Churchill was still very much interested in what was happening in Yugoslavia. On 9 September he minuted the Chiefs of Staff that, with the capture of southern Italy, munitions could soon be sent to the resistance. Although not now available, MI3b prepared a report for the Prime Minister on 12 October, no doubt on the same lines as their assessment at the end of September. Churchill was told by ‘C that the SOE had been asked to report on the provision of supplies to the resistance.
Numerous decrypts during October showed the continuing disorder in Yugoslavia and the Germans’ attempts to counter it. German army and police reports referred to the Partisan threat to the major towns of Ljubljana and Zagreb and their interruption to railways radiating from the towns. Communists were resisting the German advance in Slovenia. A major operation was launched by the Germans in the area between Zagreb, Ljubljana and the Italian border at Trieste. According to German reports, the operation resulted in the death of 3,200 Partisans by mid-November. The German commanders launched a series of operations to clear the Partisans from the remaining stretches of coast that they held, and from the Adriatic islands from which they threatened German shipping and supply routes. These operations were followed in detail by Bletchley Park. Air intelligence sent Churchill a breakdown of German efforts to capture the islands compiled from the decrypts. The German Admiral in the Adriatic reported their efforts as ‘unsatisfactory’. Decrypts disclosed that the Germans were carrying out further operations to try and keep the bauxite flowing and at least gain control of communications in Herzegovina and Montenegro.
At the end of October, Churchill was sent a further assessment by MI3b, advising him in detail on the situation and concluding that ‘the Partisans had been able to take over the initiative over practically all of Yugoslavia’. Mihailović was not mentioned except for the fact that in Montenegro some of his supporters had deserted to the Partisans as ‘the more active body’. More reports were received of Chetnik collaboration, the most significant of which was the full text of a treaty signed by one of Mihailović’s principal commanders, Lukacević, and the German Commander South East. Lukacević agreed a cessation of hostilities in his area of southern Serbia and joint action against the Partisans. A full copy of the treaty was sent to Churchill.
Maclean delivered his report to Anthony Eden on 7 November, recommending all-out support for the Partisans. This had been the view of military intelligence since at least the end of September, when Talbot Rice’s report backed the Partisans, and had also very probably been the view of MI6 for some time. The Chiefs of Staff advised Churchill on 11 November that measures to support the Partisans should be intensified. The question of what to do about Mihailović had still to be decided. Churchill took the decision to abandon him and his movement. He announced his decision to Stalin – much to his surprise – and Roosevelt at the Tehran Conference at the end of November 1943. The existence of the principal source of Churchill’s intelligence could not be revealed – hence the publicity given to Maclean’s report – although it told Churchill nothing he did not already know. In fact, the Prime Minister was better informed than Maclean, who knew little of the detail of events over a wide area of Yugoslavia or of the Lukacević treaty. In order to justify the decision to Parliament, to Allied governments, particularly those in exile, and to the press, Mihailović was told to blow up two important bridges in Serbia or lose British support. As expected, he failed to act and British liaison officers were withdrawn from the Chetniks. At the same time a delegation from Tito arrived in Cairo to negotiate with the Yugoslav government-in-exile. The delegation was able to seek instructions from Tito by radio. Dimitrov advised Tito on the negotiations and how the delegation should play its hand. A series of decrypts revealed the exact advice given, particularly on the issue of the future of King Peter. Dimitrov advised Tito ‘to show a necessary flexibility with reference to the question of the king to overcome certain difficulties on the side of the British and the Americans in the matter of their material assistance’.
