by Alastair Denniston- Code-breaking From Room 40 to Berkeley Street
A Research Section was headed by Norman Sainsbury and they were tasked with taking on new problems and making enough progress to be able to hand them over to the Language Section. Sainsbury could speak Finnish, Norwegian and Swedish, but Scandinavian countries used Hagelin systems, of which Berkeley Street had no expertise. By 1943, they were concentrating on the principal diplomatic system being used by the Free French. The bulk of the traffic was provided by Cable and Wireless Ltd.
The Commercial Section was based in Aldford House, a few streets from Berkeley Street. It was founded and led by G.L.N. Hope and the majority of his staff were women. They were dealing with some 7,000 messages per day by 1943 and almost all of them were in plain text. The German traffic dealt with things such as the sale of machinery or transactions in grain or mohair with Turkey and Spain. The traffic started to decline due to the restrictions placed on the use of codes by many governments and German occupation. The section did deal, however, with diplomatic codes which had commercial significance
All diplomatic traffic went by cable and wireless based on available routes. During WW1, the censorship organisation had provided all cable traffic. The small amount of wireless traffic was picked up by the Service intercept stations. While cable censorship ended in 1919, a Home Office order in 1921 required the Central Telegraph Office and the Cable Companies to submit all traffic handed to them to the Foreign Office for scrutiny. Censorship was re-imposed in September 1939 and at the same time, wireless traffic was steadily increasing. Traffic was picked up by Service Intercept stations kept active after WW1 by the Admiralty at Pembroke (later Flowerdown) and Scarborough, the War Office at Chatham and the Air Ministry at Waddington. The Metropolitan Police at Denmark Hill had also contributed German clandestine material from secret transmitters in embassies and legations, discovered during 1937–8.
With war looming, independent intercept stations were set up to take diplomatic traffic. These were staffed and maintained by the Post Office and operationally controlled by the Foreign Office. At the beginning of the war, two were operational, at Sandridge, Hertfordshire and Cupar (later closed down and replaced with one at Hawklaw), near Fie in Scotland. A third soon came into operation at Brora in the Scottish Highlands. They contributed Services traffic as well, being part of the Y Service. A special section, CMY (Commercial Y), was set up to centralise development and coordination, and evolved from WTC (W/T Coordination) in 1942. It eventually set up headquarters at Berkeley Street. Two more stations were set up at Whitchurch in 1943 and Wincombe in 1944.
Overseas sources were Mauritius, Ottawa, Delhi, Abbottabad, West Africa, Melbourne, Cape Town, Simons Town, Suez and Malta. Only the censorship and intercept post at Mauritius was under direct Foreign Office control, dealing with Japanese and French traffic. The rest were Service intercept centres, British and Dominion, which agreed to allocate a certain number of sets to diplomatic coverage. By July, 1945 the proportion of general traffic intercepted had reached 86 per cent and, of priority traffic, 96 per cent. Comprehensive arrangements with the US for the exchange of cable and wireless materials filled in any gaps in each side’s coverage.
The total number of translations of diplomatic and commercial messages peaked in 1943 at 14,050. All went to the Foreign Office, twothirds to MI5, one-third to the Service Ministries and 8 per cent to MEW. Staff numbers didn’t increase after 1943 and when the Japanese Section was increased, it was mainly from other Sections. AGD and his organisation were now at the forefront of British intelligence gathering. In April 1944, they decrypted a Dublin message sent the previous day and subsequent events exemplified this. As Percy Filby later recalled:
On Monday I went to see Commander Denniston and handed him the Dublin message. It was an innocuous message and Denniston appeared to wonder why I had troubled to show him it. But when I asked him to look at the date he was startled and when I told him of the results of the weekend work, he shook hands, dialled his scrambler phone and said, ‘Denniston here, may I speak to C?’ C came on the line and Denniston said, ‘C, I have some good news for you, may I come over?’ Denniston put on his hat, carefully folded the message and placed it in his inner pocket and almost ran from the building.
