Sunday, 17 February 2013


I attended a talk on the Cambridge Five given in Oxford yesterday by Professor Christopher Andrew (author of The Defence of the Realm: The Authorized History of MI5). I was keen to sit in on the talk as my new book ‘Faktion’ is currently out with publishers and Philby and Cairncross (two of the ‘Five’) feature in the story. Moreover, I hoped that Professor Andrew would speak about the mysterious ‘Elli’, the GRU agent who penetrated MI5 in the early years of the Second World War  and whose identity has never been revealed  and has been much speculated on ever since.

This is an important subject for me, given that the unveiling of the identity of Elli is the prime thrust of ‘Faktion’s’ plot and I was hoping that nothing the good Professor would say would lead me to doubt my own – somewhat radical – identification of who Elli was. He didn’t … in point of fact he hardly mentioned Elli apart from confirming – in response to my question as to how confident he was that Elli was Leo Long (his preferred candidate) – that he was ‘very confident’.

Very confident he may be, but I also think he is very wrong.

First, perhaps, a brief summary of what is known for certain about Elli (much of this coming from the defector Igor Gouzenko, a Soviet cipher clerk who gave himself up to the Canadians in September, 1945, rather than be shipped back to the USSR):

·         Elli was a man;

·         He was a GRU agent who had access to MI5 intelligence (that he was a GRU agent is important, the Glavnoye Razvedyvatel'noye Upravleniye being the foreign military intelligence directorate of the General Staff of the Red Army. The Cambridge Five were run by the NKVD, the Peoples Commissariat for Internal Affairs, Narodnyy Komissariat Vnutrennikh Del, which was the forerunner of the KGB and the secret police organisation that implemented the orders of the Soviet Communist Party. The rivalry between the NKVD and the GRU was long and bitter);

·         He had access to sensitive government policy documents (it was Elli who tipped Stalin off about the Quebec Agreement, signed in August 1943, under which the UK and the US agreed to collaborate on the building of the atomic bomb);

·         Gouzenko first heard of Elli when on a cipher course in Moscow in 1942, so Elli must have been active before this date;

·         He is described as ‘ou nego shto-to Russkoe’ … having something of the Russian about him; and,

·         He was active during the 1940’s, dropping off the intelligence radar circa 1950.

Pretty poor pickings information-wise to identify an agent who was, perhaps, the most important of all the Soviet agents – and there were a lot of them – to have penetrated Britain’s intelligence community.

And this paucity of information has, as might be expected, stimulated a great deal of controversy regarding the identity of ‘Elli’. That’s why I went to Professor Andrew’s talk: to see if I had gone wrong in my own surmising as to the identity of Elli.

Christopher Andrews claims Elli was Leo Long, who he describes as a ‘sub-agent of Anthony Blunt’. In support of this contention he cites the un-corroborated testimony of Oleg Gordievski (a Soviet defector) and the ‘fact’ that ELLI in Russian translates as the plural of the English letter ‘L’. (I am reliably informed by my in-house Russian expert – Nelli – that this is ‘bollocks’: no Russian would ever refer to double-L as Elli. And, anyway, I doubt the Soviets would have been so stupid as to give their most important agent a code-name made up of his initials). But of equal concern is that as Leo Long was a sub-agent of Anthony Blunt’s he would have been run by the NKVD. Elli was run by the GRU. No … the Leo Long hypothesis is weak to the point of unsustainability.

To my mind, the journalist Chapman Pincher has a more coherent set of arguments when he suggests that Elli is Roger Hollis – Hollis worked for MI5 from 1939, eventually becoming Director-General in 1956 (see Pincher’s Treachery: Betrayals, Blunders, and Cover-ups: Six Decades of Espionage Against America and Great Britain for details). A large part of Pincher’s ‘evidence’ supporting the ‘Hollis is Elli’ conjecture is persuasive but circumstantial. For instance, he refers to the ‘ou nego shto-to Russkoe’ description of Elli as inferring that Elli had pre-Revolutionary connections with Russia (Nelli disputes this: in her opinion the phrase is so ambiguous as to invite several, equally plausible, interpretations) and apparently the Hollis family is able to trace its lineage back to Peter the Great. But my biggest problem with the ‘Hollis is Elli’ hypothesis is that Hollis would have been too junior in the period running up to 1942 (when Gouzenko first heard of him) to be regarded as the Soviets’ most important intelligence asset in Britain (apparently Elli’s intelligence was so highly thought of that it went straight to Stalin). Moreover, Hollis’s Russian antecedents were tenuous to say the least and I am doubtful they would have stimulated gossip in the GRU’s Moscow cipher school, where Gouzenko heard the rumour.

So, if Elli is neither Long nor Hollis, who was he?

We are looking for a man who had access to important intelligence (the sort not available to junior agents in MI5); was so highly regarded that he was run separately from all the other first-rank Soviet agents in Britain; was active before 1942, leaving the stage around 1950; had pre-existing connections with Russia; AND whose identity is so sensitive – embarrassing? – that it has been very effectively protected by both the Russians and the Brits from enquiring journalists and historians ever since.

And I think I know just the man who ticks all these boxes. Elli isn’t anyone the historians or the journalists have ever flagged. Yeah, I know who Elli was.

Sorry … but you’ll have to read ‘Faktion’ to find out who my candidate is!