The FBI Laboratory: An Investigation into Laboratory Practices and Alleged Misconduct in Explosives-Related and Other Cases (April,1997) Office of the Inspector General, Department of Justice

SECTION B: THE VANPAC CASE

This case concerns four mail bombs sent in December 1989 to different locations in the southeastern United States. One killed federal judge Robert Vance in Alabama, a second killed a civil rights attorney in Georgia, and two others were discovered before they exploded. A massive investigation ensued involving the FBI and several other law enforcement agencies. The FBI referred to the case as VANPAC because it involved the assassination of Judge Vance with a bomb sent in a mail package. In June 1991, a federal jury convicted Walter LeRoy Moody, Jr. on charges related to the bombings.

In this case, Whitehurst has made numerous allegations of wrongdoing by J. Thomas Thurman of the Explosives Unit (EU) and by Roger Martz of the Chemistry-Toxicology Unit (CTU). Whitehurst complains that Thurman and Martz circumvented the procedures of the FBI Laboratory because Thurman, as the principal examiner, asked Martz to analyze material in the mail bombs even though the Materials Analysis Unit (MAU) was responsible for analyzing explosives residue. He also alleges that because Martz did not follow the protocol for residue analysis developed by the MAU, Martz reached a flawed opinion in concluding that the mail bombs contained a particular smokeless powder.

Whitehurst alleges that Thurman improperly based his opinions on the flawed residue analysis performed by Martz; that Thurman improperly testified outside his field of expertise on various matters; and that Thurman lacked a factual basis for certain testimony about the explosives used in the bombs. Whitehurst has accused both Thurman and Martz of fabricating evidence, perjuring themselves, and obstructing justice in the VANPAC case. He also has suggested that prosecutors Louis J. Freeh and Howard Shapiro, who were then Assistant United States Attorneys and who tried the VANPAC case, may have committed misconduct by offering the testimony of Martz and Thurman.

To investigate Whitehurst's claims regarding this case, we reviewed the pertinent reports prepared by the FBI Laboratory and, where available, the underlying work papers and test results. We reviewed transcripts of the testimony given by certain witnesses and the closing arguments in Moody's trial for the bombings. We also questioned agents Thurman, Martz, and Whitehurst about the case in interviews in which their answers were given under oath and transcribed. We also interviewed others involved in the case, including Director Freeh, FBI General Counsel Shapiro, and current or former Laboratory personnel James Corby, James Kearney, Roger Peele, Charles Peters, and Robert Webb.

We find no factual basis to conclude that Thurman or Martz perjured themselves, fabricated evidence, obstructed justice, or violated any FBI policies or procedures in this case. Nor do we find any evidence to support Whitehurst's claims of prosecutorial misconduct. Whitehurst did not make allegations against MAU examiner Robert Webb. Based on the documents provided by the FBI, however, we did conclude that Webb stated conclusions about the common origin of certain tape, paint, sealant, and glue more strongly than was justified by the results of his examinations and the background data. In our view, Webb did not intentionally attempt to fabricate evidence or to present biased conclusions. Our investigation of the VANPAC case also reveals several areas in which Laboratory practices or procedures should be improved. These matters are discussed further in the following sections.

II. Factual Background

In December 1989, four mail bombs were received at different addresses in the southeastern United States. One bomb killed Eleventh Circuit Court of Appeals Judge Robert Vance in Birmingham, Alabama; another killed attorney Robert Robinson in Savannah, Georgia; the third was discovered before exploding at a federal courthouse in Atlanta, Georgia; and the fourth was discovered before exploding at the Jacksonville, Florida office of the National Association for the Advancement of Colored People (NAACP).

The mail bombs had numerous similarities, which included: they were delivered in packages wrapped in brown paper, tied with string, addressed with typed red-and-white labels, and posted with stamps depicting an American flag over Yosemite National Park; they were placed in cardboard boxes that had been painted black in the inside; and each bomb included a steel pipe filled with smokeless powder, finishing nails secured to the outside of the pipe, and a detonator fashioned from a flashbulb filament with distinctive wiring and a ballpoint pen casing. The detonators from the two bombs that did not explode contained a green powder identified as high explosive primer. Three of the bombs also had welded end plates that were joined together by a steel rod through the center of the pipe.

The bombings were followed by a large-scale investigation involving the FBI, the Bureau of Alcohol, Tobacco and Firearms (ATF), the U.S. Postal Inspection Service, the Georgia Bureau of Investigation, and other law enforcement agencies. The unexploded devices found in Atlanta and Jacksonville were sent to the FBI Laboratory for analysis in December 1989, as was debris from the mail bombs from Savannah and Birmingham. J. Thomas Thurman of the EU was assigned as the principal examiner.

Thurman enlisted various auxiliary examiners in other units of the Laboratory to examine evidence. The other examiners included Roger Martz, who was then chief of the CTU. In January 1990, Martz determined that each mail bomb contained Red Dot double base smokeless powder. He also determined that a green powder found inside the detonators of the two unexploded mail bombs was a small arms primer manufactured by CCI Industries.

Walter LeRoy Moody, Jr. was identified as a suspect after ATF forensic chemist Lloyd Erwin recalled that Moody had been convicted in a 1972 case involving a pipe bomb with a design similar to that of the 1989 bombs. In February 1990, federal investigators searched a storage area rented by Moody in Chamblee, Georgia, and found a device constructed from a metal pipe that was similar in some respects to the construction of the mail bombs. Several searches of Moody's house, however, failed to reveal evidence of Red Dot smokeless powder or the type of CCI primer identified by Martz in the explosive devices.

In April 1990, a witness named Paul Sartain told ATF agents that while he was working at the Shootin' Iron gun store in December 1989, he had sold someone a four pound can of Red Dot smokeless powder and a quantity of CCI primers. Sartain later identified Moody as the person who had purchased these items.

