Last to Die

by Stephen Harding

  Already wearing their flight suits, ten or twelve pilots, including Sakai and Ryoji Ohara, ran to the fueled and armed Georges and Zekes sitting in the alert revetments. Sakai leapt into one of the latter aircraft, pleased at the unexpected chance to fly his favorite mount into combat one more time, and minutes later the fighters roared aloft, clawing for altitude even as their gear and flaps retracted.

  Wells and the men aboard his Dominator were the first to realize that the Japanese intended to oppose the recon flight. Even before 578 had started her photo run a Yoko Ku George (which the Americans misidentified as the very similar-looking army Ki-44 “Tojo”) zoomed past, several hundred feet above the B-32, then rolled inverted and reversed course, coming back at the American aircraft from the two o’clock position (off the bomber’s nose, slightly to starboard), firing as it came.39 Wells immediately turned the Dominator directly toward the incoming fighter and the nose turret gunner opened up with his twin .50s. The pilot of the George, possibly used to attacking B-29s (which had no nose turret), was apparently startled by the volume of return fire and immediately broke off the attack.40 Wells and his crew were not out of the woods yet, however. As 578 began her first photo run the aircraft’s electronic countermeasures officer, twenty-six-year-old Second Lieutenant David S. Samuelson, warned that a fire-control radar had locked onto the B-32. He immediately began jamming the Japanese signal, but several rounds of 120mm anti-aircraft fire burst close enough to rock the aircraft as it finished its run and then turned for the coast.

  Wells and his crew were not the only ones being targeted by anti-aircraft guns. At about that same time, Hobo Queen II was finishing the second of her flight lines, just north of Tokyo Bay. Though the B-32’s countermeasures operator had not detected any Japanese radar emissions, the tail gunner suddenly called out that ack-ack bursts seemed to be creeping up on the aircraft from behind. Understandably apprehensive and likely more than a little unnerved, the gunner made his announcement rather more loudly than he probably intended, for Cook—whom the Dominator’s radio operator, Staff Sergeant Robert Russell, later remembered as “real calm”—came on the intercom and said, “cut out that screaming back there. I’ll get us out of this, don’t worry about it.”41 Cook was as good as his word; realizing that determined Japanese resistance would make the planned mission impossible to complete, he immediately turned the B-32 south toward Tokyo Bay and put it into a slight dive to pick up speed. Though Japanese fighters ahead of and above the Dominator were sighted by one of the gunners, they did not attack, and Hobo Queen II was able to escape further enemy attention.

  The same cannot be said for either Harriet’s Chariot or 539, however.

  Svore and his crew had started their second flight line, just west of Chiba, when one of the gunners called out incoming enemy fighters. Harriet’s Chariot had not been fired on up to that point, but the sudden appearance of apparently hostile interceptors did not surprise the 386th’s commander. He later recalled that it simply validated his belief that the Japanese would be far more likely to intercept Allied aircraft appearing over Tokyo than they would be over other, less symbolic areas. Whatever their motivation, the enemy pilots seemed determined to inflict punishment on the B-32. Six aircraft attacked Harriet’s Chariot; they were later identified by the bomber’s crew as five army Tojos and one Kawasaki Ki-61 Hien (“Tony”), though the former were actually navy Georges.42 The first pass was made by the Tony, which started its run some 3,000 feet above and directly behind the Dominator. The fighter swooped down and pulled up astern and slightly below, and opened fire at the same time as the B-32’s rear upper gunner. The fighter took several hits, began trailing smoke, then rolled inverted and dived into the clouds below. The Georges then came at Harriet’s Chariot from different directions and altitudes, eventually concentrating their attacks on the bomber’s aft end after realizing that the tail turret was malfunctioning. Though the fighters kept up their attacks for about fifteen minutes the only damage they inflicted was a small hole in the B-32’s wing, and Harriet’s Chariot was able to make it to the sea and find cover in a layer of heavy clouds.43

