Tag Archives: ELLI

On Philby, Gouzenko, and ELLI

Igor Gouzenko

I return this month to the matter of the disclosure by the defector Igor Gouzenko of the existence of ELLI, the mystery spy within one of Britain’s intelligence agencies, and Kim Philby’s possible passing on of this information to his masters in Moscow – all occurring in late 1945. Gouzenko was a cipher-clerk in the Soviet Embassy in Ottawa, who dramatically escaped with sheafs of documents that incriminated a complex network of spies, and led, first and foremost, to the conviction of Alan Nunn May. The identity of ELLI has been one of the most absorbing of the ‘molehunt’ controversies of the past decades, and Chapman Pincher devoted a large portion of his later career to trying to prove that Roger Hollis, chief of MI5 from 1956 to 1963, was in fact the person behind the cryptonym. Peter Wright, the author of Spycatcher, was the main intelligence officer supporting this theory, which has developed a reputation that far exceeds the strength of the claims it makes.



1. The Vassiliev Notebooks

2. Odd Events in Canada

3. Menzies in Ottawa?

4. MI5’s Response

5. The News About ELLI

6. Philby’s Reactions

7. Liddell’s Reactions

8. Hollis’s Interview

9. SOE & Alley

10. Interim Conclusions


This topic is very complicated, and the archival material very fragmented. (I issue my customary health warning.) I believe the analysis calls for a very close attention to chronology and geography – part of the methodology sadly lacking in most of the literature I have read. I believe it is essential, however, that the proper groundwork be laid out in order for the inspection of the ‘ELLI’ story, as it evolved in the following decades, to be carried out properly. I thus restrict my study in this piece to the months of September to December 1945, when the ELLI hints were at their freshest, and shall pick up the subsequent interviews and examinations in episodes to come. Yet many unanswered questions remain.

My approach is as follows: I first discuss the unwitting assistance that Alexander Vassiliev’s annotations contributed to the (incorrect) notion that Kim Philby’s disclosures helped identify ELLI as a GRU asset. I then explore the situation in Canada at the time of Gouzenko’s defection that allowed MI6 – and Philby – to wrest control of the case away from MI5, and I explain why MI5 was so passive in its response, and describe the minor role in the ELLI investigation that the reputed villain Roger Hollis played. I move on to examining the way in which the few extant messages concerning ELLI were processed, and the difficult circumstances surrounding their interpretation, affected severely by Philby’s control of much of the material, and his extraordinary diversion to Istanbul at the peak of the investigation. I explore the hints that ELLI was an SOE * asset, describe the background to the relationship between the SOE and the NKVD, which leads to the way that the insight provoked Guy Liddell to search for possible wartime leakages, and some of his speculation as to who ELLI might be. That project appeared to be in full swing as the year wound down, and I draw some interim conclusions.

[* SOE, Special Operations Executive, was a sabotage organisation established in 1940. Its mission was in direct conflict with that of MI6, which was intelligence-gathering. MI6 and SIS are used interchangeably in this report.]

1. The Vassiliev Notebooks:

Alexander Vassiliev

I had discussed this topic last May, when I recognized the extraordinary sleuthing that William Tyrer had performed in winkling out further details about the interrogations of Gouzenko. Yet I detected some errors in Tyrer’s analysis, especially in his study of the KGB * reports concerning Philby, ELLI and Stalin. I had next attempted to make contact with Alexander Vassiliev, now domiciled in the United Kingdom, who had transcribed vital records in the KGB archives, but he had apparently not received my letter. I am happy to be able to report that I have now communicated with Vassiliev #, and want to clarify and expand my previous comments. I believe I raised some important questions, but I had not reflected accurately all the activity that was going on in September 1945. Something seemed incongruous to me at that time, but I had not placed my finger correctly on what it was.

[* The NKVD was the wartime name for what evolved into the NKGB, and the post-war KGB. For all intents and purposes, their names are interchangeable.]

[# Several weeks ago, a communicant overseas kindly gave me Vassiliev’s email address. I then immediately discovered that Vassiliev had in fact just posted this item on his own Wikipedia page.]

I shall, for the sake of clarity, repeat here some information that I have published beforehand. The first item of analysis is the famed reference to ELLI (actually ‘ELLY’) in the Vassiliev papers. These were transcripts of files created by Alexander Vassiliev from the KGB archives, containing information on the GRU, the Soviet Union’s military intelligence bureau, as well, and available on the Internet at https://digitalarchive.wilsoncenter.org/collection/86/vassiliev-notebooks .

Chapman Pincher presented the assertion that Gouzenko had betrayed the existence of ELLI in British intelligence as appearing in a report from Boris Merkulov, chief of the NKGB, to Stalin in November 1945, and William Tyrer echoed Pincher’s claim in his article about ELLI in the International Journal of Intelligence and Counter-Intelligence: see https://www.tandfonline.com/doi/abs/10.1080/08850607.2016.1177404) .

Yet the published Vassiliev letter states no such thing. It appears as follows (https://digitalarchive.wilsoncenter.org/document/112565) , first reproducing the salutation of Merkulov’s letter:

Comrade I. V. Stalin

Comrade V. M. Molotov

Comrade L. P. Beria

Vassiliev’s commentary next appears, in parentheses:

 [Summarizes the content of “S’s” message regarding Gouzenko and so forth. May is a Ph.D. in physics, a professor at Cambridge Univ., a GRU agent, information on atomic energy.] [Gouzenko reported on the GRU source in British intel. “Elly.”]

Inside Merkulov’s letter, the only direct citation of what Philby told his bosses runs as follows, (“S” standing for “STANLEY”, Philby’s cryptonym):

“Agent “S” reported: ‘In early November Bentley visited the Federal Bureau of Invest. (the FBI) and stated that World Tourists and the United States Service and Shipping Corporation were being used by Sov. intelligence for intel. work. What else Bentley told the FBI and which agents she knew were given up by her to the FBI, we don’t know yet. However, according to the information of agent “S”: “The FBI’s investigation of Golos’s network showed that his agents had penetrated deep into Amer. government agencies and the FBI believes that this network was controlled by the NKVD.’”

Thus the comment that “Gouzenko reported on the GRU source in British intel. ‘ELLY’” is not in the selected highlights of Merkulov’s report, but appears as an introduction in a separate pair of parentheses, looking as if it had been added by Vassiliev as editorial commentary, after the statement that informs us that what follows is a summarization of what Philby had told his Soviet handlers. If it is intended also to reflect the information received from ‘S’ that immediately precedes it, it is worth noting that the information attributed to Philby here likewise includes nothing about ELLI. Elizabeth Bentley, the subject of this report, did not defect until November 7, 1945, while Philby probably became aware of the existence of ELLI on September 13 (or soon after), and, as I shall explain, if Philby did pass on what he heard about ELLI, he would have done so almost two months earlier. The indication by Vassiliev that the letter ‘summarizes the content of “S’s” message regarding Gouzenko and so forth’is both vague and inaccurate.

Pincher also cites the comment as coming from Merkulov’s report, but uses the on-line version as his source. He is wrong. Tyrer reproduces the whole introduction in his article, but removes the parentheses. He is careless. Of course, it is very possible that Merkulov did write to Stalin about Gouzenko and ELLI, and that needs to be verified. Merkulov was, however, in the NKVD/KGB, not the GRU, and it seems implausible that he would want to lay any bad news concerning the GRU on Stalin’s plate. I cannot quickly see any other reference to the GRU in Merkulov’s communications, and Allen Weinstein and Vassiliev himself, in The Haunted Wood, suggest (note, p 105) that any reference to the GRU by Merkulov was an attempt to pass off some of the responsibility for Elizabeth Bentley’s defection to the GRU, who recruited her originally in 1936, and for whom she worked until 1938, when she was transferred to the NKVD.

Thus one might ask: if Vassiliev thought that the reference to ELLI was important enough to be highlighted, why did he not publish the original text that contained it? (I have checked the original Russian manuscript on the Wilson Center website: the texts are the same. Yet some pages are missing in all versions: the original scan of the manuscript, the Russian transcription, and the English translation). We should recall, also, that Vassiliev was not transcribing the texts surreptitiously: he had been given permission from the Association of Retired Intelligence Officers (KGB alumni) to inspect them, was well-briefed in western intelligence interests, and under no pressure. So that is why I decided to try to ask him what the significance of his commentary was.

Mr. Vassiliev kindly responded to my email, as follows:

Now, about the document. It looks like the phrase about ELLI comes from Merkulov’s letter. I used to write my comments on the margins of the pages.
There was an exchange of information between GRU and NKVD-KGB. I remember at least one document talking about someone from NKVD making enquiries in GRU. It doesn’t look like they were doing it every day due to the need-to-know principle. But in this case, before sending his letter to Stalin, Merkulov probably consulted GRU, or there was a constant exchange of information on the Gouzenko affair. And, as far as I understand, the initial info on ELLI came to NKGB from Philby

I had to re-inspect the Gouzenko documents from The National Archives, and wrote to Vassiliev:

The more I looked at this, it seemed to me that it would have been very predictable for Philby to pass on what Gouzenko said about ELLI, but that it would have been the first time he had heard the name, and he would have had no contact with him (or her – since Akhmedov said that the London ELLI was female.) But he must have passed on that nugget much earlier.

I am still intrigued by the Merkulov submission. It appears (as you say) to be a summarization or paraphrasing of what Philby reported, but it is very much in the native idiom of a KGB officer (‘Bentley told us’, ‘ ‘we believe’, the renegade Budenz’, etc.), and Philby is introduced or quoted as an aside (‘However, according to the information of agent ‘S’ . . .’).

But there are these timing issues. The letter from Merkulov that you cite is dated November 24, but the Kew Gouzenko files (and Guy Liddell’s Diary) tell us that the news about ELLI arrived on September 13, and VENONA informs us that Philby’s initial report on Gouzenko was confirmed as early as September 17. 

Thus there must be an earlier report that does not appear in your White Notebook. The November 24 missive is almost entirely consumed with the Bentley case, after Bentley’s statement to the FBI on November 7, so ELLI would have been old news by then.

I should also add Philby’s trip to Turkey on September 26, on the VOLKOV case. I had not entered that into my Chronology. He was obviously distracted for a while, and so was Merkulov.

My conclusion: that Philby or Merkulov mentioning ELLI towards the end of November would have been superfluous.”

