The Secret War

by Max Hastings

  Chi used the same staff selection methods as the British and Americans – testing recruits with crossword puzzles and mathematical problems. Successful candidates were assigned to a four-week induction course, learning the substitution and transposition methods, and suchlike. They were also urged to read a French history of cryptanalysis, and Herbert Yardley’s account of the old ‘Black Chamber’. In Chi’s 1941 heyday it employed around 3,000 people, but thereafter this strength fell, as rear-area organisations were combed for personnel who might become Frontsoldaten, or shifted to OKH/GdNA. Whereas Bletchley’s success secured ever more resources, Chi’s lack of it earned ever more scepticism. Moreover, from July 1944 it became a focus of hostile scrutiny, following the execution of its two most senior officers, signals corps generals, as plotters against Hitler. Fenner, Chi’s veteran cryptanalyst, found himself under investigation, accused by an SD officer of plotting to sabotage the Wehrmacht’s communications by recommending an inferior cipher system. Fenner said later, ‘the whole cryptographic service came under political suspicion’. Though he was eventually cleared, in Nazi eyes the department’s reputation was irretrievably tarnished.

  Britain’s Turing and Welchman, America’s Friedman and Rowlett, had no equals in the enemy’s camp. Recruitment was not assisted by the fact that many of Germany’s most brilliant brains were exiled, imprisoned or slaughtered, because they were Jews. It bears notice that three of the four members of Friedman’s original US Army codebreaking team were Jewish, together with some of the highest intellects at Bletchley. The names of Hitler’s codebreakers are almost unknown to posterity, even in their own country, while their British and American counterparts have belatedly become famous. Nonetheless, some of Chi’s men were talented by any standard save that of its enemies. Among the foremost was Dr Erich Hüttenhain, born in 1905 in Westphalia, a brilliant student who professed a passion for Mayan chronology. He left university in 1932 laden with laurels in astronomy, mathematics and physics. Thereafter he served for five years as an astronomical research assistant in Munster before being recruited as a cryptanalyst. He was quickly appointed to head a division, and set about recruiting the best brains he could identify, and ensuring that they were kept away from the front when war came. His assistant, Walther Fricke, produced a doctorate on the dynamics of the stellar system, and in 1939 had been due to take up a research post at Edinburgh University. Fricke joined Chi in 1941 knowing nothing of cryptanalysis, but proved an adept. Lt. Schubert became an expert on Russian traffic, assisted by a codebook captured by the Finns during their 1939–40 war with the Soviet Union; he was later transferred to OKH/GdNA, where he read some Red Army four-figure code material. Fenner also thought highly of Bernert, a Viennese who worked on British messages, and Döring, who was ‘always called in for difficult jobs’. There was an engineer who built ‘phase decoders’ – Willi Jensen, the nearest the organisation had to Tommy Flowers at Dollis Hill.

  Other notable personalities included Wolfgang Franz, Ernst Witt, Karl Stein and Gisbert Hasenjaeger, the last the youngest in the team, aged twenty-four. Hüttenhain recruited five professors of mathematics, including Georg Aumann, Werner Weber and Johann Schultze – together with a physician of mathematical leanings. While the British allowed most of Bletchley’s codebreakers to retain their civilian status, the Germans inducted theirs into the Wehrmacht as Gefreiters – privates first class – a loss of status which these relatively distinguished academics resented. Erich Hüttenhain had his first significant success cracking French field codes in 1938; the Germans continued to read these through 1940 and indeed thereafter, even though monthly changes were made. During the Blitzkrieg in France, army sigint decrypted relatively few important British signals, but was nonetheless able to provide German commanders with a full picture of their order of battle; to track their advance into Belgium; to read the order to their Calais garrison to hold out to the last.

  In a prominent position in Chi’s Tirpitzüfer headquarters stood a British Type-X cipher machine captured at Dunkirk, albeit without its rotors. Hüttenhain told interrogators in 1945 that ‘since Enigma was similar to Type-X, and we believe that the Enigma cannot be broken, no great effort was made to break Type-X … Enigma might be broken if a vast Hollerith complex [tabulator] was used, but this is scarcely feasible.’ The Germans created a range of intermediate technology to assist their labours, including a Roellchengerät, a 10 x 10 cylinder device built by the Foreign Ministry’s cipher department, which performed a primitive calculating function. Chi also had an electric typewriter that could handle simple letter substitution. Fenner complained after the war: ‘Mechanical scanning of perforated tapes [by Bigram devices] was always much too slow. The future belongs to photo-electric scanning.’ The British, however, had achieved miracles with intermediate technology, and in comparison with the bombes and Colossus working their marvels at Bletchley, the machinery employed by Chi and OKH/GdNA resembled a 1914 biplane alongside a 1945 jet fighter.

