taking only a few weeks. On 12 September 1938, Adolf Hitler made a speech in Nuremberg that voiced specious complaints about Czechoslovakian atrocities, aggression, belligerence, and so forth. There were
riots in Czechoslovakia. Neville Chamberlain asked Hitler for a meeting: their first session at Berchtesgaden began on 15 September. The Czechoslovakian government found itself sidelined as the Germans, aided by the British and the French, carved up their country. Its borderlands were ceded to Germany, placing Czech defences behind German lines.
While Chamberlain was settling in at the Berghof on 15 September, the dramatic events that subsequently became known as the ‘Munich crisis’ became a crisis for the Enigma code-breakers of Pyry. The Germans changed the procedure for transmitting the settings of Enigma-enciphered telegrams. Up until then, all users of Enigma machines on a particular radio network had set their machines up the same way in the morning with what they called a ‘ground setting’ and then used the same ground setting all day to encipher the six-letter indicators which Rejewski had found so useful in his virtual dismantling of the machine. The use of a single, daily ground setting had given the Poles a substantial volume of material to work with each twenty-four hours, all of which was enciphered in the same way. From now on, however, there was a different ground setting for each message. The German authorities instructed their operators to arrange the three rotors afresh for every single message. There were 17,576 ways they could do that.
And the technical problems were not the real significance of this change in procedure. According to Major Żychón’s agent, this change signified the imminence of war.
X, Y & Z Page 8