Chapter V
Early Operations

The base on nearby Tinian Island had hardly been started. Those on Guam, where the main Headquarters of the XXI Bomber Command was to be located, had not even been laid out. Communications were completely inadequate. The aircraft of the 73d Wing arrived rapidly on Saipan after mid-October 1944 and had to be double parked on hardstands. In the meantime, a shipload of supplies reached Guam, destined to become a depot. The ship had been carefully loaded so the supplies could be unloaded in reverse sequence and stacked at the depot in "combat-loaded" order. The procedure was a new and elaborate one that would give us an operating depot in a matter of weeks. Actually, fighting was still going on in Guam when the depot ship arrived, and confusion reigned supreme. The harbor master said, "I'll give you twenty-four hours to get that goddamned ship out of here."

Before I learned what was happening, the supplies were dumped in the jungle. They were never recovered. We had to provide aircraft supplies for the B-29s (themselves new and unfamiliar) by air from Sacramento, California -- 8,000 miles away! The in-commission rate of the B-29s was astonishingly high -- considering the circumstances.

As indicated earlier, the strategic concept was for the defeat or neutralization of the Japanese air forces as an "overriding intermediate objective." Thereafter, the major strategic air offensive was to be


BRIG. GEN. HAYWOOD S. HANSELL, Commanding General, XXI Bomber Command, briefs pilots before a mission over Tokyo.


launched against the war-supporting and economic systems of Japan. These systems were the "primary objectives." The plan of operations against them contemplated the destruction of certain major industrial facilities by selective target bombing.

The overriding intermediate objective and its associated target system, assigned to XXI Bomber Command by agreement of the Joint Chiefs of Staff, gave first place to Japanese factories building airplanes and aircraft engines. This assignment was not lightly conceived, for it had been learned in Europe that air superiority is essential to strategic air operations as well as to surface operations and invasions. The Joint Chiefs had been persuaded to back the air offensive, but they were looking over the shoulders of the airmen at the invasion shore.

The primary targets, to be destroyed after the Japanese aircraft industry, were of two kinds: selected targets, to be destroyed by precision bombing; and urban targets, to be destroyed by incendiary attack. Precision bombing of selected targets was the preferred method. But it was believed that small, "home-shop" type production facilities were distributed in the great cities of Japan. Those cities were known to be highly flammable. Incendiary attack of urban areas was listed high among approved target lists because of these urban shops and also because the cities were the focal points of Japanese "will to resist." The targets initially assigned the XXI Bomber Command were aircraft and engine factories supporting the intermediate objective, and they had overriding priority. They were selected targets requiring precision, daylight attack. Japanese shipping was also high on the target priority list.

It must be remembered that the Twentieth Air Force had won its right to exist only by becoming a creature of the Joint Chiefs of Staff. Strategic air leaders believed the war could be won by air power, but the official war plans of the Joint Chiefs contemplated invasion, and the Twentieth could not divorce itself entirely from that ultimate concept. Certainly that was wise in the early stages. Air power alone had never before been sufficient to force capitulation of a major nation still in full control of its own military means. What if the strategic air offensive should not be effective? The Chiefs simply had to have a backup plan. To be sure, there was some skepticism of air power, but


even if there had not been, it would have been unwise not to furnish a backup. Actually, the Joint Chiefs of Staff did give the Twentieth Air Force a priority second-to-none in the creation and launching of the air offensive, and they did direct the capture of the Marianas as a base of operations for the XXI Bomber Command.

The pattern of B-29 operations against Japanese targets was not conditioned by the limited concept of airpower's role, as a preparatory bombing operation preliminary to the basic strategy of defeating Japan by surface invasion. The first target list had as its purpose the defeat of the Japanese Air Force. This, like the defeat of the German Air Force, was an intermediate objective, a necessary preliminary to ensure and enhance the effectiveness of strategic bombing operations. No doubt that goal also helped assure the success of future ground and sea operations. But the initial primary air aims were practically the same as those in Germany -- the paralysis of the military, economic, industrial, and social structure supporting the will and the ability of the Japanese nation to wage war.

First Strikes

Plans for the maiden bombing of Japan from the Marianas called for a combined first strike with the Navy, so carrier-based aircraft would divert and absorb some of the Japanese fighter defenses. For the rest, the B-29s would have to rely upon high altitude and speed (their chief advantage) and their own defensive gunfire. The B-29 was designed as a high-altitude bomber, the first with pressurized crew compartments. It had turbosupercharged engines, was reasonably fast at high altitudes, and was heavily gunned. By operating in formation, it was expected to fend for itself against enemy fighters which would be operating at their ceiling and have little, if any, margin of performance superiority.

The first aircraft and crews to touch down on Saipan were given a little training in the Pacific area. Six short training missions were flown against Truk and Iwo Jima. In spite of all the obstacles, the XXI Bomber Command declared itself ready to meet combat commitments exactly on time -- by the middle of November.


