From August 1996 until October 1997, the FBI prefaced nearly every Flight 800 news update with the assertion that its agents were intensively scrutinizing the evidence to determine whether the initial explosion was a bomb, a missile or the result of a mechanical failure, specifically in the fuel tank. Six years of a Freedom of Information Act lawsuit against the FBI has now established that the most precise characterization of the explosion was provided by the victims themselves.

Because the reports of many witnesses described a missile-like object rising from the surface, the FBI swarmed to Long Island and took control of the investigation as a probable crime scene utilizing literally hundreds of agents. Shrapnel and other foreign objects removed from the victims' bodies were therefore treated just like bullets found at any crime scene would be. They were recognized as very precious evidence that potentially described, in excruciating detail, exactly what initiated the catastrophic breakup of Flight 800. They were therefore immediately handed over to the FBI as some of the first and key evidence defining the nature of the blasts – a possible smoking gun.

Such evidence was regarded as decisive because there was a very good chance that much of it was created and propelled by the explosions that were reported by scores of eyewitnesses.

In the intense crucible of an explosion, the debris and shrapnel created usually retain residues and mechanical and physical characteristics that define the precise nature of that explosion. The different types of explosions have unique characteristics of temperature, pressure and duration, which create recognizable "signatures" of chemical residues, fragment size and composition. Thus, if a fuel tank explosion had destroyed the plane, its artifacts – low temperature, low energy, low velocity, no significant metal fragmentation – could be confidently identified and distinguished from the high velocity artifacts caused by a bomb or missile. These could be corroborated with certainty from the residues and the smaller and shattered fragments.

In contrast to all the evidentiary debris pulled out of the ocean, the autopsy-derived evidence is also the least likely to be randomly contaminated by ocean and sludge or have any explosive residues abraded or dissolved away. Many explosive residues are known to be soluble in ocean water.

And what did the foreign object evidence reveal? One of the many peculiarities of the investigation was that the FBI never released, nor even shared inside the investigation, any of those related details.

According to the county coroner, Dr. Charles Wetli, the FBI never provided him with any of those specific forensic details. That in itself constitutes a serious irregularity in the investigation and yet its very weirdness typified much of the probe. Dr. Wetli was obliged by federal law to relinquish the autopsy evidence items to the FBI. That makes sense because of the vast and sophisticated forensic explosive expertise the FBI had at its disposal and available on-site. But Dr. Wetli also has the statutory responsibility and obligation as the chief county medical officer to determine the cause and manner of death. The FBI, by law and by convention, is required to provide such forensic evidence to the coroner. Even 10 years after the event, he cannot conclude an adequate inquest without those very pertinent facts of the case.

Even more surprisingly, the FBI did not share those particular forensic results with the National Transportation Safety Board, either. In 1998, the NTSB responded to a FOIA request about the foreign object evidence by stating that the FBI had eventually transferred all of those actual objects to them, but without any forensic or descriptive details or documentation whatsoever. The NTSB probably still has this evidence, still with no clue as to what it all signifies.

I therefore filed a Freedom of Information request in September 1998 asking the FBI for:

• the listings of the objects taken as evidence from the bodies of the victims;

• the forensic and analytical details that the FBI had derived from such objects; and

• the identification and origin of such foreign objects.

By July 2000, the FBI had only provided me with 14 pages of records in their response to this request, and there were no forensic details included whatsoever. Therefore, with the support of victim family members, aviation employees and citizen researchers concerned with safety and scientific integrity, I filed a FOIA lawsuit to force the FBI to do a better search. A six-year legal battle to obtain those forensic details eventually produced 580 pages of very responsive records, but it still only produced one page of actual forensic test results originating with the FBI.

My lawsuit argued that the FBI had not performed an adequate search. The evidence I submitted in the case initially used some of those 14 pages, and eventually the 580 pages, because those FBI records made numerous references to forensic tests and actual folders of archived responsive FBI forensic data that had not been located or released.

In the course of the litigation, the FBI testified, via affidavit, that it had in the year 2000 "searched the responsive section" of the main investigation archive kept in the FBI's New York office. The agency stated that it was designated the "Sub FF" section, and it comprised almost 2,000 pages. The main file apparently has 97 such subsections totaling about 80,000 pages in all.