In July 1990, Moody was indicted on charges that he had suborned perjury by a witness in connection with a 1988 hearing on a coram nobis petition he had filed related to his 1972 conviction. A jury convicted Moody of these charges after a trial in Brunswick, Georgia, in December 1990.

In November 1990, Moody was charged with various federal crimes related to the bombings. Venue for trial was transferred to St. Paul, Minnesota and the case was assigned to Senior Judge Edward J. Devitt. After a trial in June 1991, a jury convicted Moody on 71 separate counts. The judge sentenced Moody to seven life terms plus four hundred years.

During the trial, Lloyd Erwin, Frank Lee, and Terry Byer of the ATF testified about the construction of the four mail bombs, the 1972 bomb, and the Chamblee device. They opined that all had been made by the same person. Moody's former wife Susan McBride Moody testified that she had purchased various items at Moody's direction. The items she purchased were consistent with components used in the mail bombs. A former cellmate of Moody's, Ted Banks, testified that at Moody's request he had welded end plates onto three metal pipes that were similar to those used in three of the bombs. Paul Sartain testified that in December 1989, he had sold Moody a four-pound keg of Red Dot smokeless powder and 4,000 CCI small pistol primers.

During the third week of trial, the government presented testimony by Thurman and Martz from the FBI's Laboratory Division. Thurman testified about the construction of the mail bombs and opined that they had been made by the same person who made the 1972 bomb. Martz testified that the mail bombs contained Red Dot double base smokeless powder and that he identified CCI small arms primer in detonators from the two unexploded devices.

III. Analysis of the Whitehurst Allegations

Whitehurst did not do any work himself on the VANPAC case. In a September 5, 1994, letter to the OIG, Whitehurst detailed numerous criticisms of the testimony by Martz and Thurman in Moody's 1991 trial. Whitehurst made similar criticisms in a 26-page memorandum to James Kearney that was prepared near the time of Kearney's retirement as chief of the Scientific Analysis Section in June 1995.

In making his criticisms, Whitehurst relied on the transcripts of the trial testimony of Thurman and Martz and the closing arguments by Freeh and Shapiro. He did not review the trial exhibits or other evidence. Nor did he review any of the Laboratory reports or analytical data related to the testimony by Martz and Thurman.

For purposes of our report, we have summarized the various criticisms under the following identified headings.

Whitehurst claims that Thurman and Martz circumvented the FBI's protocols for the analysis of explosives residue when Martz analyzed the contents of the explosive devices.

Contrary to Whitehurst's position, at the time of the VANPAC case, there was no Laboratory protocol or policy requiring that all analysis of explosives be conducted by the MAU. That unit did generally perform the analysis of explosives residue and certain bulk explosives. The CTU, however, had been analyzing smokeless powders since the 1980s. This apparently occurred because Roger Martz had, within the CTU, developed techniques, including the use of the mass spectrometer, to confirm the presence of smokeless powder and to attempt to identify the manufacturer. At the time, there were no written policies describing the respective roles of the CTU and the MAU in analyzing explosives.

Martz stated in his sworn interview that the CTU became responsible for the analysis of smokeless powders sometime in the early 1980s after the CTU and MAU completed proficiency tests. We were unable to locate any FBI documents describing the results of such tests or a decision that the CTU would analyze smokeless powder. Former MAU Chief Charles Calfee, although not recalling the proficiency tests, confirmed that responsibility for the analysis of smokeless powder was transferred from the MAU to the CTU after the latter unit developed identification techniques with the mass spectrometer. James Corby, MAU chief from June 1990 through October 1995, also confirmed that the CTU was conducting smokeless powder analyses during his tenure. Others also acknowledged in interviews that the CTU was analyzing smokeless powders when the Laboratory received the VANPAC case.

Thurman stated in his sworn interview that he had not made any effort to avoid or circumvent the MAU in connection with the VANPAC case. Thurman explained that he identified what appeared to be smokeless powder particles in the evidence, and he therefore sent it on for analysis by Martz because the CTU analyzed smokeless powders. Martz gave a similar account of his initial involvement in the case. Martz also stated in his sworn interview that Whitehurst knew in December 1989 that Martz was working on the VANPAC case to analyze possible smokeless powders and Whitehurst did not at that time express any concern. Whether or not Whitehurst knew in 1989 or 1990 that the CTU was analyzing the powder, we find no factual basis to conclude that Thurman and Martz attempted to circumvent Laboratory protocols or bypass the MAU.

B. The Identification of Red Dot Smokeless Powder

After receiving the evidence, Martz determined that each device contained Red Dot double base smokeless powder made by the Hercules Corporation. Martz reached this conclusion after visually examining and measuring particles that appeared to be smokeless powder and then analyzing the substances with a mass spectrometer. The results confirmed that Red Dot double base smokeless powder was present in each device. Martz also had Fourier Transform Infrared Spectroscopy (FTIR) performed on one sample to confirm the presence of nitrocellulose, a component of smokeless powder.

In analyzing the samples, Martz followed a protocol for the identification of smokeless powder that had been used in the CTU for several years. The basic procedure was outlined in an article published by Martz and FBI examiner Lynn Lasswell in 1983. Before Moody's trial, at least one defense expert reviewed the physical evidence and the FBI Laboratory reports and agreed with the conclusion that each device contained Red Dot smokeless powder as its main charge.

Whitehurst alleges that Martz improperly analyzed the evidence because he did not follow the protocol for analysis of explosives residue developed in the MAU. The MAU protocol would have involved additional analytical tests that might have detected certain inorganic explosives or fillers that cannot be identified by the mass spectrometer. Because these tests were not performed, Whitehurst asserts that Martz, and subsequently Thurman, could not conclude that smokeless powder constituted the explosive in the mail bombs.