In early morning of November 1, 1944, an F-13A (a photoreconnaissance version of the B-29) took off from Saipan to become the first U.S. plane over Tokyo since the Doolittle raid in April 1942. The crewmembers, led by Capt. Ralph D. Steakley, insisted upon an immediate mission, even though they had just arrived from the United States. I advised a rest but they were persistent. Thank God they were. They found clear skies over Japan -- a phenomenon. Called "Tokyo Rose," the aircraft flew above the Japanese capital at 32,000 feet, photographing a complex of aircraft and engine plants just west of Tokyo and another on the outskirts of Nagoya. They shot over 7,000 excellent photographs. Before the first strike on Tokyo on November 24, 17 sorties had been flown over Japan by F-13s. Many of the missions were hampered by bad weather, but sufficient information on the location of aircraft factories was secured for the first bombing missions. Copies of the photographs were sent to General Arnold for the Joint Chiefs of Staff and to Admirals Nimitz and Halsey. Mosaics were made, strips laid out, and initial points and target approaches selected. Every crew was required to trace its photo map, mark landmarks and target runs, and then redraw them from memory -- over and over.

As the day for the combined operation against Japan approached, the Navy found itself in serious combat trouble in its involvement with the Japanese fleet and movement into the Philippines. It therefore announced it could not take part in the planned combined air operation against Japan. The Navy recommended postponing the joint strike against Japan and grounding the B-29s until it could participate. I got word of this recommendation and notified General Arnold we would be ready to conduct the air attack in November as planned, without Navy support. I did not like the idea that the B-29s could operate only with Navy assistance. If that were the case, the B-29s might better be turned over to Admiral Nimitz. However, I did have our limited number of B-29s loaded up with 2,000-pound bombs and put on alert to support the Navy in the battle of the Philippine Sea, which was then raging. I told Admiral Hoover we stood ready to offer such assistance as he might request, and we would welcome Navy officers to go with us as recognition experts. I did this for two reasons:


We were still too weak to attack Japan but genuinely wished to help. And I was worried lest Admiral Nimitz invoke the emergency clause and take command of the B-29s. He might take a long time to release command. By volunteering our services I hoped to forestall his official action. We were not called upon, but I think the gesture was appreciated and was effective.

The first planned strikes were labeled San Antonio I and II. I was to lead the first, and Brig. Gen. Emmett "Rosey" O'Donnell, Jr., 73d Wing Commander, was to lead the second. These plans were reported to Washington in detail. To my surprise, General Arnold ordered me not to lead the mission. I presumed it was because of my extensive knowledge of the Pacific campaign plans. It was concern for just this possibility that had prompted me to make a special request about a month or so earlier. Just before leaving for the Pacific, I had been summoned to Washington to be briefed on a highly secret matter. I presume it was either on the atomic bomb or Ultra. I asked if the knowledge I was about to receive would keep me from flying combat missions. The answer was, "Yes." I asked to be excused from receiving the information, because I felt it imperative that I be free to lead my command if the going got rough. This request was honored and I proceeded without the briefing.

Now, it seemed to me, was the time when I should lead. The going was likely to be rough, and there was deep concern in the command about the chances of successfully performing the mission. I decided to ignore instructions, lead the mission, and hope for the best. My hope was short lived. Two or three days after advising Washington of our plans, I was called upon by a Navy lieutenant and a petty officer who had a copy of the message and demanded a written acknowledgment of its receipt. The message had been sent through Navy channels.

At the time I thought the decision was arbitrary and ill taken. Of course, I did know of the strategic plans for the Pacific war, but plans are constantly changing, and I had only a rudimentary knowledge of the atomic bomb. But there was another factor I overlooked at the time and did not think of until much later: I was thought to be privy to the existence of supersecret intelligence and, what was much more important, to the sensitive source of such information. The story is out


now after thirty years of secrecy. It is told in fascinating detail in The Ultra Secret by the man responsible for it, Gp. Capt. Frederick W. Winterbotham, RAF. It is the story of the breaking of the most secret codes of the Germans and the Japanese. I was aware of the breaking of the Japanese code but I was actually unaware of Ultra.

How I could have forgotten this item is incredible. It had been the source of daily agony as far as I was concerned. For several months before I took over the XXI Bomber Command, General Arnold required Generals White and Kuter and me to meet in his inner office every morning. We were admitted at 0730. At 0715, General Arnold received the daily verbatim translations of high-level, secret German and Japanese messages. Before we entered, we were each familiar with the U.S. secret messages exchanged during the night. We were therefore prepared to present and discuss events pertaining to our several interests and responsibilities. However, we did not get Ultra messages directly. General Arnold would reveal the information he had just been given and demand to know what we were doing about it. We were not doing anything about it; we did not know about it. Even so, this did not save us from withering comments about our competence. Doubtless, General Arnold enjoyed this game, but it was pretty rough business to be on the receiving end.

Now that same supersecret intelligence I did not receive was returning to bite me again. There was an Allied agreement, without deviation of any sort, forbidding recipients of Ultra information and those who knew of its source from exposing themselves to capture. This may well have been the clinching argument in my case. The message from General Arnold could no longer be ignored. I designated General O'Donnell to lead San Antonio I.

I faced a very serious command dilemma. As Commanding General, XXI Bomber Command, I found myself in a severe predicament. Three nearly simultaneous events combined to make my position difficult. First was the Navy recommendation that the mission be canceled or indefinitely postponed until the Navy was ready to participate. If I accepted this, it would clearly show that the XXI Bomber Command could not operate independently but must do so solely in close concert with the Navy. If this were true, why have a


STRATEGIC BOMBING MISSIONS OVER JAPAN are reviewed on a B-29 base in the Marianas by (left to right) Brig. Gen. Emmett O'Donnell, Jr., Lt. Gen. Millard F. Harmon, and Brig. Gen. Haywood S. Hansell, Jr.