The Foreign Office allowed the messages from the German Consul in Dublin, Hempel, to go through the Leicester Square office for transmission to Berne, on the basis that it was a neutral to a neutral. This did allow the messages to be delayed for seven days so that AGD’s section could read the contents before the intended recipient. As D-Day approached, Dublin was passing possible dates and places for the landings in France. While no one at Berkeley Street knew the exact date, they were given a number to ring for any dates that were suggested in decrypted messages. This message did exactly that.
Staff numbers in AGD’s organisation in London eventually peaked at between 250 and 275 with only the Japanese Section seeing significant expansion. Translated decrypts issued in 1944 totalled 13,153 and in 1945 up to Japan’s surrender, 8,512. The Admiralty, War Office and Air Ministry received only one-third of the total product available. However, the Intelligence Exchange at BP was receiving all material and therefore intelligence from diplomatic attaché and commercial intercept was being received by Military Commands overseas by both signal and courier. A total of 72,624 decrypts were distributed to the Foreign Office between 3 September 1939 and 15 August 1945. It received, with only a few exceptions, all of the material, while thirteen other Departments each received a portion of the total, based on their particular interests. In general, ten major cryptanalytic country sections with between five and twenty-five staff produced a weekly average of 233 decrypts during the period above.
Life at Berkeley Street continued to be hazardous and the offices there were narrowly missed by V1 and V2 attacks on London as the work was relentless through 1944 and 1945 until VE Day. Suddenly, at the end of December 1944, AGD gathered his staff together. He informed them that he intended to retire and wanted to thank them for their work and loyalty to him personally. He was to be replaced by his deputy, Eddie Hastings. AGD held a dinner for some of his section heads and wished them well. He advised those that planned to stay in GCHQ, the postwar GC&CS organisation, that their lives would be different. ‘Before the war you came home and discussed the day with your wives, but now never again will you be able to share your life with your family. Some wives will not mind, but most do have some disappointment, so do think carefully before you decide.’34
Was this the speech of a bitterly disappointed man or just realistic advice from an experienced intelligence officer? In any event, AGD simply put on his hat and walked out of Berkeley Street to the Green Park Underground Station, saying nothing to his fellow workers. There was no official thank you or goodbye from GCHQ or the Foreign Office. It is not known if he was asked to nominate any of his staff for decorations but it appears that not one award was given to the 250 individuals who contributed to the success of AGD’s Berkeley Street organisation.
Chapter 6
Cut Loose
During the early months of 1945, Allied Sigint organisations began to seek new tasks as German military ciphers were continually mastered. On the night of 23/24 March, Field Marshal Montgomery led his British forces across the Rhine, and the Americans followed suit the next night. In the East, the Soviets completed the taking of Vienna on 13 April and three days later began their advance on Berlin. Military Enigma decrypts became less important and the emphasis of GC&CS shifted from BP to Berkeley Street and to diplomatic and commercial W/T traffic. AGD’s organisation processed the messages of the German Foreign Ministry, the Japanese military attaché in Berlin and Japanese diplomats in neutral countries around the world. They were also reading the ciphers of Spain, Nationalist China, the Free French and many other non-belligerents. There was also the commercial traffic processed for the MEW.1 All of the incoming and outgoing cipher telegrams, intercepted by Canada, were being sent to AGD at Berkeley Street.
The success of AGD’s Berkeley Str
eet operation and its invaluable contribution to the war effort was significant and, in the last few years of the war, may even have equalled that of Travis’s at BP. While documents providing official reports of the impact of the diplomatic product are not available, some assessment of it can be made. After June 1941, Berkeley Street revealed the authentic views of neutral countries such as Turkey. The Japanese ambassador in Berlin, Lieutenant General Hiroshi Ōshima, was admired by Hitler, and both Berkeley Street and SIS in the US were reading his traffic right through until May 1945. Berkeley Street provided the Allies (apart from the Soviet Union) with reliable and up-to-date information on the state of the German armed forces and the thinking of their high command. While other Japanese diplomats in Europe such as Yakichiro Suma in Madrid and Kurihara in Ankara provided a volume of reports equal to that of Ōshima, regular Berkeley Street weekly dispatches, discriminatingly read, provided a vital source for Allied war planners from 1943 on. Useful intelligence was also obtained from the Turkish ambassador, Tugay, and the Italians, Peppo and Quaroni.