  Given that Oppama was the primary photo target assigned to Emory Frick and the men aboard 539, it is not surprising that they received far more attention from the Japanese than the morning’s briefing at Yontan had led them to expect. As the Dominator began the initial pass over the Yokosuka area at 20,000 feet she was the target of some fifty bursts of radar-directed 120mm anti-aircraft fire, several of which were close enough to punch holes in the nacelle of the number four engine and the inboard trailing edge of the left wing. Then, even before 539 came off the photo pass a group of fighters appeared. One of them was flown by Sakai, who was surprised to find that the quarry was not the B-29 he was expecting, but a completely different aircraft. As he was noting the Dominator’s “enormous” vertical stabilizer, the other Yoko Ku pilots rolled in on her, initially from twelve o’clock high. Tracer rounds from the fighters’ guns were clearly visible in the bright noontime sky as they arced in toward the bomber. So aggressive were the Japanese airmen that Frick’s copilot, Second Lieutenant Joe E. Elliot, was certain they intended to ram the Dominator.44 That didn’t happen, but over the next twenty minutes the Japanese came at the B-32 from virtually every angle. The sky around 539 was so crowded, in fact, that Sakai had to abort his initial firing pass—from three o’clock high—when another fighter pulled in front of him.

  As soon as the attack started Frick had begun a sweeping turn to the south, toward the entrance to Tokyo Bay. His gunners kept up a steady defensive fire throughout the maneuver, observing hits on at least two of the fighters, both of which they later claimed as “probably destroyed.” As he completed the turn Frick put the B-32 into a slight dive and jammed all four throttles to their stops. Sakai, at this point waiting his chance for another pass, saw the Dominator accelerate so quickly that he wondered whether the bomber was equipped with some sort of auxiliary rocket engine. He and his wingman—most probably Ryoji Ohara—kept up the chase nevertheless, as did one or two of the other fighters, lobbing rounds at the B-32 as it pulled away from them. One by one the Japanese pilots gave up and turned back toward Oppama, with Sakai and his partner hanging on until the Dominator reached Miyake-jima. Then, fearing they might run into U.S. Navy carrier fighters, the two pilots broke off their pursuit and turned back toward Yokosuka.

  The departure of the last Japanese fighters was undoubtedly greeted by a sigh of relief from the men aboard 539. Though none of them had been injured in the aerial melee, their B-32 was somewhat the worse for wear—20mm cannon rounds and 12.7mm machine-gun bullets had holed both wings and the rudder trim tab. The damage was not serious enough to prevent the Dominator’s return to Okinawa, however, and as Frick took up a course for Yontan the men aboard 539 settled in for the long flight and the first of many discussions about the day’s events. Similar conversations were undoubtedly taking place aboard the other three B-32s, each of which was making its own way back to Okinawa.

  THE B-32 CREWMEN WERE not the only people talking about what had happened over Tokyo, of course. Terse radio accounts of the Japanese attacks radioed back to Yontan by Cook, Wells, and Svore had also sparked intense concern among senior officers at V Bomber Command and FEAF headquarters. Although the interception of the Dominators was the most serious breach of the ceasefire, it was not the only one that had taken place over Japan that day. A Consolidated F-7 Liberator on a recon mission to Yokohama had also been fired on by anti-aircraft batteries, as had two F-5 Lightnings over Kyushu.45

  The high level of command interest in what had happened to the B-32s was clearly evident when the aircraft finally landed at Yontan at about 5 P.M. The Dominators were immediately surrounded by dozens of senior officers—and reporters both civilian and military—all clamoring to know what had happened over Tokyo. As soon as the men stepped from their aircraft they were whisked off to the standard post-flight debriefing, where for the next few hours they were grilled abou
t their actions and those of the Japanese. The aviators were understandably proud that they’d been able to give better than they’d received—the Dominators’ gunners claimed a total of four “probable” kills of Japanese fighters for no losses or injuries on their own side—but they were also angry. In their minds the Japanese had once again shown themselves to be treacherous and deceitful, their apparent disregard for the ceasefire just another example of Tokyo’s perfidy.46

  For the senior V Bomber Command and FEAF officers the issues presented by the attacks that day were potentially far larger, however. The fact that these events occurred just two days before Japanese envoys were scheduled to fly to Manila as MacArthur had directed was understandably troubling for the men who’d been tasked to plan and implement the aerial portions of the occupation. Was it all just a misunderstanding? After all, in his initial messages to the Japanese MacArthur had indicated that no bombers would fly over Japan, but he’d said nothing of reconnaissance flights. Had the Japanese fired on the F-7 and B-32s because they were assumed to be bombers that might have some secret, hostile intent? Were the Japanese attacks simply the acts of diehards and renegades, they must have wondered, or did they foreshadow something far more sinister on the part of Japan’s leaders?

