A former KGB officer and current deputy in the Russian parliament, Andrei Lugovoi, who met with Alexander Litvinenko on November 1, 2006, has suggested that Litvinenko worked for MI6 and that the exiled oligarch Boris Berezovsky had a hand in Litvinenko's sensational death from radiation poisoning

According to the official version -- endlessly and uncritically recycled in the British and American press -- the British police have conclusive evidence that Litvinenko was poisoned in the Pine Bar of the Millennium Hotel by Andrei Lugovoi. It is unambiguously clear that this meeting in the Pine Bar occurred in the late afternoon of November 1, 2006, and followed, rather than preceded, an earlier meeting between Litvinenko and the Italian self-styled 'security consultant' Mario Scaramella at the Itsu sushi bar in Piccadilly.

According to the account given by Gordievsky in an article in the UK Times of London on January 20 last year -- which also quoted sources in the police, with whom Gordievsky was said to have 'worked closely' -- the poisoning took place in a fourth-floor room at the Millennium. This was attended not only by Lugovoi but by a mysterious Russian, not known to Litvinenko, and referred to as 'Vladislav' -- who was the actual murderer. This meeting is said to have taken place before, rather than after, the meeting with Scaramella.

(See http://business.timesonline.co.uk/tol/business/markets/europe/article1294705.ece.)

In addition to this Times story, it is helpful to consult Chapter 45 -- 'Suspects Galore' -- of The Litvinenko File, published by the former BBC Moscow Correspondent Martin Sixsmith in April 2007. Besides repeating Gordievsky's account, Sixsmith interviews another London-based ex-Chekist, Boris Volodarsky, to whom the police have also apparently talked, who again claims there was an earlier meeting, and states categorically that the CCTV footage of the Pine Bar meeting makes it clear that Litvinenko could not have been murdered there. This version has been further elaborated by Volodarsky in a long riposte to Epstein's story posted on the New York Sun website. Anyone seriously interested in the case should read this -- along with the fascinating if perhaps somewhat mystifying comment from Karon von Gerhke. Either Volodarsky and Gordievsky are indulging in fantasy, or they are indeed reporting evidence that the British authorities do possess but have decided to keep both from Russian investigators and from British journalists.

According to the Times story, 'Litvinenko was reportedly able to give vital details of his suspected killer in a bedside interview with detectives just days before he died on November 23 at University College Hospital.' The story quotes Gordievsky as saying that Litvinenko 'remembered the man making him a cup of tea.' Litvinenko's belief, Gordievsky tells us, was 'that the water from the kettle was only lukewarm and that the polonium-210 was added, which heated the drink through radiation so he had a hot cup of tea.'

A rather different account was however given by Gordievsky in an interview in Russian carried by Radio Liberty on December 18, 2006, which the translator David McDuff rendered into English and published on his blog.

(See http://halldor2.blogspot.com/2006/12/litvinenko-gordievsky-interview-ii.html.)

Again the actual killer is identified not as Lugovoi, but as a mysterious Russian, only this time called 'Volodya' rather than 'Vladislav'. But Gordievsky told Radio Liberty that 'everything is known', that he himself 'knew who the killer was on the fourth day', that British officials 'all know' but were 'doing it all step by step, in the correct way'.

As these accounts appear to be in tension with one another, it is easy simply to dismiss everything Gordievsky says. But this may be a mistake. The implication of Gordievsky's account is that he and British officials had a full account from Litvinenko of his meeting with Lugovoi soon after he was taken ill. And, surely, this is what one would have expected. According to a Daily Mail story last October, Litvinenko was indeed working for MI6, and Scarlett was personally involved in his recruitment -- in which case one would have expected Litvinenko or his wife to contact MI6 immediately he fell ill.

Moreover, even if one discounts the claims of the Mail, Gordievsky was, according to his own account, a close friend of Litvinenko, and we know that he is linked to the head of MI6, Sir John Scarlett -- his case officer when Gordievsky was KGB resident in London. So one would have expected that even if Litvinenko or his wife had not contacted MI6, they would have contacted Gordievsky, and he in turn would have contacted Scarlett. So British officials ought to have been as well-informed as Gordievsky suggests they were.

If one consults Chapters 9 and 10 of Sixsmith's book, it becomes glaringly apparent that he has swallowed disinformation designed to demonstrate that nobody took any interest whatsoever in Litvinenko's claims until following his BBC Russian Service interview on November 11. If moreover one searches through the archive of the Times and Telegraph papers, it appears that the first time one can find reference to Lugovoi is more than a week after that interview, when Litvinenko was close to death and it is unclear how far, if at all, he was capable of coherent communication. (He died on November 23, Lugovoi is first referred to on November 20). The information appears to have come first from Berezovsky's factotum Alex Goldfarb -- and according to Gazeta Ru, it emerged as the result of a telephone conversation between Litvinenko and Goldfarb on November 19.

It appears that at the time Litvinenko died, the police were still ignorant of the name of the hotel where Litvinenko and Lugovoi had met. Moreover, in all these early stories, the meeting with Lugovoi is indeed placed before the meeting with Scaramella -- and features both Lugovoi and a mysterious Russian known as 'Vladimir'. This fits with Gordievsky's 'Volodya', if not his 'Vladislav': but the ambiguity would be unsurprising, if in fact there had been the kind of conversation with a dying Litvinenko which Gordievsky claimed took place in the January 20 interview. What is clear is that, if the press reports are to believed, at the time when Litvinenko was almost certainly no longer capable of communicating coherently, the police investigating the case knew nothing nothing whatsoever about the Pine Bar meeting, or indeed about the existence of Dmitri Kovtun.

