The new article is an outgrowth of a scientific paper Randich published with William Tobin in 2002 which casts doubt on the FBI's technique known as bullet-lead analysis. This technique is based on the assumption that each batch of lead used to make the core of bullets has a unique chemical "fingerprint" that can be used to match it with other bullets. Thus bullet lead found at the scene of a shooting could be matched to bullets found in a suspect's gun and prove complicity in gun violence. But Randich's study of bullet manufacturing found that batches of lead are not chemically unique, and bullets from the same box could have very different chemical signatures. His work helped persuade the FBI in 2005 to stop using bullet-lead analysis in criminal prosecution.

Says Grant: "We applied the same thinking to the JFK bullet fragments that had been analyzed by a man named Vincent Guinn [on behalf of the HSCA] back in the 1970s. I knew Guinn because I took his forensic science course when I was in graduate school and it helped inspire my interest in the subject."

Guinn, now dead, had concluded that the level of a trace element called antimony in five bullet fragments taken from the JFK crime scene fell into two distinct groups chemically. Given their chemical similarity, the handful of fragments taken from Kennedy's head, Connally's body and the front seat of the limousine could have come from two -- and only two -- bullets. From 1978 to 2006, Guinn's findings heartened defenders of the Warren Commission which found that Kennedy and Connally had been hit by only two bullets.

Randich and Grant's paper found that Guinn's analysis was fatally flawed. He had assumed the chemical composition of bullet lead is consistent throughout a given bullet, a finding that Randich's metallurgical analysis refuted. Guinn also underestimated the margin of error in his measurement of antimony and wrongly discounted contradictory evidence, they said. Grant and Randich concluded that their findings "considerably weaken support for the single bullet theory."

Acoustic Hell

"Can some scattered noises in a crowded outdoor setting on a day in 1963 be recovered from an old damaged Dictaphone belt, its grooves 75 microns wide, five microns deep?" asks novelist Don DeLillo in a new introduction to Libra, his classic JFK novel. "We want to believe they can.... We want to believe we are dealing with science, not metaphysics."

Dictabelt #10
Photo: National
Archives and Records
Administration

DeLillo is talking about Dictabelt #10, a homely loop of blue plastic that now resides in the basement of the National Archives building in College Park, Maryland. It is a sound recording made by a dictation machine that was located in the radio room of the Dallas Police Department on November 22, 1963.

This souvenir of tragedy would take a two-hour season finale episode of a CSI show to sort out. It is the heart of one of the trickiest issues in JFK forensic science.

In the 1970s, some of America's top acoustic scientists studied the recording and the Dallas crime scene and asserted as fact that it contains sound impulses created by the series of gunshots fired at the presidential motorcade. In other words, this acoustic artifact is a kind of soundtrack for Abraham Zapruder's silent home movie. As the film in Zapruder's eight-millimeter camera captured the sight of gunfire hitting the presidential motorcade, this Dictabelt supposedly captured the sounds of the gunshots.

"If it's true that the sound of gunfire is captured on the recording, then it is conclusive evidence," says Paul Hoch, one of the most respected JFK researchers. "There was a conspiracy."

The tape does not contain the sound of gunfire, said five eminent scientists, in the British forensic journal, Science & Justice, in 2005. In 21 pages of closely argued scientific reasoning, physicist Richard Garwin and four colleagues said a careful analysis of the alleged gun shot sounds on Dictabelt #10 shows that they occurred approximately one minute after Kennedy was killed. They were not gunshots at all. Garwin and his colleagues could not say what created the sound impulses heard on Dictabelt #10.

Their article was a response to a 2001 article in Science & Justice which asserted the reverse: that the Dictabelt certainly captured the sounds of gunshots -- and that one of the shots came from the grassy knoll. That article, written by Don Thomas, an insect specialist at the U.S. Department of Agriculture, reviewed the findings of acoustic scientists retained by the House Select Committee on Assassinations in 1978. They concluded that the tape captured sound impulses created by four gunshots, three from the book depository behind Kennedy's limousine and one from the so-called grassy knoll in front of the motorcade. With some important caveats, Thomas says the HSCA got it right.