Chapter VI
Other Operations

The first B-29 bomber commander was Brig. Gen. Kenneth B. Wolfe, who took the XX Bomber Command to India and China and initiated operations against Japan with Operation Matterhorn. The XX Bomber Command was formed at Marietta, Georgia, where the B-29s were being built. General Wolfe was designated to head that command in November 1943. A production genius and a first-class aeronautical engineer, he literally mothered the first phases of production and modification of the B-29. Some 2,000 changes were made in the engine alone.

The XX Bomber Command Headquarters later moved to Salina, Kansas. The Second Air Force supervised its training. The 58th Bombardment Wing, commanded by Brig. Gen. La Verne G. "Blondie" Saunders, was the first element to reach operational status, and the 73d Bombardment Wing under Brig. Gen. Emmett "Rosey" O'Donnell, Jr., was scheduled to follow. On arrival in India in preparation for operations from Calcutta and advanced bases in Chengtu, China, the XX Bomber Command came under the jurisdiction of Lt. Gen. Joseph W. Stilwell. He commanded all U.S. Army forces in the China-Burma-India Theater until activation of the Twentieth Air Force. First units of the XX Bomber Command arrived in that theater during April 1944. After the XX was established there, the command conducted a "shakedown" operation on June 5, 1944,


BOMBS BURST AT THE KAWASAKI AIRCRAFT PLANT -- a strategic target located north of Tokyo. January 19, 1945.


against Bangkok from bases in India. On June 15 the XX launched a night attack from bases at Chengtu, China. The target was the Imperial Iron and Steel Works at Yawata on the Japanese home island of Kyushu.

Meanwhile, the B-29's mechanical and technical problems persisted. So in July 1944, General Arnold finally sent for the man in whom he had the most confidence, General Wolfe, who took over the Materiel Command with the primary mission of expediting production and improvements of the B-29. He was briefly succeeded at XX Bomber Command by General Saunders, and in turn Maj. Gen. Curtis E. LeMay replaced him on August 29. The original plans had earmarked LeMay for the XXI Bomber Command. However, recall of General Wolfe altered the plans, sending LeMay to head the XX Bomber Command and giving me command of the XXI Bomber Command, then in training in the Midwest.

When I first entered the Pacific Ocean Area, I was apprehensive about my command relationship. It was a Navy domain, dominated by strong-minded Navy commanders who could hardly be expected to welcome an intruder from the Army Air Forces who was independent of their operational control. But I got along well with the Navy commanders, due chiefly to the broad-minded support of Fleet Adm. Chester W. Nimitz, Commander in Chief of the Pacific Ocean Area, and the personal attitude of the Commander, Forward Area, Vice Adm. John H. Hoover whose flagship Curtiss was based at Saipan. I was uncomfortable in my relations with the senior AAF generals under Admiral Nimitz. They, quite understandably, resented this break in the chain of command. Lt. Gen. Millard F. Harmon was the senior AAF general in the Pacific Ocean Area and, under Admiral Nimitz, commanded all land-based aviation in that area -- Army, Navy, and Marine Corps. He naturally wanted control of the XXI Bomber Command too. General Arnold had sought to smooth over the situation by appointing him Deputy Commanding General for Administration and Logistics, Twentieth Air Force. But General Harmon wanted full command, including operational control -- or at least inclusion in the chain of command from Arnold to me. I had resisted this arrangement in Washington, when I was Chief of Staff of


the Twentieth Air Force, and I continued to resist it. If the Twentieth was to exist as a unified strategic air force under the direct and unbroken command of General Arnold, and with a primary strategic mission, the chain of operational command from the Joint Chiefs of Staff through Arnold to the XXI Bomber Command had to be kept direct and uncluttered.1

Under General Harmon was another senior Army Air Forces officer, Maj. Gen. Willis H. Hale, who commanded our land-based air forces in the Forward Area. Whereas my relationship with General Harmon had been tolerably agreeable, if somewhat formal, that with Willis Hale deteriorated after a confrontation on Saipan. When the second air base built for the 73d Wing of the XXI Bomber Command on Saipan proved technically unsuitable for B-29 operations, I based the entire wing at Isley Field, Saipan. I agreed to turn the other base over to General Hale's units since it was suitable for operations by other types of aircraft. When I arrived on Saipan with the first B-29, I found a half-completed base and over a hundred of General Hale's airplanes on Isley Field. Several times I requested Hale to clear the field for my impending operations. He agreed to do so but failed to move his planes. Finally, in desperation, I forced a showdown; the situation had become intolerable and threatened to prevent our first strike. Admiral Hoover offered to clear up the matter with a direct order to General Hale. But I thought it would be better if two air officers settled their problem between them. Hale moved his aircraft, then went straight back to Washington to complain to General Arnold about my "arrogant attitude." General Arnold backed me up, but I suspect the incident did me no good.

