Special Bulletin: Denis Lenihan on Sonia & Quebec

I am experimenting here by posting a fascinating piece by an Internet colleague, Denis Lenihan. I have long been looking for a forum in which discussions on outstanding intelligence conundrums could take place. My idea of an ‘Officially Unreliable’ conference never took off, and I have not found any other medium where serious, non-polemic, debates on such subjects could be undertaken. I am not certain that WordPress is the optimum home for such discussions, as I have to assume the role of editor, moderator, and host, but I want to try it out.

Denis (whom I have never met) is a dedicated analyst of intelligence matters. He describes himself as follows: “Denis Lenihan was born in New Zealand and has degrees from the University of New Zealand and the University of Sydney. Having spent his working life in Australia as a civil servant, he now lives in London. He also writes about spies on the websites kiwispies.com and academia.edu .”

I am delighted that Denis has taken the trouble to respond in depth to my piece on Sonia and the Quebec Agreement. (Readers may want to refresh their memory of it first.) I have prepared a reply to Denis, which I have sent him at the same time that this bulletin is being posted. In that way, I can deliver an unvarnished response, uninfluenced by any comments posted here, and readers who want to make any observations will likewise not be swayed by any of my pontifications. I plan to post my response here after a week or so, and Denis will of course have an opportunity to address any of my arguments.

You can respond in several ways. You can leave a short post at the end of this piece. I have to approve such postings before they are released, so, if I do not know who you are, and the message is questionable, I may not ‘Approve it’. An email address attached will allow me to contact you, if necessary. It is also possible that some messages may be trapped by the ‘Stop Spammers’ filter. Since I installed this feature last summer, I have apparently had 1,189,918 messages blocked. If you suspect I may have overlooked your message, please contact me by email at antonypercy@aol.com. (Since many readers are accustomed to my schedule of monthly postings, they may well overlook the arrival of this Special Bulletin.)

Alternatively, you may wish to send me an email with a longer message. Again, I may want to check you out if I have not heard of you before, but, after I have done that, if you want your posting to appear anonymously, I am happy to oblige. I hope to receive some comments, but, if none arrive, Denis and I can simply air our exchange. (And he has more to come.)

A colleague has informed me that Maik Hamburger, Sonia’s first-born, but last-surviving child, translator and Shakespearean scholar, died in Berlin on January 16.


Antony Percy begins by noting the claim made by Pincher in the (first) 2009 edition of Treachery that the GRU spy Sonia (aka Ursula Beurton, nee Hamburger, born Kuczynski) transmitted to Moscow ‘the details of the Quebec Agreement’ 16 days after it had been signed on 24 August 1943. Percy concludes his introductory paragraph with the words:

This article shows that the contradictions and anomalies in the accounts of the leakage of this secret leave the published claims about Sonia’s activity open to a great deal of scepticism.’

Having done his analysis, Percy concludes his article by saying that ‘a proper resolution of this affair can therefore only come from the following steps’; and he goes on to list five, ranging from verifying the existence and authenticity of a real document in the GRU archive dated September 1943 concerning the Quebec Agreement, to investigating whether the intelligence might have been gained elsewhere.

These five steps resemble counsels of perfection. It is a nice question whether if they were applied to other questions which arise in espionage/intelligence activity we would be able to reach any certainty or indeed to make any progress at all. Certainty may be beyond us at the moment so far as Sonia and the Quebec Agreement are concerned, but that does not prevent us from harvesting and studying the scraps of information which we have and developing one or more hypotheses from them.  These can be extended or modified as more information emerges. The hypothesis advanced here is that on this occasion Pincher was right about Hollis being Sonia’s source for her information about the Quebec Agreement.

A useful starting point is a close examination of the allegations about what is contained in the GRU archive in Moscow, of which there are four, in the following date order: by Bochkarev and Kolpakidi in 2002 in a book about Sonia (described as Ruth Werner) called Superfrau iz GRU (Superfrau in the GRU); by Bance (aka Dan) in a 2003 book Ultimate Deception; in the 2012 (final) edition of Treachery, in which Pincher quotes the Russian historian Dr Svetlana Chervonnaya; and in a 2016 journal article by William A Tyrer, which again quotes Chervonnaya but with one crucial addition to Pincher’s report.

Bochkarev is described by Pincher, without sources quoted, at 598 as ‘a former GRU officer’, and Kolpakidi as someone ‘who had been given some, but strictly limited, access to secret GRU records’. A Google search shows that Kolpakidi is the author of a number of books on the GRU. Pincher has the duo at 187 simply stating that ‘on 4 September, Sonia reported data on the results of the conference’. Pincher adds ‘suggesting that she sent all that she had learnt from her source’, which is clearly a non sequitur.

Bance is variously referred to by Pincher as a researcher with Moscow intelligence sources (53), having access to KGB and other USSR archival documents (202 and 207), obtaining information from former KGB officers (344-5, 386, 425, 431 and 630) and with access to GRU sources also (208 and 459).  On the other hand, Percy warns that Bance/Dan’s work ‘is a curious melange of fact and fiction that needs to be parsed very carefully’. What category the following fell into is not disclosed.  Percy quotes Bance/Dan’s ‘critical sentences about Sonia and the scientists’ thus:

‘General Groves, the newly installed head of the Manhattan Engineering District, the US codename for their atomic bomb project, agreed to the British request that a number of its scientists should work in America. Lord Cherwell, Wallace Akers and Michael Perrin, his deputy, met to decide what names to put forward to Groves, who reserved the right of refusal. Advised by two of his scientists, Mark Oliphant and James Chadwick, a list was finally agreed . . .  Word quickly spread in the scientific community as to who was on the list. Fuchs provided the names to Ruth [=Sonia], who then transmitted them to a grateful Moscow on September 4.’

As Percy notes, there was no mention of the Quebec agreement itself.  This may be the reason why this extract is not, so far as I can see, quoted by Pincher. In any event he thought that Hollis rather than Fuchs was responsible for providing the scientists’ names.

Chervonnaya is quoted by Pincher at 19 as ‘having discovered a Soviet document confirming that Sonia had sent the information about the Quebec Agreement on 4 September 1943 and that, after translation into Russian, it was taken straight to Stalin’.

Finally, Tyrer’s article in the International Journal of Intelligence and Counterintelligence in 2016 (Vol 29, no 4) quotes ‘a Russian military writer Vladimir  Lota, with access to some of the GRU files, which are off limits to other researchers’ thus:

On 4 September 1943, U Kuzcynski [Sonia] reported to the Centre information on the outcomes of the conference in Quebec. She had also learned that English scientists Pierls, Chadwick, Simon and Olifant had departed for Washington. U Kuzcynski had received this information from Klaus Fuchs…’ ‘

Tyrer gives the reference to Lota’s article, published in 2010, in a footnote and then adds: ‘Courtesy of Dr Svetlana Chervonnaya’.

So there is agreement from the supposed GRU sources that information (unspecified) was sent by Sonia on 4 September about the Quebec Agreement. Note however that the claim about the names of the scientists who had gone to the US being sent on the same day is made only by Bance/Dan, but without mention of the Conference. Bance/Dan may thus have been confused.  One gets the impression that the GRU gradually disclosed more and more about Sonia’s message: initially in Bochkarev and Kolpakidi, only ‘data on the results of the conference’; then information just about the agreement; and then – finally in Lota – information on both the outcomes of the conference and the scientists, and again with Fuchs as a source.