  The Germans relied overwhelmingly on human ingenuity to access lower Allied systems. Groups of men were assigned to work together on a given signal, cross-fertilising ideas just as at Bletchley and Arlington Hall. Fenner testified that during the early war years, when the Germans were reading America’s strip-cipher military attaché code, 80 per cent of the breaks were achieved through user errors, especially when identical messages were repeated in different codes – the familiar vulnerability that assisted the Allies. Fenner said that Soviet ciphers were secure if properly employed, ‘but if the cryptanalysts in Moscow could only see how they were used, they would be very unhappy’. This smug little observation should be judged, however, in the context of the fact that Berlin seems to have learned little of any significance to its war effort from whatever breaks it achieved.

  In 1941 the Germans transferred from France to North Africa wireless eavesdroppers with experience of monitoring British traffic, and reaped handsome dividends. On 26 June, a long report from Bletchley drew War Office attention to British operational signals decrypted by the Germans during the Cretan débâcle, some of them detailing aircraft and warship movements. The Afrika Korps considered Eighth Army’s wireless discipline very slack, and attributed many of Rommel’s 1941–42 successes to his foreknowledge of British deployments. Some of the interceptors were highly qualified linguists. One, an NCO named Schwartze, was the twenty-six-year-old son of an English mother and a half-Jewish German father. He had been educated at Cheltenham College and Merton College, Oxford, where he read law, though debarred later from practising in Germany because of his ancestry. He worked in North Africa alongside a friend named Graupe, a Berliner a few years older, who had studied in Louisiana and worked in a US factory until the authorities declined to renew his visa.

  After the war, Lt. Gen. Albert Praun described how spasms of Anglo-American carelessness enabled his officers to piece together enemy orders of battle, just as the Allies did. The sigint out-station in Athens, for instance, read a message from a British paymaster in Palestine, instructing a division being transferred to Egypt to leave behind its filing cabinets – which enabled a big red pin to be shifted on the map of British deployments in the Middle East. Later, the Germans discovered that the US 82nd Airborne Division had moved from the Mediterranean to Britain because they broke an administrative message about one of the formation’s paratroopers who was facing a paternity suit. They received warning of an impending attack in Italy by decrypting a signal about an issue of rum to the assault units. Italian codebreaking of British traffic played an important role in empowering Axis intelligence.

  Hans-Otto Behrendt, one of Rommel’s officers, wrote gleefully that in 1941–42 his chief ‘often had a clearer picture of what the British C-in-C planned than some of the British subordinate formation commanders’. The German general called 621 Radio Interception Company his ‘circus’. Until El Alamein, he boasted that no major British formation entered a battle without having been previously identified by his sigint team. The over
running of 621 Company by New Zealand troops on 10 July 1942, with the loss of almost all its personnel by death or capture, was deemed a major disaster for the Afrika Korps. When Bletchley decrypted traffic that revealed the German break into the US military attaché traffic, it achieved an important success by persuading Washington to change ciphers. Thus, on 29 June 1942 Rommel lost access to the traffic, which he himself had described affectionately as his ‘little Fellers’ – the name of the US attaché in Egypt. A German staff officer described this development as ‘a catastrophe’. The trauma felt by the codebreakers of Chi and army radio intelligence in Berlin was as great as that imposed on Bletchley by Dönitz’s introduction of the fourth rotor to U-boat Enigma. GC&CS had warned the Americans about the lethal threat posed by the Cairo traffic early in May, but bureaucratic bungling caused more Fellers signals to be sent, including highly sensitive material about defence of the Nile Delta, for almost another two months before the breach was finally closed. Colonel McCormack wrote to Washington from Bletchley, where he was visiting: ‘The mishaps … have produced in a high quarter here (and you must admit with some justification) a somewhat unfortunate impression of our own security procedures.’

