The FBI analyzes the fragments spectroscopically

Optical emission spectroscopy
      Optical emission spectroscopy (OES) is an older analytical technique for measuring the concentrations of elements in materials. About 10 milligrams of the sample is attached to one end of a carbon electrode, which is they heated to incandescence. The elements in the sample "glow,” that is, they give off light at characteristic wavelengths whose intensity is proportional to the concentration of the elements in the sample. The intensity of the various wavelengths is recorded on a glass plate. The density, or darkness, of the various lines is later measured with a densitometer and converted to concentration. Because the glass plate produced by this analysis is a spectrograph (a picture of a spectrum), the method is often called “spectrography.” “Spectroscopy” is a synonym that is easier to pronounce.
      Emission spectroscopy was the FBI’s obvious first choice for analyzing the lead fragments because the technique is sensitive and easy, and was in routine use at the FBI’s forensic laboratory in Washington at the time. The FBI apparently avoided neutron activation because they had had no experience with it. Remember that in 1964, neutron activation was a young technique. Although NAA had been around since 1937, it was not used routinely until the 1950s, when research nuclear reactors first became available.

How the FBI used OES on the fragments
      As detailed above, the FBI received the various bullets and fragments during the night of the 22^nd and the early morning of the 23^rd of November 1963. They then prepared the samples (Table 1) and analyzed them all by OES overnight. The FBI’s number-three man, Cartha “Deke” DeLoach, has written about that frenzied night of the assassination and its aftermath, as everyone in the laboratory rushed to learn as much as possible from the evidence by the next day.[1]
      Unfortunately for the FBI, emission spectroscopy is only semiquantitative. Its results will thus always contain a wide margin of error, which was made worse in this case by the small masses of some of the tiny fragments. The result was fatal—the FBI’s spectrographic results could not differentiate the fragments well enough to say anything other than that the fragments were broadly similar in composition. This said nothing about origins.Both the FBI and the Warren Commission downplayed this indecisive result: no formal report was made available, nor was the FBI spectrographer in charge of the analyses, John F. Gallagher, ever called to testify on the subject. Mr. Gallagher did appear before the Commission, however, on its last day of taking testimony, but was asked solely about his neutron-activation of paraffin casts from the hands and cheek of Lee Harvey Oswald,[2] not about his spectrographic efforts.[3]
      The only reference to the spectrographic results came earlier, in the testimony of FBI ballistics expert Robert A. Frazier. He alluded generally to Gallagher’s spectrographic results in a tiny part of his overall testimony. Frazier said only: “…it was determined that the lead fragments were similar in composition.”[4] When asked by Commission Counsel Arlen Specter whether the three tiny fragments recovered from the rear floor (CE 840) could have come from the large front-seat fragment CE 567, Frazier responded affirmatively. But he added that this idea could not be proven, and Specter did not question him further. Two months later, FBI director J. Edgar Hoover reaffirmed the limitations of the spectrographic analysis in a letter of 8 July 1964 to J. Lee Rankin, General Counsel of the Warren Commission:

“As previously reported to the Commission, certain small lead metal fragments uncovered in connection with this matter were analyzed spectrographically to determine whether they could be associated with one or more of the lead bullet fragments and no significant differences were found within the sensitivity of the spectrographic method.”[5]

Thus the FBI and the Warren Commission were stymied in their first chemical attempt to link fragments to bullets.
      Several features of the weights listed in Table 1 should be noted. First, the weights refer to fragments as received by the laboratory, not after any analysis. Second, weights of fragments remaining will not usually match those of the original specimens, because small portions of the fragments are removed for analysis along the way. Third, repeated analyses will cause the weights of specimens to decrease with time. Fourth, the total weights of specimens plus fragments will not necessarily remain equal to the original weights, because cleaning the surface of specimens before analysis always removes some material. Fifth, weights of fragments analyzed by emission spectroscopy were not supplied to the writer of the memo noted at Table 1, apparently because they could not be found in the original FBI file. These weights remain unknown.
      The FBI did not release the results of their spectroscopic analysis—the Warren Commission was forced to take their word that they showed nothing other than broadly similar concentrations. The critics, of course, thought that the FBI was hiding something, possibly as part of a broader cover-up. Harold Weisberg got a brief table of data after 17 years of pressure and lawsuits, and George Michael Evica apparently got the same information by requesting it directly from Clarence Kelley, then director of the FBI. The “table” is nothing more than some notes on a piece of paper, evidently written by the FBI analyst. The results are semiquantitative at best, as is emission spectroscopy in general. One of the pages of “data”[6] gives no numerical values at all, only symbols like —, 0, tr, +, ++, and +++. These symbols correspond roughly to “not measured,” “not found,” “trace,” “low concentration,” “medium concentration,” and “high concentration.” Another page gives some very rough numerical concentrations, many of which are ranges or limits. The coarseness of these results can be seen in Tables 5–7 below, where I have combined the data and grouped them in various ways.

Table 5. The FBI’s optical emission spectroscopic data grouped by analytical run 
(numerical data in ppm).

      This table contains all the FBI’s OES data. Note that the best result has a range of 90–800 ppm, which is nearly an order of magnitude. The other numerical data are limits (e.g. £500 ppm) or ill-defined approximate values (~400 ppm). The FBI was right—virtually nothing can be determined about the origins or the groupings of the fragments from these data.

Table 6. Two examples of (ir)reproducibility in the FBI’s OES data (numerical data in ppm).

      Another good test of an analytical technique is how reproducible its replicate analyses are. Table 6 above shows that the OES data are essentially irreproducible. This is another reason why no conclusions can be drawn from them.

Table 7. The OES data grouped by bullet according to the later NAA results
(numerical data in ppm).

      Can any level of similarity between fragments from the same bullet (as grouped by NAA) be seen in the OES data when the fragments are arranged by group? In other words, can you see any similarity when you group the fragments the way they really are? Again the answer is no, as seen in Table 7 above. The OES data are essentially useless.

Ahead to George Michael Evica
Back to Using Chemical Data
Back to NAA and the JFK Assassination
Back to NAA