The great anomaly in the material is that Chervonnaya was evidently aware in 2010 or shortly thereafter that Fuchs had been named by Lota as Sonia’s informant (at least so far as the names of the scientists were concerned), but she evidently failed to tell Pincher. He recorded in the 2012 edition of Treachery at 629-30 that he was ‘indebted’ to Chervonnaya for many reasons, one being that she provided information about ‘extracts from the continuing works of Vladimir Lota, who appears to be the only person granted any access to GRU records’ (but see his descriptions of Bochkarev and Kolpakidi above). This seems inexplicable.

Percy suggests that, in the Lota extract quoted above, the force of ‘also’ and ‘this information’ concerns only the departure of the scientists and not the Agreement, and I think this is right – although some confirmation about the accuracy of the translation would be helpful.  He goes on to suggest that Fuchs was indeed the source of the information about the scientists and that he passed this information on to Sonia in mid-August. This too seems right: why would Sonia wait until September 4 to pass this on to Moscow?  Quoting GRU archives, Pincher has Fuchs meeting Sonia in mid-August and not again until November, so this fits with the suggestion that the information about the scientists was sent in mid-August, shortly after she received it.

There is another argument against Fuchs being the source of the leak about the Agreement, and that is that nobody on the Soviet side has ever given him the credit for it. Both Fuchs and Sonia eventually admitted their roles in getting and transmitting atomic secrets. If Fuchs had been responsible for the leak about the Agreement, why not go beyond merely admitting it and boast about it, especially Sonia?

Percy proposes a number of other candidates as the source of the leak about the Agreement and while they have some plausibility they all suffer from the same defect: there is no or no apparent Oxford connection. Unless the GRU is running and has run some disinformation campaign, the material about the Quebec conference was sent by Sonia from Oxford. This argues overwhelmingly for an Oxford connection for the source. London candidates would surely have used the more secure services of Colonel Simon Kremer, a GRU officer at the Soviet Embassy, as others did. Hollis was based in Oxford at the relevant time.

The list of alternative candidates proposed by Percy might need to be expanded in the light of admission made by Churchill in the House of Commons on 13 April 1954, when the subject of the Quebec Agreement came up. A Google search of the Agreement shows that he said that

‘My telegram [from the context, about the Agreement] was addressed to the Deputy Prime Minister and the War Cabinet, but it may well be that, owing to the great respect with which the words “Tube Alloys” were treated, it slipped out at some point or other.’

Having been the Deputy Prime Minister in 1943, Attlee came to Churchill’s aid:

‘May I ask the Prime Minister, was it not a fact that it was thought best to keep knowledge of this matter in the hands of a very few people and that the War Cabinet was informed that there had been talks and agreements with the United States Government on this matter, but that they were not, as a matter of fact, informed of the details at all, or what the agreement was, but simply that some agreement had been come to, and the matter rested there? ‘

Churchill did not disagree. So members of the War Cabinet, who might have inferred from what little they were told what the agreement was about, need to be added to the list of possible sources – if they have an Oxford connection. Churchill plunged on in the exchange in the House of Commons, getting deeper into the mire, so that Attlee was moved to ask plaintively:

‘Would not the right hon. Gentleman think, on reflection, that it would have been better not to have referred to the matter, but to have left it where it was?’

Another gap in the Percy analysis is the role of Michael Perrin, whose role as described by Pincher is dismissed thus:

‘Pincher made some imaginative jumps in promoting the thesis that Hollis would have gained access to the information [about the Agreement] through his colleague at Tube Alloys, Roger [sic] Perrin.’ [Thank you, Denis. I meant Michael. I have corrected my text. Coldspur.]

Perrin gets no fewer than 32 mentions under his name in the index to Treachery, and Pincher establishes a strong personal and professional relationship between Perrin and Hollis. Both were sons of bishops, both had been at Oxford, they were the same age and Perrin had a country cottage at Henley, conveniently close to Oxford. He was originally an industrial scientist at Imperial Chemical Industries (ICI), and became an assistant to the ICI Director of Research, Wallace Akers; both became involved in a project to build an atomic bomb, known in the UK by the cover name of Tube Alloys. On Pincher’s account at 131, ‘Perrin had overall responsibility inside Tube Alloys for security and intelligence, being cleared for access at all levels’; and his entry in the Oxford Dictionary of National Biography says that he ‘was made secretary to the main Tube Alloys committees’.  He and Hollis seem to have become first involved professionally in 1941, when they cleared Fuchs for access to secret information in Tube Alloys. They went on to clear other scientists, some of whom were also spies, in a partnership which on Pincher’s account lasted at least until 1951. Hollis became ‘the natural liaison man with Tube Alloys’ and ‘was to become recognised as MI5’s atomic expert’, being appointed in 1945 as consultant and adviser to the Official Committee on Atomic Energy (Perrin was a member) which reported directly to the Prime Minister. (Treachery 203, 341).

On Pincher’s account at 186-7, Perrin was ‘a most important figure in the talks leading to the Quebec Agreement’; and he was also heavily involved in the preliminary negotiations with US representatives in London. Although Pincher gets muddled about the chronology, I think he is right to suggest that Perrin would have kept Hollis informed of progress with the Agreement, as more scientists – who needed vetting by MI5 – were to be sent to the US after it had been signed. There is also the highly relevant point, made by Pincher at 203, that Hollis had form for being pushy when it came to demanding information to which strictly speaking he was not entitled. Pincher recounts at 405 what the GCHQ officer Teddy Poulden told him about Hollis’ behaviour after the Petrov defection. Hollis repeatedly pressed for details which Poulden declined to provide, as Hollis had no need to know. Hollis then went to Poulden’s superior ‘who upheld the decision and congratulated Poulden’. Thus had it been necessary, Hollis might well have heavied Perrin about the Quebec Agreement.  There are no imaginative jumps here, only reasonable inferences.

The other factor pointing to Hollis comes from Percy’s own writings on Sonia’s Radio, a ninepart epic (plus envoi) also on coldspur.com, during which (at Part VII) the author felt compelled to issue health warnings to his readers. At heading 4 in the envoi -‘Exploiting the National Archives: “Traffic Analysis”’ – Percy makes the very useful observation that while the focus on fresh material in the Archives has been on drilling down, a more arduous but rewarding approach is ‘scouring the archives horizontally’ so that one may find links among disparate cases. What follows is an attempt to adopt this approach with Hollis to detect patterns of behaviour. It looks at his other activities as they bear on atomic spies and on Sonia during the war, when the Soviet priority was discovering what the West was up to with atomic research so that it could develop its own bomb, as of course it did; thus putting into perspective the Quebec Agreement. 

1. Atomic Spies

  • Fuchs

May 1941: Cleared by Hollis for secret work for the first time (Treachery, 130)

August 1941: MI5 received information that Fuchs ‘was well-known in Communist circles’; this was passed on by MI5 with the observation by Hollis that ‘while it was impossible to assess the risk of leakage of information, any leakage would be more likely to lead to Russia rather than to Germany’ (132)

October 1941: An officer on Hollis’ staff suggests that the Ministry of Atomic Production be warned of Fuchs’ communist connections; no action taken (136)

June 1942: MI5 told the Home Office that it had no security objection to Fuchs becoming a British citizen (159)

Late 1942: MI5 learned that while in Canada Fuchs had become a close friend of Hans Kahle, who was well known to MI5, having been noted by it in 1939 as ‘said to be running the [Russian] espionage system in this country’; rated by Hollis as ‘not significant’ (160)

Late 1943: Cleared by Hollis for secret work for the second time (191); reservations not mentioned

January 1944: Cleared by Hollis for secret work for the third time (192)

February 1946: Fuchs’ name discovered by Royal Canadian Mounted Police in the diary of Israel Halperin, a communist who had befriended Fuchs; information passed to Hollis, but not conveyed to MI5 in London (231-2)