Whitehurst is correct that the analytical tests performed by Martz may not have detected certain substances. Martz, in his interview with us, acknowledged that if trace amounts of certain inorganic materials were present, they conceivably would not have been identified by the tests he performed. Martz noted, however, that he physically examined the evidence, including debris from the bombs, and did not observe traces of other possible explosive components. In retrospect, we think a more comprehensive analysis might have been desirable, particularly given the scope of investigative efforts otherwise made in the case. The possibility that one or more of the devices may have contained other explosives or fillers, which were not identified by visual or microscopic examination, does not mean that the conclusions by Martz or Thurman about smokeless powder lacked a factual or scientific basis.

Although we do not believe that Thurman or Martz perjured themselves or fabricated evidence with regard to the presence of smokeless powder in the mail bombs, this case does illustrate an area in which the Laboratory's procedures should be improved. As noted above, during this time, both the CTU and the MAU were analyzing explosives. There was no clear delineation of the respective responsibilities of each unit. Moreover, because the units did not share a common protocol, the tests might vary depending on which unit received the evidence.

Whitehurst complains about several aspects of Thurman's testimony concerning the explosives used in each of the mail bombs.

Early in his testimony, Thurman discussed factors that affect the strength of a pipe bomb. With respect to the particular explosive used, Thurman noted, within the smokeless powder family there's two types of low explosives that we deal with more than anything else, and that's single base low explosive, and a double base low explosive. Thurman then said that a single base low explosive does not have nitroglycerin, while a double base explosive does. Whitehurst complains that [t]here are no such things as double and single base explosives, and that if Thurman was referring to smokeless powders, his remark is inaccurate, because those are not the explosives the FBI deals with most. Thurman apparently misspoke in using the term explosives rather than smokeless powder in this context. Thurman would have been accurate in saying that within the smokeless powder family, single and double base powders are the ones we deal with more than anything else.

Discussing the Birmingham device, Thurman testified that the main charge was double base smokeless powder and that the detonator contained a high explosive. These statements, Whitehurst asserts, rest on conjecture. We disagree. As noted above, Martz found Red Dot double base smokeless powder in debris from the Birmingham bomb. Thurman relied on this fact and the observable characteristics of the bomb debris to conclude that double base smokeless powder was the main charge.

With regard to the detonators, Martz did not identify primer in the debris from either the exploded Birmingham or Savannah devices. The FBI Laboratory reports, which were produced to the defense at trial, reflect this fact. Thurman's conclusion that these bombs also utilized a detonator containing a high explosive rested primarily on similarities in the debris indicating that the bombs had detonators constructed from pen casings, a distinctive wiring system, and an initiator devised from a flashbulb. Thurman also noted that a mockup device, which included a high explosive detonator, had been detonated by the FBI and the resulting fragmentation was similar to that observed in the exploded bombs.

Thurman did not, in our view, fabricate evidence in opining that the Birmingham and Savannah detonators contained a high explosive. He did have a reasoned basis for that opinion. Thurman did not in his testimony or reports state that analytical tests had confirmed the presence of primer in the exploded mail bombs. The defense attorney could have explored the basis for Thurman's opinion on cross-examination. Instead, the defense did not dispute that each device contained double base smokeless powder and a high explosive detonator. This may have reflected a tactical decision, since the defense attorney attempted to raise doubt in the mind of the jurors not by arguing that the 1989 bombs were different from each other, but by arguing that Moody could not be connected to these bombs because Red Dot smokeless powder was not found in Moody's house and the 1989 bombs were different from the 1972 device.

Whitehurst is correct insofar as he suggests that the type of explosives residue analysis performed by the MAU might have confirmed the presence of primer in the exploded devices. Martz acknowledged in his interview with us that, in retrospect, it would have been desirable to have examined the exploded devices for traces of primer residues. This again shows that the Laboratory should have clearly delineated which unit would analyze explosives residue and what tests would be performed.

Another complaint by Whitehurst is that Thurman lacked a basis to testify that the characteristics of the metal fragments from the Birmingham and Savannah bombs indicated there was a low-explosive main charge initiated with a high explosive detonator. Such testimony, Whitehurst says, is fabricated evidence and has no basis in fact. Whitehurst maintains that the observable characteristics of the metal pieces could have been caused by [a]ny number of other energetic materials. We find that Thurman had some basis for his statements. They reflected his personal experience observing exploded devices and the results of the FBI's detonation of the mock-up device modeled on the bombs sent to Savannah, Atlanta, and Jacksonville.

Whitehurst further asserts that Thurman incorrectly stated that the cut-off between high explosives and low explosives is where the shock wave travels at more than 3,000 feet per second. Thurman's statement is technically incorrect. See Attachment C, infra. We note, however, that it is not uncommon for bomb technicians or persons working in the field of explosives ordnance to distinguish high from low explosives by the explosive's velocity. The technical error here was inconsequential.

With respect to the Birmingham device, Thurman was asked on direct examination whether he had been able to reconstruct the bomb. Thurman said he had, and then agreed that he had been able to do so to a high degree of scientific certainty. Whitehurst asserts that Thurman perjured himself because he lacked scientific training and he knowingly and purposely had circumvented the FBI's protocol for the analysis of explosives residue. We do not agree with these accusations. Thurman's comments about being able to reconstruct the Birmingham device were preceded by fifteen pages of testimony about that device. When the prosecutor subsequently asked Thurman if he had been able to reconstruct the device to a high degree of scientific certainty, no objection was made to the possibly ambiguous nature of the question or to Thurman's qualifications to respond. By answering affirmatively, Thurman did not in our opinion perjure himself or intentionally misrepresent his background. As explained above, Thurman did not circumvent an FBI protocol in the analysis of the mail bombs.

Whitehurst also alleges that Thurman relied on conjecture in testifying that the location of a metal rod in the debris of the Savannah bombing indicated that the pipe had been full of powder. At the trial, Thurman stated that in photographs of the crime scene, he saw the metal rod on a desk at the scene of the explosion. He observed that the rod's essentially intact condition suggested that the rod was at the center of the bomb and had dropped straight down after the explosion. He further opined that the rod's location indicated that the pipe in which it was contained had been totally full of powder, so that you have got equal pressure all the way around this threaded rod.

