By George Lardner Jr.
Washington Post Staff Writer
Monday, March 26, 2001; Page A03
The House Assassinations Committee may have been right after all: There was a shot from the grassy knoll.
That was the key finding of the congressional investigation that concluded 22 years ago that President John F. Kennedy's murder in Dallas in 1963 was "probably . . . the result of a conspiracy." A shot from the grassy knoll meant that two gunmen must have fired at the president within a split-second sequence. Lee Harvey Oswald, accused of firing three shots at Kennedy from a perch at the Texas School Book Depository, could not have been in two places at once.
A special panel of the National Academy of Sciences subsequently disputed the evidence of a fourth shot, contained on a police dictabelt of the sounds in Dealey Plaza that day. The panel insisted it was simply random noise, perhaps static, recorded about a minute after the shooting while Kennedy's motorcade was en route to Parkland Hospital.
A new, peer-reviewed article in Science and Justice, a quarterly publication of Britain's Forensic Science Society, says the NAS panel's study was seriously flawed. It says the panel failed to take into account the words of a Dallas patrolman that show the gunshot-like noises occurred "at the exact instant that John F. Kennedy was assassinated."
In fact, the author of the article, D.B. Thomas, a government scientist and JFK assassination researcher, said it was more than 96 percent certain that there was a shot from the grassy knoll to the right of the president's limousine, in addition to the three shots from a book depository window above and behind the president's limousine.
G. Robert Blakey, former chief counsel to the House Assassinations Committee, said the NAS panel's study always bothered him because it dismissed all four putative shots as random noise -- even though the three soundbursts from the book depository matched up precisely with film of the assassination and other evidence such as the echo patterns in Dealey Plaza and the speed of Kennedy's motorcade.
"This is an honest, careful scientific examination of everything we did, with all the appropriate statistical checks," Blakey said of Thomas's work.
"It shows that we made mistakes, too, but minor mistakes. The main thing is when push comes to shove, he increased the degree of confidence that the shot from the grassy knoll was real, not static. We thought there was a 95 percent chance it was a shot. He puts it at 96.3 percent. Either way, that's 'beyond a reasonable doubt.' "
The sounds of assassination were recorded at Dallas police headquarters when a motorcycle patrolman inadvertently left his microphone switch in the "on" position, deluging his transmitting channel with what seemed to be motorcycle noise. Using sophisticated techniques, a team of scientists enlisted by the House committee filtered out the noise and came up with "audible events" within a 10-second time frame that it believed might be gunfire.
The Warren Commission had concluded in 1964 that only three shots, all from behind, all from Oswald's rifle, were fired in Dealey Plaza as the motorcade passed through. But the House experts, after extensive tests, found 10 echo patterns that matched sounds emanating from the grassy knoll, traveling carefully measured distances to nearby buildings and then bouncing off them to hit the open motorcycle transmitter.
They also placed the unknown gunman behind a picket fence at the top of the grassy knoll, in front of and to the right of the presidential limousine. The House committee concluded that this shot missed, and that Kennedy was killed by a final bullet from Oswald's rifle. Thomas, by contrast, believes it was the shot from the knoll, seven-tenths of a second earlier, that killed the president.
The NAS panel, assigned to conduct further studies after the committee closed down, said in 1982 that the noises on the tape previously identified as gunshots "were recorded about one minute after the president was shot."
The NAS experts, headed by physicist Norman F. Ramsey of Harvard, reached that conclusion after studying the sounds on the two radio channels Dallas police were using that day. Routine transmissions were made on Channel One and recorded on a dictabelt at police headquarters. An auxiliary frequency, Channel Two, was dedicated to the president's motorcade and used primarily by Dallas Police Chief Jesse Curry; its transmissions were recorded on a separate Gray Audograph disc machine.
The shooting took place within an 18-second interval that began with Curry in the lead car announcing on Channel Two that the motorcade was approaching a triple underpass and ended with the chief stating urgently: "Go to the hospital." What seemed to be the gunshots were picked up on Channel One during that interval.
The NAS panel pointed out that Dallas County Sheriff Bill Decker could be heard on both channels saying, ". . . Hold everything secure . . ." seemingly about a half-second after the last gunshot on Channel One. Curry had already told everyone on Channel Two a minute earlier to go to the hospital. As a result, the Ramsey panel concluded that the supposed gunshot noises came "too late to be attributed to assassination shots."
What actually happened was that Curry issued his "go to the hospital" order right after the first shots were fired, wounding Kennedy and Texas Gov. John Connally. The final bullet was fired in almost the same instant that Curry uttered his command. A minute later, Decker, riding in the same car with Curry, grabbed the mike and issued his orders to "hold everything secure."
The NAS experts made several errors, Thomas said, but their biggest mistake was in using Decker's words to line up the two channels. They ignored a much clearer instance of cross talk when Dallas police Sgt. S. Q. Bellah can be heard on both channels, asking: "You want me to hold this traffic on Stemmons until we find out something, or let it go?"
Those remarks come 179 seconds after the last gunshot on Channel One and 180 seconds after Curry's order to "go to the hospital" on Channel Two. When Bellah's words are used to line up the two channels, Thomas found, the gunshot sounds "occur at the exact instant that John F. Kennedy was assassinated."
How is it, then, that Decker's remarks on Channel One come a full minute after Curry's on Channel Two and yet a half-second after the last gunshot on Channel One?
"It's a misplaced bit of speech," Thomas said in an interview. "An overdub. The recording needle for Channel One probably jumped. You can hear Decker giving a whole set of instructions on Channel Two, but on Channel One, you get only a fragment, '. . . hold everything secure. . . .' "
According to Thomas, the NAS panel made other mistakes: in calculating the position of the grassy knoll shooter, in fixing the time of that shot and in stating the Channel Two recorder had stopped when it hadn't. In all, Thomas said, the chances of the NAS panel having been right were 1 in 100,000.
House committee experts James Barger, Mark Weiss and Eric Aschkenasy, have always held firm to their findings of a shot from the knoll. Similarly, Ramsey, as chairman of the NAS panel, said last weekend that he was "still fairly confident" of his group's work, but he said he wanted to study the Science and Justice article before making further comment. He said he did not recall the Bellah cross talk.