WRECKAGE of JAPANESE DIVE BOMBERS lines the runways on Tinian in the Marianas, March 1945.


separate Twentieth Air Force chain of command? Since coordination would be needed, and it could be furnished only by the Commander in Chief of the Pacific Ocean Area, Admiral Nimitz, why not place the XXI Bomber Command under his control? This would almost certainly destroy the strategic air war against Japan as a war-winning grand strategy -- one in which I and my fellow airmen fervently believed. I had hastened to notify General Arnold that the XXI Bomber Command stood ready to discharge the mission without Navy assistance. Any equivocation now would place the Twentieth Air Force and the strategic air war against Japan in serious jeopardy.

Second, I received a disturbing message from General Arnold. Members of General Arnold's staff and at least one top-level AAF field commander, to whom my plan (San Antonio I) had been passed for comment, voiced grave doubts that we could carry out our mission. Gen. George Kenney predicted the planned operation would result in disaster. General Arnold forwarded these expressions of doubt and wrote in his own hand the comment that he was inclined to agree with the skeptics. The contention was that the airplanes lacked the necessary range and furthermore the Japanese would "shoot them out of the air." General Arnold did not direct me to abandon or modify the mission. Rather, he put me on record as having been warned. He concluded with the statement that he had high respect for the critics of the mission, but said if I were convinced of its feasibility and were determined to carry on, then I was at liberty to do so. He left the decision up to me and said if I chose to go ahead, he wished me luck. The effect was chilling. The warning was coming from the very people from whom I had expected firmest support.

Third, I received a handwritten letter from my senior wing commander, General O'Donnell, stating that he, too doubted if his unit could fulfill the mission. He suggested abandoning the daylight attack and substituting a night one against some area target. In a private meeting with him, I pointed out that, the night operation he proposed would not accomplish the mission with which I was charged. I was determined to see the mission through, and if he was unwilling to lead his wing in this operation, I would turn it over to someone who was. (Brig. Gen. Roger M. Ramey, my Deputy, was anxious to head


it.) O'Donnell assured me he was willing but had felt constrained to vent his doubts and worries. I recognized his right and obligation to give me, in private, his honest opinion. If the mission failed and he had not warned me of his true convictions, I would have been justified in reproaching him. On the strength of this, with his agreement I destroyed his letter. I explained that if the mission succeeded, the letter would be a black mark on his record that would be hard to live down. If we failed, the onus would be entirely on my head, since I had been warned from other quarters.

It was quite true that until the time for takeoff of San Antonio I, the XXI Bomber Command had never flown a formation as large as a squadron, a distance as far as Tokyo and back, and had not flown against any enemy opposition. But the potential impact of the mission on Pacific strategy and the future of the Air Force extended far beyond the XXI. The Army Air Forces, at the Joint Chiefs of Staff planning and command level, had been advocating primary reliance upon the decisive effectiveness of the air offensive, with provision for an invasion of the Japanese mainland only if the air offensive proved inconclusive. This viewpoint did not mesh with Army and Navy planning. To admit at this late juncture in November 1944 that the air offensive could not even attack its intermediate objectives, would have grave repercussions indeed. The whole command structure of the Twentieth Air Force as a worldwide command, reporting directly to the Joint Chiefs in a role parallel to that of the U.S. naval fleet, was in delicate balance. To subject it to reexamination resulting from a major degradation of capability would have had serious after-effects. To those who believed the air offensive was not only the most effective avenue to victory in the Pacific but also the cheapest in terms of American lives, abandoning the planned mission would be a disaster almost as great as the tactical disaster of failure might have been. Still, there was no denying that the decision to carry out the plan was extremely risky.

I thought I understood why General Arnold had sent me this message. Disaster on the first mission of the XXI Bomber Command would have altered Pacific strategy and would have delayed recognition of coordinate air power by many years. Since it seemed highly


probable such a disaster would ensue, the ill effects would be less severe on the future of the Army Air Forces if the responsibility were borne by a subordinate field commander. Arnold had warned me, and I had chosen to go ahead in spite of the warning. His was not an unreasonable precaution to take under the circumstances. I decided to go on with the mission and so notified him.

Rosey O'Donnell asked for a change in the operational plan. I had set up two initial points, on opposite sides of the target, and planned for two converging bombing runs to confuse and divide the enemy air defenses. It called for a complicated maneuver. Rosey suggested the plan be simplified by using only one axis of attack, with Fujiyama volcano, west of Tokyo, as the initial point. I approved the change.

During preparation for the first strike, a delegation of congressmen visited my command. I was extremely busy. We seemed to be operating under one of Murphy's laws, "Anything that can go wrong will go wrong." I quartered the visiting congressmen in my own pyramidal tent, perched on a promontory jutting out into the sea. I issued each of them a mess kit and invited them to join me in the chow line and to wash their own kits afterward as I was doing. I must admit the chow was simply awful, but we were living on a shoestring and spending all our energies on training and preparation for the mission. I devoted as much time to the visitors as I felt able and told them of our general plans, problems, and expectations. I did not go into the details of operations and tactics. We kept those pretty close to the vest because security leaks would endanger our success and cost lives among our combat crews.