  After reviewing all the information presented by the returning Dominator crewmen and evaluating the most current intelligence reports, the men from V Bomber Command and FEAF apparently decided that the day’s events were some sort of aberration, for they made what appears in hindsight to be an otherwise inexplicable decision. Despite what had happened to the four B-32s that day, they decided to dispatch four more Dominators to the Japanese capital the following morning.

  It was a decision that would cost one young American his life.

  CHAPTER 5

  A DESPERATE FIGHT

  THE JAPANESE FIGHTER ATTACKS on the Dominators involved in the August 17 mission over the Kanto region confused and angered the American airmen involved in the engagement. More important, the interception of the B-32s caused senior leaders all the way up the chain of command to wonder whether the incident had been an anomaly or was the first sign of Japan’s decision to repudiate its earlier agreement to accept the Allied surrender terms and continue the war. But there was another, more immediate concern: the appearance of the Japanese fighters had forced the Dominators to cut short their aerial-photography mission, and news of the attack had also caused the cancellation or recall of other reconnaissance flights over the Home Islands. Far East Air Forces headquarters still needed to know which Japanese airfields would be usable for incoming occupation troops, and that meant that reconnaissance aircraft would have to be sent back over the Tokyo area despite what had happened on August 17.

  In determining that additional recon flights were necessary, the V Bomber Command planners decided on a multi-mission approach. Aircraft from several units were tapped to participate, including both the 20th Recon and the 386th Bomb Squadron. The former would send four F-7s over the area south and west of Tokyo, while the latter would put a quartet of Dominators back over the northeastern part of the Kanto region. Though the decision to send B-32s back to Japan on August 18 may have been a nod to the Dominator’s demonstrated abilities as a stable, long-range camera platform that could defend itself when necessary, it was more probably simple expedience—the big bombers were available, capable of performing the task, and not otherwise engaged.

  The mission plan called for four B-32s to photograph airdromes to the east of greater Tokyo in Chiba Prefecture. These included several of the bases that had not been imaged on the seventeenth because of the fighter attacks, as well as the army fields at Shimoshizu and Kioroshi. The B-32s were to cover a roughly rectangular area of some 600 square miles, flying two miles apart and “mowing the lawn” at 20,000 feet. Planners estimated that it would take the four machines approximately an hour to complete their photo runs—if they were allowed to complete their task without interference.

  Whether the Japanese would oppose the B-32s and other reconnaissance aircraft was, at that point, an open question. As strange as it may seem, there had apparently been no attempt to communicate with the Japanese regarding the August 17 attack, either by FEAF or by MacArthur’s staff in Manila. It is always possible that someone of exalted rank in one of the headquarters determined that the fighter interception of the B-32s had been the last hurrah of a small band of renegade pilots who had since been brought to heel, but even the idea that no one attempted to verify that assumption before sending multiple crews back into what was quite possibly still the lion’s den is staggering.

  Moreover, the number of Japanese interceptors involved in the initial incident and the fact that the nature of the photo-recon mission had required the individual B-32s to fly too far apart to offer each other mutually supportive defensive fire logically argued that the August 18 missions by the Dominators and the 20th Recon aircraft should be escorted by friendly fighters. Iwo Jima–based P-51 Mustangs and P-47 Thunderbolts of VII Fighter Command were more than capable of rendezvousing with the B-32s and F-7s off the Japanese coast and accompanying them during the photo runs, yet no such escort was to be provided.

  It is certainly possible, of course, that the mission planners decided not to send fighters along on August 18 out of an excess of caution. They may have reasoned that the arrival over Japan of large, four-engine bomber-type aircraft escorted by fighters would be more likely to prompt an aggressive response than it would be to prevent one. After all, during the previous two years the appearance over Tokyo of exactly the same sort of mixed bomber-fighter formations had heralded imminent death and destruction, and the Japanese might well assume that in retaliation for the August 17 attacks on the B-32s the Americans had decided to renew at least limited air assaults on the Home Islands.