Indeed, Keith Jeffery’s authorised history of MI6 appears to confirm Philby’s earlier communication on Gouzenko. On page 657, Jeffery writes: “A signal on 17 September from Moscow to Krötenschield, Philby’s controller in London, confirmed that information from ‘Stanley’ (Philby’s Soviet cover name) about ‘the events in Canada . . .  does correspond to the facts’.” This was clearly VENONA traffic, as can be confirmed from the archive. Yet would Philby have been aware of ELLI that soon? Probably not. A further message, dated September 18 (a Tuesday) refers to ‘a meeting last week’, which would put it, at the latest, as Friday, September 14. If Philby received the news on the Thursday, he would have had to arrange, at very short notice, a rendezvous with Krötenschield (also known as KROTOV), which might have been a difficult task to accomplish unless he had some very efficient  – but risky – intermediary working for him. Could Philby have been receiving information from another source –  as Peter Wright in fact suggested? And why, in any case, was Philby the master of ceremonies in this business? To answer those questions, I shall have to examine the Chronology very carefully. But first: Philby’s inappropriate control of the situation.

2. Odd Events in Canada:

As a Dominion, and part of the British Empire, Canada fell under MI5’s bailiwick when it came to intelligence matters, not MI6’s. Yet, by a strange mix of ill luck, inattention, lack of forcefulness, and sheer incompetence, or possibly by virtue of a highly secret project, MI5 allowed MI6 – and Philby, as head of the latter’s new Section IX Counter-Intelligence division – to hijack the direction of the response to Gouzenko’s defection. The official historians have been extraordinarily negligent in reporting this anomaly. In his Secret History of MI6, Keith Jeffery wrote (p 657): “Philby was the principal point of contact for MI5, who naturally had a direct interest in the case.”  Christopher Andrew, in Defend the Realm (p 346) avoids the issue, but enigmatically explains Hollis’s being sent to Ottawa in the following terms: “The fact that Gouzenko had defected in a Commonwealth capital, rather than foreign territory, meant that the Security Service, rather than SIS, had the lead role in responding to it”. Yet he studiously avoids discussing the fact that MI5 did not take a ‘lead role’. He subsequently ignores the strife until he describes Roger Hollis’s eventual complaints about Philby’s meddling on February 19, 1946. Gillian Bennett, former Chief Historian of the Foreign and Commonwealth Office, echoes Jeffery’s comment, rather lamely, and incorrectly, in her article The CORBY case: the defection of Igor Gouzenko, as follows: “Since British subjects were implicated in Gouzenko’s revelations, the case was of primary interest to the Security Service”.

Yet MI5 had more than ‘a direct interest’: it was primarily responsible for counter-espionage on Commonwealth territory. Its mission was, however, confounded by i) the absence of its regular representative in Ottawa; ii) the relationship between Canada’s Department of External Affairs and Britain’s High Commissioner and the Foreign Office; iii) the resourcefulness of Peter Dwyer, who represented MI6 (and secondarily, MI5) in Washington; iv) the energies and preferences of William Stephenson, the head of British Security Coordination in New York, and the highly secure communications links that he controlled, v) the inattention of various MI5 officers in London, and vi) the direct influence of Stewart Menzies, MI6 chief. Kim Philby was able to exploit all these factors.

MI5’s Representative:

MI5 had maintained a representative on Canadian soil in the person of Cyril Mills, an MI5 officer who had worked on the Double Cross team, handling GARBO. He had been sent to Ottawa in 1943 to manage Operation WATCHDOG, an attempt to turn a German spy into a double agent. Yet, as luck would have it, at the end of the war, he had been demobilized, and sailed from Canada at about the time that Gouzenko defected. Guy Liddell records his arrival in London on September 19. MI5 presumably had not planned to replace Mills with any other officer, thinking that, with the war over (and many of its personnel returning to civilian life), it could afford to retrench. It did not have its own representative in the United States at this time, for such matters as liaising with the FBI. Peter Dwyer represented both MI6 and MI5 until Dick Thistlethwaite was appointed in 1947.

Cyril & Bernard Mills

Cyril Mills had pointed out the deficiencies in RCMP intelligence to his bosses in London. Dean Beeby writes, in Cargo of Lies (p 195): “Since December 1942 he [Mills] had been a window on Canadian security for MI5 and MI6, and his reports were alarming. Canada’s intelligence services were in a desperate state, he warned, beginning with the RCMP (witness its clumsy handling of the Watchdog case) and extending through the three armed forces. Mills’ repeated warnings perhaps help explain the speed with which British intelligence officers arrived in Ottawa to ensure Gouzenko would be in capable hands.” Yet the facts show that they were not able to right the ship properly. Liddell had not taken Mills’s warnings seriously enough.

The Department of External Affairs:

Norman Robertson & Mackenzie King

Canada’s undersecretary for the Department was Norman Robertson, described by Amy Knight as ‘a close adviser’ to Prime Minister Mackenzie King. It was Robertson who, on the evening of September 6, brought the news that Gouzenko had presented himself to the Minister of Justice, and who suggested to an alarmed Mackenzie King, someone very anxious not to upset the Soviets, that the RCMP (Royal Canadian Mounted Police) offer Gouzenko protection. Jeffery reports that, on September 8, Robertson decided to cable Alexander Cadogan, the British Permanent Under-Secretary for Foreign Affairs, in a missive introduced as ‘strictly personal’. (Bennett echoes this version of the story, describing Cadogan as ‘Robertson’s counterpart’, which may have been strictly true, but was a relationship that did not pay homage to the protocols.) This message described Gouzenko, his defection, and the nature of the material he had exfiltrated, without identifying him. Yet the message was sent as originating in New York, and concluded, rather elliptically, that ‘the investigation is proceeding in consultation with Stephenson and F.B.I.’

This would all be highly irregular, if Robertson had not informed the High Commissioner for Canada, Malcolm Macdonald, and the FBI had been consulted before British intelligence were involved. Some historians have used this to suggest that Stephenson and an FBI representative were already in Canada, and the circumstances continue to be a subject of much controversy. The following day, however, Macdonald, writing ‘further to my telegram of September 8’, perhaps indicating that Robertson had admitted privately his error in protocol, and that he, Macdonald, had always been in charge, sent another cable to Cadogan, and was able to inform him that the cipher clerk worked for the GRU, and even to name the British atom scientist Alan Nunn May as an agent of Soviet military intelligence. It might not appear immediately obvious how or why this information was passed from Ottawa to New York so expeditiously: presumably the security of Foreign Office cables was not considered as secure as the New York-London channel, but an open telephone link between Ottawa and New York might have been regarded as safe As I shall show, however, BSC to London communications, which were used for all traffic, were able to take advantage of a highly convenient geographical location in Canada. In any case, the initial point of contact was Cadogan.

Peter Dwyer:

It was probably because of the vacuum in British intelligence in Canada that MI6’s representative in Washington, Peter Dwyer, was despatched to Ottawa. Yet even his involvement retains some measure of controversy. While Jeffery ignores Dwyer’s role completely, in Nigel West’s history of MI5, the author suggests that, owing to Mills’s departure, the news of Gouzenko’s arrest was sent to Dwyer in Washington (i.e., not New York), whereupon Dwyer immediately sent a message to his bosses in London, where it was routed to Philby. This appears not to have been the case, however.  Other sources indicate that Dwyer had promptly flown to Ottawa. William Tyrer (using David Stafford’s history of Camp X), claims that Dwyer flew to Ottawa ‘immediately’ after Gouzenko mentioned that ’British citizens were involved’ (thus on September 8 or 9, presumably), but that would mean that Robertson (or Macdonald) contacted Dwyer directly in Washington before informing anybody else, which sounds highly unlikely. Macdonald would have felt completely capable of handling the implications himself.

A far more mysterious picture emerges from Amy Knight’s How The Cold War Began, however, where she cites records that indicate that, on September 6, Robertson was conferring at his home with ‘an eminent officer of the British Secret Service’. Speculation centred on this identity referring to William Stephenson, or even Stewart Menzies. Knight more safely plumped for Peter Dwyer as the enigmatic figure. In her text, she refers to a telegram to London from Macdonald, dated September 10, which refers to one of Stephenson’s men ‘who has been here for the last three days and who knows all the facts’, identifying her source as a later item from September 25, in KV 2/145. Indeed, s.n. [serial number] 3A confirms this, but Knight initially leaves the question as to why Dwyer was in Ottawa on September 6 as simply a possible coincidence. (I shall investigate Knight’s story in more depth later in this piece.)

David Stafford’s Camp X reinforces the claim that Stephenson was already in Ottawa at the time of Gouzenko’s defection, and that he summoned Dwyer (from Washington) and his fellow officer Jean-Paul Evans (from New York) on Saturday September 7. They were then briefed by George Glazebrook, from the Canadian Department of External Affairs, in a highly clandestine manner at the Chateau Laurier Hotel in Ottawa.

I recommend that readers turn to Knight’s book for a discussion as to what might have brought Dwyer to the scene, and whether Gouzenko had had contact with British intelligence before he sought asylum. Since Gouzenko struggled initially on being heard, and only narrowly escaped being re-captured by Soviet Embassy officers, it raises all manner of questions as to why the RCMP was not better prepared to accept the fugitive. Yet, if Dwyer was indeed on the scene, it also poses questions about the prior involvement of MI6. (I shall explore these matters in the last part of this section, as well.)

William Stephenson:

William Stephenson

In any case, Stephenson took charge. He saw himself as a much more natural associate of MI6, whose interests he primarily represented at BSC during World War II, than of MI5. Bennett writes: “All the high level CORBY messages were transmitted through SIS channels, between Stephenson’s BSC headquarters in New York and the SIS Chief in London. Stephenson, a Canadian, determined to enhance his own position and provide valuable leverage in his dealings with Ottawa and Washington, insisted on this.” The implication, however, is that low level (detailed?) messages could be sent over other media, or outside Stephenson’s control. Indeed, a host of messages, for example the transcripts of the exchanges that Gouzenko brought with him, must have been too voluminous to be sent to New York (or wherever BSC’s transmission facility was situated: see below) for encryption and transmission, and were passed directly from Ottawa to London, presumably by air in the diplomatic bag.

