The Bletchley Park Codebreakers (32 page)

Colonel McCormack, who was at Bletchley to help assess the situation for Arlington Hall, cabled back on 13 May, also urging compromise. He noted that it was indeed ridiculous to think of routing to Washington all raw traffic and attempting to duplicate all of the ancillary reference and index material, built up at GC&CS over three years, that was needed to tackle the German traffic: ‘if [Colonel Preston] Corderman [the head of Arlington Hall] wants his people to learn what makes this operation tick,’ McCormack wrote, ‘he had better send them here to learn it, because they never on God’s green earth will learn it from anything that Arlington will be able to do in any foreseeable future’. On the British side, Gordon Welchman urged moderation, too; while ‘the idea of a separate “E” organization being built up from scratch elsewhere seems to us to be absurd … on the other hand it does seem logical that the Americans should take some hand in work on “E” and we certainly need help’.

On 17 May the ‘BRUSA Agreement’ was completed, providing for full American participation – in Britain – in the interception and solution of Army and Air Force Enigma traffic, though there were still thorny details to be worked out, and only towards the end of 1943 were arrangements finally settled for Telford Taylor to select decrypts for transmission to Washington directly from Bletchley Park.

There were still occasional clouds over the relationship. The Americans remained suspicious that the British were going to exploit the new intimacy to break American codes or pry out uncomfortable secrets. Colonel McCormack noted at one point with some alarm that the Navy was apparently supplying the British Admiralty with the Navy Department’s summary of State Department cables:

The British were equally suspicious about American interest in code materials from countries such as Iraq that remained within the British sphere of influence. The BRUSA Agreement of May 1943 discreetly sidestepped the issue of diplomatic traffic; it called for full co-operation and complete exchange of cryptanalytic results and intercepts, but mentioned only the military and air forces of the Axis powers and the
Abwehr
, the German intelligence service. Informally, a good working relationship was forged between Arlington Hall and Denniston’s diplomatic section at Berkeley Street. But the British drew a line at providing Washington with copies of neutral countries’ diplomatic messages sent via cables that the British controlled. An American liaison officer from Arlington Hall reported in November 1944 that Denniston ‘frequently gets the impression that we are utilizing the war to exploit British cryptographic knowledge’ in fields unrelated to actually winning the war. The United States, for its part, ceased sending Berkeley Street information about Latin American countries’ codes in September 1944. And when Arlington Hall began work on Russian diplomatic codes in 1943, it went to great lengths to conceal the fact from the British, ensconcing the Russian section behind a plywood partition and keeping it off limits from the British liaison officer, Geoffrey Stevens. Only at the end of the war did America and Britain let each other know that each had in fact been working on Russian traffic – projects codenamed Venona in the United States and Iscot in Britain.

In the end, however, what made the relationship work and endure was that both sides truly did gain; indeed when it came to breaking naval Enigma traffic during the last two years of the war, neither side could have done it alone. The contribution that the US Army contingent in Britain made under the BRUSA Agreement was not insignificant – at Bletchley Park and its outstations the US Army supplied manpower to operate bombes and to perform all of the other tasks associated with reading Enigma traffic. But a much more vital contribution was made by the ability to harness US industry to the urgent task of building the desperately needed four-wheel bombes. That would never have
happened without the British conceding some operational control to the US Navy, the move they had so long resisted. The US Navy’s 100 four-wheel bombes operating in Washington largely took over the entire job of decrypting the U-boat traffic, codenamed ‘Shark’, by autumn of 1943. The British four-wheel bombes, by GC&CS’s own admission, had never functioned as intended, and on 24 March 1944 GC&CS cabled to Washington conceding the four-wheel bombe field to the Americans. In fact, the US Navy bombes worked so well that by spring 1944 they had a huge, excess capacity to spare. Not only were they handling virtually all of the U-boat traffic; about 45 per cent of the US Navy bombe time was being devoted to German Army and Air Force problems being cabled from GC&CS. (By contrast, the US Army relay bombe never did much, and Arlington Hall’s Enigma section mainly focused on experimental and theoretical studies. The relay bombe was eventually adapted for ‘dudbusting’ – recovering the initial rotor setting on Enigma messages whose indicators had been garbled – and Bletchley Park did cable Arlington Hall some problems to be solved on the machine, and it did the job. But it was considered slow and inefficient, and other machines and methods did these jobs just as well, and often better.)

