One of the surviving engines—the only one still producing full power when the crippled B-17 reached the ground—fell from its mountings when an RAF winch began pulling the bomber away.
The men standing beside the aircraft had just watched its crew walk away from a wheels-up landing. Oil and aviation fuel were leaking from the Fortress. Two engines had already been lost during the return from France, another had been failing, and the flight-control system had suffered damage so severe that loose control cables could be seen hanging inside the fuselage. Yet the bomber had crossed the English Channel and reached British soil.
Then the winch took hold.
According to a 1992 account published by the 398th Bomb Group Memorial Association, the engine that had still been delivering full power released from its damaged mountings and dropped to the ground as the aircraft was being moved. The machine had survived the crossing with less structural and mechanical margin than even the men who flew it may have understood.
The aircraft was not an anonymous Flying Fortress from a wartime legend. It can be identified.
It was B-17G serial number 43-37892, carrying the 600th Bomb Squadron code N8-L. On August 1, 1944, it flew in the lead formation of the 398th Bomb Group on Mission 59 against Villaroche, France. The formation chart places Colonel Frank P. Hunter Jr. aboard as air commander and Captain Gene L. Douglas as the pilot of aircraft 7892-L. The group's aircraft-loss record identifies the same airplane, the same date, Mission 59, and its crash landing at RAF Merston after battle damage.
The most remarkable part of the flight was not simply that a four-engine bomber came home after losing two engines. B-17 crews had done that before, and some would do it again.
The remarkable part was how little conventional control Douglas had left.
Keith W. Anderson, Douglas's regular copilot, was riding in the tail position during this lead mission. Decades later, Anderson remembered looking forward through the damaged airplane and seeing control cables hanging in the waist section. Douglas did not trust what would happen if he disconnected the automatic pilot and tried to fly normally. The damaged Fortress therefore continued toward England largely through the automatic-pilot system, trim adjustments, engine power, and extremely careful handling.
That detail also corrects one part of the way stories like this are often retold.
A B-17 did not depend on hydraulic boost for its primary flight controls in the way a later heavy aircraft might. Its main control system relied on cables and mechanical connections. The B-17's hydraulic system operated equipment such as the brakes and cowl flaps, while the landing gear and wing flaps were electrically operated. A surviving 1944 Army Air Forces pilot-training manual for the Flying Fortress is held by the Smithsonian, and technical descriptions of the aircraft confirm that division between the electrical and hydraulic systems.
The danger aboard 43-37892 was therefore more direct than the phrase “hydraulics were gone” suggests.
The cables by which a pilot normally moved the airplane had been shot apart.
Douglas was trying to bring a wounded 65,000-pound-class bomber back from occupied France while depending on systems never intended to replace all normal control for a long emergency flight.
The U.S. Naval Institute has described one reason the Flying Fortress gained its reputation in simple terms: “the B-17 was able to sustain considerable damage and continue to fly.” The experience of 43-37892 on August 1, 1944, showed how far that statement could sometimes be stretched.
Yet the bomber's strength alone does not explain why everyone aboard survived.
The rest of the explanation begins several hours earlier at Station 131, Nuthampstead.
The mission began with an ordinary delay
August 1 was a Tuesday.
For the 398th Bomb Group, the day's target was Villaroche, an airfield near Melun, south of Paris. Allied armies had been ashore in Normandy for nearly two months, but much of France remained under German control. American heavy bombers were still striking airfields, transportation systems, industrial targets, and other military objectives across occupied territory.
The 398th had only begun combat operations in May 1944. Its crews were based at Nuthampstead in Hertfordshire, north of London, and the group was already becoming accustomed to the strange rhythm of Eighth Air Force operations. Men could spend hours preparing for a mission, sit through briefing, put on heavy equipment, walk toward their aircraft, and then discover that weather or operational changes had delayed everything.
One crewman left a particularly useful record of August 1.
Second Lieutenant Charles J. Mellis Jr., a copilot in the 603rd Bomb Squadron, kept a wartime diary and pilot log. His entry for that date identifies his eighteenth mission as Villaroche. Mellis wrote that the scheduled 9:30 a.m. takeoff had been delayed twice, each time by an hour. During one delay, some of the men played touch football on the hardstand.
That small detail is worth remembering because it places the mission back inside ordinary human time.
Several hours before a B-17 would cross the Channel with engines failing and control cables hanging loose, American airmen were passing time by playing football beside their bombers.
Nobody on the ground could know which airplane would return damaged.
Nobody could know which crew would be forced to decide whether to abandon an aircraft over France.
Nobody could know that one Flying Fortress would end the afternoon lying on its belly at an RAF fighter field far from its home station.
