The Magic Bullet: Even More Magical Than We Knew?

Gary Aguilar and Josiah Thompson


Among the myriad JFK assassination controversies, none more cleanly divides Warren Commission supporter from skeptic than the “Single Bullet Theory.” The brainchild of a former Warren Commission lawyer, Mr. Arlen Specter, now the senior Senator from Pennsylvania, the theory is the sine qua non of the Warren Commission’s case that with but three shots, including one that missed, Lee Harvey Oswald had single handedly altered the course of history. [Fig. 1]

Mr. Specter’s hypothesis was not one that immediately leapt to mind from the original evidence and the circumstances of the shooting. It was, rather, born of necessity, if one sees as a necessity the keeping of Oswald standing alone in the dock. The theory had to contend with the considerable evidence there was suggesting that more than one shooter was involved.

For example, because the two victims in Dealey Plaza, President Kennedy and Governor John Connally, had suffered so many wounds – eight in all, it had originally seemed as if more than two slugs from the supposed “sniper’s nest” would have been necessary to explain all the damage. In addition, a home movie taken by a bystander, Abraham Zapruder, showed that too little time had elapsed between the apparent shots that hit both men in the back for Oswald to have fired, reacquired his target, and fired again. The Single Bullet Theory neatly solved both problems. It posited that a single, nearly whole bullet that was later recovered had caused all seven of the non-fatal wounds sustained by both men.[1]

Figure 1. CE #399. Warren Commission Exhibit #399, said to have caused both of JFK’s non-fatal wounds and all five of the Governor Connally’s wounds, is shown in two views, above left. Arlen Specter theorized the bullet had followed a path much like the one shown at right. (National Archives photo)

But the bullet that was recovered had one strikingly peculiar feature: it had survived all the damage it had apparently caused virtually unscathed itself. The shell’s near-pristine appearance, which prompted some to call it the “magic bullet,” left many skeptics wondering whether the bullet in evidence had really done what the Commission had said it had done. Additional skepticism was generated by the fact the bullet was not found in or around either victim. It was found instead on a stretcher at the hospital where the victims were treated.

Mr. Specter’s idea was that, after passing completely through JFK and Governor Connally, the bullet had fallen out of the Governor’s clothes and onto a stretcher at Parkland Hospital. But it was never unequivocally established that either victim had ever lain on the stretcher where the bullet was discovered.[2] Nevertheless, studies done at the FBI Laboratory seemed to unquestionably link the missile to Oswald’s rifle, and the FBI sent the Warren Commission a memo on July 7, 1964 detailing how it had run down the bullet’s chain of possession, which looked pretty solid. According to the FBI, the two hospital employees who discovered the bullet originally identified it as the same bullet six months later in an FBI interview

That a bullet, fired from Oswald’s weapon and later identified by hospital witnesses, had immediately turned up on a stretcher in the hospital where the victims were treated struck some as perhaps a little too convenient. Suspicions it had been planted ensued. But apart from its peculiar provenance, there was little reason in 1964 to doubt the bullet’s bona fides. But then in 1967, one of the authors reported that one of the two hospital employees who had found the bullet, Parkland personnel director O.P. Wright, had told him that the bullet he saw and held on the day of the assassination did not look like the bullet that later turned up in FBI evidence. That claim was in direct conflict with an FBI memo of July 7, 1964, which said that Wright had told an FBI agent that the bullet did look like the shell he’d held on the day of the murder.

For thirty years, the conflict lay undisturbed and unresolved. Finally, in the mid 1990s, the authors brought this conflict to the attention of the Assassinations Records Review Board, a federal body charged with opening the abundant, still-secret files concerning the Kennedy assassination. A search through newly declassified files led to the discovery of new information on this question. It turns out that the FBI’s own, once-secret files tend to undermine the position the FBI took publicly in its July, 1964 memo to the Warren Commission, and they tend to support co-author Josiah Thompson. Thompson got a further boost when a retired FBI agent, in a recorded telephone interview and in a face-to-face meeting, flatly denied what the FBI had written about him to the Warren Commission in 1964.