Co-operation between the Allies on German diplomatic codes led to important cryptanalytic results, too, particularly as a result of the intensive and successful research effort at Arlington Hall to harness IBM machines to the task. The sharing of recoveries on Japanese naval codes produced more progress than either side could have achieved alone, especially during the early years of the Second World War when both were strapped for manpower. And just as the Allies’ agreement to centralize German Army and Air Force Enigma work at Bletchley Park was able to draw on the considerable accumulated British expertise in translating this traffic and correctly interpreting it for intelligence value, so the concentration of work on Japanese Army codes at Arlington Hall in the last two years of the war was able to utilize the huge American investment in IBM machines and methods; by 1944 Arlington Hall was not infrequently reading a coded Japanese Army message before its intended recipient was, thanks to a nearly automatic system of decryption using IBM equipment and a number of special-purpose machines.

It took a blend of individual personality, high politics, and even a certain amount of skulduggery to break down the natural suspicions
that prevented co-operation between agencies that, by their very nature, are prone to be secretive to the point of paranoia. Neither side was quite willing to admit how much they would gain through co-operation. But once the door was broken down there could be no turning back, for both sides could manifestly see how much they would lose through going it alone.

14

MIHAILOVIĆ OR TITO? HOW THE CODEBREAKERS HELPED CHURCHILL CHOOSE

JOHN CRIPPS

Yugoslavia was invaded by the Axis on 6 April 1941. Its armed forces surrendered unconditionally eleven days later: it was then divided up between the occupying powers – Germany, Italy, Hungary and Bulgaria – with the largest part, although subject to military occupation by Germany and Italy, being declared the Independent State of Croatia. King Peter and the Yugoslav government fled, arriving in England in June to be feted as gallant heroes. This image was bolstered later in the year by the British government and the media when it became apparent that a major uprising had broken out in Yugoslavia. There were two major resistance movements fighting the Axis. One, the Chetniks, was led by Draza Mihailović, a regular Yugoslav army colonel who had decided to fight on; the other, the communist Partisans, was led by Josip Broz, who operated under the
nomme de guerre
of Tito.

The British Government, at the time short of good news, and the Yugoslav government-in-exile, gave Mihailović their enthusiastic backing and he was appointed Yugoslav Minister of War in January 1942. However, lack of resources and more important commitments elsewhere meant that support from Britain for the Chetniks was limited almost entirely to words rather than deeds. It was not considered appropriate to give succour to the Partisans who were described in a report by the Director of Military Intelligence, Major General Francis Davidson, to Winston Churchill in June 1942 as ‘extreme elements and brigands'. But, less than eighteen months later, the Prime Minister decided, despite the outright opposition of Britain's ally, the Yugoslav government, to withdraw all support from Mihailović and to supply the Partisans with materiel to enable them to prosecute their resistance. This assisted the Allies by tying down Axis forces which would otherwise be deployed in other theatres of the war, but had the inevitable consequence that Yugoslavia would become a communist state after the war. How did this volte-face in policy come about and what information provided the basis for this astonishing change of policy?

