Three days after the inauguration, a B-52 came apart over Wayne County.

January 23, 1961
35.3394° N, 77.9606° W
Seymour Johnson Air Force Base, Goldsboro, North Carolina.

The MK 39 Mod 2 bomb did not possess adequate safety for the airborne alert role in the B-52.

PARKER F. JONES, SAME MEMO
One simple, dynamo-technology, low voltage switch stood between the United States and a major catastrophe.

By the end of the 1950s, the Strategic Air Command had a problem it could explain in minutes. A Soviet missile could reach an American bomber base in about half an hour. A B-52 parked on a runway could be destroyed before its crew got the engines turning. A B-52 already in the air could not. So the Air Force started keeping some of its bombers airborne around the clock. They flew long loops with live weapons in their bomb bays. Whatever happened on the ground, some part of the force would still be flying, and it could still answer.

The weapons were thermonuclear. The Mark 39 was a long steel casing with a parachute pack in the tail. Each one carried the explosive force of hundreds of Hiroshima bombs. It was built to arm itself through a chain of devices once it left the aircraft: a pull-out cable, timers, a pressure sensor that read altitude, and a switch the crew controlled from the cockpit. When every link in that chain closed, the bomb would fire. When one stayed open, it would not.

Engineers at Sandia, the laboratory in New Mexico responsible for making these weapons safe, had spent years asking what would happen if the links closed by accident. A fire. A crash. A wiring fault. They wrote memos. The bombers kept flying.

Seymour Johnson Air Force Base sits on the southeast edge of Goldsboro, in the flat tobacco and cotton country of eastern North Carolina. The town had lived alongside the base since the Second World War. Airmen rented houses in town, and their children went to the local schools. The big jets came and went over farmland at every hour, and people stopped looking up.

On Friday, January 20, 1961, John F. Kennedy was sworn in as president in Washington. That Monday, January 23, a B-52G took off from Seymour Johnson on an airborne alert mission. It had a crew of eight, and it carried two Mark 39 bombs.

The aircraft commander was Major Walter Tulloch. The crew included a third pilot, First Lieutenant Adam Mattocks. His place on the flight deck had no ejection seat. On a B-52, the men at the ejection stations could fire themselves out of the aircraft. A man without a seat had to get out through an open hatch on his own, fighting the wind and the spin. Crews knew what that meant, and nobody expected to need it.

The flight was routine for hours. Late that night, the bomber met a tanker for aerial refueling, the job that kept airborne alert possible. During the hookup, the tanker crew saw fuel streaming from the B-52's right wing and told the bomber. The refueling stopped. The leak did not.

The crew reported it and got instructions. The bomber was losing fuel faster than it could burn it. The plan changed to getting the aircraft back to Seymour Johnson. The wing had other ideas. As the B-52 came down through the dark toward the base, the right wing gave way. The aircraft rolled and began to break up.

Tulloch ordered the crew out. The men in ejection seats fired. Mattocks climbed up to the overhead hatch and threw himself through it into the slipstream. His parachute opened. He came down alive. Men in his position were not expected to survive a bailout from a falling B-52. He did.

Five of the eight lived. Three did not.

The bomber came apart over farmland north of Goldsboro, near a crossroads community called Faro. The wreckage burned across the fields. As the airframe broke open, the two bombs fell free of the bomb bay, and the bombs did what their design told them to do when they left an aircraft. They began to arm.

The first bomb's parachute deployed. It drifted down slowly and came to rest in a field with its nose in the dirt and its parachute tangled in a tree. When the recovery teams reached it and worked through its mechanisms, they found that three of its four arming devices had operated. The pull-out had happened. The timers had run. The pressure sensor had registered a fall. The only link still open was the switch controlled from the cockpit, the one the crew would have thrown on purpose in a real attack. It had not closed. The bomb had not fired.

The second bomb's parachute did not open. It hit a wet field at high speed and broke apart. The pieces drove down into soft, waterlogged ground. A crater was all that showed on the surface.

An explosive ordnance disposal team was sent in to find it. The unit included a young lieutenant named Jack ReVelle. It was winter in a swampy field, and the hole filled with groundwater as fast as the men could dig. Pumps ran constantly. The team worked in mud and cold water, pulling up fragments of a hydrogen bomb one piece at a time. They recovered the weapon's arm/safe switch. It was found in the armed position.

They kept digging. The parts of the bomb that held its thermonuclear material had gone down deeper than the rest. The team went after them until the water and the depth made it too much. The heaviest piece, the uranium-bearing secondary stage, was never brought up.

The public account in 1961 was short. The Air Force said a bomber had crashed, that it had been carrying weapons, and that there had been no danger of a nuclear explosion. Newspapers printed it. People in Faro knew soldiers had spent weeks in a farm field. Most knew little beyond that.

The Air Force bought an easement on the land where the second bomb had gone in. The farm stayed with its owner and the fields went back to crops, under one condition: nobody could dig below a set depth on that ground.

Airborne alert kept going. B-52s with hydrogen bombs aboard kept flying their loops over the Atlantic, the Mediterranean, and the Arctic. In 1966 a B-52 collided with a tanker over Palomares, Spain, and dropped four hydrogen bombs. Two of them spread plutonium across farmland when their conventional explosives went off. In January 1968 a B-52 caught fire and crashed onto the sea ice near Thule Air Base in Greenland with four more.

In 1969, a Sandia supervisor named Parker F. Jones went back to Goldsboro on paper. He was responsible for nuclear weapon safety at the laboratory. He went through the record of what the two bombs had done that night: which devices had operated, which had not, and why. He wrote it up in a memo. He titled it "Goldsboro Revisited, or: How I Learned to Mistrust the H-Bomb," after the Stanley Kubrick film that had come out five years earlier.

His conclusions were plain. The Mark 39 had not been safe enough for the job it was given. On the bomb that came down under its parachute, the last thing between Wayne County and a thermonuclear detonation had been a single low-voltage switch, the kind of device a short circuit could have closed. He noted that such a short was not hard to imagine in a bomber breaking up and burning.

The memo was classified. It stayed that way.

In the years after Goldsboro, the weapons laboratories redesigned how nuclear bombs were made safe. Newer weapons were built so that heat, crushing, and electrical faults in an accident would disable them rather than arm them. Engineers who pushed for those changes pointed to the accident record, and Goldsboro was near the top of it.

The Air Force's public position did not change. For decades its statements on Goldsboro said the weapons had not come close to detonating.

In 2013, the journalist Eric Schlosser obtained the Jones memo through a Freedom of Information Act request while researching a book on nuclear weapons accidents. The Guardian published it in September of that year. It was the first time the laboratory's own assessment of that night had been read in public, fifty-two years after the fact and forty-four years after Jones wrote it.

No one connected to the Goldsboro crash faced criminal charges. The cause was recorded as a structural failure of the right wing after the fuel leak.

Airborne alert with nuclear weapons ended after the Thule crash in 1968. The Mark 39 was retired from the arsenal.

Adam Mattocks came home and lived for decades afterward. For years the few published accounts of the crash described him as one of the only men known to have bailed out of a B-52 through the top hatch and lived.

In 2012 the State of North Carolina put up a highway historical marker on the road near Eureka, a few miles from the crash site. It reads "Nuclear Mishap."

The easement is still on the property. The field is farmed every season. Under it, below the depth the plow is allowed to reach, is the part of the second bomb the Air Force could not dig out.