Change of Command

About mid-January 1945, a delegation from General Arnold's office arrived at my headquarters at Guam. Brig. Gen. Lauris "Larry" Norstad, Twentieth Air Force Chief of Staff, bore a message from


Arnold: There was to be a major change in organization and command. Combat elements of the XX Bomber Command were to be transferred to the Marianas as soon as bases could be made available and operations from Chengtu could be discontinued. Though I had known of the plans for movement of the XX Bomber Command out of China, I was not aware of its imminence. The urgency stemmed in part from the insistence of my old friend Lt. Gen. Albert C. Wedemeyer, USA, who had replaced General Stilwell as the new commander of U.S. forces in the China-Burma Theater. Wedemeyer, strongly urged by General Chennault, requested that the B-29s be withdrawn from China as soon as possible. The XX was absorbing supply tonnage urgently required by the Fourteenth Air Force and other forces in China. The repeated requests were directed to the Joint Chiefs of Staff who gave them a sympathetic ear. General Arnold agreed.

There were other reasons for expediting the change. The XX Bomber Command was operating under numerous disadvantages. It expended about one-seventh of its flying hours in attacks against the enemy and absorbed the other six-sevenths in furnishing its own logistics, that is, transporting gasoline and bombs over the Hump from India to the forward bases in China. Furthermore, the China-based B-29s could not reach the most vital targets in Japan.

When General LeMay had taken over the XX Bomber Command in India, he quite independently arrived at the same decision that motivated me in retraining the 73d Bombardment Wing. He set about transitioning from area night bombing to daylight precision bombing from defensive formations. He also began lead crew training just as I had done in the Marianas. But his logistic problems were so severe that it was almost impossible to establish effective operations against Japan itself. The radius of action limited strikes to the southern island of Kyushu, Japan.

The XX Bomber Command, while headed by General Wolfe, had attacked the Imperial Iron and Steel Works at Yawata on Kyushu in a night raid on June 15, 1944. Only forty-seven of seventy-five B-29s arrived over the target, and damage was not extensive. Shortage of fuel at the advanced bases prevented the launching of another full-scale


strike for several weeks. On July 7, a force of eighteen B-29s launched a small night raid against a number of targets in Japan, including Sasebo, Nagasaki, Omura, and Yawata. On July 29 the XX Bomber Command attacked the coke facilities at Anshan in Manchuria. Twentieth Air Force had directed a daylight precision attack with at least one hundred B-29s. In carrying out the operation, only sixty of the Superfortresses got over Anshan, and the effects were not satisfactory. The next operation was shifted to the oil refinery at Palembang, in the Netherlands East Indies, from advanced bases in Ceylon.

This attack, launched on August 10, 1944, entailed a 3,800-mile round trip. Results, unfortunately, were poor. On the same date, the XX Bomber Command sent a small force to attack the Nakajima engine works on Kyushu. The night attack was unsuccessful. B-29s struck Yawata in daylight on August 20. The losses from combat and operational causes were heavy: 14 out of 61 B-29s. Again, the outcome was disappointing. On August 29 the XX Bomber Command sent 108 B-29s against Anshan, under General LeMay's command and with his participation. This time there was considerable damage. The attack was repeated on September 26 in daylight using 12-plane formations, but cloud cover obscured the target. The logistic troubles grew so severe that the XX had to confine its operations mainly to targets outside Japan itself. Among them were: Okayama aircraft assembly plant on Formosa, October 14 and 16; Einansho Airdrome, Formosa, October 17; Omura aircraft factory, Kyushu, October 25; Rangoon, Burma, marshaling yards, November 3; Singapore Naval Base, November 5; Omura aircraft factory, November 11 (Nanking, China, was actually attacked because of weather at the primary target); and Omura on November 21, in daylight. The overall effectiveness was disappointing, caused by an intolerable logistic situation, unfavorable weather, and early training problems with a new and untried airplane.