Berkeley Street continued to read Japanese traffic even after the BRUSA agreement which allocated this work to the Americans. Many BJs were sent to Washington by Churchill, some of which were annotated by him, for the personal attention of Roosevelt. Japanese decrypts from their Naval Attaché gave estimates of Germany’s total monthly production of front-line aircraft, which could then be compared with Allied intelligence estimates.2 The BJs remained important until early 1944, when it became clear that there was no possibility of an Axis victory.
Berkeley Street’s primary job was to provide BJs for Churchill, the JIC and the Foreign Office as well as processing plain language and encoded traffic for the MEW. This information was used to plot the economic progress of the war and to set strategic priorities. In 1943, up to a third of ‘C’s’ daily delivery to Churchill consisted of BJs. When he was away, they were sent to him in summary form but were available on his return. While Enigma traffic continued to serve the needs of the military and COSs, BJs served geopolitics and war strategy. Diplomatic intercepts reported the changing course of the war.
In mid-1944, discussions were initiated about a future peacetime GC&CS, and Travis set up a small planning group to make recommendations. It was led by Gordon Welchman, one of AGD’s early ‘professorial’ recruits in 1938 and included Harry Hinsley, who had worked on Naval Enigma and would play a key role in post-war Anglo-American-Commonwealth Sigint discussions, Edward Crankshaw, who had been responsible for wartime Sigint discussions with the Soviets, and Hugh Foss, a GC&CS veteran who had been in Washington working with US naval cryptanalysts on Japanese ciphers. Welchman’s group produced a paper on 17 September 1944, which recommended the creation of a more centralised ‘Foreign Intelligence Office’. They also called for a comprehensive body dealing with all forms of Sigint, including a modern signals organisation which exploited the latest communications technology. It would in effect become a modern ‘Intelligence Centre’ controlling all British interception work.3
Not all senior figures at BP agreed with Welchman and his group. John Tiltman produced a paper in October 1944 arguing for GC&CS to be absorbed into SIS under Menzies to create a single intelligenceproducing service.4 However, in January 1945, the Chairman of the JIC, Victor Cavendish Bentinck, suggested that GC&CS should remain under the overall direction of the Chief of SIS while remaining a separate organisation. It would be provided with its own budget along with the other secret services as part of the ‘Secret Vote’, the strangely named intelligence budget.5 In the end, it was Travis who would decide on the shape of GC&CS after VJ Day. His new organisation would be divided into five groups, Technical, Traffic Analysis, Cryptanalysis, Intelligence and Cipher Security, each run by his key subordinates with a total staff of around 1,000 civilians and 100 military staff.
Meanwhile, the legacy of AGD’s Berkeley Street organisation lived on and in October 1945, an Allied Mission was sent to Rome to ensure security of the Italian Diplomatic Ciphers. It consisted of the Head of the Italian Section at Berkeley Street, a secretary and an officer from the ASA’s Italian Section in Washington. It operated as a subsection of the Allied Control Commission until the end of the war in May 1945. Its brief was to keep a watch on Italian diplomatic communications and their security and ensure that the appropriate intelligence and government departments in both Britain and the US were kept informed about Italian ciphers and lines of communication.
On 22 October 1945, Travis arrived in Ottawa to meet with the JIC on his way to Washington to discuss post-war management of Ultra and other forms of special intelligence. According to Canadian sources, he reported that GC&CS was being absorbed into a General Signal Intelligence Centre and it would manage the product of all intercepts.6 Travis also reported that he believed that commercial codes and ciphers (those of business and industry) were not seen as legitimate targets for peacetime Sigint. While Menzies’ position was always secure under Churchill’s stewardship, after Churchill lost the election of 26 July 1945, his control over Ultra was loosened. While remaining as Director-General of Sigint (as did his successor, Sir John Alexander Sinclair) he had increasingly less time to be involved in it. In January 1945, postwar planning was taken over by William F. Clarke, a veteran of Room 40 and GC&CS’s naval sections. He argued for a separate organisation under either the Chiefs of Staff or the Cabinet Office, operating as a third secret service.
