The Defence of the Realm

by Christopher Andrew

  In September 1968 the second CAZAB meeting, this time in Washington, reviewed general progress in the search for high-level penetration of Western intelligence.72 In the case of the Security Service, since there was in reality no high-level Soviet penetration left to uncover, there was inevitably no significant progress to report. For the conspiracy theorists, however, the difficulty in finding proof of penetration was evidence of how successful it had been.

  The witch-hunt within the Security Service, though the details were known to few, made Peter Wright increasingly unpopular. At a conference of senior Security Service officers in 1969 at the Sunningdale Civil Service College, Wright was taken aback by the ‘bitter attacks’ on him and some of his colleagues in D1/Inv. ‘How do I go into the office’, he asked FJ, ‘facing this level of hostility?’ ‘That is a price you have to pay for sitting in judgment on people,’ FJ replied.73 The FLUENCY Working Party continued, however, to press the DG to authorize the investigation of Hollis. Though deeply sceptical of the case against his predecessor, FJ finally agreed in the summer of 1969. As the DRAT and PETERS cases proceeded,74 Angleton’s advice continued to be sought on both.75

  Stella Rimington later concluded that Wright and Angleton ‘fuelled each other’s paranoia’.76 Angleton, however, provided very little hard information for the FLUENCY investigations. A Security Service review of CAZAB concluded in 1969: ‘We have not received any intelligence which seems to match the bait produced by Angleton in earlier discussions, when he said that C.I.A. had information relevant to United Kingdom interests which they were unable to pass, in the absence of an indoctrination forum.’ The Service considered, however, that there had been a general gain in the closeness of liaison on counter-intelligence among the five intelligence allies:

  The . . . CAZAB networks, although having something of a precedent in the VENONA exchanges . . . are probably unique in conception, at least in peacetime. There is acceptance by the participating organisations of the threat from penetration and from disinformation and the professions of complete frankness appear to be reflected in the actual exchanges of intelligence.

  . . . The frank discussions on the threat, the exchanges of ideas on the methods to combat the threat have so far been valuable. Now that the proceedings of the Commonwealth Security Conference tend to be inhibited by the large number of delegates in attendance, it has been useful to include Australia, Canada and New Zealand in this intimate circle. It has resulted in them devoting more resources to work on disinformation and penetration.77

  In May 1970 FJ agreed to invite Golitsyn to Britain and give him access to material from Security Service files in order to identify the penetrations which he claimed to be able to uncover.78 The operation was run by an MI5 officer who had met Golitsyn while SLO in Washington. Helped by an assistant, the officer transported documents from Leconfield House for inspection by Golitsyn at a safe house on the south coast. Golitsyn was first booked into a hotel in Bournemouth but walked out, convinced he was being spied on by building workers. He was then put into a rented house near Christchurch where he was looked after by an A4 officer and his wife. His paranoid tendencies made him a nightmare to handle; he insisted on drinking only Perrier water for fear of being poisoned, and would go out only at night. After a while he asked to move into the family home of the former Washington SLO, where his behaviour alarmed both the MI5 officer’s daughter and one of their neighbours. The officer found the whole time-consuming operation a complete waste of time.79

  From the summer of 1969 to the summer of 1970 the Security Service DRAT and PETERS investigations proceeded in tandem.80 The Mitchell case was finally closed after he was questioned in FJ’s office in August 1970.81 No evidence of guilt emerged and the DG told Mitchell he had been cleared.82 John Day (K7), who took part in the questioning, later recalled that afterwards, presumably in a half-hearted attempt to make amends, the DG took Mitchell to lunch at his club. Day, who was also present, remembered the meal as, unsurprisingly, ‘rather a strain’.83 The investigation of Hollis, which eventually involved more than fifty interviews (including two with Hollis himself), also turned up no credible evidence since there was none to find.84 FJ’s personal embarrassment over this miserable episode was reflected in his failure to inform either the Home Secretary or the Prime Minister that Hollis was under investigation. By contrast, three ministers had been told about the investigation of Mitchell while it was still in its early stages.85

