The Defence of the Realm

by Christopher Andrew

  It was noticed that the bottom left-hand drawer of his desk had recently been unlocked. On examination this drawer, like all the others, contained a thick layer of dust, but this one also contained marks as if a small object on legs had been placed in the drawer and subsequently removed. No satisfactory evidence has yet been advanced to explain these marks.26

  Peter Wright, inevitably, saw a sinister significance in the marks in the dust. They had, he suspected, been made by a KGB camera given to Mitchell to photograph Security Service documents.27

  When the new Director of D Branch, Malcolm Cumming (who had talent-spotted Dick White almost thirty years before), was indoctrinated into the PETERS case on succeeding FJ in June 1963,28 he discovered that Martin had fed him some of the bogus ‘barium meal’ intelligence, which he had passed on to Mitchell believing it to be genuine. Martin saw Director D’s irritation at this discovery as evidence of a character defect – ‘a bitter pill for a vain man to swallow’. The new Director D’s lack of enthusiasm for the Mitchell investigation, though in retrospect a sign of balanced judgement, was proof in Martin’s less balanced view that he was simply not up to the job of directing counter-espionage.29 By August, a fourth Security Service officer, Hugh Winterborn (A2), had joined the PETERS investigation. All, according to Martin, had no ‘serious doubts about Mitchell’s guilt’, but feared that ‘we would not be able to produce evidence sufficient for a prosecution except by successful interrogation.’30 Martin insisted that the interrogation take place before Mitchell’s retirement and that both the ‘CIA and FBI should be told at once so that in the weeks of climax we could make our moves in concert with them.’ Hollis, however, was ‘not yet convinced either that the case was strong enough to justify interrogation or that it was necessary to inform the Americans at all’. Martin, Wright, Winterborn and a colleague then decided between themselves to force Hollis’s hand. At a Saturday-morning meeting at Hollis’s house, each in turn made a personal statement saying that he would resign from the Service unless the Americans were told. Hollis gave no immediate reply, but seems to have decided over the next few days that he had little option but to give in to their demand. At the beginning of the year, less than a week before Philby’s defection, he had assured Hoover that there was no evidence Philby had worked as a post-war Soviet agent. Having so recently and so seriously misinformed Britain’s main intelligence ally, Hollis probably decided that he could not take the risk of concealing the investigation of Mitchell. After consulting White, he went to see Macmillan who agreed that he should brief the Americans soon after Mitchell’s retirement.31

  In late September, Hollis flew to Washington and was faced with the unprecedented embarrassment of informing both the CIA and FBI of the investigation of his own deputy. Martin followed a day later to explain the ‘implications for certain American intelligence sources who, if Mitchell was a spy, must be considered either compromised or provocateurs’.32 Both the CIA and FBI were deeply sceptical about the case against Mitchell – largely because the many current US or joint UK–US operations of which he was aware showed no sign of being compromised.33 Their scepticism was shared by the RCMP.34 It was a remarkable sign of the closeness of the transatlantic intelligence alliance that President Kennedy and his brother Robert, the US Attorney General, were also briefed on an investigation known to only three British ministers and to no member of Her Majesty’s Opposition.35 Their reaction is not recorded.

  By the time Martin flew to Washington, the direction of the PETERS case had been taken out of his hands. The investigative part of his section, D1, was split off as D1/Inv,36 whose head became increasingly doubtful about the case against Mitchell, partly as a result of the scepticism of the CIA and FBI. Michael Straight’s revelation of his pre-war recruitment by Anthony Blunt, followed by Blunt’s confession to Martin in April 1964,37 also provided an explanation for much of the hitherto mysterious evidence of wartime penetration. The discovery of Blake’s treachery similarly accounted for much post-war evidence.38 A D1/Inv report in March 1964 concluded that ‘on present evidence PETERS is more likely to be innocent than guilty. I think that, while continuing the PETERS investigation, we should make a determined effort to look for other candidates.’39