  However, it is far more likely that the decision not to send escorting fighters was based on a simple, operational reality: the August 15 theater-wide ceasefire had resulted in a dramatic and fairly rapid scaling back of the elaborate air-sea rescue network that had been put in place to assist airmen who, for whatever reason, were not able to make it back to their island bases after missions to Japan. By the last few months of hostilities that network had grown to include American and Allied submarines and surface vessels stationed off the Japanese coast as “plane guards” specifically to locate and retrieve downed pilots, as well as scores of “Dumbo” aircraft—B-17s, B-24s, and B-29s equipped with air-droppable life rafts—that orbited at set locations in order to respond quickly to a ditching or bailout. The rescue effort had been particularly valuable to the aircrews of VII Fighter Command, whose single-engine aircraft had been far less likely than larger machines to make it home safely if they encountered mechanical difficulties or had been significantly damaged by enemy action. The planners’ decision not to dispatch fighter escorts as part of the August 18 recon missions may therefore have been forced on them. Given the rapid scaleback of offensive air operations following the ceasefire, it would have been nearly impossible to put adequate air-sea rescue coverage together in time to support the August 18 effort, particularly because many of the rescue units were preparing their aircraft to support the coming occupation airlift. Without that coverage to and from Japan the “Little Friends,” as the Mustangs and Thunderbolts were called by those they escorted, would have been facing extraordinary and unacceptable risks even if they never encountered Japanese fighters.

  There is another, though admittedly very improbable, explanation for why the reconnaissance aircraft tapped to overfly the Tokyo area on August 18 did so without a protective fighter escort. As the 386th Bomb Squadron’s assistant intelligence officer, Rudy Pugliese, later postulated, the mission planners might intentionally have sent the B-32s and F-7s to the Kanto region on their own as a test of Japan’s willingness to actually follow through with the surrender. While the overflights certainly had a legitimate and vital intelligence purpose—in that the photos the aircraft would bring back would be used i
n planning the arrival of Allied occupation forces—Pugliese believed the missions might also have been a rather cold-blooded “fidelity test.” According to this scenario, if the Japanese did not interfere in any way with the unescorted American aircraft it would be a solid indication that the August 17 attacks had been the last desperate act of diehards rather than an indication that Japan’s announced resolve to surrender was ebbing away.1 Although this seems a highly unlikely possibility—not many senior officers would risk eight very expensive airplanes and their crews just to determine if the Japanese were still willing to shoot at them—stranger things have certainly happened in wartime.

  Whatever the reason, the B-32s and F-7s that would be dispatched from Yontan to the Kanto region on August 18 would make the trip unescorted, with only their gunners to defend them should the Japanese prove “unfaithful.”

  BY THE MORNING OF August 18 the 386th Bomb Squadron officially “owned” seven B-32s, with two more still making the long trans-Pacific journey from the United States. Of those aircraft actually present at Yontan, three were grounded either by significant mechanical problems or because they were being used as a source of spare parts to keep the other aircraft flying. The four operational machines—Hobo Queen II, Harriet’s Chariot, 544, and 578—therefore became the mission aircraft by default, despite the fact that each had a variety of what were considered to be “minor mechanical defects.” Several of these “minor” issues were to become the source of major problems in the hours to come.

  The one resource the 386th had in abundance was B-32 crewmen. The Dominator conversion courses the squadron had been running since before the move to Okinawa had produced more than enough qualified pilots, copilots, and flight engineers, and most of the other crew positions—navigator, radio and countermeasures operators, and gunners—could be readily filled by men who in most cases had racked up many hours of B-24 time before transitioning to B-32s. Because the Army Air Forces still planned to convert the entire 312th Bomb Group from A-20s to Dominators, between the end of the combat test and the beginning of operations from Yontan Tony Svore and Selmon Wells had instituted a crew-pool system intended to give as many men as possible some operational time in the B-32s. For most missions personnel were therefore drawn from daily rosters as needed, meaning that on any given flight the pilot and copilot might never have flown together, and the gunners and other crewmembers might have only a passing acquaintance. There were exceptions to this policy, however, in that at least three combat-experienced B-24 crews transitioned intact into B-32s.2