  ‘The Case against DRAT’ submitted on 18 June 1970, drafted by John Day, who was to lead the interrogation of Hollis early in the following year, unintentionally exposes how threadbare the evidence against Hollis was. It remains a shocking document – a classic example of a paper written to support a conclusion already arrived at which excludes important evidence to the contrary and turns on its head evidence which does not fit the preconceived conclusion. There was no mention of the fact that during the Second World War Hollis had been one of the members of the Service most alert to the threat of Soviet penetration of Whitehall and the armed services, even criticizing the DG, Sir David Petrie, for his lack of attention to the dangers which this posed. Hollis had also been almost the only – perhaps the only – wartime member of the Service suspicious about Blunt. Day, however, turned this evidence on its head as pointing to a KGB plot to divert suspicion from Hollis.86 Nor could the conspiracy theorists credibly explain how a series of highly successful operations against the KGB of which Hollis had knowledge – among them the defection of the Petrovs in 1954 – could have been possible if Hollis had been a Soviet spy.

  All three of the supposed leads pointing to Hollis’s guilt cited in the ‘Case against DRAT’ were later shown, as could reasonably have been inferred at the time, to refer to others. Hollis was alleged to be the most likely candidate for the ‘acting head of a department of British Counter-Espionage (or Counter-Intelligence) Directorate’ mentioned by Volkov in September 1945 before his failed attempt to defect. It should have been clear well before 1970 that by far the most likely candidate was Philby, who at the time had recently been acting head of SIS Section V (counterintelligence), whereas Hollis had been the substantive (not acting) head of F Division (counter-subversion) in MI5 for five years. (In addition Philby was known to have been responsible for ensuring the failure of Volkov’s attempt to defect.) Hollis was also identified as the most likely candidate for a ‘cadre worker’ (serving officer) in MI5 able to bring out files on Russians at will who was mentioned by the Soviet resident in Stockholm to the future defector Petrov in early 1954. It should have been obvious that the prime candidate was not Hollis but Blunt, who had confessed in 1965 that during the war he had taken MI5 files out of headquarters in a suitcase. Hollis was also alleged to be the agent codenamed JOHNSON mentioned in London VENONA decrypts of September 1945. As even Wright later acknowledged, this also was Blunt.87

  Despite his own scepticism about the ‘Case against DRAT’,88 FJ at last informed Sir Philip Allen, PUS Home Office, and Sir Burke Trend, the cabinet secretary, of the investigation of Hollis. Though pressed by Allen, the DG remarkably refused to tell either the Home Secretary, Reggie Maudling, or the Prime Minister, Edward Heath. Since Hollis had had three meetings with the last Conservative Prime Minister, Harold Macmillan, to discuss the investigation of Mitchell, it is difficult to see the justification for FJ’s refusal to inform Heath. No minister was informed of the investigation of Hollis until Harold Wilson was told in August 1975; he too had not been informed of the suspicions against Hollis during his first administration.89 In keeping with the terms of the CAZAB counterintelligence collaboration, FJ did, however, inform his US allies. During a visit to the DCI, Richard Helms, at CIA headquarters in November 1970:

  He said that he did not suspect DRAT of having been a spy. He added, however, that others in the Service did not share his view and that in the near future DRAT would be interviewed. Because he might be thought to be prejudiced, to which he admitted, he did not intend to conduct the interview pers
onally.90

  In 1970 the FLUENCY codeword was changed.91 (The new codename remains classified.) FJ finally authorized the interrogation of Hollis by Day and a colleague in K7 in February 1971, but gave Hollis advance notice and refused to seek HOWs for letter and telephone checks.92 As even Wright acknowledged, the interrogation of Hollis effectively brought the case to an end: ‘We knew we had not brought the case home.’ At the end of the interrogation, Hollis ‘said goodbye, and meant it. He travelled back to Somerset, back to his golf, and his cottage.’ FJ declared that the case was closed and that it was time to move on. Very few within the Service apart from Peter Wright disagreed. Wright, however, had lost most of his influence. He later wrote in his memoirs that, though he did not formally retire until 1976, his retirement really began on the day that Hollis’s interrogation ended: ‘What came after was mostly going through the motions.’93 Hollis died in 1973.