On December 18 the XX Bomber Command conducted an operation that was significant from several points of view. Prompted by General Chennault, General Wedemeyer directed the XX to attack a theater target: the port facilities at Hankow, China. General LeMay


objected, citing his command relationship as a part of the Twentieth Air Force. General Wedemeyer then appealed to the Joint Chiefs of Staff, citing the provision for emergency use of the Twentieth Air Force by the theater commander as a matter of urgency. The urgency existed all right. The Japanese were launching a major but unsuccessful drive toward the B-29 base areas. The Joint Chiefs agreed, and General LeMay performed the mission, dropping incendiaries from medium altitude. Eighty-four B-29s dropped more than 500 tons of incendiaries on Hankow, setting huge fires that burned for 3 days. General Chennault, who urged the use of incendiaries against the city, described the attack as "the first mass fire-bomb raid" by the B-29s and contended it was the precursor of the massive urban incendiary attacks against Japanese cities.

By year's end, neither the XX nor the XXI Bomber Command had shown real results or approached the destructive power inherent in the B-29. But in considering the future, there was a vast difference between the XX in China and the XXI in the Mariana Islands. The XX could never hope to reach true effectiveness, so long as it had to fly all its fuel, bombs, and supplies over the Hump from India. The XXI had enormous potential. Given time to perfect its tactical performance and the growing might of the mounting accretions of new wings, it held tremendous portent for the future. I believe that portent could have been attained either through selective targeting or incendiary urban destruction, if given time for training in bombing intelligence collection and for the arrival in quantity of APQ-7 radar bombing equipment.

Night incendiary attacks on Japanese industrial areas in 1945 were contemplated in the original plans for deploying the XXI Bomber Command. However, such operations were to be undertaken solely as a last resort, and only if precision bombing proved unfeasible or failed to do the job. And the night attacks were initially scheduled to take place after selected bombing had knocked out the Japanese aircraft and engine factories and had destroyed the vital industrial targets. The complication of conducting selective bombing in daylight over targets obscured by cloud cover was expected, and the greatly improved radar bombsight (AN/APQ-7) was rushed to completion. It


was hoped it would permit all-weather bombing of selected targets (day or night). Since the sights were not instantly available, their use would take time. One of the XXI's wings, the 315th, had been equipped with this new bombsight, allowing more accurate day or night bombing in spite of cloud cover. The rest of the XXI's units used the less efficient AN/APQ-13, and they were partially trained for radar bombing of those area targets that rendered a good radar return.

Time, however, was not on the side of the XXI's Commander. General Arnold wanted and demanded measurable results at once. His judgment was heavily influenced by bomb tonnage instead of target destruction. Actually, many more tons of bombs could be dropped at night using radar bombing than in daylight. By day the force had to fly in formation and operate at high altitude to defend itself against Japanese fighters. These strictures reduced the bombload. Moreover, the rate of opportunity for daylight operations was heavily restricted by weather over the targets, which was very hard to forecast. Bad weather was the rule, and cloud obstruction was about the only weather feature that could be anticipated with any degree of assurance. At the start of the campaign, target locations were hard to determine and radar maps had not yet been prepared. Hence, the radar bombing was not adequate to put the bombs on selected industrial and economic targets.

In all fairness to General Arnold, he cannot be blamed for his impatience and his inclination to measure strategic air attacks in terms of tonnage and sorties. He was under constant pressure and criticism from his associates on the Joint Chiefs of Staff and from higher authority to explain what his Twentieth Air Force was accomplishing. It is exceedingly difficult to measure and evaluate the results of selective target bombing; in fact, we were unable to assess the real effectiveness of such operations both in Germany and Japan until after the war. To be sure, it is possible to report the destruction of a factory, but it is hard to estimate that destruction in terms of depletion of enemy industrial support for a specific set of economic or military needs. There is always the chance the enemy has found some substitute method of meeting those needs. It took the tremendous efforts of the U.S. Strategic Bombing Survey immediately after the war


to evaluate those effects in Europe and in Japan. On the whole, these survey reports showed that selective targeting was far more destructive than we thought at the time of attack.