The corollary of Stephenson’s action is how Cadogan responded to all this traffic. It surprises me that his first instinct was not to alert David Petrie, director-general of MI5, but to pass on the communications to Stewart Menzies, the MI6 chief, who reported nominally to the Foreign Office. Perhaps this again suggests that the Foreign Office and MI6 knew beforehand of something going on, and the outcome was that Menzies immediately delegated all responsibility to Kim Philby, his blue-eyed boy who headed Section IX. Andrew’s history contains no mention of BSC, Stephenson or even Cadogan in this strange breach of protocol.

3. Menzies in Ottawa?:

I believe I first read about the possibility of Stewart Menzies’s being in Ottawa at the time Gouzenko defected in Amy Knight’s book. Citing Canadian government archives, she wrote that, on the evening of September 6, Robertson was conferring at his home with an ‘eminent officer of the British Secret Service’. She briefly discussed the possibility that this could have been William Stephenson, but followed this up by quoting the diary of Mackenzie King, who very explicitly recorded, the same day: “The head of the British Secret Service arrived at the Seignory Club today. Robertson was going down to see him tonight.” The following day, King ‘noted that he had authorized Robertson to telephone Stephenson in New York’, and the day after he inserts the comment that ‘Robertson said that Stephenson and the FBI representatives would be here tonight’. Yet, despite this apparently unequivocal evidence, Knight rejected the notion that this mystery visitor could have been Stewart Menzies, concluding that ‘the top British intelligence officer in North America’, Peter Dwyer, was a more likely candidate. Yet even the plodding and unimaginative Mackenzie King would not have misidentified his visitor so poorly.

David Levy, in Stalin’s Man in Canada, is another author who has investigated these matters, and he even followed up the dossier that King had instructed Robertson to create, in order to verify (despite what King’s September 8 entry declared) whether the person was in fact Stephenson. In a later volume, Fred Rose and Igor Gouzenko: The Cold War begins, Levy backtracked somewhat, but then muddied the waters by asserting that Peter Dwyer (whom he incorrectly dubbed ‘an MI5 man’) was at that time ‘in faraway England’. Yet Levy appeared unconvinced that the personage could have been Menzies, and he may have been relying on what Keith Jeffery, the authorised historian of MI6, wrote to him, in an email in October 2010, that there was no record of Menzies travelling to North America ‘much before 1949’ – an odd way of formulating a response, it must be said. Thus Levy’s final judgment was to sit on the fence: “Could it have been something Robertson invented to get round the prime minister’s reluctance to hold onto Igor Gouzenko?”

Stewart Menzies

For some reason, both Knight and Levy, who cite John Bryden’s Best Kept Secret (1993) as a source for information on Menzies’s visit, dismiss what this Canadian journalist has written. I came to Bryden’s book late in the cycle, but it is quite a revelation. It informs us that Menzies was indeed a member of the Seigniory Club, ‘one of the most exclusive private resorts in North America’. Bryden provides evidence that a conference on HYDRA, the communications centre for BSC, which was situated at the nearby SOE training ground, Camp X, was about to be held, and that George Glazebrook, Canada’s security chief in Washington had written to his predecessor, Thomas Stone, on September 3, about the imminent meeting with Menzies. HYDRA had been a vital cog in the secret communications network of the Allies: as Bryden wrote: “These vast, overlapping networks were made possible by the direct telekrypton cable links between Ottawa, Washington, and London, backed by HYDRA, the British Security Coordination transmitter at Camp X, plus similar American and British transmitters in the Pacific.” Its future was to be discussed.

Thus not only is there a substantial reason for Menzies’s paying a highly secret visit to Ottawa, one can also understand how smoothly the secure communications between New York and London were able to be achieved. HYDRA was a powerful and flexible wireless receiver/transmitter that routed all confidential traffic between the Americas and Great Britain. Telekrypton (also known as Rockex) was specialized teletype equipment that enciphered and deciphered telegrams for MI6. BSC In New York used it to send messages to Berkeley Street in London, via Arlington, Virginia, and another telekrypton machine was located at Camp X, where an automated system for the transformation of wireless/teletype messages was created. (Hence the highly efficient exchange of information between New York, Ottawa and London.) In addition, Bryden writes that Peter Dwyer ‘sent his own reports back to London to Kim Philby rather than to BSC.’ (His source, however, is the not entirely reliable Peter Wright in Spycatcher.) In a note he adds, describing Stephenson’s part in the scheme: “He appears to have arrived on the scene several days later and then only to provide Dwyer with a telekrypton machine for secure communications with New York for onward transmission by BSC cable to London.” 

Yet the fact of an alternative conduit is confirmed by a remarkable entry in Liddell’s diary for September 11: “There is a serious [sic: ‘series] of telegrams running between Robertson of External Affairs and Cadogan, another between Security Coordination and SIS.” [my italics] Thus, if Dwyer were communicating privately with Philby, a whole bunch of messages must have existed that may never have seen the light of day, even though Liddell (and others, presumably) knew about their existence. And evidence exists that many messages were not only weeded from the archives (or not even submitted to them), as gaps in the telegram sequence numbers show, but were also concealed from MI5. Someone has annotated, on cable CXG832 from Menzies to Stephenson dated September 18, in reference to Menzies’s answers to Stephenson’s questions from CXG317 (not on file): ‘not available to MI5’.

(Parenthetically, an additional advantage of this set-up is that Gouzenko was taken into protective custody, and housed at Camp X. Thus, as he revealed information about Soviet code and cipher systems, it proved highly efficient for the passing on of such insights in a highly secure fashion to cryptanalysts in Arlington and to those at the Government Code and Cypher School, at Bletchley Park and at Berkeley Street.)

But how long was Menzies in Ottawa? Did he have an alibi? One consideration that should be entertained is the fact that Cadogan was the recipient of the first few messages, not Menzies himself. Robertson and Macdonald would have been advised by Menzies, if he had still been in Canada, that Cadogan was the appropriate addressee, as custodian of MI6 affairs until Menzies returned to the United Kingdom. And maybe Menzies did decide that he should hotfoot it back on one of the regular RAF flights that transported VIPs across the Atlantic at that time, so that he could take charge of matters from the correct location. On September 10, Macdonald is still contacting Cadogan.  On September 12, however, Stephenson is responding to a cable earlier that day from Menzies (CSS) himself, so the head of MI6 was by then back in his seat.

Yet another wrinkle in the affair occurs on September 10. In a telegram to Cadogan, jointly composed by Robertson and Macdonald, the latter write: “You will doubtless have seen telegram from Stephenson to ‘C’ reporting inter ALIA our present knowledge scientific side of espionage activities.” A handwritten annotation suggests that that message should be found at s.n. 6a, yet 6a contains a message from Menzies to Stephenson (CXG826), dated September 15, advising him of Hollis’s departure the next day, but referring to an earlier message of September 12 (CXG817), not on file, but to which Stephenson had replied the same day. As the interrupted sequence of telegram numbers shows, several messages have been weeded: perhaps some false information has been inserted. Maybe Stephenson sent a telegram to Menzies knowing that he would not yet be in his office to receive it. Moreover, it may be significant that, after September 14, messages between Cadogan and Macdonald were sent through Menzies, rather than directly to each other.

Liddell’s Diaries provide some clues as to Menzies’s whereabouts. He provides some fascinating entries about him in mid-September. Two occur on September 13 after he, Marriott and Philby draft a telegram, for Menzies’s approval, to be sent to Ottawa concerning Nunn May, recommending that the spy be allowed to leave Canada (so he could make his rendezvous with his Soviet controller in London). Curiously, Liddell adds the comment that Menzies agreed to the terms ‘over the phone’, which sounds a rather casual way of checking such an important document. The second reference runs as follows: “When I saw C the other day at the JIC he told me that it has been decided that he should be the co-ordinating authority between SIS and SOE.” Indeed, the entry for September 11 appears to confirm Menzies’s attendance at the meeting, but is couched, again, in extraordinary language: “C. who was present seemed to agree to our accepting responsibility for SIME [Security Intelligence Middle East].” Why on earth would Liddell bother to record that Menzies was actually ‘present’ at a meeting when he transcribed what he said? How could it be otherwise? And why did he not insert this conversation in his diary entry for September 11, rather than adding it as an unrelated item two days later?

Thus there remains a distinct possibility that Menzies did not return until September 12, and that Liddell and other senior officers in MI5 (and officials elsewhere) knew about his mission, and provided cover for him. If so, that would explain MI5’s collective lack of enterprise in the whole Gouzenko business, knowing that Menzies was intimately involved with the details of the case, and familiar with its cryptographic implications, and how it therefore let MI6 manage the more conventional aspects of it (e.g. the treatment of Nunn May) until it was too late. And all the highly secure Telekrypton processes could not keep the information out of the hands of the Soviets.

Perhaps Stephenson and Menzies were both in Ottawa already. There seems to be evidence of a deliberate attempt to muddy the waters, although the private testimony from the diary of the rather naïve Mackenzie King concerning the authorization to Robertson to telephone Stephenson in New York might be considered the most reliable factoid to work on. I leave it to more expert analysts to shed better light on this mystery.

4. MI5’s Response:

September 1945 was a fraught period for MI5 senior officers. Apart from the challenge of the exodus of many competent officers to civilian life, the charter of MI5 was coming under government inspection. The versatile civil servant Findlater Stewart was in the middle of another investigation, and his questioning occupied much of the time of Petrie, Liddell and others. Dick White was still in Germany, but was not immune to Stewart’s study, and he communicated with Liddell over the telephone on the implications of possible sharing of resources with MI6, and the latter’s growing ambitions.

During this period, the Director-General, Sir David Petrie, appears occasionally in Liddell’s Diaries. He was rather disillusioned, even demoralised, over Soviet infiltration by this time, as John Curry’s History suggests, and looking forward to retirement. No doubt he was preparing the ground for who his successor should be, but Liddell (head of B Division) did not get the nod: the announcement of the appointment of police officer Percy Sillitoe was made in November, in preparation for his arrival in early 1946. Liddell kept Petrie informed of the Gouzenko business, but Petrie did not react with authority. He was aware of the treachery of Alan Nunn May (whose identity was immediately revealed by Gouzenko’s disclosures), but rather curiously, on September 13, informed Liddell that he would rather ‘knock him off in Canada’ than bring him over to the United Kingdom. Certainly there is no evidence that Petrie protested to Cadogan, or Menzies, or even Attlee, that MI5 should be in control of the case.

