The problem was serious and a solution was far out of reach. But Kowalewski was not far from the scene. Having returned from Japan, he was back in Poland in 1925, meeting with Ciężki and his boss Pokorny and, crucially, his philosophy of code-breaking prevailed. Perhaps bringing on some young mathematicians might produce new insights for a new type of cipher? In May 1926, Ciężki was given authority to ‘seek information and data concerning the possibility of employing professors of mathematics that are fluent in German to do some strictly confidential work.’4 It was time to get hold of those professors and get them working.
Poland in the 1920s already had an honourable place in the study of mathematics. The brightest stars in the mathematical firmament might, arguably, be Germans. But an important region that had once been under German control was now Polish and immediately after independence the Poles had established a technical mathematical university in Poznań, in the heart of Wielkopolska, where the older generation all had fluent German. There – the principal city of Maksymilian Ciężki’s youth – is where he would find his professors.
Professor Zdzisław Krygowski had come to Poznań in 1919 to join the mathematics faculty. His specialism was the elliptical function, but he was not too specialised to try something new, even if it was rather a way out of his main field; after all, Krygowski also wrote about Frederic Chopin and his family. More importantly, he was not too grand to teach his students. Cryptography could be regarded as a case of applied permutation theory, for shuffling letters around is exactly that. Ciężki had no difficulty recruiting Krygowski to participate in a long-term project to crack Enigma. His job would be to train the cryptographers, something which could be fitted comfortably into the existing programme of algebra lectures at twelve noon, with general seminars in the evening at six.
On 30 November 1927, a detailed, costed proposal for a cryptology course was drawn up. Just over a year later, on 15 January 1929, the course began. At this stage, Ciężki couldn’t be too explicit about what was going on, so the whole thing was dressed up to look and feel like extension work for keen students, no different from the routine evening seminars at 6 p.m. As well as Krygowski, the tutors were the Polish Army’s head of cryptology Franciszek Pokorny, Maksymilian Ciężki himself, plus Ciężki’s comrade from their teenage days, Antoni Palluth. The textbook was General Givierge; the subjects were substitution ciphers and double dice. So there was nothing secret in the course, nothing secret at all, except that it was, secretly, as much a selection process as it was tuition. Twenty-three mathematics students were accepted for the two-month course, which covered traditional subjects based on the textbook. Palluth and Ciężki each gave forty-two hours of tuition. Twenty students completed the course, of whom five achieved one of the top two grades. All the students, except four who failed, were put on the mobilisation list as cryptographers in case of hostilities.5
The course had been a success. The challenge for the radio intelligence section of military intelligence was to keep the budding code-breakers on board. Since these bright, technical students were mainly based in or around Poznań, far from radio intelligence, they were likely to go off and find interesting jobs and all the investment of those two months of intensive training could evaporate. The answer was to move radio intelligence to Poznań. And that made a good deal of sense, for it was in Poznań that interception of German signals was taking place. Ciężki was authorised to set up a branch office of the Biuro Szyfrów in the city and to hire the best students to work for him.6
• • •
Winters in Warsaw have a tendency to be cold. Not that they were particularly cold in the 1920s, but January 1929 was wintry enough for people to want to stay indoors. It was Saturday and the job of processing international parcels – which Shakespeare called fardels – was going on, as usual, at the central railway station. The customs label said that this particular parcel contained radio equipment, which seemed credible enough given the size and weight of the box. Then the German Embassy phoned up. There had been a mix-up. Terribly sorry, and all that, but this fardel, this box, it really wasn’t supposed to be in the post at all. Could it please be found and sent back to Germany? It was actually quite important.
All that to-do seemed far-fetched for an everyday package and the customs officer decided – to draw from a different play by Shakespeare – that the Germans were protesting too much. Perhaps someone ought to be told. Meanwhile, the Germans were fed some flannel: as you know, it’s the weekend, not to mention the terrible weather, nothing is really possible before Monday, but it will of course be number one priority, as and when, et cetera, et cetera. The customs officer’s next call was to Polish military intelligence. And, as the box had radio equipment in it, they called in the radio men.