The group went back to Washington and wrote a scathing denunciation of my administrative arrangements. They also had interviewed members of my command who were from their home districts. Most of the responses were understandably flavored with apprehension about the forthcoming campaign. This problem was a forerunner of the massive difficulties that were to fall upon future American commanders in the field during the Korean and Vietnam Wars.

The first mission was laid on for November 17. At the final briefing before dawn, I made a short talk to the crews:


Stick together. Don't let fighter attacks break up the formations. And put the bombs on the target. If the bombs don't hit the target all our efforts, risks, worries, and work will be for nothing. That's what we're here for. If we do our job, this is the beginning of the end for Japan. Put the bombs on the target. You can do it.

Crews took their stations in the early dawn. The long line of B-29s formed up on the taxiway that led to one end of the runway. To extend our range, we were carrying an extra tank of gasoline in the front bomb bay [gross take off weight was about 140,000 pounds]. At the last moment, the wind, which had consistently blown down the runway, died down. We needed that wind badly to get off with our heavy loads. Then the wind reversed direction and freshened. It was impossible to taxi to the other end of the runway because the long line of B-29s could not be reversed. We could not use the runway itself to taxi down and then take off by successive airplanes, because the taxi time would have been excessive, causing long delays in assembly, and burning up too much of our precious fuel.

There was no choice. I had to call off the mission. I hated to do so. We had built up to a psychological climax. Delay would play on fears and apprehension. It was one of the hardest decisions I had to make, but it was one of the luckiest. In a few hours, a typhoon hit Saipan and lasted six days. Then it traveled north toward Japan on our route to our targets. It left our base a shambles and a sea of mud. If the typhoon had arrived a few hours later, we would already have taken off -- and found no way to make a landing on return. The B-29s sat on their hardstands, two deep, fully loaded, for a week. The orders had been distributed. The thought of a security leak was a nightmare. A daily weather flight followed the typhoon northward. To my great grief, one of them was lost and never heard from again.

On November 24, 111 B-29s of the 73d Bombardment Wing, XXI Bomber Command, took off on the trip toward Japan. They represented over 90 percent of the B-29s on Saipan. Some of the crews had arrived less than a week before, and their first takeoff was for Tokyo. Each lift-off was an ordeal. As noted earlier, the B-29 was originally designed for a gross weight of 120,000 pounds. By urging and pleading, we had convinced the engineers at Wright Field to raise the


ORDNANCE SPECIALISTS LOAD BOMBS on a Boeing Superfortress on Saipan, Marianas, November 1944.

BOMBER CREWS LISTEN ATTENTIVELY to the pre-mission briefing for the XXI Bomber Command's first mission to Tokyo, November 1944.


A TWENTIETH AIR FORCE B-29 SUPERFORTRESS runs up its engines before taxiing out for take-off on a mission over Tokyo, December 1944.

GROUND PERSONNEL WATCH A BOEING B-29 take off on the first bombing mission from Saipan to Tokyo, November 1944.


allowable gross takeoff weight of the B-29s to 132,000 pounds. Now, to carry every gallon of gas that could be pumped aboard, they were taking off at 140,000 pounds! A faltering engine would spell the end for any aircraft.

Primary target for the B-29s on San Antonio I was the Musashino aircraft plant of the Nakajima Aircraft Company on the outskirts of Tokyo. The secondary targets and "last resort" areas were the docking facilities and urban areas of Tokyo. A total of 277.5 tons of bombs was delivered. Seventeen bombers turned back because of fuel problems, and 6 missed their bombing runs due to mechanical troubles. Flying between 27,000 and 33,000 feet, the bombers picked up a 120-knot wind over Japan, giving them a ground speed of 445 miles per hour. Twenty-four planes bombed the Nakajima plant on the outskirts of Tokyo, and 64 unloaded on the Tokyo dock areas. Only 1 B-29 was lost in combat. U.S. gunners claimed 7 enemy fighters destroyed and 18 probables. Final count for the XXI Bomber Command listed 2 B-29s destroyed, 8 damaged by enemy action, 1 man killed, 1 missing, and 4 injured. After the war, records indicated that 48 bombs had hit the factory area: 1 percent of the buildings and 2.4 percent of the machinery were damaged; 57 persons were killed and 75 injured.

The weather at the target had been far from favorable, and the bombing results left much to be desired. But losses were small, and the operation was completed despite the hazards and obstacles. Not the least of the hazards was the return flight to base. The mission lasted twelve to fourteen hours. Landing was at night with no runway lights and only smudge pots along the single runway strip. The next nearest landing strip was at Kwajalein, over a thousand miles away. If a B-29 splattered itself on the runway, the rest of the aircraft behind it, with nowhere to land, would be all through.

Perhaps in hindsight, the decision to launch the offensive in the face of such adverse conditions and recommendations seems reckless, and the results the product more of good luck than sound judgment. But this first great gamble proved the feasibility of the assault. Momentum, confidence, and improved efficiency would come with experience and numbers. In my opinion, if the decision had been to "stand down" San Antonio I and substitute a night attack against


some urban areas, the result would have been catastrophic, particularly as regards confidence in and continuation of the Twentieth Air Force.