Thus it was left to Liddell to handle affairs. Liddell, too cerebral, too contemplative, in his Diaries consistently betrayed his lack of drive by confiding in them what he ‘personally believed’ on controversial matters of policy, as if he did not have the courage of his convictions to try to persuade other persons and bodies of their correctness. He received the news of the Gouzenko defection with concern, but not alarm, and was certainly not provoked into trying to take charge. The situation was ‘murky’: he regrets that he no longer has any MI5 representative in Ottawa.  

Ironically, it is Philby himself who eventually presses Liddell to authorize Roger Hollis to travel to Canada to take over the case. Liddell wanted to send out Herbert Hart, since Hollis was on holiday, but Marriott and Philby were insistent that the mission belonged to the anti-communism expert. Petrie consented: Hollis returned from leave on September 14, left from Prestwick on September 16, and arrived in Montreal the following day, so he hardly had time to have his laundry done, let alone be briefed properly. Dick White did not return to the UK until mid-October, where he immediately jumped into the Findlater Stewart project. Moreover, White had wedding plans. He married the communist Jane Bellamy on November 28.

Yet Hollis was not the ideal choice. He managed F Division (‘Subversive Activities’), with Section F2 responsible for ‘Communism and Left-Wing Activities’. Hugh Shillito (F2B & F2C) had been the real expert (as the Sonia business showed), but in September 1945 he had left MI5, probably in frustration at the obstinacies of MI5’s senior management, and Hollis’s disparagement of his efforts. Jane Archer would have been an excellent candidate, but she still worked for MI6, and, as a woman, would have faced enormous challenges with the RCMP and Canadian intelligence. The situation reinforced the fact that the split of counter-espionage and domestic subversion that Petrie had introduced in 1941 had not worked well: he and Liddell were again discussing how to re-unify the functions while the Gouzenko business was going on. Liddell confided in his diary entry for September 26 that he ‘wanted to get a proper Russian section going as early as possible’, implicitly admitting the flaws in MI5’s Soviet counter-espionage set-up. Yet speed did not appear to be of the essence. On November 30, he recorded having a talk with Dick [White] and Hollis on the formation of a Russian section, ‘the necessity of which Roger is now convinced, subject to DG’s approval’. Since the section would be in Hollis’s F Division (as a subsequent December 6 discussion with Petrie disclosed), that was a strange reaction on Hollis’s part. The nucleus would be Marriott and Serpell. Liddell, as the senior officer responsible for counter-espionage, should have travelled to Canada himself.

So Hollis was rushed off to Canada. On September 19, when Liddell contacted Hollis chez the RCMP, reporting on Nunn May’s arrival, he did implore Hollis to establish direct communication through that channel, even requesting ‘discontinuance of use of any other channel’, as the present situation was ‘unsatisfactory and causes delay and uncertainty’. The phrase about ‘discontinuance’ has been crossed out, however. It was too little, too late, and poor Hollis did not have the clout to remedy the situation, as he implied in a telegram to White and Liddell on the same day. Thus, on September 20, Hollis’s messages were still being sent to Menzies, under Stephenson’s codename (48000). Hollis wrote (September 24) that he and Stephenson assumed that all ‘CORBY’ (the codename for Gouzenko) messages were being sent to MI5, and, if that were not the case, offered to copy them all, and re-send them, adding, however, ‘but this seems ridiculous waste of time and effort justifiable only if you are meeting insoluble obstruction in London’. This message was sent by One-Time Pad from the RCMP to MI5, so maybe MI6 never saw it.

MI5’s overall response was passive and weak, and it appears that that behaviour led to the agency’s experiencing ‘insoluble obstruction’. Whether this was simply in character, or whether it occurred because of other arcane knowledge is a debatable subject, but Liddell’s private observations are ambiguous. One important diary entry for September 25 is worth quoting in full, however, since it shows that Liddell was more comfortable confiding his grievances to his journal than he was about remonstrating in the right places:

            “I had a talk with Marriott & xxxxxx [Philby?]. Later on Marriott brought over a file of telegrams some of which were dated the 22nd Sept. and on which action was required. This is a typical example of inefficiency and the kind of thing that results when two offices are handling the same subject. I said that I did not wish to upset Stephenson or make Roger Hollis’s task more difficulty [sic] but that quite frankly I could see no possible reason why the Security Coordination should be having a finger in the pie at all. The matter was purely one between ourselves, RMCP and the FBI. If we wanted guidance on matters of higher policy we could get it ourselves from the F.O. In fact we had already done so. Stephenson is apparently kicking up the idea of our communicating direct with the RCMP and cites the British High Commissioner as supporting his view. This of course is typical of Stephenson. He came into the case through External Affairs and having set himself up as the Great Panjandrum does not now wish to be knocked off his perch. Everything that he does or does not do is a matter of personal prestige and the organisation has to suffer accordingly.”

It sounds, however, that Liddell was not aware of the possible presence of Stephenson in Canada when the scandal erupted. It is surprising that he complains here about BSC’s interference, but not about that of MI6.

5. The News About ‘ELLI’:

While the primary focus in the flurry of messages that week concerned Nunn May (and other agents, including, rather confusingly, another Ottawa-based agent named ‘ELLI’, namely Kate Willsher), the existence of an agent in London surfaced from what Gouzenko revealed to his RCMP interrogators. William Tyrer has pointed out that the first reference to ELLI seems to be September 13, since Liddell responded, on September 23, to a telegram of that date in the following terms: “Ref. your CKG 301 of 13.9.45 – do not consider that ELLI could be identical with UREN.” Tyrer points out that CKG 301 is missing from the Gouzenko file, and that its succeeding items (303 & 303) have had information on ELLI redacted. [Unfortunately, Tyrer provides a source for this item as s.n. 27A of KV 2/1425, when it is in fact to be found in KV/ 2/1421. As he rightly points out, the Gouzenko files are ‘a shambles’. They need someone to compile a register of them, tabulated by number and source, so that a proper assessment of the chronology could be more easily gained.]

Tyrer makes two rather problematic assertions in this section of his analysis. The first is that ‘the existence of ELLI would have been telegrammed to London’ at the same time that the activities of Nunn May were described (i.e. September 10). Yet there is no evidence that ELLI was mentioned at that time: that is pure speculation. Moreover, Tyrer then claims that the fact that ‘MI5 in London knew about ELLI on or before 13 September’ is indicated by Liddell’s telegram responding to the message of September 13. How MI5 could have learned the contents of a cable before it was even sent is not explained by Tyrer, and his account ignores the perennial delays that were occurring between thought and reception at this time. It is true that Liddell first saw ‘CORBY’ telegrams on September 11 (since he records Kim Philby’s bringing them over with him), but he regards them as ‘somewhat corrupt’, and his lengthy diary analysis concerns itself solely with Nunn May and the latter’s prospective meeting with his handler in London. There is no mention of ELLI. Nor is there any when he discusses the case with Marriott and Philby two days later.

Yet researchers are indebted to Tyrer for finding another important text in the Canadian National Archives that corresponds approximately to the timing of the dispatch of the ‘UREN’ message. It was dated September 15, and Tyrer reproduces it as follows:

            Alleged Agent in British Intelligence

            CORBY states that while he was in the Central Code Section in 1942 or 1943, he heard about a Soviet Agent, in England, allegedly a member of the British Intelligence Service. This agent, who, was of Russian descent, had reported that the British had a very important agent of their own in the Soviet Union, who was apparently being run by someone in Moscow. The latter refused to disclose his agent’s identity even to his headquarters in London.

            When this message arrived it was received by a Lt. Colonel POLAKOVA who, in view of its importance got in touch with STALIN himself by telephone.

Now this text raises some provocative questions. Who interrogated Gouzenko? Is what he told his interrogator the same message as was sent to Liddell on September 13? Did it go to both Liddell and Philby? What is the significance of the references to ‘British Intelligence Service’ and British agents in Moscow? Why was the hint not picked up with more urgency? If someone in Ottawa (presumably Dwyer) had two days earlier already made a link between ELLI and Uren, surely he must have been acting on more specific pointers to SOE and Uren, who was working for SOE when he was convicted of spying? Yet the most important conclusion to be drawn from this message is that a spy within the service had revealed to Chichaev (the NKVD-SOE representative in London )in 1942 or 1943 that George Hill (the SOE-NKVD representative in Moscow) maintained an agent in Moscow, and that, even though Hill’s bosses had requested that Hill identify him, Hill had refused. Yet Gouzenko does not name the agent as ELLI here.

The reference to POLAKOVA is highly significant, however. POLAKOVA – sometimes POLYARKOVA – was a major in the NKVD (with a GRU background) who instructed PICKAXE agents at the school at Kushnarenko, outside Moscow. * PICKAXE was the project shared by the NKVD and SOE for sending Soviet-trained subversive agents from UK soil into Nazi-occupied Europe (see below). If POLAKOVA received the message, it confirms that the informer was attached to SOE in some way.

[* Maria POLIAKOVA – known as ‘VERA’ – was a significant figure in Soviet espionage. She set up the Swiss section of the Rote Kapelle in 1937-38, handing over to SONIA. When Allan Foote was sent on his final mission (before ‘defecting’ back to the British), it was ‘VERA’ who gave him instructions, and it was Foote who informed MI5 of her identity. In 1945, therefore, the name would probably not have meant anything to Liddell & co.. She was presumably on loan to the NKVD for training of Soviet agents for SOE, and stood in for Ossipov, Hill’s opposite number in Moscow, when the latter was travelling. Why she would have been identified as masculine is puzzling.]