It has been alleged by a number of commentators that ‘a conspiracy’ at SOE Cairo, having revealed the successes of the Partisans to Churchill, seduced him on to a path that was to lead to all-out support for the Partisans. It has been suggested that James Klugman, a communist activist and KGB agent, who joined SOE Cairo in 1942, was the ‘agent of influence’ and prime mover behind the decision to back the Partisans rather than the Chetniks. However, in response to these ‘wild accusations’, Ralph Bennett, a former duty officer in Hut 3 at Bletchley, and Sir William Deakin and others later put forward the view that Sigint had provided the facts which persuaded Churchill, on military grounds and military grounds alone, to choose Tito in place of Mihailović. It is now indeed clear from the decrypts that the Prime Minister was well aware of what was happening before he saw Deakin and Keble in Cairo and, while interested in what they had to say, was not manoeuvred into ultimate support for the Partisans by anything that emanated from SOE Cairo.
Churchill addressed the House of Commons for the first time in six months on 22 February 1944. He dealt with the situation in Yugoslavia at length. He was unable to justify the decision by reference to the decrypts and the advice he had received based on them, and therefore referred to reports received from Deakin and Maclean. In his peroration he advised that:
Our feelings, here, as elsewhere, I should like the House to see, follow the principle of keeping good faith with those who keep good faith with us, and of striving, without prejudice or regard for political affections, to aid those who strike for freedom against Nazi rule and thus inflict the greatest injury on the enemy.
With these few words, Churchill publicly dismissed Mihailović and the Chetniks, and embraced Tito and the Partisans. The Partisans continued to harry the enemy, although the Germans were able to keep the bauxite flowing and to keep major communications routes open, allowing their forces in Greece to complete an orderly withdrawal in 1945. The Partisans won the civil war and seized power in the immediate aftermath of German surrender. They hunted down Mihailović and captured him. After a show trial he was shot in 1946.
15
TRAFFIC ANALYSIS: A LOG-READER’S TALE
JAMES W. THIRSK
Introduction
In Chapter 15, James Thirsk introduces us to the art of traffic analysis - the study of signals to gain intelligence from them without actually reading them.
Traffic analysis using Heer and Luftwaffe signals got off to a very slow start at Bletchley, largely because GC&CS had no suitable intelligence team until Enigma decrypts became available in January 1940. The services were ill-equipped to carry out traffic analysis in the first months of the war, and could not even determine whether one major radio net carried Heer or Luftwaffe traffic. When Ultra established that it was a Luftwaffe net, it was a shock to MI8 to find that virtually all of the Army’s fixed intercept receivers had been intercepting traffic which was really an RAF responsibility.
Intercept stations had two functions: providing ‘wireless telegraphy intelligence’ (WTI, as traffic analysis was then known), and gathering intercepts for GC&CS. But since no one had decided which should take priority - and therefore who controlled intercept tasks - the stations were almost rudderless. When Josh Cooper, the head of Air Section, suggested that an RAF station should take Enigma, the head of the RAF Y service told him ‘My Y Service exists to produce intelligence, not to provide stuff for people at Bletchley to fool about with.’ A sub-committee of the ‘Y’ Committee eventually decided in mid-1941 that WTI and cryptanalysis formed an indivisible whole, and that there was no conflict between them. But the debate rumbled on.
As Cooper pointed out, deriving intelligence about the enemy’s intentions from traffic analysis must be carried out very carefully indeed, since it is ‘a difficult and dangerous art’: cryptanalysts who make false deductions can produce little or nothing, but ‘anybody with a flair for detective work can produce theories based on undecypherable traffic and nobody can contradict him’.
Good traffic analysis in that wider sense could too easily be thwarted by efficient defensive and offensive radio deception. As the Allies were to find out, the fact that radio traffic did not increase significantly before the Ardennes offensive in December 1944 did not mean that the Germans were not massing the troops to launch it.
In May 1943 a senior member of GC&CS told William Friedman, the legendary American codebreaker, that inference ‘solely based upon radio studies is of doubtful value’, but ‘intelligence concerning the enemy’s W/T network and procedures is extremely important’. James Thirsk deals with the latter kind of traffic analysis, and describes the roles of units such as Hut 6’s Fusion Room in building a full picture of the complex German radio nets. The humble intercept operators also played an important part in traffic analysis. Their analytic input could be of crucial importance, especially when the Germans made major changes in communications security, as with the introduction by the Luftwaffe of a new call sign book on 1 April 1944, just weeks before D-Day.