Late 1946: Fuchs appointed to Harwell Atomic Energy Research Establishment: member of Hollis’ staff suggested that Fuchs should be re-examined; Hollis decided ‘Present action should be confined to warning [Harwell] about the background of Fuchs’; cleared for the fourth time (263)

October-November 1946: two MI5 officers express the view that Fuchs may have been a Soviet agent; Hollis notes ‘I, myself, can see nothing on this file which persuades me that Fuchs is in any way likely to be engaged in espionage or that he is anything more than anti-Nazi’; cleared for the fifth time (264)

Early 1947: Hollis was over-ruled and Fuchs investigated for three months; there was no suspicious behaviour and the investigation ceased; shortly thereafter, as turned out later, Fuchs resumed his spying activities, one inference being that he had been warned of the investigation (265)

November 1947: after a review involving Hollis, decision made to notify the Supply Ministry if it approached MI5 that ‘We consider that the records indicate only that Fuchs held anti-Nazi views and associated with Germans of similar views and we think the security risk is very slight’; cleared for the sixth time (266)

1950: Fuchs arrested and charged; false statements supplied by MI5 made to court by the Attorney-General; ‘can be little doubt that Hollis had been involved’ in preparing the statements (chapter 42)

False statements made by MI5 to the Prime Minister, who repeated them in the Parliament; ‘can be little doubt that Hollis, as a prime source, was involved…’ (chapter 43); see also Percy Misdefending the Realm, chapter 9;

June 1950: Hollis attends a conference in Washington with US and Canadian officials to discuss security standards; Fuchs case discussed and Hollis’ performance described by Pincher as ‘a virtuoso display of brass nerve, cool dissimulation and power of persuasion without a whisper of apology’, but he failed to convince the FBI (chapter 47)

It is only fair to note that in Misdefending the Realm, chapter 9, Percy argued that Hollis was too junior to have made many of these decisions alone and that his superiors must also have been involved. Perhaps; but a copy of letter composed and signed by Hollis on the Beurtons’ MI5 file (KV6/41) displays his authority as far back as 1944. It is written to the US Embassy in response to an inquiry about the activities and identity of Rudolf Albert Hamburger’s wife and two children, then residing in the UK. (Hamburger had been Sonia’s first husband, and he had recently been arrested for spying.) While Hollis said that Sonia’s brother Jurgen ‘was a Communist of some importance’ and her four sisters ‘have all come to notice in a Communist connection’; she herself ‘appears to devote her time to her children and domestic affairs. She has not come to notice in any political connection, nor is there anything to show that she has maintained contact with her first husband. There is no doubt that she herself has some Communist sympathies…’

  • the same pattern was repeated with other atomic spies such as Engelbert Broda (94, 187, 305, 344, 613, 626), who got little attention from Percy.

2. Sonia

1942: MI5 interrogated a veteran communist, Oliver Green, who confessed to working for the GRU and ‘disclosed that several agents in Britain were transmitting directly to Moscow by radio…’; he added that ‘he had been told that the Moscow Centre had a spy inside MI5’; no action was taken on either allegation (Treachery 178-181)

January 1943: as the result of a request made by MI5 London (H Shillito) via the MI5 officer in Reading, the Oxford City Police made some local inquiries about the Beurtons at George St Summertown Oxford, where they were then residing. In forwarding the police report to MI5 London, the MI5 officer in Reading noted that ‘the most interesting point appears to be their possession of a large wireless set, and you may think this is worthy of further inquiry’; no further action was taken, and there is no indication of the reason for the inaction; Shillito worked in Hollis’ section; Hollis’ initials appear on the file as late as 1947; the Beurtons’ MI5 file KV6/41

1944: Hollis wrote his notorious letter to the US Embassy – see above.

‘For most of the war’ (Treachery 141) messages of Soviet interest intercepted by the Radio Security Service (RSS) were sent initially to an MI5 officer and an MI6 officer; Kenneth Morton Evans, who worked for the RSS and then MI5, told Pincher that the full details were given to Hollis in MI5 and Philby in MI6 and there was ‘no great enthusiasm over them’; James Johnston of the RSS told Pincher that ‘he and his colleagues had intercepted messages from an illegal transmitter in the Oxford area, which he later believed to be Sonia’s, and had submitted them to MI6 or MI5…they were returned with the reference NFA (No Further Action) or NFU (No Further Use)’ the decisions being made by Hollis and Philby; ‘…[Sonia’s] station continued to work, off and on…It must be a mystery as to why she was not arrested’.

In July 1947 MI5 obtained a warrant to intercept the telephone of Sonia’s sister, Barbara Taylor and her husband Duncan, who were ‘suspected of receiving communications from agents of a Foreign Power’. A postal and cable check was also instituted, and some physical surveillance undertaken.  The original number intercepted turned out to be incorrect, but the second number – at the Lawn Rd flats in Hampstead, where Barbara’s father lived – was correct. (It may be that Barbara and her husband moved in with her father after her mother died). The first part of the MI5 file on the Taylors (KV2/2935) records summaries of telephone calls made and received over the next three months, all the checks being removed in mid-September after the interview with Sonia mentioned below.  One striking feature of the summaries is the apparent total absence of any compromising telephone calls, while another is the lack of contact among some of the family members. Despite it being the father’s telephone, he rarely used it. There are no calls to or from Sonia in Oxford. By coincidence, the girls’ father was visiting Sonia in Oxford in September when she and her husband were interviewed by Skardon and another MI5 officer, Serpell.  Tellingly, he made no contact with Barbara in London by telephone after the interview occurred.  On his return to London late on September 15th he got Barbara to ring her sister Bridget and ask her to come and see him the following day.

All this suggests that the family were aware that Barbara’s telephone was ‘off,’ as the expression has it – that is, they knew it was being intercepted.

On Pincher’s account at 277-8, the letters etc of Sonia and Bridget were also intercepted at this time, as was Bridget’s telephone (but not Sonia’s). Pincher read all the intercepted information but it ‘yielded nothing whatever of intelligence value’.  Funny, that.

All this put together shows that Hollis:

– through his connection with Perrin, was well-placed to know the details of the Quebec Agreement;

– was in Oxford at the relevant time;

– via horizontal analysis, was contemporaneously engaged in other activities protecting atomic spies and protecting Sonia and other members of the Kuczynski family.

In the present state of knowledge, Hollis is the prime suspect for Sonia’s source of information about the Quebec Agreement.

Obiter dicta

There is a remarkable entry on ‘Tube Alloys directorate’ in the Oxford Dictionary of National Biography: it mentions the contributions of Klaus Fuchs and Alan Nunn May but omits the fact that both were spies.

In discussing Tyrer’s suggestion that the GRU was waiting for the death of Hollis’ widow before naming him, Percy says that ‘I have not been able to track the birth or death of Edith Valentine Hollis, nee Hammond..’, Hollis’ second wife.  A Google search shows that the University of Oxford Gazette of 27 September 2018 recorded the death on 5 July 2018 of Edith Valentine Hollis (née Hammond), aged 98, in the St Hilda’s College obituary list. This clears one obstacle so to speak to naming Hollis (if he be Elli); another, his son and only child by his first wife, died in 2013; but the son had children and it may be that the GRU reluctance to name names extends to Hollis’ grandchildren.

Percy writes that ‘One of the objections to the claim that Roger Hollis was ELLI has been the fact that Soviet intelligence experts have reportedly expressed bewilderment at the proposition, and no evidence has appeared in Russian archives equating Hollis’s name with that cryptonym.’  Of the ‘Soviet intelligence experts’, how many have been GRU people, Hollis having been claimed to be GRU rather than KGB? 