Tyrer assumes that this message is serial 2a in the Gouzenko file, noted, on page 30 of the report in KV 2/1420, as being extracted for placement in ELLI’s Personal File 66962, but I am unconvinced that we can rely on this.

i) First, an examination of the response indicates that it was sent by Liddell ‘for HOLLIS’, responding to CXG 323 of September 16, and Item 4 is the line that runs ‘Reference your CXG 301 of 13.10.45’. Yet Hollis did not arrive in Montreal until September 17, then moving on to Ottawa. The telegram is addressed to R.C.M.P, for Hollis’s attention. Thus someone with, or attached to, the RCMP sent the original.

ii) The message has been sent by use of the One-Time-Pad over MI5’s traditional link, and it has been annotated ‘Copy sent to SIS’, suggesting strongly that it was not sent directly to SIS for transmission, but that SIS was kept informed. Again, it indicates a more private correspondence between MI5 and the RCMP.

iii) The Canada-based representative, if coming to a conclusion that ELLI might be UREN, reveals a familiarity with British intelligence, if he came to the conclusion that the meagre hints provided by Gouzenko pointed to SOE, but he appears not to be addressing the substance of the September 15 message. Ormond UREN was an officer in SOE, of Scottish-Cornish background, who had been convicted in 1943 of passing secrets to ‘Dave’ Springhall of the CPGB. There was nothing ‘Russian’ in his background, and he would not have known about any SIS or SOE agents in the Soviet Union. Likewise, if Liddell had seen the message of September 15 at this time, he might have pointed out the obvious anomaly. Why, on September 23, would he not have referred to the September 15 information unless the reason was that it had not yet reached his desk?

The conclusion must be that a simpler statement, probably hinting at SOE’s wartime relationship with the NKVD, and perhaps the role of the Soviet military attaché, must have provoked the ‘UREN’ analysis. The sender specifically selected as a probable candidate an SOE officer whose espionage was known. Moreover, Liddell knew more than the Ottawa communicant did in order to be able to discount Uren, but had almost certainly not yet seen the September 15 message.

The CXG series of messages were sent care of BSC in New York to SIS, probably originated from Dwyer, and thus would have arrived on Philby’s desk first. It is highly unlikely, however, that Dwyer interviewed Gouzenko directly. In Molehunt (p 37), Nigel West claims that Gouzenko ‘made allegations . .  . to Peter Dwyer’ about ‘a valuable Soviet spy inside British counterintelligence’, and that his assertions were later ‘reexamined in extraordinary detail’, but later (p 75) West states that Dwyer and Jean-Paul Evans flew in to Ottawa, but ‘neither of them ever actually met Gouzenko face to face’. West’s account suggests that John Leopold translated their questions and then reported Gouzenko’s answers (with Mervyn Black some time later assuming Leopold’s role as translator). West has the substance of this message about ELLI’s background surfacing only in 1981, thus confirming the existence of the withheld document in the Canadian National Archives. Gouzenko then gave an explanation to the Times, but for some reason changed the notion of ‘British Intelligence’ to MI5.

Peter Wright is also unreliable.  He reports in Spycatcher (p 281) that, in 1965, he went over the Gouzenko transcripts again, and also describes the defector’s testimony as referring to what his co-worker in Moscow, Liubimov, told him, that there was ‘something Russian’ about ELLI, the use of a dubok, and the fact that the spy could ‘remove from MI5 the files which dealt with Russians in London’. Yet these two last ‘facts’ do not appear in the September 13 telegram. Liubimov was not named there. Wright then writes that Liubimov ‘showed him [Gouzenko] parts of the telegrams from the spy’, which cannot strictly have been accurate, as the information from ELLI would have been packaged by his handler, and not sent by ELLI himself (or herself).

It is presumably this telegram that Wright and Pincher refer to as the ‘Elligram’ (see Pincher, pp 205-206, & Wright, p 188), although Wright’s recall of it appears to draw from the September 15 telegram as well as new information appearing in Hollis’s message of November 23 (see below). Thus Pincher’s claim that Dwyer ‘quickly’ sent a fuller telegram ‘containing all the details about ELLI’ must be questioned. But then Pincher is wildly off the mark. He has Hollis at the centre of things when the Gouzenko story breaks, with his friend Philby conversing regularly with him. “By September 10, Hollis had known most of Gouzenko’s revelations”, he writes, next indulging in vague speculation about Hollis’ negligence in not taking the warnings about Nunn May’s rendezvous seriously enough, and his sorry attempts to divert suspicion from himself. Yet Pincher overlooks the fact that Hollis did not return from his holidays until September 14. He was completely out of the picture. Pincher’s account is pure fantasy.

The more careful Amy Knight also badly misrepresents Hollis’s involvement. She declares (p 137) that ‘Gouzenko’s information about “ELLI” was first conveyed during his interview with MI5’s Roger Hollis (with the RCMP present), who visited Gouzenko shortly after the defection’. Yet we know that references to ELLI appeared before Hollis’s return from holiday, and that he did not meet Gouzenko until late November (an encounter that Knight describes as his ‘second meeting’). She does, however, bolster the fact of the confused messages by citing papers from the Canadian Security Intelligence Service that indicated that Gouzenko could have been referring to two separate agents in his depositions.

What seems conclusive, however, is that several messages about ELLI were sent from Ottawa to Philby, that not all of them reached MI5, and those that were forwarded often were subject to delay. What is also critical to note, however, is that, in September, the only published observation about ELLI that Liddell makes is on the ‘UREN’ one of September 13. Yet he does not mention this observation in his Diaries (there is no obvious redaction here), and he does not react to the content of the more detailed ‘Russian connection’ message of September 15 until over a month later. I shall analyse the phenomenon later in this piece.

6. Philby’s Reactions:

Kim Philby

One might suppose that what Gouzenko had to say about ELLI could have been contained in one statement, but it appeared that it came out in dribs and drabs. One might also conclude that Philby saw every message that was issued from Ottawa about ELLI, but we cannot assume that either. As early as September 11 (namely before the first identifiable message on ELLI), Philby was preparing a report for his boss, Menzies, who, in turn, had to give Prime Minister Attlee a briefing on September 13. This was the same day that Philby introduced the case to Liddell. Jeffery quotes from his report, using an unnumbered archive in the Foreign Office papers, but makes no reference to ‘ELLI’. At that stage, Philby was probably relying on the fairly high-level report from Robertson that did, nevertheless, contain details about Nunn May. It was in a covering note to this report where he recommended that an expert in Soviet espionage be sent out. [Jeffery rather ambiguously writes: “He suggested Jane Archer or Roger Hollis from MI5”, where the syntax is unclear. Archer was still working for Philby at this time.] Philby was presumably not then aware that Hollis was on holiday, as Liddell pointed out to him on September 14, unless, of course, his recommendation was made out of devilry, in the knowledge that Hollis had been out of the picture, and would thus not be a very efficient investigator.

Thus it is difficult to determine exactly what ‘facts’ Philby passed on to Moscow to allow a confirmation of his findings by September 17. Had hints to ELLI alarmed him, or was he merely passing on the threat from the initial Nunn May revelations? In The Philby Files (p 239), Genrikh Borovik quoted a report from September that included the following: “Stanley was a bit agitated himself. I tried to calm him down. Stanley said that in connection with this he may have information of extreme urgency to pass to us. Therefore Stanley asks for another meeting in a few days. I refused a meeting, but I did allow him to pass urgent and important material through Hicks [Burgess].” Given that September 13, when the longer telegram from Canada was composed, it seems highly unlikely that it would been received, decrypted, and sent to Philby in time for him to internalize it, and arrange a meeting whereby Krotov could have likewise composed and sent a telegram so expeditiously. Indeed, Amy Knight states with confidence that this was a separate message from later in the month. Yet, if Philby were ‘agitated’, it might have been because of the ‘UREN’ message, since Philby had a strong link with SOE, having set up with Burgess its training programme at Brickendonbury Manor.  He worked there for George Hill [see later], who had established the Russian section of SOE, after working in MI6’s D Section. Hill was sent out to Moscow as SOE’s representative to the NKVD in October 1941. The September 13 telegram, however, with its ‘Russian’ link might have been interpreted as drawing attention to someone else.

Tyrer (representing Wright’s opinions) also states that, after Philby received a telegram from Dwyer (that of September 13?), his communication, via Krotov, must have mentioned ELLI, and that when the KGB checked with the GRU to confirm what Philby had passed on, and sent the confirmation message on November 17, the existence of ELLI as a GRU spy was the subject of their response. The text runs as follows:

“The Chiefs have given their consent to the confirmation of the accuracy of your telegram concerning STANLEY’s data about the events in CANADA in the neighbours’ sphere of activity. STANLEY’’s information does correspond to the facts.”

There is no mention of ELLI: the facts may merely have described Nunn May and the associated network only. As I set out earlier in this piece, it would have stretched the limits of time and space for a message from Dwyer on ELLI to be created, encrypted, transmitted, decrypted, distributed, and analysed by Philby, and then a meeting set up with Krotov, whereafter a similar bureaucratic procedure occurred between London and Moscow, the KGB then checking with the GRU, gaining approval (presumably from Stalin), and lastly compiling its response for encryption, transmission and decryption  – all in the space of five days (September 13-17), with a weekend in the middle, and across multiple time zones!

By similar analysis, since he did not reply to the ‘UREN’ suggestion until September 23, Liddell might have known nothing about ELLI at the time Hollis left (September 16). Of course, Hollis might have quizzed Dwyer on his arrival in Ottawa, but his surviving messages circle exclusively around the Nunn May business, and its considerable political implications. Hollis assuredly did not have a chance of seeing Gouzenko himself during this visit, and he had returned to London by September 28. And it was during the last week of September that Liddell noticed that a thick file of telegrams on Gouzenko was not being processed in a timely fashion.

And then, astonishingly, Philby was taken out of the picture for a while. On September 19, he learned that a potential defector, Konstantin Volkov, had contacted the British Embassy in Istanbul, Turkey, and had far more damaging stories to tell than Gouzenko could ever have imagined. Philby immediately informed Krotov, was thus consumed with the case, and eventually flew out to Istanbul on September 26, the day after a highly sedated Volkov (and his wife) had been abducted by the KGB for torture and execution in Moscow. Philby returned to London on October 1. Whether Philby delegated any of the Gouzenko work to Archer is not clear: Jeffery does not even discuss the matter. The impression that Liddell gives, however, is that there was a breakdown in communications.

The outstanding conclusion from Philby’s involvement here is that there is no evidence that Moscow Centre confirmed ELLI’s existence as a GRU asset. That suggestion appears to have been inserted by Vassiliev, on the basis that, since Gouzenko worked for the GRU, then ELLI must have likewise have done so.

7. Liddell’s Reactions:

Guy Liddell

For a while, the dribble of information on ELLI not unaccountably dries up. The Gouzenko archive (KV 1421, s.n. 35a) then shows a cryptic and incomplete reference, dated October 16, in Telegram No 533, sent with some urgency (‘MOST IMMEDIATE’). Its text runs as follows:

A. CORBY states that cover name for ?all foreign ?intelligence or counter espionage services is ZILONE repeat ZILONE meaning green in Russian.