San Antonio II was staged on November 27 with the same target objectives. The crews of the eighty-one B-29's that flew the mission found Tokyo completely covered by clouds, so the bombs were dropped by radar on the secondary targets. The Japanese were provoked into trying to halt the bombing. From Iwo Jima they made one-way suicide air raids on Isley Field, our base on Saipan, destroying some B-29s. The Japanese had realized that their home islands were indeed susceptible to sustained attack and that their fighters could not turn back the B-29s.

Improving Successive Missions

The accomplishment of these first two missions, with very light losses, was an achievement in itself. The fact that bombing results were only fair could be overlooked in view of the proof that the force could indeed reach its targets and return to base. But succeeding missions made it clear that bombing accuracy would have to be greatly improved. Two methods were adopted to achieve that end -- training in visual bombing techniques and the introduction of radar to assist and even substitute for optical bombing in attacks against selected precision targets.

A "lead crew" school was set up, and one B-29 from each squadron was set aside for training. This was rather a drastic move. We were desperately short of B-29s, and withdrawing at least twelve B-29s for training out of about one hundred aircraft in operational condition was critical weakening of the force. But there was no use sending bombers to Japan at all if they could not destroy the targets.

The other method to improve bombing simply made good sense. The AN/APQ-13 radar bombing equipment was used to supplement the bombing run with the Norden optical sight. Initial points were chosen that afforded good radar images. As each formation approached the initial point, the turn to the bombing run was made from observations by the radar-bombardier. He next conducted a radar


bombing run, using the target if possible, or an offset aiming point if necessary. The radar-bombardier could then assist the optical-bombardier in setting up his bomb run. In this manner, it was possible to set into the optical sight the drift angle and rate of ground speed and to have the crosshairs of the optical sight aligned approximately on the target. Then if the target became visible through breaks in the clouds, only minor adjustments of the optical sight were needed. Although the bombs could be released on the radar sight if the clouds obscured the target from visual adjustment, the accuracy was inferior to visual sighting. For acceptably accurate radar sighting, we would have to prepare radar maps of the targets and determine precise locations of targets with reference to good offset aiming points.

Several events with a sobering impact on me occurred during the first ten days of operations. The first of these took place when Japanese fighters commenced strafing attacks on our Saipan air base and parked B-29s. They had come down from Iwo Jima on a one-way suicide mission and were completely undetected by the Marine outfit that was supposed to furnish our air defense. One such attack began around noon on November 27. I jumped in my jeep with Col. John B. Montgomery, my Chief of Staff, and headed for the field. As we came up a rise onto the flying field, I found myself looking straight into a Japanese fighter that was strafing the area. I brought the jeep to a halt and sought shelter under it. Quick as my reaction had been, it still was not quick enough. Colonel Montgomery was already there.

Fortunately, the wing was out on a mission over Japan and there were only a few B-29s on the ground. But several of these B-29s were bombers which had aborted the mission and were on the hardstands fully loaded with bombs and gasoline. The main warning radar, which the Marines should have installed on a hill, was still in crates. There had been ample time for installation, but the Marines simply did not expect suicide attacks. General Arnold was understandably outraged. He had warned of this possibility but suitable action was not taken to provide defenses.

Another incident had to do with Operation Memphis One, which was returning to Saipan when an intense tropical storm hit the island. It was about 2030 and the rain was so heavy that, standing in the


improvised tower, I could not see the dim smudge pots outlining the single runway. Over eighty B-29s were approaching the field. The air was full of calls saying, "number such-and-such B-29, I am approaching Saipan. Visibility zero. I am out of gasoline. Request instructions." The tower operator was a noncommissioned officer who earned my highest admiration. He was calm and issued instructions without a hint of panic. The only thing we could do was hope the storm, which had arrived suddenly, would depart with equal dispatch, and that the fuel gauges which read zero did not really mean it. We were fortunate. The storm did clear. The B-29s landed in rapid succession. I realized then that I really was quite helpless. The real Commander of the XXI Bomber Command was a noncommissioned officer who was functioning superbly as the tower operator. The best help I could give him was to keep out of his way and avoid interfering with him.

In another incident one evening, an air raid warning sounded and all lights were doused. A Japanese twin-engined bomber made a couple of passes. Roger Ramey, my Deputy Commander, and I were standing on the runway watching the antiaircraft bursts when the bomber suddenly reappeared at very low altitude. A couple of B-29s had been hit and were burning brightly. They lit up the sky, and the oncoming Japanese aircraft was clearly visible. I think we noted something about it at just the same instant, because we took immediate and identical action. It was making a low-level strafing attack down the runway we were standing on. There was no place to go. We hit the pavement with great force at just the same time. Tracers from the ground defenses were pouring into the Japanese bomber but it continued on course. Then, as it approached the end of the runway, it swerved slightly and plowed into the ground. The pilot evidently had been killed. The bomber hit with a roar about a hundred yards from us and was engulfed in flames. Just as we were rising to our feet, there was a violent explosion. Evidently it still had bombs aboard.