AVA was the obvious place to send the peculiar box for analysis. Ludomir Danilewicz took the call and lugged the parcel round to the AVA workshop. It was fortunate that Antoni Palluth was in the office and knew all about ciphers, because what was in the box was clearly no sort of radio. First of all, it had a keyboard like a typewriter, but instead of a normal roller and type, the machine had a set of light bulbs which lit up letters on a translucent panel. If you pressed down on a key, one of the lights would come on, but for a different letter than the one you pressed. If you pressed the same key again, a different letter again would light up. It was a ciphering machine. It was the Enigma.
Palluth and Danilewicz spent the whole weekend prising apart the different components of the Enigma machine to see how it worked. Then they carefully put it all back together and put it back in its box. Then the box was parcelled up again in its brown paper. The Germans were told their package had been found and all was well.
Though no doubt Danilewicz had found it interesting to dismantle a piece of precision cryptographic technology, all that kerfuffle and dissembling over the weekend had, unfortunately, taught the Poles absolutely nothing. The thing was an example of the commercial variety of Enigma machine which had been available on the market for years. But what they did learn from this episode was that the Germans had become very sensitive – very sensitive indeed – about the Enigma. And that meant that it was right to give priority to the effort to prise open the secret of the military Enigma machine.
• • •
In 1929, the reorganisation of the Polish Army, initiated by Józef Piłsudski in 1926, reached the Intelligence Corps. Piłsudski’s reform movement was called Sanacja, which meant something like ‘Clean-up’. His primary concern was the creation of a lean and efficient army. The marshal was sweeping cobwebs out of the General Staff. Empty posts were left unfilled. Cosy staff jobs were axed. At first, radio intelligence seemed to have escaped the marshal’s gaze, perhaps because he remembered its usefulness in the Russian war. But perhaps not. Ciężki’s name came up for a stint in regimental duties, just at the time his boss Franciszek Pokorny was being reassigned to other duties himself.7 The broom had reached their corner now. Piłsudski himself, however, became interested in the question of appointing Pokorny’s successor, wondering whether Kowalewski would be willing to come back in. But Pokorny recommended a different man, someone from the old pre-war Austrian Army whom he’d known when they trained together in military school.8
Gwido Langer was a Pole who had grown up in the Austrian partition and served in the Austrian Army. He was born in 1894 and, at the age of 17, enrolled in the Austrian military academy at Wiener Neustadt. He graduated on 1 August 1914 with the rank of lieutenant, four days after Austria had declared war on Serbia and on the day that Germany and Russia declared war. Gwido Langer picked up two wounds on the Eastern Front, a promotion and a medal, but his involvement in the Great War came to an end with capture in 1916. At the end of the war, he joined the Polish Army in Siberia. During the Bolshevik war he was captured again, but managed to escape, travelling over 1,000km to rejoin the Polish forces. He was a career army officer, he was a disciple of General Max Ronge, the wartime head of Austrian military intelligence, and i
n 1929 his new job was to squeeze the best out of the signals intelligence section.9
The Austrian Army man had, in a way, the perfect background to lead the signals intelligence section of military intelligence. Austria’s pre-war relationship with Serbia had created a thirst for intelligence and in the Balkan Wars the Austrian military command had set up a code-breaking unit. As tension grew, the Austrian capability grew as well. Both Russian diplomatic traffic and intercepted Serbian material were being read. The act of state-sponsored terrorism that took place outside Moritz Schiller’s café in Sarajevo on 28 June 1914 was just the most provocative example of Serbia’s devious behaviour and the Austrian code-breakers knew exactly who was behind it. That intelligence may have given Austria the confidence of the righteous in how they reacted.