RE
As the train steamed into Quorn and Woodhouse station in Leicestershire I wondered whether I had been wise to volunteer for the Intelligence Corps. ‘Never volunteer’, the old soldiers used to say. But I read one day in Army Council instructions that ‘Men with suitable qualifications are required for transfer to the Intelligence Corps’. No more details were offered. Several months later, after two interviews in London, here I was, in April 1942, arriving alone at the little railway station with all my worldly goods, including a heavy kitbag, gas mask, steel helmet and a small suitcase full of books. Outside the station, a young woman from the Auxiliary Territorial Service (ATS), the women’s section of the army, was waiting for me in a jeep. I had been told to report to Beaumanor, a large country house less than a mile from the station. As the jeep passed through the imposing gateway, winding its way down a long drive, bordered by ancient trees, I wondered whether I was arriving at the secret headquarters of British Intelligence.
In the entrance hall of this large Victorian mansion, built in the Elizabethan style, I was greeted by Lieutenant Rodney Bax. We sat on a long sofa on a landing at the top of the main staircase. ‘I expect you’re wondering what you’ve let yourself in for, Bombardier Thirsk?’ he said. After a few enquiries about my civilian job and my army service, he explained that the unit I had joined was a branch of MI8, known as the Central Party, and that its job was to analyse German and other wireless signals or ‘traffic’. He explained that at many places all over Great Britain, hundreds of men and women, trained in the Morse code, listened and wrote down on pads of paper everything they heard on different wireless frequencies. These records were called logs and our job was to study this traffic, hoping to construct pictures of German Army and Air Force formations. ‘We do not deal with naval traffic,’ he added. ‘The intercept operators, as we call them, also write down messages on a separate pad which are in cipher,’ he said. When I enquired whether the ciphers were broken, he told me that there had been occasional successes with simple ciphers but that we were not concerned with cryptography. He was a good liar. It was not until nearly ten months later that we log-readers were told that the ‘impregnable’ German Enigma machine ciphers were being regularly decrypted. Lieutenant Bax told me that one of the intercept stations was at Beaumanor, buried away in huts and buildings in the grounds. Some logs came from other intercept stations by despatch rider.
At Beaumanor, we received no training in the art of log-reading. ‘I’m afraid that we have to throw you in at the deep end,’ said Bax. ‘But for a few days you will be working alongside Corporal Newte.’ During the following month, I gradually picked up the skill of scanning the logs and picking out the important items. Each separate unit in a German Army or Air Force group used an identifying label, known as a call sign, when sending or receiving traffic. This consisted of a mixture of letters and figures, usually three in number: for example TR7, VLU, 4BK. To outwit enemies intercepting traffic, call signs were changed daily, according to a printed programme.
The German Army and Air Force during the Second World War used five printed call sign books which they named B, C, D, E and F. Each book contained thousands of call signs, arranged in columns and rows so that a station, knowing the row and column which had been allocated to it, could select its correct call sign for the day. Book B, known to the log-readers as the ‘Bird book’, was used by the German Air Force from the beginning of the war until 1 April 1944, when it was superseded by Book F (Fox). The Central Party, during 1940–1, before it moved to Beaumanor, had already largely reconstructed the Bird book, by studying the call signs over a whole year. We used a well-thumbed copy daily to identify stations on our networks. It was always missing from the control table in the log-room and a familiar cry would be heard again and again: ‘Has anyone seen the Bird book?’ A German copy was captured in Libya in December 1941. The other call sign books C, D, E and F were not used so frequently by the Germans. When the German Army began to encipher call signs in November 1944 we were unable to identify networks by predicting changes of call signs until March 1945, when a copy of the German instructions for enciphering call signs was captured. This information was also of great value to the intercept stations, enabling them again to identify networks even if they had changed frequencies.