On the other hand, statistics of tons of bombs dropped and of sorties flown are easily compiled, seem factual and specific, and are impressive. Photographs of burned-out cities also speak for themselves. And "time" had become an obsessive compulsion -- the time for invasion of Japan. Washington placed great stress on a quick end to the war, emphasizing that this carnage must not go on a single week longer than necessary to achieve victory. There were obvious weaknesses in this thought. The "carnage" would be enormously increased by an invasion. Casualties on both sides would be immense. Was it worth it? Was time itself all that important? Or were the casualties the more momentous consideration, once victory was assured? Some of this fixation on time mirrored the military habit of thought. And some of it doubtlessly stemmed from restless impatience among the American people and a desire to get this war over with and resume normal living.

The overriding priority of targets assigned to the XXI Bomber Command called for destruction or neutralization of the Japanese aircraft engine and airframe factories. One of the ironies of war is that in the early months of its operations the XXI actually did accomplish this mission, though the results were not then apparent. Its performance was surprisingly good, but unfortunately we could not prove it until after the war. I am sure General Arnold did not understand what the XXI Bomber Command had gone through or had achieved.

Since the XX Bomber Command was to be discontinued from Chinese bases, Arnold wanted to change the command and deployment setup at once. The proposed change appealed to him as a sensible step to greatly enhance B-29 performance, as well as to relieve the pressure from General Wedemeyer. The China Theater Commander was vigorously protesting to the Joint Chiefs that part of the Air Transport Command Hump tonnage was going to the XX at a time when he and the Fourteenth Air Force needed it all.

General LeMay, now senior to me, would take over the XXI Bomber Command within the month. I was offered the XX Bomber


Command scheduled for transfer to the Marianas; thereafter, I could become Vice Commander under LeMay. I did not wish to accept. I knew and respected LeMay as an able and competent bomber commander. I did not think he needed another bomber commander as deputy. I returned to the United States on January 20, 1945.

I was fortunate in having gifted and able associates in the Twentieth Air Force. Brig. Gen. Roger M. Ramey, my deputy, was not just a fine military associate but a fine friend. There were others who filled this dual role, among them being Col. Cecil E. Combs, Col. John B. Montgomery, and Comdr. George C. McGhee, USN. Such associations and friendships are a priceless boon that helps compensate for the pains and disappointments of wartime duty. Colonel Combs, who had been my Chief of Combat Operations in Washington, became Deputy Commander of the 58th Bombardment Wing when it arrived at Tinian. On my urgent recommendation, Colonel Montgomery, my Chief of Staff of the XXI Bomber Command, served General LeMay in an important capacity. Montgomery was a fine planner and manager, an expert pilot and navigator, and one of the best bombardiers in the Army Air Forces. Monty and I flew practice bomb runs against the Japanese-occupied island of Rota, to improve my understanding of the bombing technique and its problems. I was glad it was I who made out Monty's efficiency report on bombing, not the other way around.

Commander McGhee, a reservist, was my Navy Liaison Officer. The Navy repeatedly offered to replace him with senior, more experienced, regular professionals, but I resisted all offers. I knew I had an exceptional man. George was a competent Navy officer with interests and knowledge that covered a very broad range. I found his judgment invaluable in discussions of many facets of our problem in the Pacific. His responses and observations were stimulating and perceptive. He was that priceless combination -- a practical intellectual. After the war, George McGhee became a member of the State Department Policy Planning Council and later Ambassador to the Federal Republic of Germany.

There was another military associate for whose friendship I am especially grateful. When I assumed command of the XXI Bomber


Command I started searching for an aide. I set forth my specifications to the Chief of Personnel. I said I wanted a young first lieutenant or captain who was intelligent, alert, hardworking, good-humored, tolerant, courteous, loyal, and trustworthy. Besides, he should be a top-notch four-engine airplane pilot with enough guts to keep his hands off the controls when I was flying. Personnel produced 2d Lt. Ray L. Milne, who filled every one of those specifications. He was a perfect aide so far as I was concerned, and he became a cherished friend.