B. Agent referred to by CORBY in 534 was referred to as working in ZILONE.

Handwritten annotations indicate that a copy of this message has been placed in the ELLI file (s.n. 4a).

Again, this is enigmatic. First of all, the telegram number precedes the item that it refers to. Second, there is no record of No. 534 in the file. Third, the construction ‘ZILONE’ is rather inaccurately formulated. The Russian word for ‘green’ would more properly be transcribed as ‘ZELYONNY’, which makes one question the Russian – and maybe the English – credentials of the interpreter. Stafford questions the ability of both men who ran the RCMP’s intelligence network: its head was Inspector Charles Rivett-Carnac, and his assistant was one John Leopold, who was the only direct contact with Gouzenko, Unfortunately, Leopold may not have been very accomplished. Stafford writes: “But the Czech-born Leopold knew little about Soviet intelligence, and his Russian was far from perfect.” That may explain some of the early misunderstandings over what Gouzenko said. Moreover, Gouzenko suspected that Leopold was a Soviet agent, and thus may have been reticent to open his mind while Leopold was the only translator. At some stage, Mervyn Black was brought in to help with translation, but Amy Knight does not state when this occurred.

So why was the revelation that the agent that Gouzenko had identified worked in counterintelligence suddenly that urgent? Had that fact not been communicated in September? ZILONE could presumably refer to either MI5 or MI6 – but also to SOE, since the Soviets made no distinction between SOE and MI6, which may have been significant. It might seem that someone in London had raised a question, and that Gouzenko wanted to clarify that his ‘Central Code Section’ handled traffic from all British intelligence services.

In any case, further messages start to appear. On October 24, Liddell reports in his Diary that John Marriott has brought more messages over, including ‘a further telegram about the agent known as ELLI who is alleged to hold some high position in British Intelligence’. (This is the first reference to ELLI in the unredacted part of Liddell’s Diaries.) Tyrer reproduces the text, but suggests that these telegrams were sent during Hollis’s second visit to Canada. This cannot be true, since, a week earlier, Liddell had written that Hollis personally brought him in another telegram from Canada (which was not ELLI-related), and the two of them had visited Petrie on October 18 to discuss the case. It sounds as if Liddell is describing the infamous Telegram 534, as he cites the claims that ELLI was working for British Counter-Intelligence, with the now notorious reference to ‘5’, which, especially now that we know of Leopold’s deficiencies, are highly ambiguous. “As CORBY’s theories are only based on scraps of information picked up here and there there is not much to work on,” he wrote, continuing: “It is possible that in mentioning the figure 5 he is referring to the five people who formerly signed JIC reports”, and he goes on to suggest that, as with the KING case, ‘it does not follow that because information is high-grade it comes from a high-grade officer’.

Hollis in fact sailed out of Southampton for Halifax, Canada, via New York, on October 22, and, according to Liddell, was ‘still there’ on October 30, although, with a five-day cruise, and an overland journey to Ottawa, Hollis could not have arrived until October 28, at the earliest. The next incident occurs on November 5, when Marriott shows Liddell ‘some recent telegrams’ on the subject of ELLI. As did Tyrer, I quote the text of Liddell’s diary entry in full:

“CORBY has been re-interrogated and refers to an incident when the Soviet M.A. [Military Attaché] in London referred to information that he had received from ELLI relating to a British agent in Russia. As the only organisations that can possibly have been running a British agent in Russia are SIS, SOE or the British Military Mission, it seems unlikely that ELLI could have any connection with ourselves. Nobody in fact knows anything about any agent in Russia. I should doubt very much whether there was one. The above does not necessarily throw any doubts on the bona fides of CORBY who may have got the story wrong.”

We should note that no mention of Hollis appears in this Diary entry: the ‘recently’ is irritatingly vague, so it may have been coincidental, or even antedated Hollis’s arrival. Tyrer categorises Liddell’s comments as ‘perplexing’, since Gouzenko had reported this information earlier (the September 13 telegram), but it would more probably indicate that Liddell had not seen that original telegram, or even that what he referred to was indeed exactly the same text, unaccountably held over for a month, and explained away by Philby’s absence in Turkey. Yet it is a very important reference, because it introduces the role of the ‘Military Attaché’, and thus partially explains M5’s lack of enthusiasm for an aggressive follow-up, as well as serving to prompt some personal reflections by Liddell himself.

Again, it is possible that some information was garbled. When Hill made a visit to London in October, he informed Liddell and White that he had been subject to provocation in Moscow, when the NKVD tried to set him up by sending him a man who had worked for him in 1920 (see Liddell’s October 5 diary entry). Despite Hill’s complaint, and the man’s being removed from the National Hotel, he made another attempt, to Hill’s exceeding annoyance. Thus both the time and circumstances of Hill’s ‘agent-running’ may have become distorted and misrepresented – a confusion over the pluperfect tense, perhaps: ‘ran’ versus ‘had run’? (There is no pluperfect tense in Russian.) Might ELLI have informed Chichaev that Hill had once run an agent in Moscow, after which Chichaev told Moscow Centre that Hill ran an agent there?

On the other hand, Hollis was still profoundly occupied with the political ramifications of the Gouzenko case. He was moving in exalted circles. On November 9, Liddell wrote: “Roger is to meet the PM, the President and Mackenize King in Washington, if required”, with rather shocking discussions scheduled on the atomic bomb, ‘and its handing over to the Russians or to the Security Council’. Meanwhile, Liddell was still focussed on the SOE connection, and the possibility of leaks in Moscow. He met with Archie Boyle (who had been Director of Security for SOE) on November 16, to discuss the ELLI case, and SOE’s set-up in Russia, where the highly dubious George Hill had been sent as chief SOE representative in October 1941. Quite a long entry appears in the Diary, in which Boyle is recorded as expressing ‘his grave suspicions about George Hill, and also about one George Graham whose real name is Serge LEONTIEFF, a White Russian.’ “The two are very close and one always backs up the other. Archie says he cannot understand how a man like Hill can possibly be acceptable to the Russians unless they are getting some sort [of] quid pro quo, the more so since they banished his mistress to Siberia and then brought her back after a certain delay.” (These comments echo what Liddell had written in his diaries about Hill back in 1943.) Hill was now on the Control Commission, and had recently told Boyle that he was about to make a private visit to Russia.

George Hill

At least Liddell started to dig more deeply. “I am getting the personal files for all the representatives of the SOE mission. Neither Hill nor Graham of course really fit the bill since the only concrete piece of evidence by CORBY is that he deciphered two telegrams indicating that ELLI was in London and worked through the Soviet M.A.” Yet George Hill, despite his dubious past, had been approved as SOE’s representative in Moscow in September 1941, and in 1942 and 1943 (the years that Gouzenko had referred to), was responsible for the Soviet end of the collaboration, in which Britain and the Soviets were supposed to cooperate in planting Soviet saboteurs in Nazi-controlled Europe. In Stalin’s world, of course, all foreigners were considered ‘spies’. But Hill might seriously have been assisting the NKVD. In his memoir From the Red Army to SOE, Len Manderstam described Hill as a ‘triple agent’, and accused him of supplying ‘a great deal of important information’ to the NKVD. Hill wrote a shameful piece of Stalinist propaganda in favour of the control of eastern Europe in his final report from Moscow when his position was wrapped up. Moreover, there should have been some controversy over ‘George Graham’ of the Intelligence Corps, whom Hill declared he had selected as his aide, and who accompanied him to Moscow. It seems that Leontieff had taken British nationality in 1933, but, as a White Russian, he would have been treated with utter scepticism by the Soviets – unless they had possibly planted him in the UK, or thought that they could exploit him once he was in Moscow.

For some reason, Liddell appears much more concerned about security problems with the mission in Moscow (not his area of responsibility) than he is about breaches in London. The NKVD was reading all of Hill’s postal communications, and the mission had initially not been provided with encrypted wireless support! Even Kim Philby, in My Silent War, wrote about the leakages from the Mission in Moscow, and the Russians’ ‘delight’ with Hill, a disclosure that must be inspected at some future time. Yet Liddell should have been focusing on security exposures on British soil.

It should be remembered that the period from late 1941 to 1943 was characterised by some wary attempts by MI6 and SOE to exchange intelligence with the NKVD, and even engage in shared subversive operations, where Soviet saboteurs were trained by SOE, and then parachuted into various European countries (Operation PICKAXE). (Attentive readers of coldspur will recall that the Radio Security Service, RSS, detected illicit use of wireless by Soviet operatives at the SOE training-centre at Brickendonbury Manor.) The head of the NKVD mission in London was the flamboyant but demanding and ruthless Colonel Chichaev. By most accounts he arrived in London, in the spirit of a cooperating alliance, in November 1941, but others indicate that he had arrived much earlier, and was perhaps acting as Gorsky’s substitute during the 1940 recall. For instance, Barros and Gregor, citing Russian archives in Double Deception, assert that, on May 14, 1941, Chichaev reported on the interpretation that Kim Philby gave him of Rudolf Hess’s arrival in Scotland (i.e. when the Soviet Union was technically still in an alliance with Germany), even though Gorsky had returned to London in November 1940. Such a claim must be treated cautiously.

For most of the time, as The Storm and the Shield (the Mitrokhin archive) indicates, Chichaev, as a secondary legal rezident in London, worked in parallel with the NKVD’s Gorsky, who continued to handle Philby & co. until he was transferred to the USA in 1944, and was replaced by Krotov. Thus Chichaev, who, unlike Gorsky, declared his role openly to the British authorities, must also have been the Military Attaché cited in the telegrams. He performed a dual role in dealing with SOE (overtly) and communist informants within foreign government-in-exile (covertly). It would have been quite natural for British diplomatic and intelligence personnel to have been meeting him openly during the period when Gouzenko describes ELLI as being active. Liddell describes a meeting that an unnamed officer had with Chichaev in July 1943 (see below).