Thurman did not base his opinion that the Savannah device was full of powder on any analytical test results. We do not agree, however, that his opinion was therefore merely conjecture. Thurman reasoned that if the rod connected the plates through the middle of the pipe, and the rod was found at the center of the explosion, the pipe must have been filled with powder so that the rod was at the center of the explosive force. According to Thurman, when the FBI detonated its mock-up bomb in a model of Robinson's office, the connecting rod was again found at the center of a desk. There was a reasonable basis for Thurman's opinion, and the defense attorney could have explored the basis for that opinion on cross-examination.

With regard to the Atlanta device, Whitehurst asserts that Thurman fabricated evidence on the witness stand when he testified that black particles on the recovered detonator were Red Dot smokeless powder. This accusation lacks any factual basis. The analysis done by the CTU identified Red Dot smokeless powder on the Atlanta detonator, as was noted in the Laboratory's March 3, 1990, report. Thurman relied on these results in his testimony.

Whitehurst makes several claims that Thurman testified about matters beyond his training or qualifications. Such testimony, Whitehurst maintains, violated FBI Laboratory policy. Related arguments made by Whitehurst are that Thurman testified about certain matters without supporting scientific tests or he improperly testified about results reached by other examiners.

To evaluate these arguments, several background points must be kept in mind. The FBI Laboratory did not at this time have any expressly stated policy concerning the permissible scope of an examiner's testimony. The common understanding within the Laboratory was that examiners should be careful not to stray outside their expertise, a point that reportedly was emphasized in the moot courts that were part of the examiner qualification process. During our investigation, many examiners told us that in testifying they had sometimes been asked to read into the record conclusions reached by other, non-testifying examiners. This generally was viewed as acceptable so long as the testifying examiner was careful not to comment further as to matters on which he or she lacked personal knowledge.

In the VANPAC case, Thurman, as the principal examiner, received dictation from other auxiliary examiners which he in turn incorporated into the Laboratory reports. He presumably was the examiner with the best overview of the work done by the Laboratory in the case. The defense received copies of Thurman's reports, which set forth the findings made by different units within the Laboratory. The attorney who led Moody's defense agreed before trial that Thurman could testify as a summary witness about the results of work done by certain auxiliary examiners. Thurman similarly understood from the prosecutors that he would testify about results reached by certain other examiners. Moreover, we note that an expert may properly draw on personal experience or common sense in forming opinions, and a conclusion is not necessarily improper because it is not based on a scientific test.

Whitehurst complains that Thurman testified outside his expertise in stating that a white sealant material or RTV was rubbery and spongy at room temperature and that nails found in the debris were bent by the explosion. The comment about the RTV appears to have been properly based on Thurman's own personal knowledge. In stating that the nails had been bent from the explosion, Thurman drew a common sense inference from the presence of bent nails among the debris. Similarly bent nails were found in the debris from the mock-up device detonated by the FBI. We do not think Thurman's statements about the RTV or nails were improper.

Thurman also testified that the use of welded end plates in the devices would create more pressure within the bomb and that the detonator would have been placed inside at least hours after the welding was done. We think the first statement is unobjectionable and was properly based on Thurman's experience. Thurman noted in our interview that it is not uncommon for pipe bombs to explode by simply blowing off their end caps, leaving the pipe itself intact. One could reasonably conclude that the use of welded end plates would cause more pressure to build up before the pipe exploded. Thurman's statement that the detonator would have been placed inside the pipe at least hours after the welding reflected his view, which seems merely common sense, that no one would place the high-explosive primer into a hot metal pipe.

Thurman testified that certain testing had been done by the Serology and DNA units, and that the results were negative in that no traces of saliva were found. These statements were consistent with the underlying Laboratory reports. We do not agree with Whitehurst's contention that Thurman violated Laboratory policy by testifying on these matters.

Similarly, Thurman testified about paint and tape found in the devices. During that testimony, Thurman noted that the Laboratory had determined that 2-inch wide, tan plastic tape and black paint found in the devices were from the same source or manufacturer. In this regard, Whitehurst asserts that Thurman was simply fabricating evidence to suit his hypothesis that all the bombs were made from the same source.

Thurman in fact was testifying based on the analytical work and dictation of MAU examiner Robert Webb. Webb, an experienced examiner in the MAU, examined several items of evidence during the VANPAC investigation. In examining packaging tape, black paint, RTV, and glue found in the devices, Webb followed an unwritten protocol that included microscopic examination, so-called wet chemical analyses, analysis with Fourier Transform Infrared Spectroscopy (FTIR), and Pyrolysis Gas Chromatography (PGC). Based on these techniques, Webb concluded that packaging tape in each device came from the same manufacturer and the same batch or lot, that black paint in each device had physical and chemical characteristics indicating it came from the same manufacturer, that RTV sealant in each device had physical and chemical characteristics indicating it was from the same manufacturer and originated from the same batch or lot, and that glue in three of the devices had physical and chemical characteristics indicating it came from the same manufacturer.

Thurman did not fabricate evidence or otherwise testify improperly about the paint and tape analysis insofar as it was based on Webb's dictation. Webb had described his conclusions about the comparison of samples of paints, adhesives, and tape in auxiliary examiner dictation dated March 19, 1990. Thurman incorporated this dictation verbatim into the FBI Laboratory report dated April 2, 1990. As part of our investigation, Webb reviewed Thurman's testimony about the paint and tape and observed that it was consistent with Webb's dictation.

Whitehurst also has maintained that the conclusion that the black paint came from the same manufacturer is flawed because data do not exist to allow one to say that two samples with a similar chemical composition necessarily came from the same source. A similar criticism could be made concerning the conclusions that the 2-inch wide tape and the RTV sealant came from the same batch or lot. When asked in our investigation about his conclusions, Webb maintained that in his experience, the battery of tests he employed would reveal some differences if paint samples did not come from the same manufacturer or if the tape had been made in different batches or lots.