I made my decision to resist remaining with the XXI Bomber Command as Vice Commander under the stress of surprise and emotion. But I still think it was the proper step. I had every confidence in General LeMay. He had been the outstanding group commander when I headed the 1st Bombardment Wing in the early and crucial days of the Eighth Air Force. When I returned from England to the United States to be the Air Staff member of the Joint Plans Committee of the Joint Chiefs of Staff, LeMay stayed on, was promoted, and became senior to me. I knew him well enough to realize he needed no second string to his bow. He did not need a second in command, and I would have been unhappy as a figurehead. Furthermore, it is not a good thing to replace a commander and leave him in a subordinate position in his own outfit.

The first three months for the B-29s in the Marianas helped lay the groundwork for the much larger bombing offensive against Japan during 1945. If it is conceded that initial periods are likely to be the most difficult ones, then that of the XXI Bomber Command was marked with reasonable success. It can not be denied, however, that such success was accompanied by a full measure of good fortune. It might so easily have been a period of disaster, seeing that our first operations were from uncompleted bases. If our aircraft had returned to find our single, partially completed runway blocked by a crippled B-29 (or the base closed by one of those intense tropical storms that came our way), the whole force could have been lost. All in all, I think it was a good beginning. Its predominant pattern was woven on the theme of selective target destruction.


Operations In November 1944, the Joint Chiefs of Staff issued a new target priority list setting forth target systems for the XXI Bomber Command in this priority: (1) Japanese aircraft industry, (2) Japanese industrial areas, and (3) Japanese shipping. Our schedule of operations -- all against aircraft and engine factories, except shakedown missions against Iwo Jima -- were as follows:

Tokyo November 242
Iwo Jima December 7
Nagoya December 13 & 18
Iwo Jima December 24
Tokyo December 27
Nagoya January 5
Tokyo January 9
Nagoya January 14
Akashi January 21

Thirteen missions were flown in fifty-six days, or an average of one every four and one-half days, counting Iwo Jima.

Seven of the ten primary targets prescribed by the Joint Chiefs of Staff were specific aircraft and engine factories. Then an urban area system was prescribed, followed by "shipping." After these three top priorities were three secondary target systems: coke, steel, and oil. It is surprising that these vital selective target systems should have been assigned a priority below Japanese urban industrial areas.

When I left Washington as Chief of Staff of the Twentieth Air Force, a change in strategic policy set in. The policy I had espoused, and which I believe was generally accepted, was in this vein:

  1. Strategic Objective: To force Japan to acknowledge defeat and to accept our terms of surrender.
  2. Primary Air Strategy: To achieve the strategic objective by applying strategic air power. More specifically:
    1. To destroy the effectiveness of the Japanese Air Force to the degree


      where it should be incapable of offering a serious threat to our own bases and forces, or of offering effective hindrance to our strategic air offensive. To approach this objective by destroying Japanese sources of air power, using selective targeting and precision bombing.

    1. Through precision bombing, to destroy the war-making industrial structure of Japan by demolishing selected targets and systems vital to the war effort.

    2. Again, through precision bombing of selected targets, to destroy and undermine the social and economic structure of the Japanese state, by selection and annihilation of essential structures and systems indispensable to the organic functioning of the Japanese nation.

    3. To prepare for and, if necessary, to carry out urban incendiary attacks as a last resort.

  1. Secondary Air Strategy: To support a surface invasion of the Japanese home islands if the air offensive failed to achieve its purpose.

I was in full agreement with this emphasis on selective targeting and precision bombing. I had been one of the authors of this policy at the Air Corps Tactical School. I had seen it work well in Europe and had devised the plans for it in that theater. I believed in it. But, after I left Washington to prepare and direct the XXI Bomber Command to carry out this strategic concept, a switch in strategic interest became apparent. In the communications I received from Washington, there was repeated reference to and stress upon incendiary urban attack. I do not know if this change was brought about by General Norstad, General Arnold, or the Committee of Operations Analysts.

Perhaps it was General Arnold. Unknown to me, Arnold harbored a lively interest in incendiary urban attack. On April 5, 1944, he wrote General Spaatz about the proposed U.S. Strategic Bombing Survey and included these remarks: "Of particular interest to me would be some idea as to the most effective mixture of high explosives and incendiaries against heavily built-up areas." At any rate, the Committee of Operations Analysts gave incendiary urban attacks a high priority.

