That meant 22 lawyers in the room - five of senior counsel - as well as former Justice Brian Martin, who is conducting the inquiry. These, and 10 more public servants with professional reasons to be there, are sucking in the limited air at public expense. So too, if not at public expense, are half a dozen journalists and spectators.

The fresh team was not there to push things along, of course. Like most of the other teams, other than counsel assisting and the Eastman team, it was demanding sharp limitations on the scope of the inquiry, threatening a trip to the Supreme Court for a restraining order if Martin did not comply.

A Full Court is already reserved in its judgment on earlier efforts on behalf of the Director of Public Prosecutions, John White, designed to effectively shut down the inquiry.

Martin is being brave in pressing on regardless in an effort to make sure that the inquiry - commissioned about 1½ years ago - comes to a conclusion some time.

Martin is now hearing from an expert scientific witness from Northern Ireland, Dr James Wallace, whose time in Australia is limited. He has already rescheduled hearings on scientific evidence several times, chiefly to meet demands of a key forensic witness at the trial, Bob Barnes. Reorganising matters again would cause even more expense and delay.

Freckelton appears for Barnes, who was once put up as an expert on firearms and explosives residue by the Victoria Police forensic sciences laboratory. Barnes, who has a diploma in metallurgy, but whose membership of a string of professional associations allows him to put any number of initials after his name, flew to Canberra on the night of the assassination. It was his work that identified the murder weapon as being of a particular type, then a particular rifle. He was the one who linked microscopic gunshot residue vacuumed up at the murder scene to residue in the boot of Eastman's car. The case against Eastman was circumstantial. This was what most closely linked Eastman to the murder.

In the years since, the work and the reputation of Barnes have come under fierce attack. Eastman's jury was not told Barnes had been sacked from the forensic laboratory for alleged professional and scientific misconduct, and that his work had been stringently criticised by Victorian judges.

One expert on the role of the scientific witness is Ian Freckelton, SC, now, by coincidence appearing for Barnes. He summed up some of the criticisms of Barnes in an academic paper in 1997.

As a senior police-employed forensic scientist, Barnes ''had a disturbing history of giving biased, pro-prosecution evidence,'' Freckelton said. His evidence in one case had highlighted the dangers posed by an expert witness ''whose opinions are not readily reviewable or are subtly skewed in favour of the side calling him''.

Further, ''Barnes as a senior scientist had been known for his refusal to submit his findings to peer review and had an inclination to give evidence that was both overstated and ambiguous but favouring the prosecution''.

Similar allegations - and more - are made against Barnes in the Winchester case. Last week counsel assisting, Liesl Chapman, SC, foreshadowed a fundamental assault on his evidence and findings at the Eastman trial, as well as on his general credibility. She has expressed doubts about whether some of the tests Barnes said he did were performed at all. She will be leading evidence from other experts saying that critical findings were not supported by evidence. The biggest doubt is whether the evidence proves gunshot residue in Eastman's car was the same as found at the murder scene.

Freckelton's first tack was to say that evidence Chapman wanted to lead was beyond the jurisdiction of the inquiry. Being overruled on that, he asked the inquiry to adjourn to allow him to appeal the ruling in the Supreme Court.

Martin, concerned with getting

on with a case that has been repeatedly delayed, refused. But pressed, he agreed to put reporting restrictions on criticism of Barnes by other experts, at least pending action by Freckelton to take an appeal.

The press and public are not excluded, but cannot report proceedings, or see affidavits. This is based on Freckelton's argument that his client's reputation could be trashed in public for weeks, without his having a right to be heard in reply. Some others might remark that almost all that can be said is anyway already in the public domain.

Freckelton, with suppression orders in place and able to claim preoccupation with the inquiry itself, has no reason to rush to file his papers.

Martin's caution is understandable. He already has considerable material which could lead him to express doubts about the safety of a conviction. On what we know, there's more to come. Some evidence taken is quite disturbing - a reason why some parties are trying so hard to prevent criticism of them being made.

That criticism seems likely to embrace the conduct of the trial judge, some decisions of the prosecution team, some of the expert evidence and the quality of the police investigation, including police tactics said to be designed to make Eastman blurt out a confession.

Martin has yet to turn his mind to the findings he must make. But if his inquiry is aborted by the courts - by challenges from parties whose costs are being paid by the ACT government, one can expect that he would be shouting loudly about the potential for serious injustice to Eastman, still behind bars 18 years later.

That would embarrass ACT Attorney-General Simon Corbell into doing what he should have done in the first place. That's convening an open-ended executive, rather than judicial, inquiry into a potential miscarriage of justice, unable to be aborted by technical and jurisdictional points. Had Corbell done this in the first place, perhaps with the present inquiry personnel, the case would probably be over now, and at a good deal less cost to taxpayers.

It costs much more to run a very narrow inquiry when interested parties can take most of the time parsing the terms of reference, and can run to courts, at public expense and without accountability, whenever they get a ruling they do not like.

Jack Waterford is editor at large