  In the course of 1964, Martin’s relations with his colleagues became increasingly fraught. As even Wright later acknowledged, Martin was both temperamental and obsessive, and ‘never understood the extent to which he had made enemies over the years’.40 On the Tuesday after Whitsun he was summoned to the DG’s office, where he found Hollis ‘choking with anger’, though, in Martin’s view, ‘it seemed an artificial, rehearsed anger for it was not reflected in his eyes.’ Hollis told Martin that he was ‘a focal point for dissension in the Service, and in D Branch in particular, and that he could not tolerate this any longer’. Martin was suspended for two weeks.41 After further disagreements over the next few months, including a complaint by D1/Inv that Martin was undermining his authority, in November he was once again summoned by the DG. Hollis told Martin he had made D1/Inv’s life ‘a misery’ and was ‘at the centre of all the unrest in the office’. The directors had originally intended to dismiss him but had decided instead to second him to another section of the intelligence community for a period of two years.42

  With Martin’s transfer, Peter Wright emerged as the Security Service’s leading conspiracy theorist. His investigations, however, uncovered nothing against Mitchell save for personal eccentricities which provided no evidence of treachery.43 A later investigation by the former cabinet secretary Lord Trend concluded: ‘. . . Mitchell’s curious behaviour is reasonably explicable on the assumption that it represented the natural reaction of a highly strung and rather “odd” individual to the strain of working for a DG with whom he was increasingly out of sympathy.’44

  After leaving the Security Service, Martin continued to co-operate with Wright in their pursuit of imaginary traitors within the intelligence community. By the beginning of 1964 both were convinced that Hollis, not Mitchell, was the most likely suspect. Martin persuaded himself that Hollis had engineered the investigation of Mitchell in order to throw them off the scent and ‘protect himself’.45 During the early stages of their investigation of Hollis, Wright narrowly avoided dismissal from the Service. Before Hollis’s retirement in 1965 he asked an A1 operations officer to stay behind one evening, then told him, ‘There’s a drawer in the DG’s office I want to look into.’ ‘But I can’t do that,’ the officer replied. ‘Oh yes, it’s all above board,’ Wright improbably assured him. The officer later recalled that he agreed to open the drawer, only to discover that it was empty: ‘I thought, “What the devil is he on about?” ’46 Had Wright’s break-in been discovered, he might well have been sacked and the Service would have been rid of its most troublesome and conspiratorially minded member. Hollis’s empty drawer remains as a striking visual symbol of the baselessness of the allegations against him.

  In November 1964 a joint Security Service–SIS working party codenamed FLUENCY, chaired by Peter Wright (but excluding Martin), was directed to examine all available evidence of penetration of both Services.47 The report of the FLUENCY Working Party to Hollis and White on 28 May 1965 concluded not merely that both Services had been penetrated by Soviet intelligence but that the penetration continued.48 D3 (responsible for counter-intelligence research and collation) simultaneously sent a note to Hollis indicating that the DG himself was under suspicion.49 There followed a meeting between Hollis and Wright on about 11 June which was unprecedented in Security Service history. According to a later note on file by Peter Wright, Hollis asked him, ‘Why do you think I am a spy?’ They went on to discuss two of the leads which Wright believed made the DG a suspect.50 On 5 July 1965 Hollis and Sir Dick White agreed that the Working Party had ‘established a prima facie case for penetration of British Intelligence which requires further investigation’, and formally instructed the FLUENCY Working Party to continue an inquiry which the DG knew was bound to make him the fir
st British intelligence chief in modern times to be investigated on suspicion of treason. Golitsyn remained a malign influence on the inquiry. Though admitting that ‘details of penetration of [British] intelligence were particularly tightly held in the KGB’ and therefore mostly unavailable to him, ‘as a result of his intensive reading of KGB files over a period of sixteen years he was certain that British Intelligence on both sides [MI5 and SIS] had been continuously and widely penetrated by the KGB.’51 Wright’s faith in Golitsyn was unaffected by such inconsistencies in his behaviour as the fact (reported by SLO Washington) that, despite claiming to fear assassination by the KGB, he dined in New York restaurants frequented by Soviet officials.52 The root cause of the conspiracy theories which did such damage to the Security Service, apart from the unbalanced judgement of a minority of its officers, was the lack of good defector intelligence on Soviet penetration of the quality later provided by Oleg Gordievsky, which would have demolished the case against Hollis and Mitchell. Tragically, the most important KGB defector apparently able to provide leads on Soviet penetration, Anatoli Golitsyn, so far from dispelling the myths which obsessed Wright and the head of the CIA’s Counter-Intelligence Staff, Jim Angleton, encouraged their conspiracy theories.