Colonel Chichaev

The vitally important aspect of Chichaev’s status, however, is that, despite being represented as the ‘military attaché’, he was appointed by, and communicated with, the NKVD, not the GRU. (He had no contact with Sonia, for instance). Thus any SOE asset who provided intelligence would have been approved and acknowledged by the NKVD in Moscow. Even though Gouzenko (of the GRU) heard about ELLI, and reported his existence, it did not mean that ELLI was a GRU spy. Intriguingly, Amy Knight, in a footnote (pp 331-332) concludes, using a reference in Nigel West’s and Oleg Tsarev’s Crown Jewels that stated that Philby had reported the existence of an MI6 spy in Moscow called ‘TEMNY”, that ‘If it were not for the fact that Gouzenko’s “ELLI” was a GRU agent with a Russian background, this piece of information would point us straight to Philby as the ELLI suspect’. With two possible agents at large, and the fact that ELLI was an SOE-NKVD spy, the whole question remains up in the air.

8. Hollis’s Interview:

Meanwhile, the RCMP applied pressure on Hollis to extend his stay in North America, and return to Ottawa. Chapman Pincher wrote that Hollis interviewed Gouzenko on November 7, but that cannot be correct. In mid-November, Hollis was still occupied in explaining to London the reasons for the delays in publicizing the Gouzenko case, and Elisabeth Bentley’s confessions to the FBI created fresh turmoil when she named Cedric Belfrage (of BSC) as one of her spies. It was not until November 21 that he returned to Ottawa from New York to have his first interview with Gouzenko. Again, we are indebted to William Tyrer for persuading MI5 to release the telegram that he sent on November 23 – and which presumably provoked Liddell’s flurry of meetings with Military Intelligence officers (see below).

The full text of that message may be read in Tyrer’s article in the International Journal of Intelligence and Counter-Intelligence, so I paraphrase its contents here. It was clearly Hollis’s first exposure to Gouzenko, who made a good impression.

            i) Gouzenko had himself deciphered two telegrams from London, one stating that ELLI was going over to the dubok method, and another that the attaché in Moscow would not reveal name of British agent there.

            ii) Gouzenko’s colleague Liubimov told him in 1943 that ELLI was a high-level counter-intelligence officer, and a member of an important committee. The number ‘5’ had an association.

            iii) Koulakoff told Gouzenko a high-grade agent was still working in the UK in 1945, but it may not have been ELLI.

            iv) Gouzenko told Hollis he was not aware that ELLI had reported two incidents of theft of papers from Military Attaché in London.

            v) Hollis asked Gouzenko about the nature of information ELLI provided: was it about German war dispositions, political matters? Gouzenko could not say.

What are we to make of this? It is all rather a muddle. If Leopold was still in charge of the interpretation process, it must have been a difficult encounter, and Hollis must have wondered what fresh advances he could bring to the proceedings. What does Point i) mean? That George Hill had told Chichaev that he was running an agent in the Kremlin, but would not reveal his name? That would be the height of irresponsibility. Point v) is revealing, however. It strongly suggests that Hollis had MI14 on his mind. MI14 was responsible for analysis of German military operations and Leo Long had joined the section late in 1941. In 1944, he and Anthony Blunt had been discovered removing ULTRA decrypts to give to the Russians, so Hollis might naturally have wondered whether Long was ELLI. (That, incidentally, was the conclusion that Christopher Andrew came to in Defend the Realm, under the ‘guidance’ of Oleg Gordievsky.)

Point iv) is superficially puzzling: where did this item come from (was the anonymous Military Attaché the victim of a theft?), and did Gouzenko really speak up only in response to something Hollis knew? Did matters get garbled in translation? Yet I believe that this is one of the most significant items in the telegrams, something that has not received its due attention until now, and one that helps explain Liddell’s concurrent and subsequent actions. The full paragraph runs as follows:

            “CORBY told me that he did not know that the two incidents of the theft of the papers from Military Attaché in London and attempt to Telephoto his office were reported by ELLI.

First of all, this is clearly information that Hollis provided to Gouzenko, although the wording is slightly ambiguous. (Gouzenko may have known that the theft incidents had been reported, but not that ELLI was the source.) The intelligence must have been communicated to Hollis recently, while he was in Canada, else he surely would have tackled it with more urgency. So why were Hollis and MI5 so confident that this leakage could be placed at ELLI’s door? How could they have verified that ELLI provided any information unless either a) they knew who ELLI was, and had interviewed him or her, or b) they had access to an insider on the Soviet side who could confirm that such information could be traced to ELLI? Because of the ambiguities of the transcript, we cannot be sure whether they assumed ELLI was the source because of the close connections between the Military Attaché that Gouzenko had pointed out beforehand (and noted by Liddell in his diary), or because ELLI had been directly identified by a source in Moscow. Yet Hollis and Liddell knew that ‘the theft’ had been reported, presumably because MI5 had been involved in the exploit.

So what were the circumstances of the theft? A vital clue may be found in the memoirs of George Hill, Maia Shpionskaya Zhizn (My Life as a Spy), published in Moscow in 2000. They are cited by Dónal O’Sullivan in Dealing With the Devil, who also indicates that the Hoover Archives at Stanford University in Palo Alto preserve a copy of Hill’s unpublished 259-page manuscript titled ‘Reminiscences of four years with the NKVD’. When writing about Colonel Ivan Chichaev, NKVD’s representative in London between 1941 and 1945, O’Sullivan writes: “According to Russian accounts [in fact an introduction provided by the Russian editors to what turned out to be a reprint of Hill’s 1933 memoir, ’Go Spy the Land’: coldspur], British Intelligence attempted a ‘burglary’ of his residence to discover secret documents but found nothing as Chichaev had deposited them in the Soviet Embassy’ (Hill, p 37). Chichaev, unusually for such a functionary, established a private residence at 54 Campden Hill Court in Notting Hill, so that it is surely the house that is being referred to. (Rather incredibly, Colonel Gubbins, the head of SOE, lived in the same building.) Yet, if they found nothing, had ‘a theft’ occurred?

Thus the sequence of events would appear to be as follows: MI5 believed that secret documents were being passed to Chichaev (or had, perhaps, even planted them on him). They broke into his house in an attempt to find them, and to catch him red-handed. Chichaev was warned by an inside source of the planned raid, and thus moved the documents into a safer haven, in the Embassy. Chichaev reported the incident, and ELLI’s contribution, to Moscow. That information reached Hill, who may have passed on the information to his bosses in SOE. Alternatively, he may have been the source of the suspicions of Chichaev’s espionage. When he made a visit to the UK in November 1943, and had a meeting with Liddell, the question of the surveillance of Chichaev came up, and Hill requested that any evidence of possible espionage be reported to him, so that he might advise the NKVD of such complaints.  Moreover, Anthony Blunt, assistant to Liddell, could conceivably have been an alternative, as responsible for the leakage.

MI5’s knowledge of Chichaev and the PICKAXE operation is worthy of a separate study. SOE employed an officer, John Senter, who was the liaison with MI5, so he surely kept Liddell at least partially informed of what was going on. Indeed, an entry in Liddell’s diary of August 14, 1942, rather provocatively states: “John Senter came to see me and brought with him an interesting document which had been extracted from the kit of one of the Russian parachutists sent over here”. It was probably a shoddily forged ID-card for one of the members of the COFFEE mission, characteristic of Soviet efforts. But at this stage, Liddell’s further comments show him not intimately familiar with the set-up under Chichaev, whose existence he first recognizes only at this late date. The archives show, as reported by Bernard O’Connor, that, when the COFFEE team struggled in its mission, unfit and ill-equipped, its members sought to defect, and on September 1, 1942, MI5’s Seddon and Wethered were brought in to consider the plea.

The complex relationships between Stephen Alley [see below], Chichaev and Hill – and indeed Philby, who worked with Chichaev, too, and the British Ambassador to the Soviet Union, Clark Kerr, who protected Hill – are too involved to explore here, and will have to be examined another time. Yet the conclusion must be that the Soviet Military Attaché in the ELLI incident was indeed Chichaev, that it involved SOE and SIS (between whom the Soviets made no distinction), and that, while Liddell was following the whole business closely, Roger Hollis had nothing to do with it.

Assuredly, this exchange would provoke some heated discussion over the years, which will be the subject of a later analysis. Suffice it to say that Hollis was a bit out of his depth at this stage. He had been focusing on high-level political strategy, dominated by the Allies’ vacillation over publicising the Gouzenko affair, and subsequently the fall-out from the Bentley revelations. He faithfully reported all that Leopold translated for him, but it cannot have made much sense to him. Knight cites some enigmatic handwritten notes written by the RCMP, summarizing the interview, that indicate possible confusion on Gouzenko’s part between information given to him in 1942 by Liubimov in Moscow and just recently by his successor Kulakoff in Ottawa, and which confirmed that there could be two agents described in his testimony. But the muddle may have been the fault of the interpreter/translator, and those who recruited and managed him.

Hollis had a reservation on the Clipper to return to the United Kingdom on November 26, and on his arrival was no doubt happy to pass the responsibility for Gouzenko over to Liddell, who, as has been shown, ran – and sometimes ambled – with the ball thereafter. As late as December 6, Petrie was still having discussions with Liddell about creating a new section in Hollis’s F Division to deal with Soviet espionage. At least, Liddell’s immediate meeting with the Director of Military Intelligence, de Guignand, [see below] showed some level of urgency. Because of this SOE story, however, and perhaps after speaking to Hollis immediately on the latter’s return, Liddell next indulges in some startling speculation.

9: SOE & Alley:

On November 16 Liddell had arranged to speak to one of George Hill’s closest friends and colleagues, a man called Stephen Alley. Nigel West’s Historical Dictionary of British Intelligence includes a short thumbnail sketch of Alley: “A veteran Russian-speaking Secret Intelligence Service (SIS) officer, Major Alley was based in Paris between the wars and worked with Sidney Reilly. During World War 1, he had operated for SIS in St. Petersburg and had been evacuated in February 1918. After the war, he served in MI5 for three years and then moved to Paris, where he ran a business trading in commodities.” Yet the truth would turn out to be a bit more complex.

Stephen Alley

Liddell’s entry runs as follows: “I threw a fly * over Alley about George Hill. He said he had known him for many years and that he regarded him as a charlatan. He had in fact employed him on behalf of Imperial Tobacco in the Balkans but that he had been too expensive even for them. He had also used him in the old days to make contact with various MPs like Commander Kenworthy who seemed to be a spokesman for the Soviet Govt. in the House of Commons. There were periodical meetings in the form of luncheon parties which were arranged by Hill.”

[* This is presumably an angling term, used to indicate a lure being cast in the direction of a fish. I was not familiar with it, but it crops up regularly in Liddell’s Diaries.]