We find that Webb's conclusions about the tape, paint, RTV, and glue were stated more strongly than was justified by the results of his examinations and the background data. As a general matter, we question the validity of Webb's working proposition that the examinations he performed would have necessarily revealed some differences if the materials had come from different manufacturers (or different batches or lots for the tape and RTV). At the time of the VANPAC case, neither Webb nor the FBI had a data base to confirm that black latex paints, RTV, glue, and tapes like those involved in the samples did in fact differ among manufacturers in terms of their chemical composition and physical characteristics. Moreover, the tests that Webb performed had not been validated by the FBI or, to our knowledge, any other laboratory, with regard to their ability to successfully determine if samples actually came from the same source. In these circumstances, the methods employed by Webb would allow an examiner to conclude that samples could have come from the same source or manufacturer, but not to opine that they necessarily did.

Webb's conclusions about the common origin of the different samples also seem overstated in light of differences in the results from certain analyses he performed. More specifically, the PGC chart for the black paint from the Jacksonville device contains a peak not observed on the PGC charts for samples from Atlanta and Birmingham; the FTIR chart for a sample of glue from the Atlanta device contains a peak that is absent from the FTIR results for glue from the Jacksonville device; the PGC chart for a clear glue sample from Atlanta has a peak absent from the PGC charts for another sample from Atlanta and a sample from Jacksonville; the FTIR chart for a sample of RTV from the Savannah device has a different pattern than the FTIR charts for samples from Atlanta, Jacksonville and Birmingham; and the PGC chart for a sample of RTV from Birmingham has a peak absent from the PGC charts for samples from Atlanta and Jacksonville.

With regard to the comparison of the 2-inch wide tapes, charts could not be located for analyses done on samples from Atlanta and Jacksonville. The FTIR charts for the tape adhesive from the Birmingham and Savannah devices exhibit several differences. The notes that we reviewed do not explain how Webb reconciled these differences with his ultimate conclusion that tape found in each of the four devices had come from the same batch or lot. When we interviewed Webb about these differences, he said that they may reflect contamination, variations due to sample preparation, the fact that tests were run on different dates, or calibration. Webb acknowledged that certain differences in the test results for the tape and other items he examined are significant enough to require further explanation, but he did not retract the conclusions he reached in 1990 about the common origin of the identified samples.

The differences noted above do not in themselves establish that samples of a particular substance, such as paint or tape, did not have a common origin. Such differences, however, appear to preclude the firm conclusion that the samples came from the same source or manufacturer (or batch or lot). Our questions about the differences in the test results remain unresolved, in part because the case files do not include all the pertinent charts or complete notes explaining the basis for the ultimate conclusions.

We conclude that Webb did not intentionally attempt to fabricate evidence or to present biased conclusions in his work on VANPAC. It appears that Webb's unit chief reviewed and approved his conclusions about the intercomparison of paint, adhesives, and tape. More significantly, Webb also did analytical work and prepared dictation that identified differences between certain samples. For example, he concluded that the white glue found in the Birmingham device did not match samples from the other devices. He also concluded, as was stated in the FBI reports, that certain glues and tape seized from Moody's residence and storage area did not match samples from the explosive devices. Such a match would, of course, have been very incriminating.

The comparison of tapes, paints, and adhesives in VANPAC does illustrate several areas in which we think the ASCLD/LAB accreditation process should improve the quality and consistency of the Laboratory's work. To become accredited, the Laboratory will have to assure that there are written, validated procedures for standard analytical techniques and examinations. Such protocols did not exist within the Laboratory for the types of analyses done by Webb during the VANPAC case. Accreditation will also require the Laboratory to provide for the review of reports to confirm that examiners' conclusions are reasonable and within the constraints of scientific knowledge. In order to become accredited, the Laboratory will also be required to maintain a case record that includes all the notes, worksheets, charts, and other data that support the examiner's conclusions. Such complete information was not included in the files we reviewed for the VANPAC case and several other matters that were the subject of our investigation.

Whitehurst also complains that Thurman improperly testified outside his expertise with respect to metallurgical matters. Thurman testified that a rod found in the debris from the Savannah device had been stretched as a result of the explosion. On cross-examination, he admitted that the metal people in the Laboratory could sometimes identify metal filing residues from files or grinding wheels and compare them to other metals. Defense counsel then elicited Thurman's acknowledgment that no metal residue was found in grinding wheels or files seized from the defendant that could be compared to debris in the bombs.

Thurman's testimony about the effects of the explosion on the rod was based on his visual inspection of the rod. The Laboratory reports did not indicate that any analytical test had been performed to confirm that the explosion caused the stretching of the rod. We do not believe that Thurman testified improperly in opining that the rod had stretched as result of the explosion. If the defense attorney had wished to explore the basis for Thurman's comments, he could have done so on cross-examination.

We also think Thurman responded properly to the questions that were posed on cross-examination concerning metallurgy. If he in fact believed that the tests he was asked about could be performed, and if to his knowledge no metal debris was found on wheels and files for comparison purposes, we think he was obliged to respond as he did. Notably, in responding to these issues, Thurman was conceding points the defense wished to develop, which further belies the allegation that Thurman was determined to perjure himself or fabricate evidence to secure a conviction.

Whitehurst also asserts that during the cross-examination, Thurman improperly testified outside his expertise concerning paints, tool marks, DNA analysis, smokeless powder, and the analysis of primers. With respect to paints, Thurman was asked whether you would have the capability of matching the paints, if black paint had been found at the defendant's properties. Thurman responded, I would expect so, yes, sir. Given the reports Thurman had received in the case from examiner Robert Webb, we see no basis to criticize Thurman's response.

The defense counsel later asked Thurman to explain what a useful gripping tool mark would be. Thurman noted he was not a tool mark examiner, but said he would try his best. He then explained how some tools will leave identifying marks that allow a particular tool to be matched with a marked object. Thurman then acknowledged that no identification had been made in the case with respect to certain tools seized from Moody.