That evening was replete with mental impressions as well. As soon as the attack started, the base was aswarm with thousands of men eager and determined to see what was going on. They were completely beyond control. Two B-29s were burning like torches, and there were other B-29s nearby. Men pushed and pulled to get the other sixty-ton


monsters away from the fire. The B-29s were fully loaded with gasoline. Some of them had bombs aboard in preparation for tomorrow's mission. No one seemed to know which of them had bombs and which had not. Succor came from an unbidden and unexpected source: the engineers. They appeared on the scene with their massive bulldozers and earth-moving equipment. They pushed the flaming carcasses aside, piled dirt on them, and rode over them until they had crushed out the fires. It was the most amazing sight I have ever seen. No one knew for sure that there were not bombs in those flaming masses. Fifty-caliber ammunition was going off like firecrackers. The scene was an animated illustration out of Dante's Inferno. Engineers were riding in bulldozers through flames reaching high in the air; enormous monsters of steel were burying other monsters which threatened to lash back and blow up at any moment. I still do not know whose idea it was for the engineers to take this action. It certainly was not mine. I really think it was a spontaneous reaction from the engineers themselves. I have always regretted that I did not get them a unit combat citation. It certainly was heroism far beyond the call of duty.

The next three months (November 1944 through January 1945) were frustrating, to say the least. Schools worked hard to train the lead crews, determined to improve bombing accuracy. Enormous efforts were made to upgrade maintenance. The depot had to start all over again, and in the meantime the air supply from Sacramento had to be improved. More missions were run against Japanese engine and aircraft factories. But the weather was a terrible opponent, and there was no intelligence of its movements. Japanese fighter opposition was desperate but not deadly, at least in comparison with German fighters. Air kamikaze-ramming tactics were tried with some success. Morale was a critical problem. The airplane engines were still unreliable. Aircraft disabled from combat or other causes were 1,500 miles from friendly territory. The crews had the choice of drowning or bailing out over Japan, to be executed by maddened Japanese. The U.S. Navy contributed tremendously to morale by stationing rescue submarines at intervals along the route. Their performance was superb. One submarine entered Tokyo Bay in daylight and picked up a B-29 crew


A B-29 FLIES OVER THE HARBOR OF SAIPAN -- headquarters of the XXI Bomber Command.

WITH NO. 4 ENGINE FEATHERED, a Superfortress lands on Saipan after a successful mission against Tokyo.


A REINFORCED CONCRETE STRUCTURE at the Musashino aircraft engine plant escaped the brunt of the AAF bombing attacks, which severely damaged the weaker buildings.

BRIG. GEN. HANSELL visits with Col. Seth S. Terry of XXI Bomber Command.


DESPITE THE THREAT OF EXPLOSION, aviation engineers ram their bulldozer into a burning B-29, pushing dirt over the flames. The near-disaster was caused by a low-level Japanese attack on the Saipan airfield, November 1944.

A BULLDOZER PULLS THE TAIL OF A SUPERFORTRESS away from the burning wreckage following the Japanese raid on Saipan, November 1944.


right under the guns of the Japanese. From November 1944 to August 1945, 600 Twentieth Air Force flyers were saved in open-sea rescues. The system involved Navy flying boats, B-17s (Dumbos) carrying droppable lifeboats, B-29s (Superdumbos), and submarines, all under Navy control. The system was largely the achievement of my Navy Liaison Officer, Comdr. George C. McGhee.

On December 13, 1944, 74 B-29s of the 73d Wing received credit for doing significant damage to Japanese aircraft plants. Most of the bombers carried 500-pound general purpose bombs, while others were loaded with incendiary clusters. The primary target was the Mitsubishi engine plant at Nagoya. Photographs failed to show all the damage. Later reports disclosed that engine assembly shops and auxiliary buildings were destroyed or damaged. A total of 246 people were killed and 105 injured. Aircraft engine production capacity was reduced from 1,600 to 1,200 per month. The Mitsubishi No. 4 Engine Works no longer made parts. The Japanese also began transferring plant equipment to underground facilities. It was the most destructive mission to date for XXI Bomber Command.

The order for succeeding missions was for maximum strikes against top-priority targets by high-altitude precision bombing when weather was acceptable. When it was not, secondary targets were to be hit and time was also given to single-aircraft night operations, collecting weather data, and radar bombing. Bombing results remained hard to assess due to cloud cover, and were deliberately played down by XXI Bomber Command Headquarters. I wanted to build a reputation for credibility in the XXI's reports to counterbalance the known tendency to exaggerate. Our whole energy was devoted to improved effectiveness and accuracy.

During November and December 1944, forty-nine photo missions were flown and, by January 1945, thousands of photographs of Japanese targets had been taken. These missions performed five functions for the XXI Bomber Command: photo reconnaissance for target and strategic intelligence, weather reconnaissance, radar scope photography, lead-crew training, and nuisance raids.


The Aerial Mining Campaign

Preparation for aerial mining operations against shipping in Japanese home waters was likewise started during this early period. Later, Dr. Frederick M. Sallagar, while with the RAND Corporation (but a member of my 4th Operations Analysis Section in 1944), completed a postwar study, Lessons From an Aerial Mining Campaign (Operation STARVATION). Dr. Sallagar showed clearly that the aerial mining of Japanese inland waters by B-29s of the Twentieth Air Force's XXI Bomber Command was a tremendous success and contributed enormously to the fall of Japan. He further noted a reluctance on the part of senior Air Corps officers to enter into this operation with the enthusiasm it deserved. As Commanding General of the XXI Bomber Command, I was one of those reluctant leaders. But there were reasons for my reluctance.