Denis Lenihan


January 2020

Coldspur’s Response to Denis

I thank you for this comprehensive commentary. I have learned much. I shall address your points in the sequence you presented them.

Despite what I would describe as our shared inquisitive devotion to discovering the truth behind some perplexing intelligence conundrums, I believe we might have a slightly different approach on methodology. Your commentary leads up to the conclusion ‘Hollis is the prime suspect for Sonia’s source of information about the Quebec Agreement’. Yet I would counter that we have no reliable evidence that Sonia ever provided details of the Quebec Agreement to Moscow, whether by direct transmission (as Pincher claims), or by any other medium (such as via the Soviet Embassy in London). And the claims that are made are riddled with inconsistencies.

You suggest that the absence of any documentary evidence should not obstruct our search, questioning whether, in such circumstances, ‘we would be able to reach any certainty or indeed make any progress at all’. But I would rejoin that it does not constitute ‘progress’ to build further hypotheses on such a shaky foundation, however compelling the circumstantial evidence – which itself is constructed on known friendships and alliances, with no unambiguous evidence. For there exists no external phenomenon indicating Stalin’s awareness of secret agreements made at the Quebec Conference beyond what he was told by Churchill and Roosevelt. The event is not like, say, the acknowledged distribution of Ultra secrets through the Rote Drei in Switzerland, or the long concealed Burgess-Berlin mission to Moscow in 1940.

To recap the official major communications between the war leaders in those months in 1943:

August 7: Churchill delivers a telegram to Clark Kerr, informing Stalin of the ‘Quadrant’ conference, which Clark Kerr passes on to Stalin the next day.

August 8: Stalin writes to Roosevelt, apologizing that his time at ‘the front’ (an habitual excuse) has prevented his replying to Roosevelt’s invitation to hold a Summit.

August 9: Stalin replies to Churchill, with similar excuses, expressing his desire to delay the Summit meeting.

August 18: Churchill and Roosevelt write a joint message to Stalin, saying they expect to be in conference for about ten days, and stress the importance of a Summit meeting soon.

August 22: Stalin writes to Roosevelt and Churchill, expressing irritation over the armistice negotiations with the Italians.

August 24: Churchill and Roosevelt are taken aback by Stalin’s sharp tone.

August 24: Stalin sends a more conciliatory message.

August 25: Churchill and Roosevelt send Stalin a summary of the Quebec conference, outlining the various European initiatives (but obviously excluding anything about atomic weaponry, ‘the Quebec Agreement’).

September 2: Churchill and Roosevelt send Stalin a preview of events in Italy.

September 4: Roosevelt writes to Stalin, endorsing idea of politico-military discussions at State Department level.

September 5: Churchill writes to Stalin, stating he wants to restrict discussions to military matters.

September 8: Stalin writes to Roosevelt, saying he would prefer the military-political Commission to be held in Moscow.

September 9: Churchill and Roosevelt write to Stalin, informing him that Eisenhower has accepted the unconditional surrender of Italy.

September 12: Stalin writes to Churchill and Roosevelt, looking forward to the meeting of the Commission on October 4.

What is remarkable about this is that Stalin appears to have offered no opinions on the ‘open’ aspects of the Quebec conference, either of approval or disapproval, being more concerned about the situation in Italy. Yet the suggestion is made that he received details of FDR’s and WSC’s other plans, as the GRU archive cited by Pincher (p 17) claims to indicate. As I have stated before, however, the statement in the archive that ‘there was no word about the fact that they had made an additional secret agreement about the use of nuclear weapons’ does not make sense. It must have been added as some commentary or report later.

Pincher then goes on to say that Stalin was in a ‘prickly, suspicious mood’ when he met Churchill and Roosevelt in Tehran on November 28. Apart from suggesting that Stalin learned about the Agreement from Sonia, Pincher also suggests that the dictator may have been upset about the delay of the invasion until the spring of 1944 (probably leaked by MacGibbon). Stalin, however, knew in June 1943 about the delay of the Second Front until spring 1944, since he complained about it in a message to Roosevelt on June 11. Stalin’s annoyance may have been provoked by information on the Quebec Agreement that he received recently from another source. We simply do not know. Pincher is not a reliable source, and he sometimes piles on his arguments unnecessarily.

What this points to is that no exterior behaviour of Stalin can be attributed to the fact that he learned of the Quebec Agreement from Sonia in early September. All we have are the claims of various GRU/KGB officers or hangers-on, all with privileged but restricted access to secret records, who make assertions about Sonia’s revelations, while none of them can reproduce a document that confirms the event, its timing, its origin, or its precise contents.

I think your analysis of the anomalies in Pincher’s account of the contributions from Chervonnaya and Lota is very sound. I might wonder whether your statement: “Unless the GRU is running and has run some disinformation campaign, . . .” was made with tongue in cheek, as I can hardly imagine the GRU (or KGB) doing anything else – just like MI5, in fact. That is what secret intelligence services do. Your additional evidence about the 1954 exchange between Attlee and Churchill is extremely useful, and I am very grateful to you for your coverage of Michael Perrin. I admit to having taken the Hollis-Perrin connection a bit too casually, and I think the 32 mentions in Treachery which you cite are worthy of deeper analysis. If there were any stronger evidence that Hollis was ELLI (or a mole with another cryptonym), the Perrin link would be absolutely vital.

I recall that Perrin, who was deputy to Lord Portal when the latter was appointed Controller of Production (Atomic Energy) at the Ministry of Supply, was the individual to whom Fuchs made his full confession, an experience that distressed Perrin. And Perrin also came strongly to Peierls’s defence later in 1950 (see Frank Close’s Trinity, p 395) after the FBI and MI5 had begin to cast doubts on Peierls’s loyalties. Early in 1951, Peierls had the ominous conversation with Portal, followed up by the letter to him that I discussed on coldspur a couple of months back, where Peierls admitted his connection with the assassin Leonid Kannegiesser. Shortly after this, both Lord Portal and Michael Perrin left their posts. I wonder whether those events had anything to do with the Peierls ‘confession’? Perrin must have been devastated, and he did not die until 1988, so must have witnessed the attacks on Hollis.  Another trail to pursue.

Your ‘horizontal analysis’ of Hollis is also very welcome, although I would prefer that it be based on the original National Archives records rather than Pincher’s unsourced references to them! I have been through your citations, and checked them against my Chronology, and believe that they are all sound, although one must point out those occasions where Hollis took a more positive line in trying to counteract Soviet espionage or subversion. Pincher does not dwell on these. Again, a topic for further study, especially when it comes to the matter of whether his arguments were countered or condoned by other MI5 officers.

I am glad you drew attention to Engelbert Broda, as well. I had not spent much time on him, but I have recently read Paul Broda’s Scientist Spies, and it is evident that he was one of the key contributors to Soviet Russia’s access to atomic secrets. I have now acquired the Kew files on him and his wife.

When it comes to surveilling Sonia and Green, as you now, I am of the opinion that senior officers in MI5 and SIS were complicit in a scheme to allow the spies to carry on their undercover work, and that Hollis was thus not a lone wolf trying to protect Sonia. And yes, I did notice the obituary of Hollis’s widow last year. If the GRU were sensitive enough to want to spare their agents’ widows, that time has now passed. As you point out, Roger Hollis’s son, Adrian, died in 2013.