On December 18, the day of our first reasonably successful attack on the aircraft facilities at Nagoya, I received a directive to launch a full-scale incendiary attack on Nagoya. This was a blow. I had been


sweating blood in my efforts to make the 73d Wing a respectable precision bombing outfit, with very moderate success. We were just beginning to overcome the predilection for night area bombing, and we were just starting to show some improvement in bombing accuracy, both visual and radar. Now we were ordered to reverse our painfully achieved progress in accuracy and turn to area bombing. It was no good trying to attain real accuracy with the incendiaries. Their imprecise ballistic characteristic precluded any accuracy in delivery even if the sighting performance should be perfect.

Though in General Arnold's name, the directive had been signed by General Norstad. I protested directly to Arnold. I pointed out I had "with great difficulty implanted the principle that our mission is the destruction of primary targets by sustained attacks using precision bombing methods both visual and radar." I did not contend we had achieved an acceptable measure of success in this attempt, but I did assert that diversions from our determined efforts would impede a progress that was beginning to be encouraging for the future. General Norstad replied for General Arnold that the aircraft industry still had overriding priority and the fire raid was "simply a special requirement resulting from the necessity of future planning."

Future planning? Was the switch to area urban bombing already under way? The change to area urban incendiary attack, when it finally came, can not be laid directly at General LeMay's door. Its initial support came from Twentieth Air Force Headquarters. And it had begun with the selection of urban targets, after a revised report on Far East economic objectives was written and issued in October 1944 by the Committee of Operations Analysts. By that time, I had departed Washington for Saipan and was no longer in a position to influence strategic target selection. The report listed these cities as vital Japanese urban industrial areas to be considered for incendiary attack: Tokyo, Yokohama, Nagoya, Kobe, Kawasaki, and Osaka.

The Committee of Operations Analysts contended that the air offensive against Japanese urban areas would cut deeply into Japanese war production by (1) direct physical damage to major and feeder plants, (2) destruction of finished items and materials in process, (3) disruption of internal transportation and services, and (4) reduction of


labor efficiency. Cities were specified as preferred targets, superseding economic and industrial systems. This list showed a sharp departure from earlier strategy. The selected industrial "primary targets" still contained aircraft factories. But iron and steel (to be disrupted by attacking coke ovens) and oil (the petroleum industry) were all dropped to "secondary targets," below urban areas. Shipping ( presumably to include aerial mining) remained a top priority as it had been in the Committee of Operations Analysts' recommendation a year before. However, antifriction bearings and the electronics industry had been dropped, for reasons that I do not to this day understand. And the electric power and rail transportation systems had not been revived from their first rejection.

Since I had not yet accomplished my first-priority task -- destruction of Japanese aircraft and engine plants -- I was not immediately affected by this change. I continued to pursue selective bombing.

Refining Pacific Strategy

On December 1, 1944, the Joint Chiefs of Staff issued a revised memorandum describing U.S. Pacific strategy. It read:

The United States Joint Chiefs of Staff have adopted the following as a basis for planning in the war against Japan. The concept of operations for the main effort in the Pacific is:

A. Following the Okinawa operations to seize additional positions to intensify the blockade and air bombardment of Japan in order to create a situation favorable to:

B. An assault on Kyushu (Island) -- in order to establish a tactical situation favorable to:

C. The decisive invasion of the industrial heart of Japan through the Tokyo Plain.

General Marshall was generally acknowledged to be the author and proponent of this strategy.

Dr. Sallagar reviewed the U.S. Pacific strategy in Lessons from an Aerial Bombing Campaign. He discovered that:

To the Army, the JCS endorsement of naval blockade and strategic


bombardment merely meant that the Navy and the Air Force should be allowed to apply their favorite methods of warfare, provided that these preliminary operations were used to soften up the enemy in preparation for the invasion and did not interfere with the major objective.

But my chief aim as Commanding General, XXI Bomber Command, was unchanged. That is, the "intermediate objective of overriding priority" was still the Japanese aircraft industry. Moreover, the overall statement of military strategy for the strategic air war did not specifically countermand the initial statement of the strategic air objective -- to destroy Japan's capability to support the war.