Although it had all gone horribly wrong in the end, at the beginning the war went Austria’s way. The Eastern Front was long. Very long: the Austrian sector alone stretched 600km from south of Warsaw to the Romanian border. So the Russians had resorted to radio to organise themselves. After the defeat of Tsar Nicholas II by the Germans at the Battle of Tannenberg, partly attributed to Russian plain-language radio communication, the Russians quickly recognised their mistake and encrypted their signals. A steamroller of sixty Russian infantry divisions headed into Poland, straight for the border with the Austrian Empire. To avoid being completely overwhelmed, Austria had to fight an intelligent war. ‘What joy, as radiogram after radiogram came to us in plain speech! What even greater joy, as several phrases were given in cipher. My trained cryptanalysts fell on to these puzzles with enthusiasm,’10 crowed General Ronge. Forewarned of Russian movements, the Austrians could target their response precisely to points of Russian weakness and achieve disproportionate success. With the Austrian counter-attack in the harsh mountain winter, the steam went out of the Russians and the roller ground to a halt. On 6 December 1914, the Russians called a retreat. The achievement was down to the Austrian Radio Interception Service. Ronge’s trained cryptanalysts, like many of his Intelligence Corps and many of the fighting men, had come from the Austrian-controlled partition of Poland.
The situation inherited by Gwido Langer in Polish signals intelligence in 1929 was not perfect. The generals were preparing for the previous war: the focus of interception was Russia, with the radio masts and direction-finding equipment primarily directed eastwards. Moreover, the system was static and inflexible and would not be able to cope with a war of rapid movement. It would take Gwido Langer a year or two to get his programme for reform into motion, but the first step was obvious. Radio intelligence was, apparently, understood by the marshal and it was made into an autonomous division in 1930. The next step would be to centralise the code-breaking and reintegrate it into intelligence. One important existing positive in the situation, however, was that Ciężki was confirmed in his position as head of the German ciphers section.
• • •
1930 was an interesting year for military intelligence and not just in Poland. In France, another country with an honourable tradition of radio-based code-cracking from the Great War, it had dawned on the intelligence service that an integrated approach to signals intelligence would be wise. Sure, there was a ciphers section – the Section du Chiffre – but the officer responsible for the section’s successes during the war was now on the reserve list and writing textbooks. And what remained of the section was no match for the mechanical ciphers which Germany had started to use in the mid 1920s. A new division was created, Section D of the Service de Renseignements, the intelligence division of the Deuxième Bureau of the French Army General Staff, which, when understood practically, meant that on 1 November 1930 the French Army had a dedicated team of one active and two reserve officers focused on interception and decryption, with specific responsibilities for obtaining intelligence on foreign cipher and exploiting enciphered intelligence.11
The sole officer on the active list was one who may, to borrow from the title of the book he later wrote, be described as the greatest enigma of the war. Gustave Joseph François Marie Alfred Bertrand weaves in and out of the Enigma story in a most unpredictable way and his uncanny skill for survival attracts suspicion as much as admiration. To begin with, Bertrand’s military career was about as unorthodox as Gwido Langer’s had been old school. When the Great War broke out, Bertrand enlisted, aged 18, as a private soldier and was sent off to fight in the hopeless Gallipoli campaign. Bertrand was wounded, but along the way he discovered an interest in radio intelligence, signals interception, direction-finding and the unravelling of enemy secrets through the mysteries of code-breaking. Bertrand was spotted by Colonel Bassières, number two to the former French cipher chief (now turned text-book author) Marcel Givierge, and a post-war career in the army beckoned. By the end of the war, Bertrand had joined the victorious French Army on a permanent basis, with an officer’s commission, and by 1926 he had succeeded in his wish to be transferred to the Section du Chiffre. In 1930, Bertrand became the sole active officer of the new, dynamic Section D. The philosophy of this new section was built around the old-style techniques of spying: good agents, clandestine meetings and the timeless lubricant of success deals, a plentiful flow of cash. Bertrand’s section was a business. It was about buying and selling, with a brisk trade in the code books of foreign powers.