You echo Pincher’s account that Morton Evans of RSS informed him that ‘messages of Soviet interest’ were passed on by him to Hollis and Philby. But this cannot be true. Morton Evans could not have detected that undeciphered messages were in Russian, or from a Soviet spy. His procedures did not involve sending such messages to Hollis or Philby, who would have had no idea what to do with them. He may well have invoked Liddell and Robertson to determine, if the locations had been triangulated by direction-finding apparatus, whether the transmissions originated from Double-Cross agents. If the location could not be pinpointed accurately, he would have sent out mobile direction-finding teams, but an officer from MI5 B1 section would have had to accompany them.

Your fascinating observations on the September 1947 telephone intercepts of Sonia’s sister Barbara, where you draw attention to the fact that the Kuczynski family appeared to be aware that their telephone calls may have been listened to, made me go back to Sonia’s Radio, Chapter 6. I relate there Sonia’s own story that Alexander Foote warned her that she was in danger by contacting Ullman, the Austrian who had introduced them, and was now living in London. Of course, the GRU might have insisted that she present that anecdote in Sonyas Rapport, to conceal the real source of the leak in MI5 . . .  

Having returned to my original text of Sonia and the Quebec Agreement, I also wanted to add a few comments. I may have given the impression that the Nazi spy JOSEFINE (representing the attachés in the Swedish Embassy) had passed on to her masters the details of the Quebec Agreement. As David Kahn indicates, in Hitler’s Spies, the information passed on was the basic military stuff  – still very useful to the Germans, and evidence that many lips were talking too loosely.

And MI5 was given the same information – though six days later than when JOSEFINE passed it on. Liddell had, however, known about a ‘Quebec decision’ as early as August 27. He later records that he (and others) were briefed on the outcome of the Quebec Conference by (Gilbert) Lennox, on September 7. Lennox is a figure who crops up regularly in Liddell’s Diaries, although Nigel West declines to provide any details for him (apart from listing him as ‘Operations’), and sometimes confuses him with Gordon-Lennox. We owe it to Christopher Andrew who, while omitting to provide an entry in the Index for Lennox, informs us (on p 235 of Defence of the Realm) that Lennox was MI5’s liaison officer with military intelligence, and had joined the service at the outbreak of war, on the recommendation of Jane Archer and Dick White. Why Lennox should be the first recipient of the news, rather than Petrie, is not clear, although the emphasis on military (rather than political) matters would point to his involvement. Liddell understandably says nothing about the Quebec Agreement, and there may later have been another more restricted chain of communication via Petrie to him and others officers (such as Hollis), given the requirements to vet the list of scientists who would shortly be selected.

Yet I remain unconvinced that Perrin would have received a copy of the approved Agreement that early. Graham Farmelow, in Churchill’s Bomb, writes that Lord Cherwell did not see it until mid-September, and then complained about the terms. It would have been quite reckless, and unnecessary, for Perrin to tell Hollis anything before the first informal meeting of the Combined Policy Committee (September 8), before Churchill had returned (September 19) and before the UK scientists (led by Chadwick) returned to London (late September) to lay the proposals before Sir John Anderson’s council. There may, however, have been a leak before the parties left for Canada, since the draft of the agreement was prepared by Sir John Anderson and Churchill while Stimson and Bush were in London, and Perrin contributed to the proceedings by having private talks with Stimson at the time (Gowing, pp 167-168). Did Perrin perhaps inform Hollis of what was then being called the ‘Tube Alloys Agreement’ before Churchill and his party left for Canada, and then one of them took the imaginative step of calling it the ‘Quebec Agreement’? Or did Peierls give a hint to Fuchs when he received the call (see below)? Until we see the precise text of the supposed message passed on to Stalin, this is all very speculative.

One vital aspect of all this that does not appear to have been covered: why did Pincher not inyerview Perrin, and ask him about the Quebec Agreement? I can find no indication in Treachery of any correspondence or meeting between the two, which is utterly incredible. And yet, on page 130, Pincher writes that the American biographer of Fuchs, Robert Williams, had a meeting with Perrin on November 12 1985, and immediately afterwards ‘lunched with me at the Randolph Hotel in Oxford and told me about it [the clearing of Fuchs’s file by Hollis and Perrin]’. In Klaus Fuchs, Atom Spy, Williams refers to correspondence from Perrin of May 10 and August 14, 1984, and the interview with him on November 12. Why could Pincher not be there himself? Why did he need an American intermediary to communicate with Perrin? Why was Perrin willing to speak to Williams, but not apparently to Pincher? Did Pincher perhaps suppress any contact, as it did not help his cause?

Lastly, I want to inspect a number of Pincher’s claims in his chapters in Treachery about the Quebec Agreement (1 and 23), in light of our recent discussions. (Some of this analysis has already appeared in my coldspur segment.)

Chapter 1:

P 16: “Churchill therefore kept the Quebec Agreement and its details secret to himself, a few trusted aides, and the chiefs of staff of the British Armed Forces. . . .  Several documents now in the British National Archives testify to the extraordinary extent of the measures taken to prevent any unauthorized persons having any knowledge of its details.”  I do not know to which papers Pincher refers. He lists only three files on the Quebec Agreement in his list of Archival Sources, FO 800/540 (a copy of the Agreement), FO 115/4527 (from 1951), and PREM 8/1104 (from 1949). It is unlikely, anyway, that a procedure to keep something secret would itself be registered. And for Perrin to act in contravention of such a procedure would therefore be highly irregular. (I should add that the text of the Agreement itself still refers to ‘the Tube Alloys Project’. I don’t know when it was first called ‘the Quebec Agreement’.)

Pp 16-17: “Yet the Russian archives have now shown that on Saturday, 4, September – only 16 days after the signing – Sonia, sitting in Oxford, supplied the Red Army intelligence Centre with an account of all the essential aspects of the Quebec Agreement, along with ancillary details, sending them directly to Moscow by radio.” There is no evidence of this. There were no ‘ancillary’ details outside the text of the Agreement itself.

P 17: “The GRU archives record: ‘On 19 August 1943, in a secret personal message to Marshal Stalin, Roosevelt reported about their agreed plans for the surrender of Italy and other matters but there was no word about the fact that they had also made an additional secret agreement about the use of nuclear weapons.’” This commentary, if it does indeed exist, is an historical observation, made later, and not an original archival entry.

P 17: “What Stalin regarded as his allies’ perfidy inevitably affected his attitude when, on 28 November, he met Churchill and Roosevelt in Tehran to discuss both the war and the postwar situation.” This is pure speculation. Stalin would have had a chance to react earlier; there were many other reasons for him to express annoyance; alternatively, he may have heard about the Quebec Agreement from other sources in October or November.

P 18: “In 2006, the release of the private papers of Vyacheslav Molotov, Stalin’s most trusted deputy, revealed that, on 15 October 1943, another British spy had supplied details of the early plans for Operation Overlord, the Anglo-American assault on Normandy. . . . They included the estimate that the ‘second front’, which should reduce pressure on the German forces, was unlikely to be attempted before the spring of 1944. Small wonder that Stalin was ’prickly and suspicious’ at Tehran.” Stalin knew about the timing of Overlord in June 1943, as The Kremlin Letters show.

P 18: “Whether Sonia appreciated the scale of her achievements is unknown because she never mentioned it to anyone and, probably on Soviet orders, withheld it from her memoirs.” Why?

P 18: “Her [Sonia’s] Quebec coup was extraordinary for another reason: she was so heavily pregnant that she gave birth only four days later.” Indeed.