A Bullet is Found at Parkland Hospital

The story begins in a ground floor elevator lobby at the Dallas hospital where JFK and John Connelly were taken immediately after being shot. According to the Warren Commission, Parkland Hospital senior engineer, Mr. Darrell C. Tomlinson, was moving some wheeled stretchers when he bumped a stretcher “against the wall and a bullet rolled out.”[3] He called for help and was joined by Mr. O.P. Wright, Parkland’s personnel director. After examining the bullet together, Mr. Wright passed it along to one of the U.S. Secret Service agents who were prowling the hospital, Special Agent Richard Johnsen.[4]

Johnsen then carried the bullet back to Washington, D. C. and handed it to James Rowley, the chief of the Secret Service. Rowley, in turn, gave the bullet to FBI agent Elmer Lee Todd,[5] who carried it to agent Robert Frazier in the FBI’s Crime Lab.[6] Without exploring the fact that the HSCA discovered that there may have been another witness who was apparently with Tomlinson when the bullet was found, what concerns us here is whether the bullet currently in evidence, Commission Exhibit #399, is the same bullet Tomlinson found originally.

The early history of the bullet, Commission Exhibit #399, is laid out in Warren Commission Exhibit #2011. This exhibit consists of a 3-page, July 7, 1964 FBI letterhead memorandum that was written to the Warren Commission in response to a Commission request that the Bureau trace “various items of physical evidence,” among them #399 [Fig. 2].  #2011 relates that, in chasing down the bullet’s chain of possession, FBI agent Bardwell Odum took #399 to Darrell Tomlinson and O.P. Wright on June 12, 1964. The memo asserts that both men told Agent Odum that the bullet “appears to be the same one” they found on the day of the assassination, but that neither could “positively identify” it. [Figs. 2, 3]

Figure 2. C.E. 2011. Chain of possession of #399 (FBI Letterhead Memo Dallas 7/7/64)

Positive identification” of a piece of evidence by a witness means that the witness is certain that an object later presented in evidence is the same one that was originally found. The most common way to establish positive identification is for a witness to place his initials on a piece of evidence upon first finding it. The presence of such initials is of great help later when investigators try to prove a link through an unbroken chain of possession between the object in evidence and a crime.

Understandably, neither Tomlinson nor Wright inscribed his initials on the stretcher bullet. But that both witnesses told FBI Agent Odum, so soon after the murder, that CE 399 looked like the bullet they had found on a stretcher was compelling reason to suppose that it was indeed the same one.

However, CE #2011 included other information that raised questions about the bullet. As first noted by author Ray Marcus,[7] it also states that on June 24, 1964, FBI agent Todd, who received the bullet from Rowley, the head of the Secret Service, returned with presumably the same bullet to get Secret Service agents Johnsen and Rowley to identify it. #2011 reports that both Johnsen and Rowley advised Todd that they “could not identify this bullet as the one” they saw on the day of the assassination. # 2011 contains no comment about the failure being merely one of not “positively identifying” the shell that, otherwise, “appeared to be the same” bullet they had originally handled. [Figs. 2, 3]

Thus, in #2011 the FBI reported that both Tomlinson and Wright said #399 resembled the Parkland bullet, but that neither of the Secret Service Agents could identify it. FBI Agent Todd originally received the bullet from Rowley on 11/22/63 and it was he who then returned on 6/24/64 with supposedly the same bullet for Rowley and Johnsen to identify. Given the importance of this case, one imagines that by the time Todd returned, they would have had at least a passing acquaintance. Had it truly been the same bullet, one might have expected one or both agents to tell Todd it looked like the same bullet, even if neither could “positively identify” it by an inscribed initial. After all, neither Tomlinson nor Wright had inscribed their initials on the bullet, and yet #2011 says that they said they saw a resemblance.

Figure 3. Last two pages of 7/7/64 FBI memo to Warren Commission, as published in C.E. #2011. Note that FBI states that both Dallas witnesses said #399 looked like the bullet they found on 11/22/63.

And there the conflicted story sat, until one of the current authors published a book in 1967.

Two Different Accounts from One Witness

Six Seconds in Dallas reported on an interview with O.P. Wright in November 1966. Before any photos were shown or he was asked for any description of #399, Wright said: “That bullet had a pointed tip.”

“Pointed tip?” Thompson asked.

“Yeah, I’ll show you. It was like this one here,” he said, reaching into his desk and pulling out the .30 caliber bullet pictured in Six Seconds.”[8]

As Thompson described it in 1967, “I then showed him photographs of CE’s 399, 572 (the two ballistics comparison rounds from Oswald’s rifle) (sic), and 606 (revolver bullets) (sic), and he rejected all of these as resembling the bullet Tomlinson found on the stretcher. Half an hour later in the presence of two witnesses, he once again rejected the picture of 399 as resembling the bullet found on the stretcher.”[9]
[Fig. 4]

Figure 4. In an interview in 1966, Parkland Hospital witness O.P. Wright told author Thompson that the bullet he handled on 11/22/63 did not look like C.E. # 399.

Thus in 1964 the Warren Commission, or rather the FBI, claimed that Wright believed the original bullet resembled #399. In 1967, Wright denied there was a resemblance. Recent FBI releases prompted by the JFK Review Board support author Thompson’s 1967 report.

A declassified 6/20/64 FBI AIRTEL memorandum from the FBI office in Dallas (“SAC, Dallas” – i.e., Special Agent in Charge, Gordon Shanklin) to J. Edgar Hoover contains the statement, “For information WFO (FBI Washington Field Office), neither DARRELL C. TOMLINSON [sic], who found bullet at Parkland Hospital, Dallas, nor O. P. WRIGHT, Personnel Officer, Parkland Hospital, who obtained bullet from TOMLINSON and gave to Special Service, at Dallas 11/22/63, can identify bullet … .” [Fig. 5 - Page 1, Page 2]

Whereas the FBI had claimed in CE #2011 that Tomlinson and Wright had told Agent Odum on June 12, 1964 that CE #399 “appears to be the same” bullet they found on the day of the assassination, nowhere in this previously classified memo, which was written before CE #2011, is there any corroboration that either of the Parkland employees saw a resemblance. Nor is FBI agent Odum’s name mentioned anywhere in the once-secret file, whether in connection with #399, or with Tomlinson or with Wright.

Figure 5. Declassified FBI memo reporting neither Tomlinson nor Wright could identify “C1” [#399] as the bullet they handled on 11/22/63.
[Page 1, Page 2]

A declassified record, however, offers some corroboration for what CE 2011 reported about Secret Service Agents Johnsen and Rowley. A memo from the FBI’s Dallas field office dated 6/24/64 reported that, “ON JUNE TWENTYFOUR INSTANT RICHARD E. JOHNSEN, AND JAMES ROWLEY, CHIEF … ADVISED SA ELMER LEE TODD, WFO, THAT THEY WERE UNABLE TO INDENTIFY RIFLE BULLET C ONE (# 399, which, before the Warren Commission had logged in as #399, was called “C ONE”), BY INSPECTION (capitals in original). [Fig. 6]

Convinced that we had overlooked some relevant files, we cast about for additional corroboration of what was in CE # 2011. There should, for example, have been some original “302s ” – the raw FBI field reports from the Agent Odum’s interviews with Tomlinson and Wright on June 12, 1964. There should also have been one from Agent Todd’s interviews with Secret Service Agents Johnsen and Rowley on June 24, 1964. Perhaps somewhere in those, we thought, we would find Agent Odum reporting that Wright had detected a resemblance between the bullets. And perhaps we’d also find out whether Tomlinson, Wright, Johnsen or Rowley had supplied the Bureau with any additional descriptive details about the bullet.

Figure 6. Suppressed 1964 FBI report detailing that neither of the Secret Service agents who handled “#399” on 11/22/63 could later identify it.

In early 1998, we asked a research associate, Ms. Cathy Cunningham, to scour the National Archives for any additional files that might shed light on this story. She looked but found none. We contacted the JFK Review Board’s T. Jeremy Gunn for help. [Fig. 7] On May 18, 1998, the Review Board’s Eileen Sullivan, writing on Gunn’s behalf, answered, saying: “[W]e have attempted, unsuccessfully, to find any additional records that would account for the problem you suggest.”[10] [Fig. 8] Undaunted, one of us wrote the FBI directly, and was referred to the National Archives, and so then wrote Mr. Steve Tilley at the National Archives. [Fig. 9]

On Mr. Tilley’s behalf, Mr. Stuart Culy, an archivist at the National Archives, made a search. On July 16, 1999, Mr. Culy wrote that he searched for the FBI records within the HSCA files as well as in the FBI records, all without success. He was able to determine, however, that the serial numbers on the FBI documents ran “concurrently, with no gaps, which indicated that no material is missing from these files.”[11] [Fig. 10] In other words, the earliest and apparently the only FBI report said nothing about either Tomlinson or Wright seeing a similarity between the bullet found at the hospital and the bullet later in evidence, CE #399. Nor did agent Bardwell Odum’s name show up in any of the files.

Figure 7. Letter to Assassinations Records Review Board requesting a search for records that might support FBI’s claim that hospital witnesses identified #399.

Figure 8. ARRB reports that it is unable to find records supporting FBI claim Parkland Hospital witnesses identified #399.

Figure 9. Letter to National Archives requesting search for additional files on C.E. #399.

Figure 10. Letter from National Archives disclosing no additional files exist on C.E. #399.

[editor's note: Dr. Aguilar followed up in 2005 with the National Archives, asking them in letters dated March 2 and March 7 to search for any FBI "302" reports that would have been generated from CE399 being shown to those who handled it. On March 17, 2005 David Mengel of NARA wrote back reporting that additional searches had not uncovered any such reports.]

Stymied, author Aguilar turned to his co-author. “What does Odum have to say about it?” Thompson asked.

“Odum? How the hell do I know? Is he still alive?”

“I’ll find out,” he promised.

Less than an hour later, Thompson had located Mr. Bardwell Odum’s home address and phone number. Aguilar phoned him on September 12, 2002. He was still alive and well and living in a suburb of Dallas. The 82-year old was alert and quick-witted on the phone and he regaled Aguilar with fond memories of his service in the Bureau.  Finally, the Kennedy case came up and Odum agreed to help interpret some of the conflicts in the records. Two weeks after mailing Odum the relevant files – CE  # 2011, the three-page FBI memo dated July 7, 1964, and the “FBI AIRTEL” memo dated June 12, 1964, Aguilar called him back.

Mr. Odum told Aguilar, “I didn’t show it [#399] to anybody at Parkland. I didn’t have any bullet … I don’t think I ever saw it even.”  [Fig. 11] Unwilling to leave it at that, both authors paid Mr. Odum a visit in his Dallas home on November 21, 2002. The same alert, friendly man on the phone greeted us warmly and led us to a comfortable family room. To ensure no misunderstanding, we laid out before Mr. Odum all the relevant documents and read aloud from them.

Again, Mr. Odum said that he had never had any bullet related to the Kennedy assassination in his possession, whether during the FBI’s investigation in 1964 or at any other time. Asked whether he might have forgotten the episode, Mr. Odum remarked that he doubted he would have ever forgotten investigating so important a piece of evidence. But even if he had done the work, and later forgotten about it, he said he would certainly have turned in a “302” report covering something that important. Odum’s sensible comment had the ring of truth. For not only was Odum’s name absent from the FBI’s once secret files, it was also it difficult to imagine a motive for him to besmirch the reputation of the agency he had worked for and admired.

Figure 11. Recorded interview with FBI Agent Bardwell Odum, in which he denies he ever had C.E. #399 in his possession.

Thus, the July 1964 FBI memo that became Commission Exhibit #2011 claims that Tomlinson and Wright said they saw a resemblance between #399 and the bullet they picked up on the day JFK died. However, the FBI agent who is supposed to have gotten that admission, Bardwell Odum, and the Bureau’s own once-secret records, don’t back up #2011. Those records say only that neither Tomlinson nor Wright was able to identify the bullet in question, a comment that leaves the impression they saw no resemblance. That impression is strengthened by the fact that Wright told one of the authors in 1966 the bullets were dissimilar. Thus, Thompson’s surprising discovery about Wright, which might have been dismissed in favor of the earlier FBI evidence in #2011, now finds at least some support in an even earlier, suppressed FBI memo, and the living memory of a key, former FBI agent provides further, indirect corroboration.

Missing 302s?

But the newly declassified FBI memos from June 1964 lead to another unexplained mystery. Neither are the 302 reports that would have been written by the agents who investigated #399’s chain of possession in both Dallas and Washington. The authors were tempted to wonder if the June memos were but expedient fabrications, with absolutely no 302s whatsoever backing them up.

But a declassified routing slip turned up by John Hunt seems to prove that the FBI did in fact act on the Commission’s formal request, as outlined in # 2011, to run down #399s chain of possession. The routing slip discloses that the bullet was sent from Washington to Dallas on 6/2/64 and returned to Washington on 6/22/64. Then on 6/24/64, it was checked out to FBI Agent Todd. [Fig. 12] What transpired during these episodes? If the Bureau went to these lengths, it seems quite likely that Bardwell Odum, or some other agent in Dallas, would have submitted one or more 302s on what was found, and so would Agent Elmer Todd in Washington. But there are none in the files. The trail ends here with an unexplained, and perhaps important, gap left in the record.

Figure 12. FBI routing slip. Note that #399 was sent from Washington to Dallas and back again, and that FBI agent Todd checked out the bullet on 6/24/64, the day it was reported the Secret Service Agents told Todd they could not identify #399. [See Fig. 5 (page 1, page 2) and Fig. 6.] (Courtesy of John Hunt)

Besides this unexplained gap, another interesting question remains: If the FBI did in fact adjust Tomlinson and Wright’s testimonies with a bogus claim of bullet similarity, why didn’t it also adjust Johnsen and Rowley’s? While it is unlikely a certain answer to this question will ever be found, it is not unreasonable to suppose that the FBI authors of #2011 would have been more reluctant to embroider the official statements of the head of the Secret Service in Washington than they would the comments of a couple of hospital employees in Dallas.


In a memo to the Warren Commission [C. E. #2011] concerning its investigation of the chain of possession of C.E. #399, the FBI reported that two Parkland Hospital eyewitnesses, Darrell Tomlinson and O. P. Wright, said C.E. #399 resembled the bullet they discovered on the day JFK died. But the FBI agent who is supposed to have interviewed both men and the Bureau’s own suppressed records contradict the FBI’s public memo. Agent Odum denied his role, and the FBI’s earliest, suppressed files say only that neither Tomlinson nor Wright was able to identify the bullet in question. This suppressed file implies the hospital witnesses saw no resemblance, which is precisely what Wright told one of the authors in 1967.

What we are left with is the FBI having reported a solid chain of possession for #399 to the Warren Commission. But the links in the FBI’s chain appear to be anything but solid. Bardwell Odum, one of the key links, says he was never in the chain at all and the FBI’s own, suppressed records tend to back him up. Inexplicably, the chain also lacks other important links: FBI 302s, reports from the agents in the field who, there is ample reason to suppose, did actually trace #399 in Dallas and in Washington. Suppressed FBI records and recent investigations thus suggest that not only is the FBI’s file incomplete, but also that one of the authors may have been right when he reported in 1967 that the bullet found in Dallas did not look like a bullet that could have come from Oswald’s rifle. 

[1] The eighth wound, JFK’s head wound, accounted for one of the bullets. And evidence from the scene and from a home movie taken of the murder by a bystander, Abraham Zapruder, suggests that a third bullet had missed entirely.

[2] Josiah Thompson. Six Seconds in Dallas. Bernard Geis Associates for Random House, 1967, p. 161 – 164.

[3] The President’s Commission on the Assassination of President John F. Kennedy – Report. Washington, D.C.: U. S. Government Printing Office, 1964, p. 81. See also 6H130 – 131.

[4] 18H800. See also: Thompson, J. Six Seconds in Dallas. New York: Bernard Geis Associates for Random House, 1967, p. 155.

[7] See Ray Marcus monograph, The Bastard Bullet.

[8] Text of email message from Josiah Thompson to Aguilar, 12/10/99.

[9] Thompson, Josiah. Six Seconds in Dallas. New York: Bernard Geis Associates for Random House, 1967, p. 175.

[10] 5/11/98 email message from Eileen Sullivan re: “Your letter to Jeremy Gunn, April 4, 1998.”

[11] Personal letter from Stuart Culy, archivist, National Archives, July 16, 1999.