  Until the summer of 1942, the Germans and the Allies were in about the same place in the struggle for intelligence in the Mediterranean: Rommel and his British counterparts knew approximately the same amount about each other. Moreover, in northern France the disaster that befell the August Dieppe raid was partly attributable to the fact that the defenders were on high alert, thanks to plentiful German sigint and humint about British preparations. Thereafter, however, the Wehrmacht lagged increasingly far behind. It became the turn of the Germans to suffer from the deficiency that had dogged British operations for the past three years: lack of hard power to exploit information. By 1943, one of Rommel’s intelligence officers noted that while they were still acquiring some good material from the Y Service and PoW interrogation, they could do little useful with it: ‘tactical intelligence was not of much use. We were just too weak.’ The Germans never secured another Allied military source as good as Bonner Fellers. Operation ‘Torch’, the November 1942 North African invasion, came as a complete surprise to Berlin because Allied wireless discipline was strict, special new naval cipher tables had been introduced, and the Abwehr supposed the ‘Torch’ convoys, though reported by its watchers in Spain, to be destined for Malta, or else for landings further east. In 1943 they were sometimes alerted to impending Mediterranean landings by references in British or American field-ciphered signals or voice traffic to colours – beaches designated ‘Green’, ‘Blue’, ‘White’ and suchlike. In the first hours of Operation ‘Husky’ in July, however, they were slow to respond partly because a fortnight earlier radio intelligence had rung false alarm bells about just such an Allied invasion of Sicily. Investigation showed that German eavesdroppers had picked up transmissions from a landing exercise in North Africa, which happened to take place on the same compass bearing as Sicily from one of their direction-finders, but a hundred miles further south.

  As for the Eastern Front’s HLS/Ost, in February 1943 the intelligence officer of XXXth Panzer Corps paid lavish tribute to the ‘outstanding’ sigint service being maintained on his formation’s front. But about that time the Russians captured a German interception company at Stalingrad, and belatedly awoke to the sophistication of its activities: thereafter, Soviet officers often broke off a voice conversation if their other parties violated security. After Stalingrad, said Albert Praun, the Red Army maintained the best radio discipline of any of Hitler’s adversaries, ‘and posed a greater problem to German direction-finders than did their Western Allies’, who became ever more careless. In the year beginning 1 May 1943, Army Group North, for instance, intercepted 46,342 Russian signals, but read only 13,312 – all at low level and many out of real time. Gen. Kurt von Tippelskirch said after the war: ‘As time went on [the eavesdroppers] had more and more difficulty coping with Russian deception measures, consisting of constant simulated troop movements.’ The Wehrmacht’s spirits were not improved by breaking an enemy signal in which a Red Army intelligence officer urged units that killed German prisoners to spare at least one for questioning. This appears to have prompted another Soviet message: ‘20 Fritzes captured, one sent back for interrogation, remainder shot.’ Praun wrote disdainfully after the war that this gave ‘a truly shocking picture of the Asiatic combat methods used by the Russians’ – he had evidently forgotten that his own army killed or allowed to starve to death two million Soviet prisoners in 1941–42, and plenty more thereafter.

  In the early summer of 1944, before the Red Army launched ‘Bagration’, greatest offensive of the war, Pavel Sudoplatov was summoned to the Kremlin with his boss Merkulov, Abakumov of SMERSh and the GRU’s chief to discuss a new twist to the long-running deception operation ‘Monastery’. Reinhard Gehlen and the Abwehr were still enthusing about the steady flow of information they received from Agent ‘Max’. The Germans’ Eastern Front intelligence chief was rash enough to tell his own high command that it might expect ‘a calm summer’. The Stavka in Moscow thus decided that the time was ripe to use Alexander Demyanov and his network to build a new edifice of disinformation to support ‘Bagration’.

  The intelligence chiefs entered Stalin’s suite in cocky mood, inspired by the tide of success attending Russian arms – Sudoplatov had recently received the Order of Suvarov for his own role. Stalin, however, with characteristic perversity, received the visitors coldly. Traditional deception ideas were played out, he said; he wanted to try something new, of immediate assistance to the Red Army. Sudoplatov, at a loss, kept cautiously silent. Abakumov simply urged placing ‘Monastery’ under his own control. Stalin then called in the deputy chief of the general staff, Gen. Sergey Shtemenko, who read an order already drafted. The ‘Monastery’ team was to pass intelligence to OKH to persuade the Germans that one of their brigades in Belorussia was cut off, but still fighting. The objective was to goad the Germans into launching an operation to break through to relieve it.