  In January 1966, a month after Hollis’s retirement, a FLUENCY report concluded that the most likely penetration agent in the Security Service was either Hollis himself or a ‘middle grade spy’, to whom a lead had been given by the Polish defector Goleniewski.53 According to Peter Wright, the new DG, Furnival Jones, dismissed the suspicions against Hollis as ‘grotesque’, and instructed that the allegations about a ‘middle grade spy’ be investigated. The Director of C Branch (protective security), Michael Hanley, a future DG, seemed a possible fit. FJ questioned Hanley himself and quickly declared him innocent. Wright agreed.54 The next candidate for the ‘middle grade spy’, ‘Gregory Stevens’ (as Wright later referred to him), was less fortunate. As during his investigation of Hollis,55 and probably other cases, Wright instructed an A1 operations officer to break into the suspect’s desk, in this case to read ‘Stevens’s’ personal diaries. The interrogation was conducted by Wright and John Day. In addition to Wright’s published account in Spycatcher (which misidentifies Day),56 later unpublished recollections survive by both Day57 and ‘Stevens’.58 Despite some discrepancies on other points, all accounts agree that, under great strain after several days’ interrogation, ‘Stevens’, for reasons he could never later fully explain, made, then retracted, a bogus confession. He also accused his interrogators of being worse than the KGB. Soon afterwards, ‘Stevens’ left the Service. Though subsequently reinstated, he never fully recovered from the trauma of having his unblemished loyalty to the Security Service called into question. His position became even more invidious when news leaked out to his colleagues that he had been grilled by Wright, and he took early retirement.59 Wright’s own reputation within the Service never recovered. As he later acknowledged in Spycatcher: ‘There was talk of the Gestapo. Younger officers began to avoid me in the canteen. Casual conversation with many of my colleagues became a rarity.’60

  After the fiasco of the hunt for the ‘middle grade spy’, the FLUENCY Working Party once again focused its attention on Hollis. As well as recommending a full-scale investigation of Hollis (who in September 1967 was assigned the codename DRAT), the Working Party also decided to make a final attempt to resolve the case against Mitchell.61 By this time US intelligence, chiefly in the person of Jim Angleton, was also involved in the investigation. The point at which Angleton finally became lost in a conspiratorial wilderness of mirrors from which he was never to escape came in 1965 when Golitsyn persuaded him of the absurd proposition that the Sino-Soviet split, one of the turning points of the Cold War, was a mere charade designed to deceive the West. For Angleton to have believed that both the Soviet Union and the People’s Republic of China could have convincingly maintained over a series of years the pretence of a bitter quarrel between them to deceive the West is proof that his obsession with Soviet penetration and disinformation had led him to lose all sense of proportion. Golitsyn, however, dismissed all agents, defectors and other Soviet sources who disputed his deception theories as part of the KGB deception – and Angleton believed him. At the time of his defection Golitsyn had not claimed that the Sino-Soviet split was a fraud. During a visit to Washington late in 1965, Peter Wright ‘asked Angleton why he thought KAGO [Golitsyn] had held back so long, if he really thought all this when he came out. [Angleton] said that KAGO was afraid of being laughed at (as indeed he was).’62

  The extraordinary scale of the Soviet deception which Angleton believed had taken hold in the West was, in his view, proof that Western intelligence had been penetrated. He was anxious to be involved in the investigation of that penetration on both sides of the Atlantic.63 On the morning of 14 March 1966 A. M. MacDonald, who had succeeded Cumming as Director D, was rung up by Maurice Oldfield of SIS (of which he later became chief) with the ‘somewhat unexpected news that Jim [Angleton] and Anatol [Golitsyn] had turned up unheralded in London’ and were leaving next day. Their visit was so secret that Angleton insisted that even the CIA London station must not be informed of their presence. He asked for an urgent meeting with Sir Dick White, Furnival Jones and some of their senior officers, and said that he brought with him a:

  brief from [the DCI Richard] Helms to discuss with certain Liaison Services the problems of disinformation and penetration and to explore whether arrangements could be made both to step up investigations in the countries which they were visiting and to make arrangements for pooling, transmitting and distributing the results.