  While ‘going through the motions’, Wright’s conspiracy theories continued to grow. He convinced himself ‘that the Ring of Five stood at the centre of a series of other concentric rings, each pledged to silence, each anxious to protect its secrets from outsiders’.94 Stella Rimington and some of her friends in the Security Service called Wright ‘the “KGB illegal”, because, with his appearance and his lisp we could imagine that he was really a KGB officer himself, living under a false identity . . .’.95 After the investigation of Hollis and Mitchell was concluded, about one hundred leads pointing to possible other penetrations – half of them provided by Golitsyn – remained. By 1973 it was generally recognized within D Branch that only five of these leads merited investigation; only one was still thought worth pursuing three years later. Between 1973 and 1978, however, the leading SIS conspiracy theorist repeatedly complained that the investigations into Soviet penetration had been improperly conducted, on one occasion appealing directly to the Prime Minister, Harold Wilson. Though he did not succeed in seeing Wilson, he put his case to the cabinet secretary, Sir John Hunt. Partly in response to the SIS officer’s complaints, the previous cabinet secretary, Lord Trend, was asked to review the investigation, and reported in May 1975 that there was no evidence to show that either Hollis or Mitchell had ever been a Soviet agent.96 Wright, Martin and the SIS officer were unconvinced. Conspiracy theory of the kind contracted by all three is an incurable condition. What J. A. Allen (Director KX) wrote of the SIS officer was equally true of Wright and Martin: ‘Involvement in counter-espionage cases induces in some a form of paranoia which causes them to find only pusillanimity, or worse, in others who see through different eyes.’97 The damage done to the reputation of the Service in the eyes of the four Prime Ministers privy to the charges against Hollis or Mitchell or both – Macmillan, Wilson, Heath and Callaghan – was inevitably substantial. The charges against Hollis in particular became the main theme of Wright’s best-selling intelligence memoirs, which succeeded in spreading the author’s conspiracy theories around the world.

  After the publication during the 1980s of books by several authors, in particular Peter Wright’s Spycatcher (which appeared in 1987), which publicized the charges against Hollis and Mitchell, a number of Service reports reviewed the earlier investigations. Those reports represent probably the most damaging indictment ever produced within the Service of any of its investigations. An officer in K10R/1, which was responsible for research into old espionage cases, concluded in June 1988: ‘In our estimation the case originally against Hollis which culminated in two interviews with him was so insubstantial that it should not have been pursued.’

  It may well be asked why belief in high-level penetration continued so long in the UK, directed first at Mitchell . . . and later at Hollis. The reasons are complex:

  (a) a lack of intellectual rigour in some of the leading investigators . . .

  (b) dishonesty on the part of Wright, who did not scruple to invent evidence where none existed . . .

  (c) the baleful influence of Golitsyn who realised in 1963 that he had told all he knew and set about developing his theory of massive and co-ordinated Soviet deception (‘disinformation’) supported by high-level penetration of all western intelligence and security services98

  Among the instances of Wright’s dishonesty in fabricating or distorting evidence to fit his conspiracy theories which came to light after his retirement in 1976 were the following:

  (i) Wright has quoted an incident in 1961 when the KGB Resident Korovin was said to have been watching television [news] on the evening of the day when Lonsdale and the Krogers etc were arrested – and showed no emotion [the implication being that he had been forewarned]. This is quite untrue. Korovin was having a party with his two deputies and they were not watching television or listening to the radio.

  (ii) Wright ‘led’ his witnesses most unscrupulously in, for example, his interviews of the Petrovs, Elsa Bernaut99 and Blunt (in one interview he suggested to Blunt that Hollis was a spy, causing Blunt considerable anxiety for his own safety).

  (iii) His tendency to select a solution, then tailor the evidence to fit it (as in his 1965 investigation of the fire at the Moscow Embassy in 1964).