Army Air Forces leaders, including me, remembered what had happened in Germany. AWPD-1 committed the American bomber units to the factories, industrial systems, and enemy air bases in Germany. But these units were split apart and half assigned to the Allied forces invading North Africa in Operation Torch. This afforded a bitter lesson: Theater commanders, accustomed to seeking victory through surface warfare, would demand and get strategic air to support their ground campaigns at the expense of strategic air objectives. The Eighth Air Force had hardly recovered from this nearly mortal blow, when the ordeal was again endured in preparation for and during the invasion of Normandy. Strategic targets in Germany were neglected far too long while air power was tied down in ground operations. Consequently, the appeal for aerial bombing of Japanese waterways found a cool reception in the first months of the XXI's operations. It looked like one more diversion to the local needs of a ground commander, and away from primary industrial targets leading to defeat of the enemy air force. So the mining program, which proved to be one of the principal achievements of the Twentieth Air Force, met with initial opposition.

Mining of rivers and harbors in the Netherlands East Indies by the B-29s of the XX Bomber Command had been one of the first


operations conducted by that command from bases in Ceylon. However, Admiral Nimitz's staff proposed a much more extensive campaign for XXI Bomber Command in Japanese home waters. In fact, the Navy's first proposal would have absorbed the total capacity of the XXI for the first three or four months of its operations. I objected to this on the ground that it was another major diversion from the chief purpose for which the command had been created and deployed. The objection was not directed against the idea of mining itself, but to the magnitude of the diversion at a time when utmost endeavor was needed to develop our primary capability.

I was not, of course, in a position to change the directive issued by the Joint Chiefs of Staff. But when General Arnold sought my recommendations on Admiral Nimitz's request, I opposed the application of so much of the XXI Bomber Command at that time to aerial mining. I recommended postponement until the force had grown and suggested that one group be charged with developing the technique and with limited initial operations. The problem was settled when General Arnold issued a directive calling for a somewhat postponed mining effort at a much reduced initial level.

Even as the problem was being discussed at high level, steps were taken to prepare for a mining campaign of some intensity. I directed the 313th Bombardment Wing, whose aircraft began to arrive on Tinian in December, to develop tactics and techniques for aerial mining. One group of the wing was designated to do this work. The XXI Bomber Command owed a debt of gratitude to the Navy personnel who assisted in adapting Navy mines to installation in B-29s and in helping develop dropping techniques and tactics.

When Maj. Gen. Curtis LeMay later took over the XXI Bomber Command,1 his decision to launch a massive mining operation was a sound one. Adequate forces had been assembled. Tactics and techniques had been worked out (many of them by LeMay). Since mining was not dependent on weather, it was possible to achieve continuity of operations. I think General LeMay did not view mining at night as


abandonment of selective targeting. Moreover, the night mining, like the night urban bombing, could be carried out regardless of cloud cover at the target. It is quite clear I could have endorsed mining as an aspect of strategic bombardment against the Japanese transportation system rather than as an auxiliary aspect of the sea blockade. I probably could have persuaded General Arnold to stretch my target priorities as prescribed by the Joint Chiefs to include aerial mining when sufficient aircraft became available. I doubt if I could have, or should have, devoted most of the command's air power to this purpose in the initial phase of strategic attack against Japan, when first priority was prescribed as destruction of the Japanese aircraft industry.

In retrospect, the actual evolution of events was probably about right. The XXI Bomber Command did attain the "overriding intermediate objective of undermining Japanese aircraft production." It did preserve its identity and structure as a separate command even though operating in an area under Navy jurisdiction. And it did retain unity of effort while subjected to constants pressure to become subordinate to other commanders and staff agencies. These achievements might have been jeopardized or at least delayed if the XXI Bomber Command had initially devoted its capacity to aerial mining, and the independence of the XXI might have been compromised or lost.

The aerial mining campaign as pursued by General LeMay succeeded beyond anyone's expectation. Fleet Admiral Nimitz said, "The planning, operational, and technical execution of the Twentieth Air Force aircraft mining on a scale never before attained had accomplished phenomenal results and is a credit to all concerned." And as Dr. Sallagar stated in his Lessons From an Aerial Mining Campaign:

The campaign was outstanding in many respects. More mines were laid in five months (over 12,000) than were dropped by all the other aircraft in the Pacific in more than two years (9,000). The "phenomenal results" mentioned by Admiral Nimitz included at least 70,000 (and possibly as much as 1,250,000) tons of Japanese shipping sunk or severely damaged. Perhaps more important, much of the surviving ship tonnage was bottled up in mined harbors for


prolonged periods while waiting for the mines to be cleared, which led to a virtual paralysis of Japan's essential maritime traffic.

We have gone somewhat ahead of the story. Yet, in view of the later undeniable success of the aerial mining campaign, I must explain my reluctance to begin an all-out campaign while I commanded the XXI Bomber Command.

Problems Faced by XXI Bomber Command

Our new equipment presented some technical problems affecting combat operations. I asked for three fixes which were attempted at Wright Field but did not materialize in time to be useful. They pertained to gunnery, weather penetration, and rescue at sea.