The British received intelligence about Yugoslavia during the war from a number of sources. Mihailović established radio contact with his government and this link provided a certain amount of information, albeit entirely from Mihailović's point of view. Reports were received from neutrals who had visited Yugoslavia and from Yugoslavs who had escaped the country. Neutral and Axis wireless station broadcasts, together with those from Radio Free Yugoslavia transmitting from the Soviet Union, were monitored and analysed, as were press reports. But as soon as the organization charged with fostering resistance movements, the Special Operations Executive (SOE), became aware of the revolt, it wanted to know more. It sent liaison officers, initially to Mihailović. But when the British became disenchanted with Mihailović, as they began to perceive his movement was less effective than the Partisans, the SOE sent missions to the Partisans. The first, led by Captain Bill Deakin, was sent to Tito's headquarters in May 1943. He was joined by Brigadier Fitzroy Maclean, a Conservative MP and former diplomat, the following September. Maclean subsequently reported to Anthony Eden, the British Foreign Secretary, and produced his ‘blockbuster report', which recommended that the British should transfer support to Tito and sever their links with Mihailović. In the
absence until recently of Sigint on Yugoslavia, and in particular of decrypts of signals sent by the
Wehrmacht's
intelligence service, the
Abwehr
, it has generally been assumed that Maclean's report was the crucial element leading to the change of policy. The vast amount of Sigint relating to wartime Yugoslavia that has now been deposited at the Public Record Office shows that this was not the case.

The German occupying forces, and to a lesser extent the Italians, had little option but to use radio for communications between their units in Yugoslavia and to both the German Army Command in Salonika and the
Oberkommando der Wehrmacht
(OKW – the High Command of the German armed forces). The infrastructure of the country was primitive and was frequently disrupted by the resistance. Radio was, so they thought, a safer and more reliable form of communicating than post or courier. Decrypts were to show that the
Abwehr
had to rely to some extent on pigeons to send messages, probably not terribly dependable but more secure. Bletchley Park was able to decrypt many of the German signals, which led the authors of its internal history of Sigint in Yugoslavia and the Balkans to write in 1945 ‘that never in the field of Signals Intelligence has so much been decrypted about so little'.

Sigint provided intelligence from many different sources for the policy-makers, not least Churchill, who referred to it as Boniface, the mythical secret agent from whom, for cover purposes, it was supposedly obtained. Before the outbreak of hostilities in Yugoslavia, Bletchley Park had broken German railway Enigma, some air force Enigma ciphers and a very little army Enigma. Some intelligence was forthcoming about German plans for the invasion of Yugoslavia, including the build-up of forces and the exact date and time of the invasion, to the extent that the British were aware when the start time was brought forward, at the eleventh hour, by thirty minutes. After the occupation the amount of decrypted material gradually increased. The army established its headquarters for the Balkans (Army Group E), initially in Athens, and then in Salonika. Bletchley regularly read messages and situation reports passing between the Commander of Army Group E and his subordinate commanders, the German Generals in Zagreb and Belgrade, together with those of the German liaison office with 2nd Italian army, all of whom used Enigma. During 1943, it was also able to read some messages passed on the Fish links (which used teleprinter cipher machines) between Berlin and
Salonika and Belgrade, and between Vienna and Salonika, as a result of the development of ‘Heath Robinson'. Daily situation reports from Army Group E to Berlin were read regularly. German naval Enigma messages were decrypted, particularly after the Italian capitulation in September 1943 when the German Navy was trying to regain control of the Adriatic islands off the coast of Yugoslavia.
Luftwaffe
Enigma was also decrypted and some Italian Air Force messages using hand ciphers were read. The
Abwehr
had established a presence in Belgrade before the invasion and its radio messages were decrypted. After the occupation of Yugoslavia by the Axis, the
Abwehr
established offices throughout Yugoslavia, which used both hand ciphers and a special version of the Enigma machine to send reports of resistance activity to their area headquarters. Many messages from
Abwehr
officers or their agents were read. Others used the
Abwehr
links including the German consul in Dubrovnik, Herr Aelbert. In addition, the Nazi Party's intelligence service, the
Sicherheitsdienst
, operated in Yugoslavia: its messages were read, although in lesser volume than those of the
Abwehr
. In order to try to keep control, the Germans used various arms of their police service, often manned by local ethnic Germans, whose reports by radio were frequently decrypted. By the autumn of 1943, after the Allies had established themselves in southern Italy, a monitoring station was established in Bari, and later on the island of Vis, off the Yugoslav coast, which enabled local traffic between Chetniks, Partisans and Croatian units to be read. This supplemented intelligence from Chetnik and Partisan radio messages, which were summarized in German reports of their own decrypts of intercepted resistance radio messages. Despite Bletchley's own use of the Boniface cover, one of its internal reports described the German efforts to disguise the source of their intelligence, somewhat smugly, as being ‘camouflaged rather transparently as information obtained from agents'.