Regarding DNA analysis, Thurman acknowledged that this was a new scientific technique and the defense attorney himself noted that Thurman was not a DNA expert. Thurman agreed that an enzyme called amylase that is in saliva can be used for DNA analysis. In response to further questioning, Thurman admitted that DNA testing could not be done on certain envelopes because no amylase was recovered from them. Defense counsel then had Thurman concede that there was no DNA match to Moody based on the envelopes.

With respect to smokeless powders and primer, Thurman admitted Moody's house had been vacuumed in virtually every conceivable area to identify minute microscopic grains of gunpowder, but none had been found. He was also asked if he recalled that the CCI primer had a unique two percent aluminum component. Thurman noted that the question concerned examinations done by Martz, and that he thought Martz would be testifying during the trial. When asked if the primer material could be obtained not only from primers but also from small arms ammunition, Thurman told the defense attorney he would have to pose the question to Martz. When asked if powders could be matched to determine if they were from the same batch, Thurman said, sometimes yes, sometimes no, and again referred the question to Martz. The defense attorney noted that primer had not been recovered from all four devices, and Thurman then agreed that the Laboratory had not been able to determine that the primer recovered from the devices and the primer sold by Sartain to Moody had come from the same lot.

On cross-examination, Thurman further admitted that the Laboratory had not been able to match a keg of Red Dot smokeless double base gunpowder obtained from the Shootin' Iron Gun Shop with the gunpowder found in the four bombs. On re-direct, Thurman noted that although batches could not be matched, the powder was of the same type and the same manufacturer. Thurman also said the CCI primer that Sartain said he sold to Moody was of the same type and manufacturer as that used in the bombs.

Thurman did not improperly testify outside his expertise or contrary to FBI policy with respect to the matters raised in his cross-examination. As noted above, Thurman as principal examiner had assembled the Laboratory reports after reviewing the dictation of the various auxiliary examiners, and defense counsel had agreed that he could testify as a summary witness. In an apparent effort to raise doubt about the connection between Moody and the mail bombs, the defense counsel sought Thurman's acknowledgment that the Laboratory had not made certain findings. If anything, Thurman might have been fairly subject to criticism if he had refused to concede the points he did.

In testifying about the DNA tests, Thurman correctly stated that there had been no DNA match to Moody, but his testimony was inaccurate in a relatively minor respect. Thurman erred in agreeing with the defense attorney that the DNA testing was based on amylase recovered from saliva. Amylase is an enzyme present in saliva, but it is not the basis for DNA analysis. Such analysis can be based on epithelial cells taken from saliva. The FBI Laboratory reports noted in separate sections that serological tests for amylase were negative and that DNA test results could not be obtained. Thurman's imprecision regarding the DNA tests illustrates that examiners must be very cautious in testifying, even as summary witnesses, outside their area of expertise.

We also think Thurman testified appropriately on re-direct in stating that the smokeless powder and primers that Sartain said he sold to Moody were of the same type and by the same manufacturer as the materials used in the bomb. Whitehurst asserts that Thurman could not have known that the smokeless powder was of the same type because Hercules manufactures Red Dot powders for sale in products other than the canister powder identified by Sartain. Whitehurst evidently believes that Thurman, by saying the powder was of the same type, misleadingly indicated it could only have come from one kind of container. We disagree with this interpretation of Thurman's testimony. With respect to the primer material, Thurman had earlier acknowledged that it was not found in all four devices. When Thurman agreed on re-direct that the CCI primer that Sartain said he sold to Moody was the same type as that used in the bomb, Thurman would have been more precise if he had said the same type as was identified in the two unexploded bombs.

Whitehurst makes two criticisms about Thurman's testimony on cross-examination which we think are best characterized as disagreements over the choice of words. Thurman agreed when the defense lawyer asked if he had, as the supervisory scientist, received all the reports of scientific examinations in the case. Whitehurst claims that this is misleading and a fraud upon the court because Thurman is not a scientist. We do not agree. Earlier in his testimony, Thurman had explained both his experience and the fact that as principal examiner he had coordinated the work done by various laboratory units on the case. Thurman could have spoken more precisely by stating again that he was the principal examiner rather than accepting the defense counsel's phrase supervisory scientist.

Whitehurst also states that Thurman incorrectly agreed with the defense attorney that smokeless powder has a dusty residue. The defense attorney phrased in everyday language how many people would describe the feel of such powders, but to be technically accurate, Thurman should have noted that they do not actually leave a residue of dust. Before Thurman testified, ATF Agent Frank Lee had agreed on cross-examination that, double base smokeless gun powder is like a dust and will adhere to walls, floors, clothing, [and] vacuum cleaner brushes. The defense attorney apparently sought to compare smokeless powder to a dust as background to his emphasizing that no traces of smokeless powder were found in the searches of Moody's house or the storage unit he rented. Both Lee and Thurman acknowledged that no smokeless powder was found in the searches.

Whitehurst contends that Martz testified in a misleading way about his education and qualifications and the role of the CTU. Martz testified that he received a bachelor's degree from the University in Cincinnati and had worked as a chemist for several years both before and after he had joined the FBI. Whitehurst complains that Martz failed to disclose that his bachelor's degree was in biology rather than chemistry and that he had never been qualified by the FBI to examine explosives residue.

Martz was not asked on either direct or cross-examination to identify his undergraduate major. Since 1980, Martz has been qualified as a forensic chemist examiner within the FBI Laboratory. Martz has stated he had 40 quarter credit hours in chemistry during college, and he worked as a chemistry technician in the CTU before becoming an examiner. With regard to the analysis of explosives, Whitehurst is correct that Martz never completed the MAU's program to become qualified by that unit as an explosives residue examiner. As noted above, however, the CTU had analyzed smokeless powders since the early 1980s. We do not think that Martz testified improperly with respect to his background or qualifications.