Pp 18-19: “The GRU archives show that Sonia also included the fact that some senior Americans, both military and scientific, had reservations about any atomic partnership with Britain. So, she had clearly been given a summary of all the atomic aspects of the Quebec Agreement, which had been circulated in secrecy in some British government department and abstracted by some high-level traitor.” This is pure speculation. It is highly unlikely that any such analysis would have been attached to the Agreement. The Agreement was exclusively about the Tube Alloys project, and hence there were no non-atomic aspects to it. The implication here, moreover, is that Perrin was a ‘high-level traitor’.

P 19: “In July 2011, the Moscow-based historian Dr Svetlana Chervonnaya reported having discovered a Soviet document confirming that Sonia had sent the information about the Quebec Agreement on 4 September 1943 and that, after translation into Russian, it was taken straight to Stalin.” We have no evidence of this serendipitous discovery.

P 19: “It seems certain that the information was delivered to her in documentary form rather than verbally [sic: ‘orally’], because on 4 September she also transmitted a complete list of the 15 British scientists who had already been selected to move to America.” This is nonsense. The scientists were not selected until two months later. If any such document is produced, it must undeniably be a forgery.

P 19: “When her coup was made public in 2002, in a GRU-sponsored book, Lota’s The GRU and the Atomic Bomb, the GRU’s Colonel General Alexander Pavlov, who vouched for its authenticity in a foreword, was at pains to point out that ‘the time has not yet arrived when still unsuspected or unproven wartime sources can safely be named.’.” I have not seen this book, but would mistrust any ‘GRU-sponsored’ publication, just as I would mistrust any ‘MI5-sponsored publication’, such as Alan Moorehead’s Traitors, or Alexander Foote’s Handbook for Spies. And, if the evidence existed, why did Chervonnaya have to stumble upon it? Since Sonia had already been named, and her (GRU-sponsored) memoir published, why was there reticence in naming names? And the ‘source’ who ‘abstracted the document’ must have been (according to Pincher) not Hollis, but Perrin.

Chapter 23:

P 185: “The culprit could not have been Klaus Fuchs. He knew about the British determination to send scientists, including himself, to America, if concord could be reached, but he would not have had access to such a particularly secret political document at Birmingham University.” This is a red herring: in his enthusiasm to indict Hollis, Pincher misses the point. Fuchs could not have passed on details of the Quebec Agreement, but, as Tyrer suggests, citing Lota via Chervonnaya, Fuchs had met Sonia and passed on details of the departure of Chadwick, Peierls, Simon and Oliphant for Washington. This is quite plausible to me. Tyrer muddies the waters by indicating that Sonia also learned from Fuchs the outcomes of the conference in Quebec. As you point out, Chervonnaya inexplicably omitted to tell Pincher about this. You do state, however, that ‘there is agreement from the supposed GRU sources that information (unspecified) was sent by Sonia on 4 September about the Quebec Agreement’, but that is not strictly true. Tyrer’s passage does not mention ‘the Quebec Agreement’, only ‘outcomes from the conference’.

P 185: “Further, the GRU archives show that after seeing Fuchs in mid-August Sonia did not meet him again until November.” So what do these archives have to say to us about the mid-August meeting? If Fuchs did indeed meet Sonia in mid-August, why would she wait until her childbirth was imminent to inform her bosses of his message about Chadwick and co.?

P 185: “Nor could the traitor have been any of the minor agents Sonia claimed to have recruited. He had to be a spy with exceptional high-level and rapid access to top-secret political information and who could safely visit Oxford and deposit a document near Sonia’s house.” But this does not describe Hollis’s role and access. He would have been totally reliant on Perrin.

P 185: “The GRU had no known spy in the Foreign Office at that time anyway . . .”  Unknown – by whom?

P 185 “  . . .according to Gouzenko, information from ELLI was so highly prized that it went straight to Stalin, as the Quebec information certainly did.”  I have not been able to verify this claim yet, though it may well be true.

P 186 “Elli was in MI5, where the director general, Sir David Petrie, would have received a copy of the Quebec Agreement because of his responsibility for its security. . . . . who then decided which divisional heads needed to receive it.” Well, maybe. If it were that sensitive, it would have been much more likely that information was passed down orally. There was nothing in the Agreement, moreover, that addressed security aspects, such as vetting of scientists. That all came later.

P 186: “Because of his close connection with Michael Perrin when clearing scientists like Fuchs to work on it, he had known about the project for more than two years. He was also one of the few people who knew the names of all the scientists chosen to move to America, because he had been involved in clearing them for work on the bomb there.” The first part may well be true. But Pincher does not claim he had exclusive knowledge, and the record backs that up. The Fuchs archive (KV 2/1245-1) shows that Michael Serpell and Milicent Bagot were shocked when the Department of Scientific and Industrial Research contacted MI5 on November 17 to determine whether it had any objections to Fuchs’s departure.  (Bagot was outraged.) Major Garrett of D2, the security liaison officer with the DSIR, is involved, and states that “FUCHS’ name was not on the original list of workers going to the U.S., sent me by D.S.I.R., but I imagine his name was added at the last minute.” (This may suggest a plot to get Fuchs in under the radar, since Peierls had recommended him back in August.) Hollis does not seem to have been involved, and his F Division was not even the first to hear about DSIR requests. The documents were flying around everywhere: Otto Frisch had to be granted naturalization papers in two days.

P 187: “Clearly, Perrin knew the nature of the Quebec Agreement in advance. He also knew that the agreement had been signed because he had been given the ‘all-clear’ to dispatch the first batch of scientists He is also likely to have confided in Hollis to explain the reasons for the haste.” The first sentence is partially true, although the fact that something called ‘The Quebec Agreement’ might be signed, or even the understanding of what issues Roosevelt and Churchill would agree on, is not. The second is ridiculous, as the experts arrived in Quebec just before the signing, and they were not ‘the first batch of scientists’. It is extremely unlikely that Hollis would have had to be informed of their departure: they were all British citizens.

P 187: “The text of Superfrau states: ‘On 4 September, Sonia reported data on results of the conference.’” Note that this says nothing about the Quebec Agreement.

P 187: “On 4 September, Sonia also transmitted a list of the atomic scientists chosen to work in America. Where did she get it? The fact that the first batch of scientists had all been cleared in advance is shown by the haste with which they arrived in America. Hollis was involved with their clearance.” This is nonsense. The four who left in August had all been to the USA before. They were citizens of the UK (including Peierls), and did not need any clearance. Their visit was temporary. As Gowing reports, Wallace Akers, confident that an agreement would be signed, on August 10 cabled to London, with Anderson’s approval, that Chadwick, Simon, Oliphant and Peierls should leave immediately, for purposes of imparting information. Yet Peierls would have had an opportunity to pass on to Fuchs the reason for his journey.

P 188: “Early September was also the time when Hollis and Perrin were actively engaged together as a matter of urgency in going through the motions of vetting more of the scientists for despatch to America.” Again, this is pure fantasy. No names had yet been submitted. Where does Pincher get this information from? (But was Pincher suggesting that Perrin was a co-conspirator?)

P 188: “As the leak was of such massive proportions, the supplier of the information is also likely to have been someone who could be confident that neither Sonia nor her transmissions were under surveillance. That factor reduces the candidates to a small number, of whom Hollis was certainly one, as the evidence of the reliable officer Kenneth Morton Evans has confirmed.” We have discussed this off-line. Morton Evans was by no means a reliable witness, and fed some absurd stories about RSS passing the transcripts of encrypted wireless messages to Hollis and Philby. Something to be investigated and analysed separately.