The organization responsible for investigating all enemy and neutral communication during the Second World War was known as the Y service. It was responsible for monitoring or intercepting signals, for direction finding (DF) and for the decryption of low-grade ciphers and for plain language traffic. The intercept operators often became so familiar with the ‘fist’ of a German transmitting in the Morse code that they would make a note on the log, telling us that the German operator today was the same man who had used a different call sign yesterday. One added a note: ‘Italian operator’ on the log, having allegedly recognized a Latin rhythm in the Morse transmission.
Although some stations intercepted foreign radio signals before the war, many more were needed to cope with the flood of traffic as soon as the war started. Men and women in the Post Office, familiar with the Morse code, having worked in the old telegraphic service, were recruited. But, as many more were needed, it was necessary to train hundreds and later thousands of men and women. The women in the WAAF and the ATS far outnumbered the male soldiers and airmen. Some civilian men and women also joined the service.
The intercept operator’s job was arduous: working around the clock on a three-or four-shift timetable, they sat for hour after hour with headphones, transcribing the Morse signals as they arrived. Often atmospheric conditions were poor, stations drifted from their frequencies, other transmitters were heard overlapping and signals faded. Great concentration and patience were essential and the operators had no idea whether their efforts were of any value or not. Beaumanor was one of the largest intercept stations. There were a number of camouflaged buildings, one of them disguised as a cricket pavilion, and spread around the grounds were the wireless masts bearing the aerials. By the end of the war, more than a thousand intercept operators worked at Beaumanor; there were many other such stations in different parts of the country.
At the top of each log, written in pencil on sheets from red printed pads, appeared the operator’s initials or codename, the date, the frequency covered, the time of interception and the call signs indicating who was calling whom. Below were written any preambles to messages and the ‘chat’, which was the name given to everything heard on a frequency other than ciphered messages. ‘Keying’, usually in the form of a series of letter Vs in the Morse code (…-) was used to maintain contact between stations in the intervals between messages. Also appearing on the log were messages in the Q Code, an international three-letter code with the first letter always the Morse for Q (- -.-). The Germans in the main kept to the standard international meanings; for example, QCB meant ‘you are causing delay by answering out of turn’. QSA was ‘What is my signal strength?’ The reply QSA 5 meant that the signa
l strength was good, and QSA 1 that it was poor.
The logs were passed to the Central Party, where they were sorted by frequency and allocated to the different groups of log-readers, each studying a section of the German Air Force and Army networks. We never used the terms ‘traffic analyst’ or ‘traffic analysis’: we were log-readers reading logs. At the beginning of a shift, each log-reader would collect the logs for his or her networks (there could be as many as a hundred) from a pigeon-hole. The logs were first checked to see if there were any messages in clear German. The next job was checking that all the logs bore the same frequency and were therefore recording the traffic of the same German group or network. Sometimes a log-reader might be studying two or three smaller groups using different frequencies. The Bird book was then used to check all the call signs on the logs, equating those in use yesterday with those in use today.
The most common type of operational group in the German Army or Air Force was in the form of a star (Stern in German). This was a group of stations with a headquarters (control) and two or more outstations. In some cases, the outlying stations were allowed to contact each other direct. Another type was in the form of a circle (Kreis in German). Here, three or more stations, usually of equal status, communicated with each other. The log-readers drew network diagrams on proforma sheets to show the stations using a particular radio net with lines drawn between various stations that communicated with each other and arrows depicting each message and the direction in which it was being sent. In the mornings, headquarters often checked outstations in turn to make sure that they were awake and alert. We were asked by the cryptanalysts to look out for messages at regular times in the mornings or evenings. These were usually situation reports, which often had standard opening phrases like ‘Morgenmeldung’ (morning report) in cipher. This could help the cryptanalysts to decrypt the message by providing a crib. Below the network diagram we wrote notes about the activities of the stations on the radio network during the past twenty-four hours. We also noted any unusual flow of traffic, whatever the direction, and any changes from normal procedure.
The Bletchley Park Codebreakers Page 30