“The Commissar Vanishes’ (Stalin & Yezhov)

So what was Alley up to, so easily accessible by Liddell? It appears that he was still (or again) working for MI5 at this time. He does not appear in either West’s or Andrew’s account of MI5, as if he had been sanitized out of the picture, like Stalin’s commissar. John Curry’s internal history, however, shows him, in April 1943, as working in E Division, as E2, responsible for Alien Control of Finns, Poles and the Baltic States. Intriguingly, Liddell has a diary entry for July 19, 1943, where he records ‘that [XXXX] had made contact with Chichaev  . . .  who had considerable background in Finland and Riga’. The name of that redacted person could well have been Alley: the entry concludes by noting that XXXX is confident he can handle Chichaev.

In any case, it would appear that Alley had been a senior officer in MI5 throughout the war, and carried on afterwards. Another current Liddell diary entry, for September 11, 1945, recorded that Liddell had spoken to Alley about the arrival of Colonel Einthoven, who was the head of Dutch security in the Ministry of Justice. Alley was making ‘the necessary arrangements’: he was a loyal and trusted servant to Liddell. Richard Deacon records that Alley (unlike Dick White and Roger Hollis) attended Liddell’s funeral in 1958. Moreover, Alley had ‘something Russian in his background’. He had been born, in 1876, in a house at the Yusupov Palace in Moscow, and, after receiving his school education in Russia, he moved to King’s College in London, gained a degree in engineering at Glasgow University, before returning to Russia where (some claim) he was involved in the murder of Rasputin, and an attempt to rescue the Romanov royal family. Even more intriguing is the statement made by Michael Smith in Six (based on Alley’s unpublished memoir held at Glasgow University) that he was fired from MI6 because he refused to carry out an order to assassinate Joseph Stalin.

I have tried to contact the archivist at Glasgow University to determine whether this memoir can be made available, but, as so often happens with such inquiries, I have received no reply. The name and profile of Alley resonate, however. The close phonology, the Russian background, the presence in MI5 during the war, the connection with Hill and SOE in Moscow all make a lot of sense. So how does Liddell complete his diary entry? “ELLI=ALLEY is I think too fantastic to merit any serious thought.”

All those agents, or putative spies, with the double liquid consonant! Hollis, Mitchell, Ellis, Costello, Leo Long, Hill, Alley, Elliott – even Liddell himself, whom the author John Costello suspected, as he laid out in Mask of Treachery. And then Akhmedov said that ELLI was female – Evelyn McBarnet? Ray Milne, née Mundell? One longs for some traditional solid English names, like Hodgson or Winterbotham, who could immediately be dismissed from the inquiries because of the illiquidity of their surnames. It is, however, Liddell’s raising the possibility that Alley might be ELLI rather than his dismissal of the idea as preposterous that intrigues me more. It indicates that Alley at least fitted the profile of what could be deduced about the agent/informer. Did someone suggest it to him, or did he come up with it himself? Was the idea expressed outside his diaries? We may never know, but at least, for a while, MI5 officers were considering seriously whom Gouzenko might have been pointing towards.

Liddell was not finished yet. On November 21, he noted that he dined with Archie Boyle and Darton, and discussed the SOE Mission in Moscow. He added in parenthesis ‘See note in front of diary’, but that is not to be found. And then a very significant entry for November 27 needs to be cited in full:

Air Commodore Archie Boyle

“I saw the DMI [Director of Military Intelligence, Freddie de Guignand, who had replaced John Sinclair in September 1945] and told him about the ELLI case. He sent for the current files of telegrams between the British Military Mission in Moscow and London which only covered a period of 3 months. All back files are sent to the Record Office at Droitwich. He is sending for those covering the years 1942-43 so that we can go through them. There was nothing in the current files to show the Mission was running an agent. DMI also sent for a list of officers who had served on the mission during the relevant period. He found a Capt. Chapman, who he is going to see. He will merely ask him whether at any time the Mission had run agents and if so whether he recollects any request from London for the identity of such an agent. This seems to be as far as we can go at the moment.”

This exchange is puzzling. The mission in Moscow was designed as one of coordination with the NKVD over the running of saboteurs in Nazi-controlled territory: it was not an intelligence-gathering exercise (although the NKVD thought otherwise), and attempting to develop agents would only have incurred the additional wrath of MI6. De Guignand surely knew that. So had Hill really reported to his bosses in London that he was running an agent, but had concealed his identity? Or was it a bogus claim he made to Chichaev, to impress him? And how did ELLI learn that fact? George Hill should have been an obvious source to shed light on affairs.

In any case, Liddell followed up on December 28: “I spoke to Archie Boyle and told him that I had seen DMI to whom I had spoken about HILL. DMI was considerably worried and was anxious to know who backed HILL in WO [War Office]. Archie said that HILL’s file which would be with the Military Secretary, would give the answer. Archie saw no objection to my discussing the whole affair of HILL with the FO.” Furthermore, on January 4, 1946, Liddell had another meeting with de Guignand on the ELLI case. Someone called Jimmy Way was detailed to speak to Hollis and ‘arrange to get the names of all people who handled JIC and JIS as well as Military Mission documents, at the material time’. Such names would then be passed over MI5 records.

Thus 1945 came to a close. In mid-December, Liddell had learned that he had been overlooked as Petrie’s successor, and was obviously disappointed. He had to re-apply himself to the tasks at hand. What happened to his SOE inquiry in 1946? Was George Hill picked up for questioning? What was Alley’s relationship with Hill? What was going on with the White Russian George Graham, aka Leontieff? Why did MI5 start to think, in 1951, that ELLI might be Philby? And how was Gouzenko’s testimony picked up in later years? I shall inspect these questions in a later report.

10. Interim Conclusions:

  • ELLI was an SOE asset providing information to Chichaev, the unnamed Military Attaché, who worked for the NKVD. The appearance of Polakova, the PICKAXE trainer, is a strong reinforcement of this theory. Because ELLI’s story was revealed by a former Central clerk who was assigned to the GRU, it has been wrongly assumed that ELLI must have been a GRU agent. Vassiliev’s unintentionally misleading account has done much to reinforce this misconception.
  • MI5 was sluggish. It should have demanded control, sent out an expert dedicated to the case, and ensured that a qualified interpreter was used. The confusion over the translations and transcriptions is unpardonable.
  • Hollis was not central to the inquiry. He was on vacation at the time of Gouzenko’s defection, and his mission in North America was to handle the high-level political implications of Nunn May’s actions. He was not introduced to the ELLI case until late, and he was the wrong man for the job.
  • There is no evidence that Philby actually referred to ELLI in his messages, and no detected confirmation by Moscow Centre that ELLI was a GRU asset. Philby’s references to GRU were probably in relation to Gouzenko and Nunn May. He surely saw the information, but never thought it pointed to him.
  • ELLI might well have been Alley, who was active in 1942-43, and knew Hill well. He was still around in 1945. His role may have been exaggerated. ELLI’s name may have been changed after the events of 1945, or he/she may have been taken out of service. The anomalies of the dubok and the committee remain. Gouzenko (or his translator) may have confused multiple events and personalities. Feeble and obvious attempts have been made to excise Alley’s name from the record.
  • SOE had shown gross laxity in allowing Hill to appoint ‘George Graham’ as his aide de camp. This may have led to the security exposures in Moscow that Philby and Boyle refer to.
  • Pincher’s account is mainly fantasy. Hollis was out of the picture when the Gouzenko story broke, and he had no business dealing with Chichaev and the SOE. Amy Knight’s much respected work also gets the story wrong.
  • The analysis reveals the relative slowness of everything, compared to today, and the primacy of Chronology and Geography. Cross-Atlantic travel, even by air, was laborious, and wartime passage between London and Moscow (or Kuibyshev) especially arduous; notwithstanding the advances of HYDRA and Rockex, the process of the composition of messages, encryption, transmission, decryption, recryption, transmission, decryption, distribution, analysis, and further routing was long-winded; arranging contacts between spies and handlers had to be undertaken very cautiously; other business processes, such as arranging meetings, and gaining approval and signatures for decisions and messages, were much more challenging in in a pre-electronic age.


Gouzenko files at TNA (KV 2/1419-1429)

Guy Liddell Diaries at TNA (KV 4/185-196; KV 4/466-475)

The Unresolved Mystery of ELLI, by William Tyrer (in International Journal of Intelligence and Counterintelligence)

The CORBY case: the defection of Igor Gouzenko, September 1945, by Gill Bennett (from FCO publication From World War to Cold War)

How the Cold War Began, by Amy Knight

Defend the Realm, by Christopher Andrew

The Secret History of MI6, by Keith Jeffery

The Crown Jewels, by Nigel West and Oleg Tsarev

The Security Service 1908-1945, by John Curry

The Secret History of SOE, by William Mackenzie

The Secret History of British Intelligence in the Americas, 1940-45, by Gilbert Highet, Tom Hill & Roald Dahl

MI5, by Nigel West

MI5: 1945-1972, by Nigel West

Molehunt, by Nigel West The Sword and the Shield, by Christopher Andrew and Vasili Mitrokhin

The Haunted Wood, by Allen Weinstein and Alexander Vassiliev

Mask of Treachery, by John Costello

Treachery, by Chapman Pincher

Spycatcher, by Peter Wright

The Storm Birds, by Gordon Brook-Shepherd

The Philby Files, by Genrikh Borovik

Fred Rose and Igor Gouzenko: The Cold War Begins, by David Levy

Stalin’s Man in Canada, by David Levy

Best Kept Secrets, by John Bryden

Dealing With the Devil, by Dónal O’Sullivan

Sharing Secrets with Stalin, by Bradley F. Smith

Camp X, by David Stafford

Cargo of Lies, by Dean Beeby

Six, by Michael Smith

Double Deception, by James Barros and Richard Gregor

Churchill & Stalin’s Secret Agents, by Bernard O’Connor

From the Red Army to SOE, by Peter Manderstam

Trotsky’s Favourite Spy, by Peter Day

My Silent War, by Kim Philby

To Spy the Land, by George Hill

The Greatest Treason, by Richard Deacon

Historical Dictionary of British Intelligence, by Nigel West

(Recent Commonplace entries can be found here.)


Filed under Espionage/Intelligence, General History, Geography, Management/Leadership, Politics, Technology