P 188: “Sonia’s receipt of super-secret information of such consequence supports the contention that the GRU had posted her to Oxford specifically to service a source who had high-level access and operated in that area.” It is true that the decision to send Sonia to the UK occurred in the same month that most of MI5 was posted to Blenheim (October 1940), but Sonia did not know where in England she should go until she arrived in Lisbon. Milicent Bagot reported, however, that Sonia’s father, who was lecturing in the city, took up temporary residence in Oxford on December 13, 1940. An area for further investigation!

What this boils down to, for me, is that Pincher was a fantasist, and a highly unreliable chronicler. That may be attributable to his obsession over Hollis, but he may also have been fed false information that would bolster the case against Hollis, as a diversionary tactic. None of this is to deny that Hollis showed a pattern of highly irregular behavior, but it may have been due to incompetence rather than malignance. And he was not alone in MI5 in his (selective) indulgence towards Communism.

Tony Percy, January 2020

Response by Denis Lenihan, February 20, 2020

It was not necessary for Perrin, much less Hollis, to have received a copy of the Quebec Agreement at an early stage. As you say when commenting on Pincher p 185, ‘If it were that sensitive, it would have been much more likely that information was passed down orally’. Further, Perrin at least would have known what was being proposed, and in all likelihood would in the first instance have been told of the outcome orally (as an aside, like you, I deplore ‘verbally’ being used in this context); so that he could go ahead with making arrangements – which would have extended beyond arranging for scientists’ travel. The ‘Quebec decision’ that Liddell learned about on August 27 from Lennox presumably came by the same means?

You ask whether Perrin informed Hollis of what was then being called the ‘Tube Alloys Agreement’ before Churchill and his party left for Canada? I suspect the answer is yes, and nothing wrong with that from Perrin’s point of view, since after it was signed he needed Hollis to get cracking on clearing the next batch of scientists (and he may have been pushed by Hollis to let him know the outcome). Somebody at some early point, if the GRU is to be believed, told Sonia – after it was signed.

Also on the Quebec Agreement, you say that ‘Tyrer’s passage does not mention ‘the Quebec Agreement’, only ‘outcomes from the conference’; true, but Lota referred specifically to ‘the conference in Quebec’. If these are not references to the Quebec Agreement, as it came to be called, then to what do they refer?

Under Pincher p 187 you say that ‘It is extremely unlikely that Hollis would have had to be informed of their [the first batch of scientists] departure: they were all British citizens’. While I can’t lay my hand immediately on any evidence, I’d be surprised if British-born scientists were not also checked (or supposed to be checked) if they were going to the US for this sort of work. (Oliphant was Australian-born but in those days – until 1949 – would have had a British passport). As you point out, however, the four had all been to the US previously, so there would have been no need for any additional check on this occasion if they had been checked previously. Nunn May went to Canada, where the word of the mother country in those days would have been sufficient. Clearly Pincher got his two lots of scientists muddled, however.

You’re right about it being incredible that Pincher did not interview Perrin. He may have tried and Perrin refused. Perrin may have been willing to talk to Williams about Fuchs, but not to Pincher about Hollis. Another straw in the wind? We really do need that Pincher archive in Kings College.

You ask: ‘If Fuchs did indeed meet Sonia in mid-August, why would she wait until her childbirth was imminent to inform her bosses of his message about Chadwick and co.?’ The answer to this might be as you suggest in Sonia’s Radio that Sonia sent two separate messages: one about the scientists (perhaps in mid-August after meeting Fuchs) and another in early September having received a message from Hollis/Elli/Whoever about the conference outcome.

I note that you do not address what I think is one of my strongest points: the message about the conference went from Sonia in Oxford, not from the Embassy in London, thus pointing to an Oxford source – and a high-level one at that, not many people being in the know about Quebec.

While many of your comments about particular passages of Pincher in Treachery are right, it is with respect a bit over the top to describe him as ‘a fantasist and a highly unreliable chronicler’. He was a superb reporter with a genius for getting people to talk to him. More often than not he was right during his career; otherwise he would not have flourished in the way that he did. He may have got it wrong or got out of his depth here and there in Treachery but some of this may as you say be down to being fed false information. Like an historian, a reporter is only as good as his sources.

February 2020

Contribution by Richard Learie, February 26

Richard is another electronic colleague, living in Brussels, whom I have regrettably not yet met. He is an eager and well-read student of the tribulations of MI5, and of the efforts to unmask ELLI, and I think it is safe to describe him as a coldspur enthusiast. Here follows his contribution, for which I heartily thank him.

                                                                                                                                                Brussels 17 February 2020

Dear Tony

I offer my comments to your intriguing proposal after reading Denis Lenihan’s post. I read your comments in reply but that hasn’t changed my views which remain in summary:

  • I believe Hollis was ELLI
  • It is likely he passed the Quebec Agreement directly to Sonia

I’m not sure if most of my comments are within your scope because Sonia and the Quebec Agreement is but one of many strands to the “Who was Elli?” debate. I will start with the general debate and then add my small contribution to the detailed matter at hand.

The ELLI Debate

I have written to you before on the ELLI debate and I still consider that the weight of evidence on Hollis (despite it all being circumstantial) means that it was probably him. Clearly we know that ELLI existed so to debunk Chapman Pincher entirely is problematic for 2 reasons: if it wasn’t Hollis then who was it? – there cannot be too many possibilities/candidates!! And debunking Pincher is a long and arduous task since Treachery is over 600 pages of the case for the prosecution containing his many so called “anomalies” (50 or so from memory).

If I may say, Tony, that here we have a classic difference in styles. It is clear to me that, as stated in the Introduction to your book, you are a serious researcher, analyst, writer and historian who uses a very scientific methodology to ensure that your output is always of a high quality. In my opinion your book and website prove that this Introduction is no bluff and the detail in your work is prodigious and well thought-through. In contrast, Pincher was a journalist who did not always quote his sources and “Treachery” is a kind of character assassination rather than a piece of work “keeping all options open”. However, Pincher certainly found a lot of circumstantial evidence pointing to Hollis, although I admit that in Treachery he always seems to be trying to use evidence to prove Hollis’ guilt. That is consistent with what journalists do to provoke interest in their stories.

Nevertheless, I see the ELLI debate as a bit like solving a murder. We know the crime has been committed and we need to investigate the prime suspects. Graham Mitchell was also investigated within MI5 by his own junior officers but they quickly realised that Hollis was a much better fit and his background was ideal and similar in many respects to the Cambridge 5. There were not many candidates to investigate at all because ELLI obviously had high level access to material. Hollis’ background and inaction/incompetence would be well explained if he were ELLI. I am particularly persuaded by the University drop-out in Shanghai element to his background where he admitted to meeting Agnes Smedley who knew Richard Sorge and Sonia. It seems obvious that there would have been attempts to recruit him. He had left wing views and was at a low point in his life. Also on his return to UK he applied for jobs at the Times and the Intelligence Services – classic GRU/KGB advice!!

So I personally (for what its worth as solely an avid reader) remain in the “Hollis was ELLI” camp. As I read a lot of the espionage material from this period, I am continuously frustrated by authors who claim that the spy in MI5 story is a conspiracy theory. Clearly it is not and it’s the greatest intelligence mystery out there. Let’s hope there is a breakthrough one day.

I have not noticed that you have nailed your colours to the mast either way. Denis Lenihan’s papers on Paddy Costello presents another “anomaly”. I hope that this is “the straw that breaks the camel’s back” for you personally.

Sonia and the Quebec Agreement

Coming back to the matter at hand. I will follow the Sonia/Quebec Agreement issue with great interest.

If it was Fuchs who supplied the Agreement to Sonia, then it was under Hollis’ nose because both Fuchs and Sonia were on his radar. That would be consistent with Hollis’ inaction/incompetence strategy as an agent of influence that was to become so damaging to UK/US relations.