Martz testified that in the CTU, [w]e do chemical analyses on . . . evidence. In some cases, it may be drugs, in other cases it may be arson, or the identification of an accelerant, the identification of explosives. We do a wide range of identification of unknown chemicals. Whitehurst asserts that Martz led the trier of fact astray because the CTU never was entrusted with the analysis of explosives. This allegation again reflects Whitehurst's view that the MAU was solely responsible within the Laboratory for the analysis of explosives. The CTU, however, was at least analyzing smokeless powders, as several witnesses confirmed during our investigation.

Apart from the allegations previously discussed that the analysis by Martz was flawed because he did not follow the MAU protocol, Whitehurst also criticizes other aspects of Martz's testimony concerning smokeless powders.

On direct examination, Martz agreed that Red Dot smokeless powder came in the types of canisters represented by three exhibits. Whitehurst suggests Martz should have volunteered that the powder also comes in other types of products and containers. Similarly, Whitehurst states that Red Dot smokeless powder might be removed from manufactured ammunition. These points, if relevant, could have been developed by the defense attorney on cross-examination. Given the questions posed, Martz's responses were not improper.

Whitehurst also asserts that Martz lacked knowledge to testify about the degradation of smokeless powders and that he gave misleading testimony about his inability to match smokeless powders found in the devices with powder later obtained from the Shootin' Iron Gun Shop. Our interview with Martz left us persuaded that his remarks concerning degradation of smokeless powder had a basis in his prior work in the Laboratory, but his testimony about his attempts to compare powders was unnecessarily ambiguous.

With regard to determining whether different powder samples came from the same lot, Martz received a can of Red Dot smokeless powder that had been obtained from the Shootin' Iron Gun Shop sometime after Moody had bought powder there from Sartain. In our interview, Martz said he initially opposed attempting to determine if powder samples had come from the same lot, because he knew that smokeless powder's chemical composition changes with exposure to air and he anticipated that samples would yield different results. Martz performed liquid chromatography and gas chromatograph/mass spectrometer analysis of a sample from the can obtained from the Shootin' Iron, a sample from the Jacksonville device, and other lots of Red Dot smokeless powder from the Laboratory. Martz observed similarities and differences in his test results. This caused him to conclude, as stated in the Laboratory report dated June 6, 1990, that he could not determine whether the smokeless powder obtained from the Shootin' Iron came from the same lot as smokeless powder recovered from the bombs.

In his direct testimony, Martz stated that he examined four exhibits consisting of powder from each of the four devices and that the powder was Hercules Red Dot smokeless powder. Martz then acknowledged that he had also received a four-pound can of Hercules Red Dot smokeless powder. He then testified as follows:

Q: Did you do that?

A: Yes, I did.

Q: Could you determine anything at that point?

Later on cross examination, Martz again stated that he had tried to compare the powders but was unable to do so. The defense attorney asked, [t]hey were both Red Dot but you could not determine from your comparisons if they came from the same batch? Martz responded that he could not make that determination.

Martz was ambiguous in stating on direct examination that he had been unable to successfully compare the powders. In fact, he did compare a sample from the Jacksonville device, a sample from the four-pound can, and some known samples from the Laboratory. He should have stated more directly that he found differences and similarities when he compared certain samples. The differences, however, were insufficient to draw a meaningful conclusion as to whether the powders originated from different lots. As Martz stated in his interview with us, he found nothing in his comparison work suggesting that the samples had come from different lots. Accordingly, we conclude that Martz did not suppress exculpatory information regarding his comparison of the powder samples.

Whitehurst asserts that Martz should not have testified about the identification of primers in the detonators because examiner Roger Peele in the Elemental Analysis Unit (EAU) was responsible for primer residue analysis.

The green powder found in the detonators was sent initially to the CTU for analysis because its chemical composition was unknown. Based on analyses with infrared spectroscopy and a scanning electron microscope, Martz determined that the green powder was similar to primer materials that he had recently been analyzing. Through contacts with industry representatives, Martz learned that small arms primers made by CCI Industries were unique in having a 2% aluminum content.

In order to have the composition of the green powder analyzed further, Martz recalled that he asked Charles Peters, who then was an EAU technician, to perform inductively coupled plasma (ICP) atomic emission spectroscopy analysis of the samples. A logbook maintained by Peters indicates that he performed primer analysis for Martz in January 1990. Peters told us he could not specifically recall his work on the VANPAC case, and said he would have forwarded all his analytical results to Martz. Results of the ICP analysis have not been located, but Martz recalls that they confirmed a 2% aluminum content in the primers.

Roger Peele told us that he thought there was nothing improper in Martz's having analyzed the primer material in the VANPAC case. The EAU, Peele explained, at that time attempted to confirm the presence of gunshot primer residues on persons suspected of firing a gun by locating antimony and barium, components of primers. This analysis did not attempt to identify the particular primer or its manufacturer based on the overall composition of the primer. Peele believed that some interaction of units would have been necessary even if the unidentified green powder had first gone to the EAU, and he thought Martz properly involved the EAU by having Peters perform the ICP analysis.

On direct examination, Martz said that he had participated in a search of Moody's house in January 1991. He described his assignment as looking for residues of primer material and also smokeless powder. Martz explained that he and other agents had vacuumed the house and even pulled up a floor that Moody had replaced. Martz acknowledged on cross-examination that they had searched throughout the house, including the bags and brushes of vacuum cleaners that were there, and found no evidence of Red Dot smokeless powder.

Whitehurst complains that Martz was not qualified to search for residues. He also asserts that Martz lacked any basis to say that what was found in the search could have been affected if someone had vacuumed the area previously. We do not believe Martz lacked qualifications to conduct the search he described. Martz's statement that the results of his search might have been affected if someone had already vacuumed the scene strikes us as unobjectionable common sense.

Finally, Whitehurst complains that Martz lacked a basis to state on cross-examination that if any chemical tests had been done under the house, there would likely be detectable residue remaining on the pipes. Martz's observation was based on his experience as an agent and chemist, and we do not think it was improper because it was not supported by specific data or analytical results. In response to further defense questions, Martz said he had taken swabbings of pipes and other areas of the house looking for gunshot residues. He then acknowledged that no evidence of primers was found. The defense clearly pursued this line of testimony to underscore that although the FBI had conducted an extensive search, certain evidence was never found at Moody's house.