The gunnery equipment of the B-29 was new, formidable, and complex. Experience in operating without fighter escort in Europe had made defensive gunfire important. I had been instrumental in changing the top turret of the B-29 from a two-gun, .50-caliber pair to a set of four such guns, to meet the most dangerous of fighter attacks -- those from the front. All the guns were remotely operated from sights placed in transparent sighting blisters. A master gunner operated from a master-gunner's position in a top blister from which all the guns except the two tail guns could be controlled and fired. The gunsights and controls were ingenious and sophisticated, but highly complex. The sight was swiveled by the left-hand grip control, both laterally and in height. The range of the approaching fighter was automatically fed into the sight computer by a right-hand grip control by which the gunner sought to keep an illuminated ring in the sight adjusted to the wingspan of the approaching fighter. Each hand had to work simultaneously with, but independently of, the other. To complicate the problem further, the messages fed into the computer were accepted in terms of "rate." The "rate of turn" of the controls established the velocity, rate of turn, and rate of approach of the target. If a gunner failed to manipulate his sight smoothly and moved with a series of jerks, these rapid jerks told the computer that the velocity of the target was accelerating or decelerating wildly.

I asked that a device be designed to determine range and rate of


approach by radar and feed it directly into the sight computer. Such a device was available for the tail guns which had a somewhat limited field of fire. This would have relieved the gunner of at least half his burden. Actually, it would have provided far more relief than that. It certainly is far easier to do one thing at a time than to do two separate things simultaneously.

Another fix that I desired had to do with "stationkeeping" by radar. I worried about penetrating heavy weather fronts flying in formation. It was essential that the formations be able to reassemble easily after penetration of a weather front to furnish mutual protection. It was possible to use the APQ-13 bombing radar for this purpose to a limited degree, but I was not content about it. Finally, I asked for a floating transponder that could be tossed into the sea just as a damaged airplane "ditched." This would permit rescue aircraft to home on the transponder and quickly locate the position where the plane had gone down and, hopefully, where the crew would be rescued from their rubber rafts.

We had tactical problems as well as technical ones. In the early stages of operations, before we learned how to get the most from our engines, we were seriously constrained in terms of range. To save fuel, the first 1,000 miles or so were flown at low altitude. The climb to penetrating altitude was begun after the airplane had become lighter by the weight of the expended fuel.

General Arnold was understandably concerned about the large number of losses due to "ditchings" and failures to return for unknown reasons. I felt that losses must be expected in a highly risky wartime operation that gave promise of being decisive. Nevertheless, I undertook extensive measures to reduce them. The actual combat losses were not extreme, considering the nature of the operations and the desperate severity of the defenses. Even so, we could cut those due to inadequate maintenance and to equipment malfunction. We simply had to do everything within our power to overcome our deficiencies.

One measure, besides better maintenance and inspection, was lightening of the airplane by removing items that could be spared. A lighter aircraft had a much better chance of returning to base after battle damage or equipment failure. We stripped the plane itself of


1,900 pounds and removed one of the bomb-bay gas tanks for another 4,100, giving a total weight reduction of 6,000 pounds. Then, too, our operations were still confined to narrow deviations from the direct route to Tokyo and Nagoya. Mission after mission had to follow the same path. Our aircraft were therefore detected upon passing Iwo Jima (until its capture) and a few reporting ships. They gave warning in plenty of time for the Japanese defenses to concentrate at Tokyo or Nagoya.

The daily "command decision" whether to launch a mission the next day and against what target hinged upon the weather forecast more than any other factor. Our weather information came chiefly from a nightly B-29 flight to Japan. I had a meteorological officer who did a magnificent job under almost impossible conditions. His name was Col. James Seaver; I had known him in England. He knew perfectly well that my decision to "go" or to "stand down" depended directly upon his forecast. He also knew that his estimate was going to be better than mine, so he stated it without equivocation. He said what he thought would be the case, without hedging it with subjunctive clauses. Sometimes he was wrong, but more often he was right. I relied upon him heavily and was careful never to criticize when the weather forecast did not pan out.

The XXI Bomber Command had no special liaison unit (SLU) to receive Ultra information -- a grievous omission. I cannot understand why. Group Captain Winterbotham in Ultra Secret drops the casual statement:

In Brisbane (Australia) many of our main signals now came from Delhi, but radio blackouts were frequent. Sometimes signals came via the Australia Post Office cable, or even radio from Bletchley (England), and Japanese weather reports came up from Melbourne by teleprinter, so the SLU at Brisbane had a bit of a job sorting out what was going on.

What Colonel Seaver would have given for those Japanese weather reports! Weather over Japan was our most implacable and inscrutable enemy. Such reports received through Ultra were of great value in the strategic air war against Germany; they would have been priceless in


the air war against Japan. It seems simply incredible that no one "in the know" recognized our need, especially for Japanese weather reports, and took steps to supply me and later General LeMay with an SLU.


Table of Contents ** Previous Chapter (4) * Next Chapter (6)


1. General LeMay succeeded me as Commanding General of the XXI on January 20, 1945.

Transcribed and formatted for HTML by Charles Hall for the HyperWar Foundation