It could have been Hollis himself of course because he was in the Oxford area. Also Hollis appears to have had an excuse for visiting the area where Sonia lived because I seem to remember reading she was renting a Laski property.

Hollis was clearly a very cautious man in his MI5 career. I believe this character trait served him well and explains why he got away with it for so long.

Tony, you have contributed enormously to the debate with your epic series of posts about Radio Security matters and the Sonia mystery. I really look forward to any progress you make with the theory that MI5 aided Sonia’s arrival in the UK in order to keep an eye on her and use her for mis-information and to trap Russian spies in the UK. Have any other writers written to provide support? Your theory makes sense but it seems that it was spectacularly bungled? The interesting part would be then “was that due to incompetence or was it due to the work of ELLI (or both)”? Whichever it was a massive cover up has been carried out ever since. I believe part of this cover up was the Skardon/Serpell visit to Sonia in Oxfordshire. MI5 seemed to fear that their bungling could be exposed and clumsily tried to practice some damage limitation. Frank Close (and others) seem to think they were outwitted by Sonia. But they had so few cards to play and so little to gain at that stage except to limit further damage on UK/US relations. I don’t think they even wanted to outwit or trap her. Certainly by that stage there was no way MI5 could ever let her be prosecuted.

Finally, if Hollis was Elli (and also if Elli is revealed to be someone else) then clearly the espionage history of this period needs to be reassessed from top to bottom. But at least it would resolve a few mysteries whilst leaving MI5 to sort out the exposure of their biggest ever cover up.

Good luck with this experiment, Tony.

Best regards

Richard LEARIE

Response to Denis and Richard by Tony, February 26

Thank you, gentlemen! I wonder whether your eagerness to seal the case on ELLI means that you have been a bit too forgiving of the evidence  . . .

The questions that Denis poses about aspects of the ‘Quebec Agreement’ prompt me to suggest something else. First, we must remember that there were two parts to the decisions that were made at Quebec. There were the decisions about opening the so-called ‘Second Front’ in 1944, which were secret, but widely dispersed among military and government departments. (It was on those decisions that Liddell was briefed by Lennox: Liddell does not refer to any ‘Quebec Agreement’.) And then there was the highly confidential Tube Alloys Agreement, which concerned sharing of skills and technology in the realm of atomic weaponry and energy, between the USA and the UK (and Canada), which was so secret that neither Roosevelt’s nor Churchill’s cabinet ever knew about it.

Thus, whenever a writer refers to the ‘Quebec Agreement’, one needs to ascertain to which part of the overall decisions he or she is referring (such as when Tyrer talks about ‘outcomes from the conference’). I should like someone to point out when the phrase ‘Quebec Agreement’ first came into existence, and whether it was then used to replace the term ‘Tube Alloys Agreement’. It appears as the header in the transcript at Appendix 4 of Margaret Gowing’s history, but, on p 171, she writes that Anderson handed to Churchill and Roosevelt the ‘Tube Alloys Agreement’ for signature, and then she simply slips into calling it the ‘Quebec Agreement’. When Roosevelt and Churchill signed an aide-mémoire at Hyde Park on August 18, 1944, confirming the ‘utmost secrecy’ of the project, it contained no mention of a ‘Quebec Agreement’, but continued to refer to ‘Tube Alloys’. Furthermore, the official US Army copy of the Tube Alloys agreement (see https://avalon.law.yale.edu/wwii/q003.asp ) introduces the Articles as ‘The Quebec Conference – Agreement Relating to Atomic Energy’. The document was created from a photocopy of the British original.

It thus seems to me that, because of the ambiguities, and the need for compartmentalization, it would have been highly unlikely for any of those in the know at the time to have referred to the overall conclusions, or the separate Tube Alloys Agreement, as the ‘Quebec Agreement’. When Lota refers to ‘the conference in Quebec’, he may accurately not be referring to what has become to be known as the ‘Quebec Agreement’. When you claim, Richard, that Hollis passed ‘the Quebec Agreement’ directly to Sonia, I think you need to explain what document you believe that was, and why a highly secret subset of all the written outcomes was identified that way. Any claim for the existence of a ‘copy’ of a comprehensive ‘Quebec Agreement’ must be treated very sceptically, as no document containing all the decisions signed in Quebec exists (so far as I know). And any archival material in Moscow that actually referred to the combination ‘Quebec Agreement’ in September 1943 would reflect some tantalising ingenuity.

I repeat my assertion that, even if Perrin had divulged to Hollis the terms of the upcoming Tube Alloys Agreement, it would have been very premature to refer to it as the ‘Quebec Agreement’ before the participants had even congregated there. Moreover, it makes no sense to talk about ‘the next batch of scientists’. The four whom Akers had shipped out before the Conference were not ‘the first batch’: Vannevar Bush told Akers that his move was premature, and that the role of any British scientists on the project would have to await the meeting of the Combined Policy Committee. There was no time for any further check on the four, even if it had somehow been required. As I think we all agree, the four had already visited the United States in their scientific capacity. It was weeks before the other scientists were chosen. Hollis never had to ‘get cracking’, and was not even in the picture, as I showed in my last response.

The reason I did not address what Denis describes as ‘one of my [his] strongest points’, namely the claim that, since the message about the conference went ‘from Sonia in Oxford’, it thus pointed to an Oxford source, is that I have still to see any solid evidence that a message from Sonia about the conference was received in Moscow. Offering the inescapable conclusion that ‘Pincher got his two lots of scientists muddled’ does not bolster the case.

I concede that my characterization of Pincher as a ‘fantasist’ was a little extreme. But it was intended to be provocative. Since he embellishes what could be a very workmanlike argument with so many absurdities, his worth as a reliable witness diminishes. Denis says he had ‘a genius for getting people to talk to him’, but Pincher may never have imagined that he was being used as a conduit for any number of damaging stories about Hollis that may have been designed to distract attention from another suspect, or simply to muddy the waters while other groups maintained his innocence.

As Richard indicates, I cannot yet come off the fence on ELLI. I agree that Pincher has presented an enormous amount of circumstantial evidence that could point to Hollis’s guilt, but in a Scottish court the verdict would be ‘not proven’. The apparent failure of Pincher to engage with Perrin is, to me, a vitally important example of ‘a dog that did not bark in the night-time’. If a clincher for Hollis’s being ELLI exists, it is definitely not the anecdote about Sonia and the Quebec Agreement.

P.S. I notice that the Wikipedia entry for the Agreement still boasts Tyrer’s claim that Fuchs supplied Sonia with the details. Does anybody know who the author is? Should he or she be introduced to this discussion?


Filed under Espionage/Intelligence, General History

4 Responses to Special Bulletin: Denis Lenihan on Sonia & Quebec

  1. coldspur

    I have posted my response to Denis’s very thoughtful and useful piece. I welcome his reply in turn – and, of course, any other comments!


    February 15, 2020

  2. coldspur

    I have posted Denis’s reply. For those readers who have access to Academia.edu, I recommend your turning to Denis’s articles there – particularly a recent posting titled ‘Roger Redux: Why the Roger Hollis Case Won’t Go Away’.


    February 21, 2020

  3. coldspur

    I have posted a contribution from Richard Learie, as well as my response to Denis’s and Richard’s comments. Not intended to be the ‘Last Word’ – let the debate continue!


    February 26, 2020

  4. Denis Lenihan

    Hello Richard!
    Our host Tony is such a reasonable man that if you and I keep piling on the straws about Hollis and Elli his back will surely break! I have a further straw in preparation.


Leave a Reply

Your email address will not be published. Required fields are marked *