He has kept his secret for 60 years.

Paul Landis was one of two Secret Service agents tasked with guarding first lady Jacqueline Kennedy on November 22, 1963—the day President John F. Kennedy was assassinated. In a new book, The Final Witness, to be published in October, Landis claims to have seen something that afternoon that he had never publicly admitted before. His secret, coming to light only now, will certainly reorient how historians and laymen perceive that grave and harrowing event. His account also raises questions about whether there might have been a second gunman in Dallas that day.

After much prodding and reflection, Landis, now 88, made the decision to begin laying out his recollections for publication. Because I have written three books on presidential history, and because Landis’s publisher, Chicago Review Press, happens to be my publisher, an editor there asked me to read a copy of the galley and offer my comments, which I did quite eagerly. In fact, I was so taken with Landis’s backstory and, upon spending time with him, so drawn to the facets of his tale that are not answered in the book (whose details were first reported in The New York Times), that I probed further, maintaining a healthy dose of skepticism.

And yet, as I got to know him during more than a dozen meetings this past year, I was won over by his integrity and by the way his account of what he witnessed in Dallas—and in the grave months of American mourning that followed—remained consistent and unwavering. Over time, Landis and I became close. As a result, I am writing this assessment of his narrative (and of his motives for coming out with his story) not only as a historian and armchair investigator but as Landis’s confidant.

Twenty-three-year-old Paul Landis applied to become a Secret Service agent in 1958. He came from Worthington, Ohio, a suburb of Columbus, and had graduated from Ohio Wesleyan University 15 months earlier. A neighborhood boy, Bob Foster, who was friends with Landis’s sister, had joined the Secret Service two years before. After speaking with Foster, Landis thought being in the Secret Service sounded like the “coolest job in the universe.”

Landis was intrigued. But because he has always been slight of build, his immediate concern was whether he could meet the minimum height requirement (five feet, eight inches). During the physical exam, he stretched himself like a rubber band and, as he recalls, barely made it.

He started work in October 1959, at the time the youngest special agent, at 24. Just over a year later, John Kennedy was elected president; soon the young recruit was assigned the job of guarding the Kennedy children and, eventually, along with Special Agent Clint Hill, Mrs. Kennedy herself. Not all agents were given code names, but as a result of Landis’s new assignment, and because of his youth and boyish looks, he was eventually christened “Debut.”

Landis found himself deep in the inner workings of Camelot, coinciding with the apex of Jackie’s popularity. As an international superstar, she was the Princess Di of her era, and Landis was on hand as the media followed her every move. Landis traveled with the first lady and her daughter, Caroline, to Italy in 1962. (John Jr., her young son, remained back home.) Landis was the agent who helped speed and accompany Jackie to the Otis Air Force Base emergency facilities when she went into premature labor with son Patrick, who died two days after his birth in August 1963. That October, at the suggestion of Jackie’s sister, Lee Radziwill, a trip to Greece followed for an excursion aboard the luxury yacht of the shipping magnate Aristotle Onassis.

Then came November 22, 1963. A month after returning from Greece, Landis stood on the right rear running board of the Secret Service follow-up car, code-named “Halfback,” in the president’s motorcade as the vehicle headed from Dallas’s Love Field airport to a luncheon at the city’s Trade Mart. Landis was approximately 15 feet away when Kennedy was mortally wounded, a close witness to unspeakable horror.

That horror was compounded when the president’s limo reached Parkland Memorial Hospital, where Landis and Clint Hill tried to coax Jackie to release the president, whom, by that point, she had cradled in her lap. Climbing into the back seat area, which had been spattered with blood and brains and bullet fragments, both agents, according to their subsequent accounts, gently encouraged the first lady to let go.

As she did—standing up to follow Hill and another agent, Roy Kellerman, who lifted her husband’s body onto a gurney and raced into the hospital—Landis saw and did something that he has kept secret for six decades, he says now. He claims he spotted a bullet resting on the top of the back of the seat. He says he picked it up, put it in his pocket, and brought it into the hospital. Then, upon entering Trauma Room No. 1 (at that stage, he was the only nonmedical person in the room besides Mrs. Kennedy, and both stayed for only a short period), he insists, he placed the bullet on a white cotton blanket on the president’s stretcher.

This secret, as it turns out, may upend key conclusions of the Warren Commission, the body created by President Lyndon Johnson to investigate the assassination.

The sad fact is that Landis—though required to provide his version of events to the Secret Service (and, in a second report, to what would become the Warren Commission)—never sat for an interview before the FBI and never testified before the commission itself. He left the Secret Service months after the assassination and before the panel had finished its work and issued its report.