In 2004, under pressure from the ongoing litigation, the FBI performed a second search of those very same 2,000 pages and this time they discovered an additional 560 clearly responsive pages. To explain how they overlooked these 560 responsive pages in the first "page-by-page" search process, the FBI asserted that it had required a much more intensive "line-by-line" search. This was a disingenuous claim because most of those additional pages were very clearly responsive and did not require close scrutiny to recognize the fact.

Among the 580 pages were approximately 100 pages of evidence recovery log sheets that listed hundreds of foreign objects removed during the autopsies and given to the FBI. Those evidence recovery lists proved that the autopsy teams discovered hundreds of pieces of metal fragments, pellets, particles and other foreign bodies from the 230 victims. Also among those released records are memos referring to intensive forensic analysis and scrutiny and efforts to identify those foreign objects.

Buried among the 580 pages was one single page of FBI forensic test results relating to just one single autopsy object. It was the test summary produced by FBI explosive technicians' tests on an article of clothing. Those particular tests were negative for any "evidentiary value," that is, nothing suspicious or unusual. This test summary was apparently misfiled in this section of the archive.

The 580 pages also revealed that one sample of the autopsy evidence was sent to Brookhaven National Lab and two other items were sent to the Defense Intelligence Agency Lab. Some of the forensic results from those tests were released to another researcher, Don Collins, and can be seen at www.twa800.com. Some of these records and details of the litigation "Sephton v. FBI" can also be viewed at www.foiac.org.

Given its failure to produce the forensic results, in October 2004, Federal District Court Judge Michael Ponsor asked the FBI at least to provide an affidavit stating that the agency had performed a reasonable, good-faith search that looked in the places that responsive records could be expected to be found. Initially, the FBI agreed to provide such an affidavit, but when the court deadline arrived, the FBI was unwilling (or unable) to provide it. Despite that refusal, Judge Ponsor ruled in favor of the FBI in March 2005. He did not require the agency to broaden its search.

In March this year the First Circuit Appeals Court affirmed the decision of the lower court and in June denied a petition to review the case.

So six years of litigation has established that there was abundant evidence obtained during the autopsies that potentially defined the exact nature of the reported explosions, and that the FBI apparently now has no idea where to find that particular forensic evidence generated by their $40 million investigation.

When Flight 800 exploded, 230 people had their lives cut short and the fallout from the tragedy became one of the nails in TWA's coffin. What do those "missing" forensic test results show? Victims' family members, TWA employees and the public at large have a right to feel confident that the government's investigation was competent and scrupulous.

The NTSB has released thousands of the non-conclusive forensic details that explain why the explosion is still very mysterious. By winning its legal battle to continue withholding the most fundamental evidence, the FBI has won a pyrrhic victory, which further undermines the credibility of the investigation and the public's trust.

But there is a more far-reaching issue in this case, too. In our modern world full of complex man-made perils, we depend mightily on government scientific integrity and efficiency to inform us about such dangers and to keep us safe. And it is critical to be able to assert FOIA rights to allow the possibility of auditing government performance.

Initially, my concern was about the apparently ineffective investigation of this specific tragedy. Eventually, I became just as concerned about the vital importance of our Freedom of Information rights as one of the most effective tools we have for asserting accountability and auditing government performance in such circumstances. Those rights have been eroded over the last few years. and if the courts and the media allow this trend to continue it will be increasingly hard for anyone to find anything useful about government blunders or ever hold agencies accountable in general.

If the courts allow government agencies to perform ineffective searches and tolerate agency refusals to give reasonable assurances that good-faith searches occurred, then our Freedom of Information rights are being undermined. Given all the above, I feel compelled to now petition the Supreme Court. With the current conservative composition of the Court, my petition is not likely to sway that august body unless other concerned parties such as our "watchdog" media take up the issue as well. We will only ever have government integrity and transparency if we stand up and demand it.

Read previous installments in this series:

Part 1: "New data prove CIA 'zoom-climb' a fraud"

Part 2: "270 people saw plane shot out of the sky"

Part 3: "Richard Clarke's politicized exit strategy"

Part 4: "1 secret the Times has kept"

Part 5: "What Jamie Gorelick knew"

Related special offer:

Get Jack Cashill's groundbreaking exposé, "First Strike: TWA Flight 800 and the Attack on America"

Graeme Sephton works as a telecommunications engineer at the University of Massachusetts in Amherst. More litigation details are at www.foiac.org.