  Sudoplatov was excited by the boldness and originality of the plan. In July 1944 his deputy Leonid Eitingon, along with ‘Fisher’ the radio specialist and an NKVD team, were dispatched to Belorussia to implement it. Alexander Demyanov – ‘Max’ – informed the Germans that he had been transferred to a new assignment, in the communications department of the Red Army on the Belorussian front. On 19 August Gehlen informed his commanders, on the authority of ‘Max’, that a Wehrmacht brigade of 2,500 men commanded by Lt. Col. Heinrich Scherhorn, with some guns and a few tanks, was struggling desperately in an encirclement near the Berezina river. The size of the bait was finely crafted by the Russians: small enough to be credible, large enough to be worth an effort to save. In reality, of course, Scherhorn and 1,500 survivors of his shattered command had been disarmed and were in the hands of the Red Army; their wireless-operators were now transmitting with Eitingon and his comrades holding pistols to their heads, figuratively and perhaps literally.

  Amazingly, this deception was successfully sustained from 19 August 1944 to 5 May 1945. Although the Wehrmacht’s circumstances were too desperate to launch a ground-force operation to relieve Scherhorn, during those months a procession of transport planes parachuted to the colonel supplies, ammunition, radio equipment, cash and Polish guides tasked to lead the brigade across country towards the German lines. All this material was recovered, including thirteen wirelesses and ten million roubles, together with twenty-five Abwehr personnel. Some of the German aircraft carrying out the drops were allowed to fly safely home, to sustain the deception. To the special delight of the NKVD, on 28 March 1945 Scherhorn received a personal signal from Gen. Heinz Guderian, announcing his promotion to full colonel and the award of the Knight’s Cross. Although the Scherhorn deception was superbly ingenious, from the Kremlin’s viewpoint the results were modest, not remotely matching the contribution of ‘Monastery’ to the Stalingrad envelopment.

 
; Amid the Wehrmacht’s dire shortage of sources, its intelligence officers devoted increasing effort to determining where the Red Army was massing artillery, but this became problematic as ever more guns proved to be dummies. Gen. von Tippelskirsch said, ‘PoW interrogation was the most profitable source [of information], with the Y service second.’ The Germans had despaired of cracking high-level Allied codes, and now focused exclusively on humbler traffic. Moreover, it was always necessary for knowledge to be matched by capability. In the latter war years on the Eastern Front, Albert Praun noted gloomily that it became fruitless to locate Soviet concentrations, because the Wehrmacht and Luftwaffe could act on such intelligence only ‘inasmuch as the Germans had the means for appropriate counter-measures in this theatre’. Which was not often.

  In 1944–45, after the Germans lost the capability to carry out significant air reconnaissance, the Nachrichtenarmaufklärung voice-interception service became ever more important. Its stations abandoned interception of low-priority traffic in such places as Ireland and Spain, to concentrate on the Allied armies. They tracked many US formations through the movements of their APOs – Army Post Office numbers. The German interception station at Bergen in Norway had 150 receivers, and monitored a wealth of traffic as far afield as the continental United States. Albert Praun praised British radio discipline, but expressed bafflement that its units customarily transmitted callsigns and signatures in plain language. Voice chatter during Allied exercises proved a fertile source of intelligence.

  Erich Hüttenhain of Chi regarded cipher-making and cipher-breaking as processes that should fertilise each other, and was exasperated that the Wehrmacht ignored his warnings about the vulnerability of some of its codes, and of its teleprinter links – though not, until a late stage, of the Enigma machine. Even after a special Chi cipher security department headed by Karl Stein was created in 1942, OKW’s hubris persisted. Wilhelm Fenner said sourly after the war: ‘The high command’s view was that “Germany had won all her battles so far by using the system [it had got], and there was no need to overload the troops with new methods.”’ In August 1944, when Stein’s team belatedly urged abandonment of Enigma, the army’s resistance persisted. Fenner again: ‘There was a storm of protest whenever the army was asked to change a system.’ Compare and contrast the attitude of the US Army, which lost a cipher machine in France, and accepted the immense logistical burden of rewiring every other Sigaba in the theatre, just in case the Germans had captured the lost example – which it was eventually found that they had not. In November, Hüttenhain lectured to a military signals symposium, highlighting the vulnerabilities of Wehrmacht communications, but nothing was done to assuage his fears. Chi’s cryptanalysts had no hard evidence of British success in breaking Enigma, and in its absence were forced to make their case simply by demonstrating its theoretical vulnerability. This was not enough for Germany’s generals. Meanwhile, manufacturers dallied for years with the development of a new and more advanced Enigma rotor, the Luckenfüllerwalze, but this became fit for introduction to service only on 1 May 1945.