Landis, to this day, attests that in the first few years following the assassination, he was simply unable to overcome his PTSD from witnessing the murder firsthand. He says that the mental image of the president’s head, exploding, had become a recurring flashback. He maintains that he desperately tried to push down the memories. He also says he felt unable to read anything in detail about the assassination until some 50 years later, starting in 2014, when he began to come to grips with all that he had witnessed, suppressed, and finally processed.

Landis, two years shy of his 90th birthday, remains vigorous. He exercises daily and plays golf once a week. He works a steady job as a security guard and a kind of welcome ambassador at the Cleveland History Center.

But still, the JFK conspiracy hounds are legion, and with his new book’s publication, Landis can expect intense scrutiny. I made myself available to him as a way of helping to prepare him for what was to come.

In writing this analysis of his account, I have tried to determine if his story was possible, seen against the mountains of evidence, not to mention newly released documents collected over the years by the Warren Commission, congressional probes, countless individual authors, and a kind of industry that has evolved, in which dozens upon dozens of “researchers” trade information and, inevitably, misinformation.

Over the decades, there have been endless theories surrounding the assassination, but not one of them considered that a Secret Service agent might have brought a fully intact bullet, found on top of the rear seat of the limousine, into Parkland Memorial Hospital and placed it on the president’s stretcher. Not one.

So there is virtue in looking anew at the evidence that was collected in 1963 and attempting to draw some tentative conclusions.

My own conclusion is that Landis’s story, for several reasons, is not just possible; it in fact makes more sense than the core finding of the Warren Commission, known as the “single bullet” theory.

That theory posits that a single bullet caused all of the wounds in Kennedy’s neck as well as all of the serious injuries to Texas governor John Connally—who was sitting in front of the president at the time—including the shattering of four inches of Connolly’s fifth rib and the fracturing of a major bone in his right wrist.

Yet the bullet that Landis now claims to have discovered that morning emerged largely intact and only moderately damaged, its base having been squeezed in.

By possibly placing the “magic bullet” theory in doubt, Landis’s disclosure raises as many questions as it answers. I will try to address some of them here.

First, it makes sense to retrace the main tenets of the Warren Commission’s official version of the assassination. According to the panel’s final report, issued in September 1964, three gunshots rang out as the president’s limousine passed by the Texas School Book Depository building in Dallas. Witnesses’ auditory memory differed, their testimony ranging from two to six shots. Most, however, recalled hearing a trio of blasts.

Three spent shells, in fact, were found under a window on the sixth floor of the book depository. Nearby, partially hidden by some cartons, a rifle with a scope was discovered, a cheap Mannlicher-Carcano. Lee Harvey Oswald, the man history identifies as the lone assassin, worked in that building. The commission determined that the three shots all came from the sixth floor of the book depository.

The commission concluded that two of the three shots had hit the occupants of the limousine: One bullet had transited Kennedy’s neck and then, most probably, hit Governor Connally, and one had fatally wounded Kennedy, striking his head. (Connally survived the attack, later becoming President Richard Nixon’s Treasury secretary.)

In the view of the task force, one of the shots had likely missed the limo and, though the conjecture was inconclusive, possibly struck a nearby cement curb, sending a fragment that hit a spectator some distance away, near an overpass, slightly grazing his face.

But what of the ammunition itself? Two large bullet fragments were found in the front seat of the limo, and slivers of lead fragments were recovered from an area below the jump seat where the governor’s wife, Nellie Connally, had been sitting.

At Parkland Memorial Hospital, on the day of the assassination, an additional intact bullet was discovered on a stretcher. Through testing, commission investigators determined that the copper-jacketed, 6.5-millimeter bullet matched the rifling of the Mannlicher-Carcano that had been abandoned on the sixth floor of the depository. Testing on the bullet fragments resulted in a similar finding.

One key point to raise here concerns a fundamental underpinning of the Warren Commission report: the supposition that the retrieved intact bullet had been discovered on Governor Connally’s stretcher, not on Kennedy’s. It was from this assumption, in part, that the commission reached its pivotal conclusion: The available evidence indicated that “the bullet found on the Governor’s stretcher”—the single bullet—“could have caused all his wounds.” Over time, critics have referred to it as the “pristine” or “magic” bullet.

Moreover, if that single bullet did not cause the damage, then ballistics tests performed at the time suggest it would have been almost impossible for Oswald to have fired all three shots within the tight, multi-second time frame derived from the other main piece of evidence of the assassination: the Zapruder film, a 26.6-second home movie recorded in color by bystander Abraham Zapruder, a local clothing manufacturer who happened to have brought along his Bell & Howell 8mm camera that day and, by happenstance, captured the entire sequence of the assassination.

In his book, Paul Landis now says that when Jackie Kennedy stood up to enter Parkland, he looked over and saw that a bullet was improbably sitting on top of the rear seat of the limo, right around the spot where the limo’s detachable roof, which had been removed that day, would have otherwise been affixed to the trunk. Also, amid the blood and gore, Landis remembers, were two bullet fragments on the back seat, next to where Jackie had been sitting.

Landis contends that he reached over, picked up the lone bullet nestled in the crevice, and decided to place it in his pocket, mindful that if it were left there, precariously, it might be overlooked, pilfered by an unauthorized passerby, or misplaced once the president’s body was removed. Accompanying the first lady into Parkland, he says, he brought the bullet with him and, without conferring with Mrs. Kennedy, his fellow agents, or hospital staffers, placed it on JFK’s stretcher, thinking it needed to be with the body for the autopsy. As such, he contradicts a key linchpin underlying the findings of the Warren Commission. The bullet—as Landis tells it—was not from Connally’s stretcher.

From Landis’s description, three lines of inquiry emerge.

First, how did a largely intact bullet wind up on the ledge of the back seat, where JFK had been riding when he was shot?

Second, if Landis’s account is accurate, could Lee Harvey Oswald—who shot the president from a vantage point behind the motorcade—have acted alone, as the Warren Commission theorized?

And finally, why did Landis decide to keep this information to himself for six decades?

There is no way to know for sure how the undamaged bullet ended up on top of the rear seat. But there seem to be only two real possibilities, both of which can be inferred from the Zapruder film. One way is that an undercharged bullet, having already been lodged in the president’s back from an initial gunshot, was jolted out of his body after a subsequent shot to the head caused his upper body to be thrown violently back against the seat, bouncing off of it with great force. A second possibility is that at some point in those hectic moments, the bullet fell out of the president’s back and onto the first lady’s clothing (her white-gloved hand did brush hard against his back, around where the bullet could have been embedded at the moment of the final shot). As one can see in the Zapruder film, Jackie, at this stage, climbed onto the trunk of the speeding car, possibly to look for or retrieve a portion of her husband’s skull—or out of sheer panic to take cover from further gunshots. In fact, the section of the back seat over which she stretched corresponds to the spot where Landis says he found the bullet.

The autopsy evidence, as developed the night of the assassination, supports either one of these results.

A short recap is in order. Initially, President Kennedy was declared dead at Parkland—his body lying face-up on the table after the surgical team had performed a tracheotomy, hoping to provide needed oxygen through a ventilator to keep him breathing (which by that time was described as gasping or agonal respiration). These emergency room doctors used what they believed was an entry bullet wound in the front of the president’s neck to create the tracheotomy. They were apparently unaware of a bullet hole in the president’s back.

But later that night, an autopsy began at Bethesda Naval Hospital, near Washington, DC. During the procedure, doctors examined the president’s remains, only to discover a small bullet hole in the right shoulder, about five inches down from the top of the collar. This injury had gone unnoticed at Parkland since the president was declared dead before his body could be surveyed in its entirety. The Bethesda pathologists were puzzled when they probed the wound because it clearly was an entrance puncture, but it did not seem to have an exit wound, even though X-rays showed no bullet in the body.

In fact, the shoulder wound was shallow. Two doctors found that they could not pass more than half a pinky finger into the opening. Metal probes likewise uncovered no path of the bullet through the body.

Standing in proximity to the doctors were two FBI agents, Frank O’Neill and Jim Sibert, who had been dispatched by the bureau’s director, J. Edgar Hoover, to witness the autopsy and recover bullets or bullet fragments for the FBI lab. In their written statement, the agents discussed the frustration of the Bethesda doctors when they could not locate a bullet or exit wound for the projectile that had entered the president’s shoulder.

That night, according to the agents’ account, one of them placed a call to the FBI lab and found out that a “stretcher bullet” had been discovered at Parkland. Doctors used this information to theorize that “this accounted for no bullet being located which had entered the back region and that since external cardiac massage had been performed at Parkland Hospital, it was entirely possible that through such movement the bullet had worked its way back out of the point of entry and had fallen on the stretcher.”

The next morning, the Bethesda pathologists, as stated in their Warren Commission testimony, were told by Parkland doctors that the wound in the front of Kennedy’s neck was more than just the result of the tracheotomy they had performed. In fact, the Parkland team stated, there had been a bullet hole in the anterior (front) of the neck, and the ER staff had used that wound to create the tracheotomy. No one at the autopsy, according to FBI agents Sibert and O’Neill, had suspected there was a hole in the front of the president’s neck. With this new information, the Bethesda doctors revised their findings and assumed that the front wound was an exit for the bullet that had entered the president’s body from the back.

There were problems with this inference. The neck and shoulder had not been sectioned by those performing the autopsy to establish a bullet path. And by the time of the revelation of the front-neck injury, the president’s body had been brought to the White House to lay in repose in the East Room. (The next day, it lay in state in the Capitol rotunda.) Further, the wound in the back, according to Silbert and O’Neill, did not align with the location of the front-neck wound; such a pathway would have required a bullet traveling from the book depository, behind the motorcade, to have changed course inside the president’s body so as to exit higher up, through the neck, without hitting any bone to alter its course.

Agents O’Neill and Sibert didn’t buy it. “I do not see how the bullet that entered below the shoulder in the back could have come out the front of the throat,” O’Neill told the House Select Committee on Assassinations in 1978.

Landis’s discovery of the bullet on top of the rear seat, if true, comports with the initial finding: that the bullet had lodged superficially in the president’s back before being dislodged by the final blast to his head. It also explains the “pristine” nature of the bullet.

The genesis of the “single bullet” theory was twofold.

First, the Zapruder footage showed Kennedy reacting to the bullet that hit him in the back—and then, apparently, exited through the front of his neck (his arms spasmodically began to rise, elbows out, fists shielding his throat)—about a second or so before Connally seemed to react to his own wounding. To the Warren Commission staff, that double reaction on the part of the two men was puzzling. Given the type of weapon Oswald was using, there would have been no way for him to have gotten off two firings in such a short span of time.

Secondly, when the panel attempted to recreate the shooting in a manner consistent with the Zapruder film, FBI marksmen found that it took about 2.3 seconds to shoot, reload the bolt-action rifle, aim, and shoot again. Given Governor Connally’s reaction time, there did not appear to have been enough time for Oswald to have taken a second shot so quickly, let alone with any accuracy.

Howard Willens, an assistant counsel to the Warren Commission, and the author of the 2013 book History Will Prove Us Right, wrote about this dilemma: “If the interval between the first and second shots covered a span of less than 2.25 seconds, the time estimated to be necessary for the assassin to fire two shots, it might suggest that a second rifle was involved.”

The commission’s solution, however, championed by staff attorney Arlen Specter (who would become a US senator from Pennsylvania), was that the same bullet that hit Kennedy must have gone on to hit Connally on his right side. Connally’s second-later response was explained by the commission as a “delayed reaction” to an earlier wounding.

But, as noted, that theory depended on the single bullet having been found on Connally’s stretcher at Parkland Memorial Hospital, not on Kennedy’s stretcher.

The panel’s members speculated that the bullet, after causing Kennedy’s and Connally’s wounds, ended up superficially stuck in Connally’s left leg and must have dropped onto his stretcher once inside the hospital.

That said, the original evidence from 1963 is far from clear on this point.

The provenance of the bullet is also important in supporting or refuting Paul Landis’s purported memory. How was that bullet found? And how did it make its way to the FBI lab in Washington, DC, on the night of the assassination?

Landis’s recollection, as stated above, is that he found the undeformed bullet on top of the back seat of the limousine. “It was resting in a seam where the tufted leather padding ended against the car’s metal body,” he writes. When Jackie Kennedy stood up to follow her husband into the hospital, he saw it. He picked up the bullet, worried that souvenir seekers or others might take it or move it.

Upon arriving inside the emergency room, as stated above, he was jammed in with the first lady and a gathering horde of doctors and nurses. Standing near the feet of the president’s body, Landis left the bullet on his stretcher, as he believed it was crucial evidence and needed for the autopsy, which, under Texas law, should have taken place in Dallas.

But then a new chain of events overtook the gruesome sequence surrounding the assassination. A decision was made to transfer the president’s body, along with the first lady, Vice President Johnson, and others, back to Air Force One at Love Field. And with new tasks taking precedence for Landis—and the overwhelming national shock of the first assassination of an American president in 62 years (since the death of William McKinley in 1901)—the special agent simply never gave the bullet a second thought, he says. He had left it where someone would find it.

Landis didn’t make reference to the bullet in either of the two reports he submitted, hastily written in the turbulent days following the assassination. One short file, written two days after the funeral, didn’t even mention Parkland Memorial Hospital. A second, typed three days later—a day after Life magazine journalist Theodore White interviewed Jackie at the Kennedy compound in Hyannis Port, in what became known, famously, as the “Camelot” interview—was drafted during a time of deep shock and trauma.

That Thanksgiving, November 28—three days after the state funeral at which world leaders marched behind Mrs. Kennedy in the streets of Washington, DC—Landis and Hill traveled to Hyannis Port in a security capacity, protecting Jackie and her children. The agents had no time off to regroup or get their bearings. Sleep had eluded them. Landis had been up for practically four days straight. In the months after Lyndon Johnson was sworn in and assumed the presidential reins, Landis’s role switched from being part of the overall White House protection group to working full time for the former first lady. (Congress passed an act to authorize this service.) With this change of responsibilities, he found it hard to think of much beyond the weeks ahead. And if his thoughts did migrate back to November 22, he dwelled on the horrific scenes of the assassination, and rarely on what he says he considered a minor detail: the fact that he had picked up a bullet and placed it next to the president’s body.

The evidence from 1963 makes it fully plausible that the stretcher on which the bullet was found could have been President Kennedy’s. How so? A Parkland Memorial Hospital engineer, Darrell Tomlinson, was asked on November 22, before the president’s remains had been taken from the hospital to travel back north, to set the controls of the elevator in the emergency area—the one that had taken the wounded Governor Connally up to the second floor for surgery—so that the elevator would only be operable manually. The security team had determined that only people with official clearance would be allowed access; Tomlinson was instructed to control who got on the elevator and where they would go.

When he pushed the button to open the elevator, he later recalled, there was a stretcher in the elevator—one that the Warren Commission presumed was Governor Connally’s stretcher, returned from the surgery floor. Tomlinson testified that the stretcher had some sheets on it and a white covering on the pad, but no bullet. He moved the stretcher out of the elevator and placed it against a wall.

However, Tomlinson testified that there was another stretcher already in the hall, which had been placed in front of a men’s restroom in the corner. That stretcher had bloody sheets and some used medical paraphernalia on it.

Tomlinson said that sometime later, “an intern or doctor,” in order to use the bathroom, pushed the stretcher out of the way but failed to return it to its spot against the wall after leaving. Tomlinson roughly pushed it back against the wall, and when he did so, he claimed, a bullet rolled out from under the mat. This was clearly not Connally’s stretcher.

Arlen Specter, who had traveled to Dallas to take Tomlinson’s deposition in March 1964, was thunderstruck when Tomlinson relayed this scenario. To judge from a transcript of that conversation, Specter spent much of the remainder of his time with Tomlinson essentially trying to talk him out of his recollection, causing a distressed Tomlinson to say he just was not sure about his memory.

But the Q&A itself clearly suggests that Tomlinson, unprompted and unbadgered, had a cogent recollection that a bullet, wherever else it ultimately ended up, had come from the stretcher that had already been left in the hall in front of the men’s room.

In its final report, the Warren Commission mentioned nothing about this detail from Tomlinson’s account. Instead, the panel largely dismissed Tomlinson’s testimony, writing that even though he was “not certain whether the bullet came from the Connally stretcher or the adjacent one,” the commission “has concluded that the bullet came from the Governor’s stretcher.”

And what of the contemporaneous evidence from the witnesses who provided care to the governor in the emergency room and on the surgery floor?

When Connally was brought in from the limo, he was in great distress. He was taken to Trauma Room No. 2, where his clothes were immediately cut off. No one saw a bullet sticking out of his leg when his pants were removed. Neither was a bullet seen on the stretcher nor found in his clothing. Once stabilized, he was wheeled to the emergency room elevator and taken to surgery on the second floor.

Connally was then transported from the emergency room stretcher and placed on an operating table. No one testified to having seen a bullet on that stretcher. All of the medical paraphernalia on Connally’s stretcher was removed by a nurse before the stretcher was put back in the elevator. And the attendant who placed that stretcher onto the elevator did not recall having seen a bullet.

Darrell Tomlinson, however, went further. He said that as soon as he found the bullet, he alerted O.P. Wright, the chief of security at the hospital. Wright, in turn, gave the bullet to Secret Service agent Dick Johnsen, who was in the process of leaving the hospital with the president’s casket and Mrs. Kennedy. Johnsen wrapped the bullet in a handkerchief and placed it in his pocket. He then rode on Air Force One back to Washington.

Johnsen was assigned to sit with the casket and near Mrs. Kennedy in the back of Air Force One on the return journey to the nation’s capital. The body, the widow, and the bullet all returned to Washington on the same plane, in close proximity.

When Johnsen got back to the White House, he typed a short note to describe the bullet and how he ended up with it. “The attached expended bullet was received by me about 5 min., prior to Mrs. Kennedy’s departure from the hospital,” he wrote. “It was found on one of the stretchers located in the emergency ward of the hospital. Also on this same stretcher were rubber gloves, a stethoscope and other doctor’s paraphernalia.”

Johnsen dated his note: “7:30 p.m., Nov. 22, 1963.”

Johnsen then handed over the bullet to the director of the Secret Service, James Rowley. Rowley, finally, gave the bullet to the FBI that very night; it was signed in as the first piece of evidence in the assassination investigation, labeled with the designation “Q1.”

That evening, the initial supposition was that the bullet had come from JFK’s stretcher because the autopsy doctors at Bethesda, attempting to understand the whereabouts of the bullet that had entered Kennedy’s back, thought it might have been lodged in his back and then fallen out when rigorous chest massages were performed at Parkland.

It was only after the autopsy (and after the president’s body had been moved to the White House) that Parkland doctors told the pathologists that they had used a bullet wound in the front of Kennedy’s neck to make a tracheotomy.

Upon hearing this, the autopsy doctors tentatively revised their thesis and surmised that the bullet that entered Kennedy’s back must have exited through the front of his neck.

And so the problem started: The Warren Commission could not explain what happened to the bullet if it exited through the front of Kennedy’s neck. Howard Willens described the Warren Commission staff’s internal debate: “There was one basic question that now seems very simple,” he wrote. “Where did the bullet go after it exited the president’s neck?”

Paul Landis, in effect, answers that question: It ended up in a crevice on top of the back seat. It seems unlikely that the bullet traversed the president’s body and exited through the front of his neck.

Maybe the bullet entered the president’s back only superficially; these WW II–vintage bullets, after all, were notoriously undercharged with gunpowder. If this were the case, it might have indeed fallen out when he was violently struck with the final shot; when Mrs. Kennedy, at one point, pushed him down onto the seat; or when she clambered onto the trunk, the bullet falling off of her and onto the top of the back seat.

To be fair to the record, the Warren Commission’s findings suggested that Kennedy’s stretcher was not the stretcher in the elevator lobby because the nurses who had treated the president testified that once they’d been informed that a casket had arrived at Parkland, they had removed bloody sheets from his stretcher before moving it across the hall to Trauma Room No. 2.

But given the blood-soaked scene described by all of the medical personnel in Trauma Room No. 1, it is not unreasonable to assume that an orderly in Trauma Room No. 2 preemptively moved the president’s stretcher into the hall, stained sheets, medical paraphernalia, and all, to be further cleaned up by other attendants.

What does all this mean when considering whether Lee Harvey Oswald, as proposed by the Warren Commission, was the lone assassin? It certainly raises the stakes that another shooter might have been involved.

First, if the “pristine” bullet did not travel through both Kennedy and Connally, somehow ending up on Connally’s stretcher, then it stands to reason that Connally might have actually been hit by a separate bullet, coming from above and to the rear. The FBI recreation suggests that Oswald would not have had enough time to get off two separate shots so quickly as to hit Connally after wounding the president in the back. A second shooter must be considered.

And what about the bullet wound in the front of Kennedy’s neck? In one of the earliest critiques of the Warren Commission report, Josiah Thompson, author of Six Seconds in Dallas, proposed, not unreasonably, that the front-neck wound might have come from a bullet or bone fragment that was driven down and exited through the president’s throat from the final blast to his skull.

But there are other, darker explanations arising from the secrecy surrounding the X-rays and photographs taken at the autopsy and then not made public for decades. Jerrol F. Custer, the principal X-ray technician at the autopsy, testified in 1997 that there were several small metallic fragments in the cervical spine (the spinal region directly below the skull), which were visible in an X-ray, and that this was one of three X-ray exposures he took that night that went missing from the collection in the National Archives. This might have contained evidence of a shot from the front of the motorcade—a frangible bullet that disintegrated into tiny pieces after entry into the body. A heavy lift, for sure, but medical staffers who saw the front-of-the-neck wound before the tracheotomy believed it was an entrance wound because of its neatness.

Though some observers, down through the years, have mentioned the possibility of a second shooter, perhaps positioned on the so-called “grassy knoll” area along the route of the motorcade, neither this article nor Landis’s book has the insight or forensic expertise to hazard any new conclusions in this area. Others will have to analyze the evidence in full to see where it now leads. But, without question, this revelation will spark major debates. And that should not be surprising. The Warren Commission report, though lauded right after it was issued, has had its share of credibility detractors as time has passed. In 2013, a week before the 50th anniversary of the assassination, public opinion polls found that more than 60% of Americans believed the president’s murder had not been the work of one man, as the commission contended, but part of some kind of conspiracy.

There are many reasons for this skepticism, including a growing public distrust toward governmental pronouncements. And yet a chronic source of questioning has been the commission’s claim that a single bullet wounded both Kennedy and Connally, emerging mostly undamaged after having done so much damage.

Perhaps Landis’s revelations will offer a critical mass of additional conjecture to prompt a reassessment of the “magic bullet” theory.

The more confounding question, in my view, is a simple, human one: Why did Paul Landis decide to keep this information to himself for 60 years? The answer is complicated. But it is in many ways understandable if one considers his predicament at the time and in the years that followed.

Upon leaving the bullet on Kennedy’s stretcher, Landis explains today, he felt that he had done the right thing, expecting an autopsy and mindful of the need for the bullet to remain with the body. Like all of the Secret Service members on hand that day—and, indeed, the entire nation—he was also racked with grief and loss (to say nothing of PTSD, which was an unrecognized condition at the time). For Landis, however, a man who had been a constant presence in his life had been slain right in front of him—a man whose wife’s safety had been, in part, his own responsibility.

But soon the intensity of the moment enveloped him. Landis’s focus turned to responding to Mrs. Kennedy at Parkland; attending to the needs of the family upon the return of the remains to the capital; accompanying Jackie and the president’s brother Robert F. Kennedy during the nine-hour wait at Bethesda as the autopsy proceeded; remaining on protective duty during one of the riskiest state funerals in history; and then staying on in Jackie’s peripatetic orbit for the next several months—unable to find the time to properly attempt to cope with the raw trauma he had experienced.

In May, nearly six months after the assassination, Landis realized that the ordeal had taken its toll; concerned about his own mental health, he decided he couldn’t take it anymore. By August, at age 29, he had left the Secret Service. At the time, the Warren Commission had not issued its report, nor had Landis been interviewed for it; the public had not yet heard of the “single bullet” theory.

From that point onward, Landis now says, he withdrew from his more public-facing life, adamantly refusing to read more about the assassination. He avoided most requests for interviews by the media. He assumed, without looking at the final report, that the Warren Commission must have gotten it right.

Then, around 2010, things began to thaw. A project by Secret Service agent Gerald Blaine brought Landis out of his shell. Blaine was writing a book, The Kennedy Detail, with Lisa McCubbin, and Landis agreed to participate, mainly, he says, because his friend Clint Hill had also signed on. The agents met in Dallas to film an accompanying documentary, and Landis discovered he was not alone in feeling guilty and isolated following the assassination. But even then, Landis was not yet prepared to tell his own story.

That all changed in 2014 when Landis finally read a 1967 book on the assassination—a gift from a friend. When he began to page through Six Seconds in Dallas, by Josiah Thompson, he first saw that CE 399—the “pristine” bullet—was believed to have come from Governor Connally’s stretcher. Landis claims today that he immediately reached out to Clint Hill to tell him that he brought the bullet into Parkland and left it on the president’s stretcher, not the governor’s. He needed to correct this record.

One issue at the time was that the Secret Service was struggling with very public scandals, ranging from breaches of the White House grounds by intruders to accusations of liaisons with prostitutes in South America. And so, in deference to the troubles roiling the agency, Landis decided to remain mum.

In late 2015, however, Landis had a change of heart, he insists. He spoke with former Secret Service director Lewis Merletti, who had lived nearby, in Beachwood, Ohio, as the former head of security for the Cleveland Browns. Landis told Merletti of his secret; soon thereafter, Landis began the odyssey of carefully writing his book.

In the intervening seven years, he struggled with his conscience. His guilt, in my estimation, stemmed in part from a creeping concern that others might accuse him of having done something wrong by moving the bullet. Moreover, he must have worried, to some degree, about not having spoken out about finding the bullet in the first place—and not having sought to clarify the record more speedily once it became apparent to him that many historians and the public at large had cast doubt on the findings of the Warren Commission. Another factor amplifying his angst, I would imagine, was that the longer he remained silent, the harder it became to speak out.

Adding to his concerns, I’m sure, was the fact that Landis had been criticized—unfairly, in my view—for having stayed out most of the night before the assassination. Over the years, journalists and others have written sporadically about how “nine members” of the Kennedy Secret Service contingent went out drinking into the wee hours of November 22. A detailed report, with statements by the agents, was gathered following the assassination, and those comments are included as part of the record in the Warren Commission report volumes.

That investigation showed that around midnight on November 21, the Secret Service agents arrived in Fort Worth at the Hotel Texas with the president and first lady. Kennedy had barnstormed the state that day with events and motorcades in San Antonio and Houston. Though exhausted and famished, the agents were nonetheless used to putting in long hours in service to a youthful, hard-charging—and sometimes hard-partying—president. Double shifts were not uncommon.

Some members of the group, seeking food, were encouraged by Merriman Smith, a respected wire service reporter with United Press International, to visit a complimentary room set up by the Fort Worth Press Club in a nearby hotel. By that late hour, however, the food was gone. Several agents had a beer or two. Some had mixed drinks. Landis, for his part, professed to having had a scotch and soda.

Still hungry, the agents were told of a late-night coffeehouse, The Cellar, that might have something to eat. The Cellar was known for its “beatnik” performers—poetry-reading and guitar-playing—and a female waitstaff that wore skimpy outfits. Landis was among the agents who went to The Cellar, where, as he mentioned in one of his 1963 statements to authorities, he had two “Salty Dicks”—a local concoction that might or might not have had alcohol. Landis attested that he drank grapefruit juice, which he says today had no alcohol. At 4 a.m. CT, he says, he left to retire to his hotel.

By 8 a.m., Landis had breakfast and was asked by Roy Kellerman, the man in charge of the entire unit, to assist in protection for the president, who had decided to give an impromptu speech in front of the Hotel Texas. He later helped escort Mrs. Kennedy to a breakfast sponsored by the Fort Worth Chamber of Commerce at the hotel. He contends that he and his fellow agents that day were amped up by the heightened alertness involved with a presidential motorcade. They were likely stoked, as well, by the adrenaline of the crowds and the excitement and pomp. To a man, everyone involved who was interviewed for the Warren Commission said that all of the agents, including those who had gone out the night before, showed up on time for work, exhibiting no effects of excess drinking. Photos from that morning and day, in fact, show Landis bright-eyed and working diligently in his protective capacity.

Columnist Drew Pearson, however, got wind of the late-night excursion and in the week following the assassination published an exposé about the agents’ behavior. It caused an uproar. The incident, in fact, has continued to attract attention: In 2014, author Susan Cheever, finishing up a book called Drinking in America, reconstructed her own narrative of the evening for Vanity Fair, quoting both Paul Landis and Clint Hill.

I believe there is little reason to think that late-night alcohol consumption contributed to the agents’ response times or decision-making that day. Hill, who had been out with the others, reacted quickly in trying to get to the limo. In a six-second incident, he did not make it in time. The agent on the running board in front of Landis, Jack Ready, started to jump off himself but was called back by the agent in charge of that car, Emory Roberts, who feared that Ready might be run over and realized that Hill was already on his way toward the vehicle. Roberts had not been out the night before, drinking or otherwise. The same is true of the only agent who really had a chance to avert disaster, driver Bill Greer, who might have taken evasive action with the president’s limo once the shooting started. The agent next to Greer in the front passenger seat of the presidential limo, Roy Kellerman, likewise didn’t react in time. Kellerman had not been out drinking; he had gone straight to bed once they checked into the Hotel Texas.

Landis recognized that, despite any accusations to the contrary, there was nothing he could have done to prevent the tragedy. He also knew that he risked being criticized for having stayed out most of the night and having violated Secret Service policy by drinking any alcohol that might possibly impair him “if called upon to perform an official duty.” This no doubt contributed to his overall reluctance to come forward.

More to the point, I sense that he had an underlying guilt about what he might have done. He had found a bullet—the first piece of evidence logged into the record of the assassination of a US president—and then he went on his way, alone, in private.

He understands today how history might have changed had he told the pathologists at Bethesda that night where the “stretcher bullet” had come from—but he was not the one in the autopsy room (Kellerman and Greer were), and he had his hands full with the stream of family and mourners who arrived on the hospital’s 17th floor to console Jackie.

Landis is an upright, respected, private man. His moral authority and personal credibility have always been two hallmarks of his persona. My gut tells me that in his own way, he didn’t want to be the guy who had done a good deed under intense pressure, and then, forevermore, was raked over the coals for it. Which is how society often treats people these days. That anxiety might well have led to a sense of regret—even though his initial actions had been completely laudable.

In addition, he was in his late 20s at the time, a man whose values were grounded in those of the 1950s and ’60s. Silence and discretion, to him, had always been virtues. And he didn’t feel that it was appropriate to change his stripes and “go public”—drawing attention to his own behavior—when conspiracy theorists ran rampant, when other agents had been in the press over the years, and when President Kennedy had been killed, in effect, on his watch.

All of this, I contend, contributed to his years of silence.

But nothing, as I see it—and as Landis himself sees it—should detract from the fact that he has now come forward with his version of what happened on that dreadful day. And history will be the better for it.

James David Robenalt is an attorney and Washington Post contributor. He is the author of four nonfiction books: The Harding Affair; Linking Rings; Ballots and Bullets; and January 1973: Watergate, Roe v. Wade, Vietnam, and the Month That Changed America Forever.