Mellis's diary also provides an independent view of what happened later. He was flying elsewhere in the formation, not aboard Douglas's aircraft. He remembered the antiaircraft fire around Melun as accurate, although not unusually heavy. Then he saw what happened to the lead aircraft carrying Colonel Hunter.
Mellis believed for a time that they had lost him.
Instead, Hunter's aircraft eventually reached southern England and crash-landed.
That contemporary diary entry is important because it strips away the decades of legend that often grow around damaged-bomber accounts. The basic event was already being recorded by another airman on the day it happened.
The lead Fortress had been badly hit.
Other crews thought it might be lost.
It nevertheless reached England.
The aircraft can then be traced through the 398th's surviving formation and loss records.
Mission 59's formation chart identifies Colonel Hunter with the lead group and Captain Douglas as squadron leader aboard 7892-L. The aircraft-loss table identifies the full serial, 43-37892, its code N8-L, its 600th Bomb Squadron assignment for the mission, the date 01/08/1944, the crash landing at RAF Merston, and battle damage as the cause of the loss.
There is no need to invent a mysterious declassified file number for this event.
A real searchable artifact already exists.
The published Mission 59 formation chart identifies its archival source as File 482b, Roll #2, PDF 0582-0720. Beside that record is Mellis's August 1 diary entry, Anderson's later oral history, the aircraft-loss list, and a crew photograph taken only one week before the mission.
Together, those records reveal a flight more remarkable than the simplified version.
Gene Douglas was flying the lead ship
Captain Gene L. Douglas was not simply flying another bomber somewhere inside a large formation.
His aircraft was helping lead it.
That changed the crew arrangement.
Colonel Frank P. Hunter Jr., commanding officer of the 398th Bomb Group, was aboard as command pilot and air commander. Douglas remained the aircraft pilot. During missions when a senior command pilot occupied the cockpit, Keith Anderson, Douglas's regular copilot, could be moved to the rear of the aircraft as an observer. Anderson later explained that this was why he was in the tail on August 1.
A photograph taken at Nuthampstead on July 25, 1944, only seven days before the Villaroche mission, shows Douglas's regular crew. The identified men include Anderson, bombardier Gilbert Goldman, Douglas, engineer Alfred Burke, navigator Harry Nelson Jr., radio operator Kenneth Shutts, and several of the gunners. The 398th's photograph record states that the crew had flown 43-37892 N8-L that day and that the airplane behind them was probably the same aircraft.
The photograph makes the August flight easier to understand.
These were not names placed later into a dramatic account. They were a working bomber crew. They knew one another's habits. They knew the sounds of their aircraft. They knew what another man's voice sounded like when something was wrong.
That kind of familiarity mattered inside a B-17.
A Flying Fortress was too large and too complicated for one pilot to survive combat alone. The men were scattered through the aircraft, sometimes separated by narrow passages, equipment, bomb-bay spaces, gun positions, and noise. One crewman's view could be completely different from another's.
Douglas could see the instruments and the air ahead.
Hunter could help manage the mission from the cockpit.
Goldman had the target and bombsight in the nose.
Nelson worked navigation.
Burke watched engines and systems.
Anderson, placed in the tail for this mission, had perhaps the most frightening view of what happened after the bomber was hit.
He saw the fire.
The pilots could not see it in the same way.
That difference would matter within seconds.
Why a lead B-17 depended on its automatic pilot
The automatic pilot aboard a wartime B-17 was not an early version of modern autonomous flight. It was an electromechanical system designed to maintain controlled flight and make corrections when properly set.
During a bombing run in a lead aircraft, the bomber's automatic controls could work in coordination with the bombsight. The purpose was precision. Large, abrupt pilot movements during a bomb run would spoil the bombardier's work and could disturb the formation following the lead aircraft.
Anderson later described the normal sequence from his own experience.
During the bomb run, the lead crew could be flying through the automatic-pilot system coordinated with the bombsight. Once the bombs were released, the aircraft's weight and center of gravity changed. The pilots would normally return to manual control, retrim the aircraft, and then use the automatic pilot again as needed.
On August 1, Douglas had a problem.
He no longer knew whether that normal transition could be made safely.
The control cables had been hit.
Some could be seen hanging loose farther back in the fuselage.
If Douglas disconnected the automatic pilot and discovered that the elevator, rudder, or aileron controls no longer responded correctly, reconnecting the system might not save the aircraft before it entered an uncontrollable attitude.
That created a strange situation.
The automatic pilot had been intended to help a functioning aircraft hold a planned attitude and course.
Now it was becoming part of an emergency control system for an airplane whose mechanical connections had been damaged by flak.
Douglas could still use trim.
He could change engine power.
He could make gentle corrections.
What he could not safely do was treat the B-17 as though nothing had happened.