  (iv) His standard manoeuvre when worsted in argument of taking refuge in mystery (‘If you knew what I know’). This was a characteristic shared by Angleton in CIA. In both cases this was later shown to be a dishonest charade.100

  Though all these points were well taken, the 1988 and other investigations did not address the underlying strategic failures which gave such free rein to the witch-hunts against Hollis and Mitchell. Overawed by the discovery of the atom spies, the Cambridge spies and (to those with access to VENONA) the Soviet wartime penetration of OSS and much of the Roosevelt administration, the Service formed a greatly exaggerated view of the efficiency of the Stalinist intelligence system and paid far too little attention to defector evidence (from, for example, the Petrovs) on its numerous inefficiencies, thus making itself susceptible to conspiracy theories which took no account of the frequency of Soviet cock-ups. Secondly, the Service had no mechanism for challenging the passionately held but intellectually threadbare conspiracy theories of a disruptive minority – particularly one which was so willing to fall back on claims of superior information (‘If you knew what I know’) which it refused to expose to critical examination. Thirdly, the lack of such a mechanism was ultimately the result of a managerial failure – particularly by FJ, who, despite his deep scepticism about the DRAT investigation, failed to appoint a Team B from within the Service to review the deeply flawed evidence on which it was based. Fourthly, the introverted work culture of the Service made it deeply reluctant to call in an experienced, senior figure from Whitehall to provide a second opinion. But that, in the end, a decade too late, is what it was forced to do.

  There was thus one long-term gain from the decade of otherwise futile and disruptive FLUENCY investigations. In 1975, at the suggestion of the cabinet secretary, an independent ‘assessor’ was appointed to oversee all current and future investigations of alleged penetration of the Security Service or SIS. Having completed his inquiry into the Hollis and Mitchell cases, the first incumbent continued as assessor until his death in 1987. Successive incumbents have been appointed by the home secretary, with the agreement of the prime minister and the foreign secretary.101 Had such a system been in place at the beginning of the 1960s, it would doubtless have brought to an early conclusion the witch-hunts of the Service’s conspiracy theorists.

  11

  The Wilson Government 1964–1970: Security, Subversion and ‘Wiggery-Pokery’

  Labour returned to power in 1964 with many of the suspicions about the Security Service which had troubled it in 1945. Even among the Service’s supporters in the government there was a widespread delusion that their correspondence and telephone calls were intercepted. Tony Benn, whom Wilson made postmaster general, noted in his diary after a discussion with his future cabinet colleague Dick Crossman fifteen months before Labour’s election victory:

  Dick, who worked for
Intelligence during the war, is a fierce security man and said that, as a Minister, he would think it right that his phones should be tapped and all his letters opened. This is quite mad. I am terrified that George Wigg may be made Minister for Security and given power over all our lives.1

  Wigg, a former colonel in the Education Corps, told journalists in 1963 that he was out to ‘get Profumo’ (in his view a bad secretary of state for war as well as a security risk), and gave himself much of the credit for forcing Profumo out of office.2 Benn had good reason to be alarmed about Wigg’s future role in a Wilson government. As Wilson’s official biographer observes: ‘[Wigg’s] passion was secrets, the more malodorous the better. He was at his happiest in the twilight world of spies, counter-spies and Chapman Pincher, and viewed his fellow MPs with the same ferocious suspicion as he would have lavished on an accredited agent of the KGB.’3 What Wilson did not know was that, as well as claiming (with little justification) to be an expert in security matters, Wigg combined prurience about the sex lives of others with the use of prostitutes.4 Ironically, during the Attlee government Wigg had been on the list of ‘Lost Sheep’ pro-Soviet Labour MPs marked down for possible expulsion from the Party by its general secretary, Morgan Phillips.5 By 1964, at least as regards the Soviet Union, Wigg’s ‘Lost Sheep’ days were far behind him.

  On Saturday 19 October 1964, the day after Harold Wilson was sworn in as prime minister, the PUS at the Home Office, Sir Charles Cunningham, phoned the DG, Sir Roger Hollis, to tell him that the new Home Secretary would be Sir Frank Soskice but that Wilson proposed to transfer responsibility for security to George Wigg as paymaster general (in effect minister without portfolio). Hollis and Cunningham both saw ‘substantial objections’ to (in other words, were appalled by) Wilson’s plan, which would have thrown the whole Home Office Warrant system, on which the Security Service depended, into disarray: ‘It would be impossible for Colonel Wigg, who would not be a Secretary of State, to sign warrants.’6 In the course of the weekend, however, Wilson was persuaded to reconsider Wigg’s role – probably by the cabinet secretary, Sir Burke Trend, for whom he had enormous respect.7 On the morning of Monday 21 October, Trend summoned Hollis and Sir Dick White to his room in the Cabinet Office to inform them: