B1a’s successes owed much to the leads provided by the ‘Beavers’ in B1b, who analysed ISOS decrypts, Abwehr communications with the double agents and other intelligence relevant to the Double-Cross System.54
Headed initially by the future judge Helenus ‘Buster’ Milmo, B1b contained some of the Security Service’s most formidable intellects: among them Herbert Hart, later professor of jurisprudence at Oxford and principal of Brasenose College, and his fellow Oxford philosopher Patrick Day of New College. In August 1940 ISOS decrypts enabled the Beavers to give B1a advance warning of the imminent despatch of a new wave of Abwehr agents. The attempted expansion of the German agent network in Britain was part of the preparations for Operation SEELÖWE (‘SEALION’), the planned but never implemented German invasion of Britain. During the three months from September to November twenty-five agents landed by parachute or small boat, most of them inadequately trained and poorly equipped. The Security Service found them ‘an easy prey’,55 all the easier to detect since some of their forged identity documents were based on misleading information supplied by MI5. In August SNOW provided the Abwehr, at its request, with names and numbers for a dozen forged identity cards; two months later he was asked for more.56
B1a rapidly turned four of the twenty-five German agents into double agents. Central to the turning process was the Security Service’s wartime interrogation centre, Camp 020 (known within the Service as B1e), based in Latchmere House near Ham Common in west London and run by Captain (later Colonel) Robin ‘Tin-eye’ Stephens. Tin-eye owed his nickname (never used to his face) to the monocle which seemed permanently glued to his right eye – he was rumoured to sleep in it. Born in 1900, he had spent fourteen years in the Indian army and political service, latterly working in the Judge Advocate’s department. After returning to England in 1932, he went to Lincoln’s Inn; though (for unknown reasons) not qualifying as a lawyer, he co-authored law books as well as working as a journalist. Like many other MI5 officers, Stephens was a keen sportsman, working from 1937 to 1939 for the National Fitness Council. He was also a good linguist, fluent in French, German and Italian, with an interpreter’s qualification in Urdu and what he described as a ‘poor’ knowledge of Somali and Amharic.57He took a dim view of most Europeans. Italians were ‘undersized, posturing folk’, Belgians overweight, ‘weeping and romantic’, the French corrupt, Polish Jews ‘shifty’ and Icelanders ‘unintelligent’. Stephens had a particular dislike of Germans. But, for all his ethnic prejudices, he proved a remarkable judge of individual character. ‘National Characteristics in spies’, he acknowledged, ‘are inconclusive . . . The interrogator must treat each spy as a very individual case . . . a very personal enemy.’58
In January 1941, the London Reception Centre (LRC) was established at the Royal Victoria Patriotic School in Wandsworth to act as a screening centre for aliens arriving from enemy territory. In the course of the war, approximately 33,000 refugees passed through the LRC (known in the Security Service as B1d), where they were questioned about their methods of escape, the routes they had followed, safe houses, couriers, helpers and documentation. Their statements were meticulously indexed and crosschecked against those of their companions and earlier arrivals. Intelligence was extracted and circulated to Whitehall departments. Any inconsistencies in their stories were rigorously followed up. Though SIS and SOE officers were also present at the LRC, MI5 reserved the right to carry out the first interview. As Dick White put it, the function of the LRC was to separate the ‘sheep from the goats’.59 The goats – those suspected of being Axis agents – were transferred to Camp 020.
Camp 020’s first major contribution to the construction of the Double-Cross System was to turn, in the space of only a few days, two of the first wave of Abwehr agents, codenamed SUMMER and TATE (both Scandinavians), who landed in England in September 1940. SUMMER was caught soon after landing by parachute in the Northamptonshire countryside in the early hours of 6 September by a farmhand who found him asleep in a ditch, together with his parachute, radio transmitter, £200 in banknotes, a false identity card and a loaded pistol. During interrogation at Camp 020, Stephens assessed him as ‘a fanatical Nazi’. Within two days, however, SUMMER had agreed to work as a double agent in return for a promise that the life of his friend TATE would be spared. This was the only time in the history of Camp 020 that such a promise was made to a prisoner. It proved to be the key to the recruitment of TATE, who was dropped by parachute in the Cambridgeshire fens in the early hours of 20 September. As SUMMER had done, he landed heavily, spraining an ankle, and quickly aroused suspicion when he hobbled into the village of Willingham, wearing a smart blue suit. When arrested, TATE was found in possession of both a genuine Danish passport in his own name and a forged British identity card in the name of ‘Williams’. The interrogation of TATE at Camp 020 followed a quite different pattern from that of SUMMER. He claimed to have landed by boat from Denmark three weeks earlier, fleeing persecution by the Germans – in the process infuriating a military intelligence officer from MI9, Colonel Alexander Scotland, who was present at the interrogation.60 The sequel is described in Guy Liddell’s diary for 22 September:
I have just been told that the officer from MI9 who was present at the interrogation of TATE yesterday took it upon himself to manhandle the prisoner without saying anything to Colonel Stephens, Dick White or Malcolm Frost [all of MI5]. The interrogation broke off at lunchtime, when Colonel Alexander Scotland left the room. Frost, wondering where he was, followed him and eventually found him in the prisoner’s cell. He was hitting TATE in the jaw and I think got one back for himself. Frost stopped this without making a scene, and later told me what had happened. It was quite clear to me that we cannot have this sort of thing going on in our establishment. Apart from the moral aspect of the whole thing, I am quite convinced that these Gestapo methods do not work in the long run. We are taking the matter up with DMI [Director of Military Intelligence] and propose to say that we do not intend to have that particular military intelligence officer on the premises any more.61
Henceforth all interrogations, save on some matters of technical detail, were conducted exclusively by Camp 020 officers.62 Stephens remained completely opposed to physical violence during interrogation (a principle later restated by the post-war Security Service). Guy Liddell noted an example of his punctiliousness. One of the officers at 020 was a qualified dentist, who carried out any dental work required at the camp, which occasionally included extracting a secret ink capsule hidden in a hollow tooth. Before doing so, the dentist would first obtain the prisoner’s written consent.63
Stephens was, however, a firm believer in the use of psychological pressure to isolate the prisoner, intimidate him and demonstrate his powerlessness.64 Potential double agents were told they could either work for the British or be sent for trial and hanged. Of the 440 people despatched to Camp 020, only fourteen were executed; Stephens was disappointed there were not more.65 He was also a strong advocate of deception as a tool of the interrogator. The information given by SUMMER to gain a guarantee that TATE’s life would be spared, including details of the arrangements for them to meet at the Black Boy Inn at Nottingham, was used to persuade TATE that his friend had betrayed him. TATE, wrote Stephens, ‘lost all his previous composure, cursed “the swine [SUMMER]” and blurted out that he would tell the whole truth. He held back little.’66 After only two days’ interrogation,67 TATE began a career as the longest-serving of all B1a’s double agents, exchanging wireless messages with the Abwehr in Hamburg continuously from October 1940 until May 1945. His German controllers described him as a ‘pearl’ among agents, sent him large sums of money and had him awarded the Iron Cross, both First and Second Class.68
In the course of the Second World War, B1a, which usually had five case officers, ran a total of nearly 120 double agents. On average, five of its officers were running twenty-five agents at any one time.69 Only six double agents were German nationals. The Abwehr deployed against Britain agents of t
hirty-four different nationalities. Though only eleven of the double agents were turned by Camp 020, they included some of the most successful: among them (in addition to TATE and SUMMER) the Briton Eddie Chapman (ZIGZAG) and the Norwegian Helge John Niel Moe (MUTT).70
Though the Security Service retained the lead role in the Double-Cross System, its successful expansion depended on an unprecedented degree of collaboration within the British intelligence community, on a scale and with a sophistication not matched by any other wartime power. ISOS decrypts from GC&CS led directly to the capture of five of the twenty-three German agents despatched to the UK during 1941, identified two more and provided valuable guidance in devising the disinformation with which the double agents deceived the Abwehr.71 In December 1941 the veteran codebreaker ‘Dilly’ Knox, though terminally ill with cancer, succeeded in breaking the Abwehr variant of the Enigma cipher. The decrypts were given the codename ISK (Intelligence Service Knox) to distinguish them from the ISOS decrypts of Abwehr hand cipher messages, which had been circulated for the past twenty months. Most recipients, however, did not understand the distinction and referred to all Abwehr decrypts as ISOS. By the spring of 1942, Tar Robertson was able to state categorically, on the basis of the intelligence derived from the decrypts, that the Security Service controlled all German agents operating in Britain.72 ULTRA provided crucial evidence that the Germans were successfully deceived by much of the disinformation fed to them by their turned agents.
The Double-Cross System also involved SIS, which both recruited and assisted double agents in neutral capitals. Its first major recruit was Dušan ‘Duško’ Popov, a young Yugoslav working for the Abwehr who contacted the SIS station at Belgrade. In December 1940 Popov travelled to London via Lisbon, telling the Abwehr that he was going to collect intelligence from a friend in the Yugoslav legation but with the real intention of making contact with MI5. Codenamed TRICYCLE because of his fondness for three-in-a-bed sex, Popov became the centre of a considerable network of agents, some imaginary (invented by B1a). After the war he was given British citizenship and presented with the OBE by Robertson at an informal ceremony in the bar of the Ritz.73
Like TRICYCLE, the Catalan businessman Juan Pujol Garcia (GARBO), the most successful of all the double agents run by B1a, began as an SIS recruit. Pujol, whose experience of the Spanish Civil War had left him with a loathing of both Fascism and Communism, first offered his services to the British in Madrid in January 1941 but was turned down. He then approached the Abwehr, told them he was travelling to England, and was eventually taken on as Agent ARABEL. Pujol got no further than Lisbon, from where, claiming to be in England, he despatched to the Abwehr plentiful disinformation on non-existent British troop and naval movements, spiced with details of ‘drunken orgies and slack morals at amusement centres’ in Liverpool and the surprising revelation that Glasgow dock-workers would ‘do anything for a litre of wine’. By February 1942 Section V (counter-intelligence) of SIS had identified Pujol as the author of these colourful reports, which were decrypted by Bletchley Park. A month later SIS recruited him as a double agent.74
By cutting across the demarcation line which confined MI5 to British soil and SIS to foreign territory, the double agents necessitated closer operational co-operation between the two Services than ever before. Unsurprisingly, the collaboration did not always run smoothly. In March 1942 Felix Cowgill, the head of SIS Section V, told Liddell that he wished to bring GARBO to London to be debriefed in SIS headquarters but wanted a guarantee that the Security Service would then allow GARBO to return to Lisbon to be run by SIS. Liddell was outraged:
[Cowgill] did not wish to give [GARBO] up or to allow us to have access to him even though in all our interests it might be better that he should remain here. Fundamentally, his attitude is ‘I do not see why I should get agents and then have them pinched by you.’ The whole thing is so narrow and petty that it really makes me furious.75
Liddell won the interdepartmental battle which followed. Since Pujol’s reports to the Abwehr claimed that he and his partially mythical agent network were based in Britain, it made better sense for him to be run by the Security Service in London than by SIS in Lisbon. On 24 April GARBO arrived in London and was transferred to the control of B1a.76
Relations with Cowgill remained tense. B Division, which had hitherto believed it was receiving all ISOS and ISK Abwehr decrypts, discovered in April 1942 that Cowgill had been withholding all those which mentioned SIS agents – in all over a hundred decrypts, including some reports from Pujol. Though the Security Service had got on well with the previous head of Section V, Valentine Vivian, it found Cowgill difficult to deal with. So did some other sections of the intelligence community. Commander Ewen Montagu, the naval intelligence representative on the newly established Twenty Committee, complained to Tar Robertson that Cowgill was possessed of a ‘pathological inability to inform anyone of anything that he can possibly avoid’.77 Kim Philby told Herbert Hart of B1b ‘how he had argued with Major Cowgill that the secret abstraction of these messages from the circulation of ISOS and ISK was quite wrong, since it mutilated the total series, and in any case some of the considerable amount of information in these messages might reasonably be held to concern us’.78 Cowgill was eventually overridden by ‘C’, Stewart Menzies, who assured Masterman on 11 June that all decrypts relevant to double agents would henceforth be passed to the Twenty Committee.79
In order to run double agents successfully, B1a needed to have available a mixture of information and disinformation with which they could both impress and deceive German case officers. In January 1941 the Wireless Board (also known as the W Board) was set up to decide what to tell the Abwehr. On it sat Guy Liddell (who had succeeded Harker as head of B Division), Stewart Menzies and the three service directors of intelligence. This elevated committee, while considering broad policy issues, inevitably lacked the time to provide the detailed, sometimes daily, operational guidance which became necessary following the expansion of the Double-Cross System in the autumn and winter of 1940.80 The Wireless Board therefore quickly delegated day-to-day selection of information and disinformation to the Twenty Committee, so called because the roman numeral for twenty (XX) is a double cross. The Committee, which had representatives from the Security Service, SIS, the War Office, the three service intelligence departments, GHQ Home Forces and, when necessary, other interested departments, began meeting in January 1941 and thereafter met weekly for the remainder of the war.81
The MI5 chairman of the Twenty Committee, the Oxford history don J. C. Masterman (later knighted), was, like the two other key figures in the Double-Cross System, Robertson and Stephens, an inspired choice. Born in 1891, Masterman was considerably older than most other B1a officers, and owed his recruitment in November 1940 to Dick White, who had been his pupil at Christ Church. While a young fellow of Christ Church, Masterman was studying in Germany in August 1914, and was interned for the remainder of the war.82 He first came to the Security Service’s attention as the result of an officious postal censor who in August 1918 reported to the Chief Constable of Hampshire that Masterman’s mother had been ordering ‘suspicious books’ from a Dutch bookshop to be sent to her son during his internment. Kell was subsequently informed that the books, mainly poetry and on the origins of the war, had been ordered by Masterman himself with a request for the bill to be sent to his mother. No further investigation followed.83
As well as being an academic, Masterman was probably the best allround games player ever to join the Security Service. As an undergraduate he had won an athletics Blue. Between the wars he played hockey and tennis for England and at the age of forty-six was still a good enough cricketer to tour Canada with the MCC.84 In his reports on the Double-Cross System, Masterman sometimes used cricketing analogies. ‘Running a team of double agents’, he believed, ‘is very like running a club cricket side. Older players lose their form and are gradually replaced by newcomers.’ He compared the leading double agents to well-known cricke
ters: ‘If in the double-cross world SNOW was the W. G. Grace of the early period, then GARBO was certainly the Bradman of the later years.’85
Though there were some tensions when Masterman rejected risky proposals from the young case officers of B1a, he won the respect of all and, at least in retrospect, they conceded that his caution was necessary.86 Guy Liddell noted at the end of the war: ‘Apart from his ability he is an extremely delightful personality and has been liked by us all.’87 Masterman was also an excellent chairman. He preceded the first meeting of the Twenty Committee with what he later called ‘a small but important decision, to wit that tea and a bun should always be provided for members’:
In days of acute shortage and of rationing the provision of buns was no easy task, yet by hook or crook (and mostly by crook) we never failed to provide them throughout the war years. Was this simple expedient one of the reasons why attendance at the Committee was nearly always a hundred per cent?
The Defence of the Realm Page 35