Two other sources were identified and decrypted. Tito and the separate Slovene communist party kept in touch by radio with their masters in Moscow, the Comintern and its Bulgarian Secretary-General, Georgi Dimitrov. The volume of messages intercepted was not great but they yielded significant intelligence and continued with Dimitrov after June 1943, when the Comintern itself was dissolved. For a long time before the outbreak of war, the principal activity of GC&CS had been the decrypting of messages sent between diplomatic
missions and their governments. This continued during the war. One of the most useful sources to the British, in terms of German policy in Yugoslavia, was the link between General Oshima Hiroshi, the Japanese ambassador in Berlin, and Tokyo. Oshima frequently reported on his conversations with Hitler and the German Foreign Minister, Joachim von Ribbentrop.

It was realized at a very early stage in the war that the decrypts provided the British with a priceless asset. Every effort was made to protect its security. By 1943, those who received Ultra were far greater in number than the thirty men who were its sole recipients (outside GC&CS and MI6) in 1940. By early 1943, Ultra was being sent to twelve destinations in the Middle East alone. Some of those who received intelligence derived from Sigint were not informed of its source, but as Montgomery's intelligence officer, Brig. E.T. Williams, has written, they must have guessed that it came from wireless intercepts. Nevertheless, some of those responsible for policy, and not just that on Yugoslavia, were denied access to the Bletchley Park material. The SOE in London did not receive any Sigint from any source during the war, let alone Enigma, although its office in Cairo did receive some locally decrypted
Abwehr
material on Yugoslavia in early 1943. MI6 received the decrypts on Yugoslavia. The Military Intelligence section concerned with the Balkans (MI3b) received almost all the relevant decrypts, including diplomatic but not Comintern. The Directorate of Military Operations certainly had summaries from military intelligence. The Joint Intelligence Committee, whose task was to advise the military and the government on major matters of intelligence, had reports based on Sigint, as did the Chiefs of Staff. There were other recipients of Bletchley Park's output of Yugoslav Sigint. The Soviet agent John Cairncross passed some decrypts to the KGB when he was working in Hut 3 at Bletchley and later after being transferred to the MI6 headquarters at Broadway Buildings, which included both
Abwehr
and German Army signals from Yugoslavia. The evidence for this can be found in messages from the Comintern to Tito, giving information that was identical to Bletchley Park decrypts – it is unlikely that the source was the Russians' own decrypts. Thus Tito also benefited from Bletchley's work!

The Foreign Office received some heavily disguised intelligence reports based on Sigint, but did not receive any raw material other than diplomatic decrypts, until the autumn of 1943. Therefore, the
two principal organizations charged with the development of policy towards the Yugoslav resistance – the SOE and the Foreign Office – had either no access, or very limited access, to the decrypts. This inevitably led to much confusion in 1943 between those who wished to continue to support Mihailović – the SOE and the Foreign Office – and those who wanted to switch support to Tito – SOE Cairo, MI6, the Directorates of Military Intelligence and Operations, the Chiefs of Staff and, ultimately, Churchill himself. The dilemma, after months of argument, was only to be resolved by the Prime Minister. Churchill received his daily box of raw decrypts, which frequently included detailed information about the actions of the resistance in Yugoslavia and Axis counter measures. Churchill, using his customary red pen, underlined or ringed items that caught his attention. He was, from time to time, briefed in detail in writing and given advice by military intelligence. He received summaries of decrypts from both military intelligence and air intelligence, some of which have survived, despite being marked ‘to be destroyed'. Churchill received regular oral briefings from ‘C', and from General Davidson, of which there is no record. It must be presumed that he also received written assessments from MI6, but they have not been released.