I. The Conduct of the Prosecutors

Whitehurst has suggested that prosecutors Howard Shapiro and Louis J. Freeh may have engaged in misconduct through their presentation of testimony by Martz and Thurman or their arguments in the VANPAC case.

We find no basis to conclude that either Shapiro or Freeh knowingly presented any improper evidence in the case. Both Freeh and Shapiro said that while they worked on the case they had never heard any suggestion that there was any impropriety in how the evidence had been analyzed within the Laboratory. Nor did they ever hear any suggestion that Thurman or anyone else had attempted to circumvent the Laboratory's procedures for the analysis of explosives or that the MAU should have done certain work rather than the CTU.

With respect to the closing argument, Whitehurst notes that Freeh reminded the jury of testimony by Paul Sartain, who said he had sold Moody four pounds of gunpowder and 4000 primers in December 1989. Whitehurst asserts that Freeh did not know what kind of explosive was used in the bombs, so this evidence was not probative in tying Moody to the bombs. For reasons noted earlier, we think that the FBI examiners could reasonably conclude that Red Dot smokeless powder and a high explosive detonator were components of each bomb. Sartain's testimony noted by Freeh in closing was both relevant and, in our view, highly probative.

Whitehurst also questions the basis for Freeh's remark to the jury that some eighty nails were traveling at 13,000 feet per second in the Birmingham bomb that killed Judge Vance. When interviewed in our investigation, Director Freeh could not recall the specific basis for his remark. The comment about 13,000 feet per second may reflect that Freeh misspoke or that the court reporter misheard him. Thurman testified that the cutoff between high and low explosives was 3,000 feet per second, and Freeh may have had in mind that number in his remarks. Both Thurman's testimony and the relevant Laboratory report indicated that eighty nails were attached to the Birmingham bomb.

Whitehurst also notes that Freeh in his closing commented about black paint being sprayed inside the devices to cover up fingerprints and also asked why every color of paint was found at Moody's house other than black. The statement that paint was sprayed was incorrect, as the Laboratory reports indicated the paint was brush-applied, and Thurman testified that the Laboratory had determined that the paint was hand-applied. We have no reason to think this was anything other than an honest mistake by Freeh. Whitehurst's other complaint here is that an FBI paint examiner, and not Thurman, should have testified about the significance of the paint. As noted above, Thurman understood and Moody's lawyer confirmed that Thurman could testify as a summary witness.

Freeh also stated in his closing that Thurman had been very conservative in his testimony. Freeh noted that Thurman did not conclude that the partially constructed Chamblee device was made by the same person who built the 1972 and 1989 devices. Whitehurst asserts that the jury may have been misled by these remarks because Thurman testified incorrectly, outside his expertise, or without a basis on various matters. We have addressed the allegations about Thurman's testimony above, and we find no basis to conclude that Freeh misled or attempted to mislead the jury in his comments about Thurman.

Whitehurst also notes that Howard Shapiro referred in the government's rebuttal argument to a survey conducted by the FBI that involved some 16,000 devices in a computer data bank and 217 crime laboratories around the country. Shapiro stated that the survey had not found any other device with certain features present in both Moody's 1972 bomb and three of the 1989 bombs. Whitehurst states that [t]his evidence should be thoroughly reviewed in detail because an EU technician named Mike Fanning had told Whitehurst that the FBI's Express computer data base had been built because the EU lacked such a data base at the time of the VANPAC case.

Shapiro's remarks in closing were based on Thurman's testimony. With regard to data bases, Thurman testified that the FBI had access to essentially three data bases: the EU's collection of previous Laboratory reports; the information on bombing incidents collected by the FBI's Bomb Data Center; and the data base maintained by the ATF. Thurman later described a review the FBI had conducted to determine if there had been other bombing incidents with devices with features similar to those of the bombs involved in VANPAC. He noted that the FBI had examined its own data base as well as the data bases maintained by ATF and the U.S. Postal Service and also had sent a survey to 217 forensic laboratories. Subsequently, Thurman testified that the survey and data bases had involved more than 16,000 devices, and that no devices were identified other than those involved in VANPAC that had certain design features and were designed to be sent in the mail.

The statements by Thurman and Shapiro concerning the review of some 16,000 devices had a factual basis. We interviewed Steve Schied, an Intelligence Research Specialist with the ATF, who has overseen the Exis data base since 1975 and who reviewed the data base for the VANPAC case. Schied said that at the time of the VANPAC case, the ATF Exis data base alone included 15,921 entries. He observed that it would have been accurate to say that the FBI survey involved approximately 16,000 devices based only on the ATF data base. Insofar as the FBI also examined other data bases and surveyed other forensic laboratories, we find Thurman could properly testify that the FBI's review involved more than 16,000 devices.

IV. Conclusion

We find no basis for the allegations made by Whitehurst that Thurman and Martz obstructed justice, circumvented Laboratory protocols and procedures, perjured themselves, or fabricated evidence in the VANPAC case. We also find no support for Whitehurst's suggestion that Freeh or Shapiro engaged in prosecutorial misconduct. As explained above, there were certain areas in which we believe that agents Thurman or Martz testified ambiguously or, in relatively minor ways, inaccurately. We do not find any basis to conclude that this testimony involved knowing or deliberate misconduct.

Given the documentation we were provided by the FBI, we also conclude that the conclusions made by examiner Robert Webb concerning the origin of black paint, glue, RTV, and 2-inch wide tape were stated more strongly than was justified by the methods employed and the analytical results. We find that Webb did not intentionally attempt to fabricate evidence or to present biased conclusions.

The case does highlight several areas in which we believe the Laboratory's procedures should be improved. The Laboratory would benefit from: (1) expressly stated and agreed upon guidelines concerning the respective responsibilities of different uni