  The DG cancelled an appointment with the PUS at the Home Office, and went with MacDonald and Wright for a meeting with Angleton and Golitsyn at White’s flat. FJ and MacDonald agreed later that ‘the whole performance was somewhat extraordinary, but then Jim and Anatoly are quite extraordinary chaps.’ MacDonald noted afterwards:

  The D.G. and ‘C’ indicated very tactfully that while they accepted the facts of penetration and disinformation, they did not consider that it was therefore necessary to subscribe to the Sino-Soviet deception theory. I said that even if we agreed to differ on the Sino-Soviet split, this would not really invalidate the remainder of Jim’s proposals.

  After the meeting ended at 6.30 p.m., MacDonald and Wright accompanied Angleton and Golitsyn to Maurice Oldfield’s Marsham Street flat, where discussions continued until the early hours with a break for dinner. According to MacDonald, ‘Nothing very new emerged during these talks but Jim emphasised that he was convinced there was penetration of his own organisation . . .’64

  The willingness of non-conspiracy theorists in the Security Service such as MacDonald to collaborate with Angleton, despite rejecting his insistence that the Sino-Soviet split was a Soviet deception, reflected both their sense of the importance of the intelligence ‘Special Relationship’ and a personal regard for Angleton himself. Golitsyn too could be good company when he chose. MacDonald wrote after their visit: ‘I liked Jim and I was, of course, pleased to see Anatol again. He gave me a very warm welcome and I presented him with a book on wine, a subject in which I know he is interested, as a memento of the many good meals we had together in the past.’65 Even after Angleton entered the wilderness of mirrors, he remained personally popular with his British colleagues. As Peter Wright later recalled, ‘He drank us all under the table and played sharper poker and still sat up and argued politics hours after younger men had lost control and fallen asleep.’ He also had a well-deserved reputation as an orchid breeder.66 Angleton was a highly cultured man with an impressive range of literary interests. A BBC producer with a first-class honours degree in English literature who tried to arrange an interview with Angleton for a radio documentary in 1980 found himself being questioned in detail about twentieth-century English poetry for over half an hour on the telephone. ‘My God!’ the producer said afterwards. ‘That was a better viva than I ever had at Oxford!’67

/>   During the year following the surprise visit to London by Angleton and Golitsyn, talks in Washington between representatives of MI5, SIS, CIA and FBI, some attended by FJ, agreed on British–American collaboration in an investigation of Soviet disinformation and penetration by a small group of intelligence officers who exchanged highly classified information. Though J. Edgar Hoover rejected formal FBI membership of the group, the head of the Bureau’s Communist desk, Bill Sullivan, collaborated with it and attended its first conference in Washington in June 1967. A Security Service review two years later concluded that the collaboration had been kept so secret that the other intelligence allies – ASIO, RCMP and the New Zealand Security Service (NZSS)– were not, ‘as far as is known’, aware of its existence. All were, however, involved with the CIA, MI5 and SIS in a larger secret association with a similar agenda, codenamed CAZAB, which held its first meeting in Melbourne in November 1967.68 The meeting agreed to set up a joint counter-intelligence group (in association with the FBI) to seek to identify high-level Soviet penetration in the West, especially of intelligence agencies, and to uncover Soviet disinformation campaigns.69

  Angleton brought Golitsyn with him to Melbourne, where, according to Peter Wright, Golitsyn ‘laid special emphasis on Britain, and the many penetrations which, he claimed, were as yet undiscovered, and which only he could locate. FJ was smiling the smile he reserved for particularly tiresome people.’70 Security Service officers continued to consult Angleton personally about the search for traitors in MI5’s own ranks. In February 1968, D1/Inv wrote to the SLO in Washington, updating him on the Mitchell investigation and adding:

  I think it would be well worth discussing all this with Jim. This would, first, demonstrate that we are trying to get to grips with the case and bring it to a conclusion. Secondly it would enable you to ask for any help CIA can possibly give us in filling in the strange gap [in Mitchell’s career] between 1936 and 1938.71