The natural conclusion is that regardless of the nature of Litvinenko's business with Lugovoi, he was anxious that the police investigating his case should not know about it. This hardly gels with the widespread British assumption that the motives which drove Litvinenko to become involved with Lugovoi were purely commercial, resulting from a lack of money (for which the evidence is actually less than compelling.) Whether Gordievsky and Scarlett, or indeed Berezovsky and his people, knew about any meetings with Lugovoi, and simply did not see fit to inform the detectives handling the case, is a most interesting question. That is the assumption which is necessary if both of Gordievsky's stories are to be regarded as at least having some connection with the truth. But it is not actually a particularly implausible assumption.

If one collates Sixsmith's account with that given by Volodarsky in responding to Edward Jay Epstein, moreover, one finds further evidence in support of the latter's skepticism about the notion that the contamination trail points unambiguously to the Pine Bar and has to originate with Lugovoi. In Sixsmith's version, Scaramella was first judged to be contaminated -- but it was then declared that this was a mistake. The Itsu was contaminated -- but this turned out to be because Litvinenko and Lugovoi had eaten there during the latter's previous trip to London. However, Sixsmith trips over himself, because he says that the copier in Berezovsky's office was contaminated by Litvinenko. According to his account, Litvinenko coped emails from Scaramella before the Pine Bar meeting, not after: when if the timeline is to work, he should have been clean.

According to Volodarsky, both Scaramella and the Itsu were contaminated, but this was because Litvinenko had been contaminated at the earlier meeting when he was murdered. The inconsistencies are puzzling, to put it mildly. And, crucially, signs of contamination appearing prior to the Pine Bar meeting really do make difficulties for the official British version -- although they do not for the Gordievsky/Volodarsky version. The famous contaminated teapot is sometimes presented as clinching evidence for the official British case. But Sixsmith himself contemplates the possibility it could have been used by room service, in which case it would be compatible with the Gordievsky/Volodarsky account. Moreover, given that the teapot seems to have been first mentioned in late January, about six weeks after it was first supposed to have been identified, it is difficult to be absolutely sure it exists.

A further problem with Gordievsky's claims as reported by the Times is that according to the official version, it was only just before Litvinenko died that the results showing he was poisoned with polonium, not thallium, arrived. If Litvinenko was talking about polonium heating his tea before he died, then some of those talking to him must have been given these results earlier -- or must have known all along that polonium was involved somewhere. But then, in the light of Epstein's work, neither possibility is to be discounted, surely?

In addition to Epstein's invaluable work, another crucial contribution is the story 'Why a spy was killed' which Cathy Scott-Clark and Adrian Levy published in the UK Guardian in January.

This portrays Scaramella and Litvinenko as pivotal figures in Berlusconi's disinformation campaigns against his political opponents. It describes Litvinenko's efforts to frame Alexander Talik, supposedly to get him to become a source on the gangster Semion Mogilevich -- who by coincidence was arrested in Moscow just before the story appeared. Among the many allegations against Mogilevich is that of being behind RosUkrEnergo and its predecessor Eural Trans Gas, the middlemen in the gas trade to and through Ukraine -- which Scott-Clark and Levy fail to mention. Also of course there are the allegations of involvement in the Bank of New York scandal -- now the subject of a lawsuit by the Russian authorities. And crucially, there are the old allegations of attempted smuggling of enriched uranium, which I think come from reasonably credible sources.

It is possible that MI6 -- using private security companies as a front -- was engaged in a bona fide investigation of nuclear smuggling. Equally, it is possible that Litvinenko was trying to frame figures in the Russian elite with allegations of nuclear smuggling. Both could be the case, at the same time, and there are many other possibilities: a number of them featuring the murky underworlds of the European energy trade. Equally, Litvinenko's death could simply have been an accident, derived from incautious handling of polonium -- as Lugovoi has suggested.

The earlier meeting described by Gordievsky and Volodarsky could or could not have existed. Even if it did exist, it might or might not have taken place at the Millennium. And its existence would hardly be proof that Litvinenko was murdered at it, and still less that there is a compelling case against Lugovoi. But then again, the Gordievsky/Volodarsky account does not have the complete incredibility of the suggestion that Lugovoi chose the Pine Bar as the appropriate place to murder Litvinenko. It could indeed be that there is a cogent case against Lugovoi. But if there is, it is a case which the British authorities are afraid to make public, because -- for reasons good bad or indifferent -- candor would involve revealing a lot of information they want to keep secret.

One thing that militates against any confidence in the role of the British authorities in the affair is precisely the kind of facile black and white thinking exemplified by the article by Robert Service which Patrick Armstrong is discussing. And then there is the history of British intelligence failures over Iraq. These were intimately bound up with a propensity uncritically to swallow disinformation produces by exiles and émigrés -- of whom key figures, such as Ahmad Chalabi, were clearly conmen. Both MI6 and Scarlett were at the heart of these failures. The possibility that they have once again fallen for a bunch of conmen -- Berezovsky and his minions -- cannot, I fear, be ruled out. And, while the competence at intelligence gathering of sections of British intelligence is very much in question, there can be little doubt about their skills in manipulating the British press.

David Habbakuk lives in London, and is a former television current affairs producer. After studying economics and history at Cambridge, he lived in Mexico, then worked on the Liverpool Daily Post newspaper and the Financial Times. Subsequently he produced programs for London Weekend Television and the BBC as an independent television producer.