In January 1945, we planned a variation from the steady stream of air attacks on factories in the Tokyo-Nagoya area. The Kawasaki Aircraft Industries Co., Ltd., was the third largest aircraft production company in Japan. It had a new engine and airframe complex at Akashi, about 12 miles west of Kobe and around 100 miles west of Nagoya. Besides being the biggest facility of the company, Akashi was also the headquarters of the Kawasaki engine division. There was another engine plant at nearby Futami, approximately 8 miles west of Akashi, and one at Takatsuki (about halfway between Kyoto and Akashi), nearly 20 miles from Akashi. Akashi was, however, the key installation in the engine complex. There was an additional Kawasaki airframe plant at Kagamigahara (Gifu), just north of Nagoya. The Akashi and Futami plants were on the coast of Harimanada, an arm of the Inland Sea, and adjacent to prominent landmarks that showed up well on radar. The targets, the plants of the Akashi engine and airframe facility, were about 2 miles from the town of Akashi. The engine plant occupied 1,287,700 square feet of productive floorspace, and the adjacent airframe plant occupied 1,047,000 square feet. The total target area was slightly over 3,300,000 square feet.

Kawasaki was one of the oldest and most experienced engine manufacturers in Japan. It started under French license in 1919, progressed through various German licenses for in-line engines, and culminated in the Daimler-Benz design designated Ha-60 (Model 22) of 1,150 horsepower, the Ha-40 of 1,175 horsepower, and the Ha-60 (Model 33) of 1,350 horsepower (purchased in 1937). These were used


BRIG. GEN. HANSELL AND HIS XXI BOMBER COMMAND STAFF plan a mission against Tokyo. Staff members include: (clockwise from left) Col. John B. Montgomery, General Hansell, Col. Ralph B. Garretson, Maj. D. P. Hatch, Lt. Col. Alan F. Adams, Col. Willard R. Shephard, Col. Seth S. Terry, Col. Albert T. Wilson, Jr., and Lt. Col. James T. Seaver.

B-29 ATTACKS INFLICTED HEAVY DAMAGE on the Kawasaki aircraft factories in January 1945.


in Tony fighters which resembled the German Me-109. The Tony entered operational service in 1943 and at one time was the most potent of the Japanese army's fighters. Its service ceiling was given at 32,800 feet.

Akashi engine works also turned out Nakajima-designed air-cooled radial engines, the Ha-35 (Model 22) and the Ha-35 (Model 32), rated at 1,100 and 1,150 horsepower respectively. They were used in Oscar fighters. Akashi likewise produced a Mitsubishi-designed radial of 1,970 horsepower, the Ha-45 (Model 21) used principally in the army's Frank fighters, Ki-84-1a, thought by many to be the best Japanese fighter built in quantity during World War II. Bearing a resemblance to our P-47, it was smaller and much lighter. Its service ceiling was given at 34,450 feet.

The Akashi airframe plant made Nick, a twin-engined fighter with day or night versions. A two-seater for defense against the B-29, its service ceiling was put at 32,800 feet. The plant also assembled Randy, a twin-engined, two-slot, attack fighter much like Nick but with better performance. In both its main plants, between 1941 and the end of the war, Kawasaki put together 8,269 airframes. Overall, that company completed 10,274 engines during January 1941-August 1945. Kawasaki accounted for 12 percent of the combat engines manufactured in 1944 (the industry's peak year) and 17 percent of the combat airframes. All in all, the Akashi complex presented a lucrative precision target.

The mission of January 19, 1945, against the Akashi works contained a diversionary ruse and a tactical variation. The 73d Wing went up the well-beaten path to Nagoya, but just as it approached the coast of Honshu, the force split. Three aircraft of the lead squadron continued toward Nagoya and bombed Hamamatsu, southeast of Nagoya, at high altitude -- 35,000 feet. It dispensed "rope" to obscure enemy radar screens and to impart the belief the main force was bound for Nagoya. The rest of the force (56 B-29s) turned sharply to the left and approached Akashi. The axis of attack was selected to optimize radar bombing of the target should it be covered with clouds. Bombing altitude was dropped to 25,000/27,400 feet, about 5,000 feet below previous levels. This decision was made to improve bombing accuracy.