Good, old-fashioned intelligence work meant contacts and relationships with like-minded intelligence services. Over the course of the next few years, Bertrand built up an information exchange network, liaising with the General Staffs in Czechoslovakia, Japan, Lithuania, Poland and Britain. Gustave Bertrand could show the mettle of his Section D by scoring a few victories against its greatest enemy. The enemy was not Germany, nor even England. The enemy was the Section du Chiffre. The Section du Chiffre had a large staff: twelve officers, one under-officer and seven administrative staff. It was headed by Lieutenant-Colonel Jean Marie Antoine Edouard de France de Tersant, who signed his reports in magnificent style with the single word ‘France’.12 Unfortunately for France the man, and worse still for France the state, the magnificent reports contained absolutely nothing of interest. For the Section du Chiffre was achieving nothing. ‘The output of the Section du Chiffre was of little importance, as – be it apathy or be it incapacity – they only attacked what was easy to decrypt: everything else was put away or thrown away.’13 Perhaps Bertrand’s judgement was rather unfair, because the Section du Chiffre was also responsible for the French Army’s own codes and ciphers and its signals security. But its other duties left the field of cryptanalysis clear for Section D and Gustave Bertrand.
As winter began to bite in late 1932, Bertrand’s network was still building and two liaisons were particularly promising. First was one with the British. Earlier in the year, a Captain Tiltman had come to Paris, bearing gifts.14 John Tiltman was foremost among the cryptanalysts of World War Two, breaking codes and unravelling secrets of a bewildering variety and with astonishing intellectual agility. In 1932, though, the British were obsessing about the Bolshevik threat and Tiltman’s mission was to see whether the French could help fill gaps in the British appreciation of the Soviet Navy. The French couldn’t help, but Britain’s willingness to share suggested that a partnership on codes and ciphers might be profitable.
The other liaison that might be worth the trouble was Poland. Since 1921, a formal Franco-Polish Military Convention had been in place. This committed both General Staffs to keeping communications open, and the Poles to keeping the French informed on all current matters. An intelligence pact was a neat fit for this structure and Bertrand got the go-ahead from General Maurice Gamelin, the French Chief of Staff, to make arrangements with his Polish counterparts. So, in March 1931 Gustave Bertrand got on the train and went to Warsaw. His opposite number there was Lieutenant-Colonel Gwido Langer and the head of intelligence was Colonel Stefan Mayer. It must have been daunting for the Frenchman, for although comparable in age to Bertrand, Langer was far senior to a mere captain, and Mayer even more
so. Nonetheless, Langer spoke competent French and the two were soon at ease. Bertrand saw an Enigma machine – the commercial model the Poles had bought – for the first time and was told of the major problem with breaking the encryption it created. The German military had a modified version of the Enigma machine and until these modifications were understood, there could be no progress.
Amid all the bonhomie of the ad hoc liaison, Bertrand proposed a more structured arrangement: the French would supply all German military documents which allowed for exploitation of transcribed intercepts. In return, the Poles would give all similar documents on the USSR and all results of decryption of enciphered German Army traffic. Bertrand thought they had a deal, albeit a somewhat theoretical one, as there was no decryption of enciphered German Army traffic, at least not Enigma traffic, which is what they both cared about. From Langer’s viewpoint, the French had made only a proposal, whereas for Bertrand, both of them had made a promise and in honour it was rightly one they both should keep, or be very explicit about their inability to do so.
Despite the ambiguity the liaison was fruitful. During 1931–32, Bertrand (cover-name Bruno or Bolek) and Langer (Luc) had four meetings in Warsaw and exchanged knowhow on current problems.15 The biggest current problem for both of them was mechanical encipherment on the German military Enigma machine and neither of them had an answer.
Except that Gustave Bertrand still had a secret trick to play in the shadowy market for codes. He was going to deal a new card from the pack. The card was well used and slightly bent and one which might not withstand fastidious scrutiny, but it was at the top of the deck. Bertrand was going to play his king.
X, Y & Z Page 4