Group Captain Claude Daubeny was tasked with finding a new home for GC&CS and he recommended recombining the remnants of the staff at BP with those at Berkeley Street. The move to Eastcote in London took place in early 1946 in four main groups, and was completed by April.7 Between 1945 and 1948 the name ‘GCHQ’ was used interchangeably with both ‘London Signals Intelligence Centre’ and ‘Station X’, although the first was preferred as it gave nothing away about the nature of the organisation.8 By November, British defence chiefs argued for increasing the expenditure on intelligence, with particular emphasis on Sigint, and it was granted.9 With a larger budget approved, GCHQ staff began the move to its new base in Cheltenham in late 1952 at two locations. By now Travis, who had been fighting poor health throughout the late 1940s,10 had been replaced by Wing Commander Eric Jones. He had taken over Hut 3 after its internal problems had bubbled to the surface, in July 1942. GCHQ’s existence became widely known in the 1970s and 1980s as a result of several trials over leakage of Sigint secrets and the sacking of a number of employees as a result of industrial action.
With Travis, rather than Menzies in charge in 1945, it is hardly surprising that AGD had no role to play in a future GCHQ. He was ‘encouraged’ out of the organisation before VJ Day and officially retired on 1 May on an annual pension of £591,11 considerably smaller than what he expected. Eventually, the Sigint organisation that AGD had created no longer required BP and perhaps fittingly, it was left to his faithful personal assistant during his days there, Barbara Abernethy, to shut the now abandoned site: ‘I and a guy called Colonel Wallace closed up the place. We just closed down the huts, put all the files away and sent them down to Eastcote. I was the last person to leave Bletchley Park. I locked the gates and then took the key down to Eastcote.’12
***
After years of secrecy and intense work pressure, AGD returned to his original profession of teaching to supplement his meagre pension. He taught French and Latin for a while at Downsend Prep School near Leatherhead, but now in his 60s, he found it too strenuous. So he retired completely and, with his wife, settled down to family life in the New Forest while continuing to play golf and squash. While his years of public service were over, friendships established during that service were not. He had befriended the American liaison at Berkeley Street, Telford Taylor, and his successor, Bancroft Littlefield as well as another American, Lou Stone. Robin Denniston remembered both men spending weekends at Ashtead during the V1 and V2 raids on London. Bancroft picked and ate the Denniston’s soft fruit, while Taylor played piano duets with Robin and
beat everyone at tennis. In the evenings, Taylor and AGD drank whiskey and no doubt confided in each other about intelligence matters.
AGD was a meticulous and careful man, noting his expenses, however small, in his diary. He was generous and gave all his women friends a diary at Christmas. He got quite bad depression (which his wife called Scottish blight) which disabled him for hours. After he left BP, he didn’t get depressed or turn to drink but developed an uncontrollable lower lip quiver. He was irritable and had seemingly lost confidence in himself, perhaps even feeling betrayed by close former colleagues. Furthermore, he did not trust the bureaucrats at the Treasury or the rich and privileged who circulated on the edge of and sometimes in the middle of the Secret Services.
In late 1944, after he retired, AGD became aware of a classified memorandum by Eric Jones13 titled ‘Post-War Intelligence’. Jones’ comments were intended to inform discussions about a new post-war GC&CS. One remark in particular compelled AGD to respond: ‘It would indeed be a tragic and retrograde step for intelligence as a whole, and therefore – this is not putting it too high – for the future of the country, if GC and CS were to sink back into its pre-war position.’
He decided to document his personal thoughts on the origin and purpose of the 1919 GC&CS, its establishment under Treasury control and its development. He did so completely from memory, and Dorothy typed the manuscript on her old typewriter.14 He felt there were few who knew the true history or anything of his early work to prepare GC&CS for war. The perceived wisdom was that GC&CS was completely unprepared for war, had failed to take on the mathematical needs of machine decryption and was run by amateurs unable to cope with the officials in the Treasury, the needs of the Armed Forces or the requirements of BP’s enhanced wartime capability. AGD had kept in touch with Birch, Tiltman, Cooper and De Grey and to some extent Travis and eventually with the new Head of GCHQ, Eric Jones, who he had appointed.