It diminished the problem caused by very high winds and extreme turbulence encountered above 30,000 feet and cut down on errors that tended to be proportional to bombing height. The lower levels entailed a calculated risk in terms of fighter opposition. I relied on the ruse to deflect most of the enemy aircraft -- which it did. And I made the first move in a planned schedule of reducing altitude by successive steps to enhance bombing. This was based upon the discovery that Japanese fighter attacks, while bitter and reckless, were not as deadly against the B-29s as German fighters had been against B-17s and B-24s. I had intended to press this step-by-step lowering of altitude to sharpen bombing accuracy, until we reached a level where further reduction incurred too many losses from enemy fighters.

The January 19 mission was a magnificent success. The bombs were dropped between 1450 and 1524 in clear weather. A total of 610 500-pound bombs were dispensed on the primary targets; 275 (45 percent) hit within the plant areas, measuring roughly 1,200 feet by 4,000 feet for the engine facility and 900 feet by 2,400 feet for the airframe one. Every important building in the engine and airframe complex was hit. Nearly two-thirds of the bombs struck within the engine works. Production in both facilities dropped 90 percent and never recovered. Eleven Japanese fighters attacked, the bombers claiming 4 shot down. No B-29s were lost. The mission has been depicted as one of the most perfect examples of selective bombing in the entire war. It was among the best of which I had personal knowledge. An important side effect of the mission was to accelerate the dispersion of the aircraft industry -- a drastic move from which it failed to recoup.

The mission was, in my opinion, of great significance, the selected target being virtually destroyed. But of far more consequence in the long run, the bombing accuracy showed substantial improvement and the bomb pattern was well concentrated. The analysis of bombing accuracy by the 73d Wing Intelligence Section revealed that 46 percent of the bombs actually released at the primary target fell within 1,000 feet of the aiming point, giving a circular error probable of 1,030 feet. Nevertheless this analysis, based on bomb craters identified in reconnaissance photographs, did not account for all the bombs that


were carried to the target area. A damage report prepared by the U.S. Strategic Bombing Survey after the war painted a less favorable picture. Forty-five percent of the bombs landed in the principal target area, which exceeded 1,000 feet in radius.

As a broad approximation, the target area was closely equivalent to that within a circle of 1,490 feet radius, and 45 percent hits within a circle of that area gives a CEP of about 1,600 feet. This is not a demonstration of good bombing accuracy from 26,000 feet by the 73d Wing. Still, it was a marked improvement when compared with earlier strikes, and the bombing pattern showed a sufficient close concentration to destroy all elements of the target. Further training produced an average circular error probable of 1,250 feet, based on all bombers that did not abort for mechanical reasons.

To be sure, the visibility was excellent and local fighter opposition was minimal due to the successful ruse. Even so, the XXI Bomber Command had every reason to be elated. The intensive training program was paying off. It was clear to all, especially to the combat crews, that the XXI could destroy selected targets when weather conditions permitted visual bombing. It was an immense first step. The next would be the achievement of acceptable accuracy in radar bombing of selected targets. That, too, was on the way to attainment later in the war. This was the last mission that I laid on as Commanding General, XXI Bomber Command.

Since that time, I have assessed the situation in terms of the success of selected strategic bombardment in regard to the Japanese aircraft industry and war production industries. Using the U.S. Strategic Bombing Survey, the extensive postwar evaluation of the air war against Japan, I came to the conclusions set forth below.

The Japanese aircraft industry did comprise a selected strategic target system. Initial operations against the Japanese aircraft and engine factories were far more destructive that I judged them to be at the time. I was highly critical of our bombing accuracy. Nonetheless, that bombing was so devastating that the Japanese believed their industry doomed. They took the drastic countermeasure of dismantling their aircraft engine and airframe industry, dispersing it, and protecting it underground. The combined result of our destructive


bombing and the precipitate Japanese dispersal effectively realized the purpose of the strategic air offensive against the sources of production and supply of Japanese aircraft. Japanese aircraft production never recovered. As Dr. Robert Frank Futrell, Air Force historian, observed in his Ideas, Concepts, Doctrine: A History of Basic Thinking in the United States Air Force, 1907-1964:

Actually the B-29 attacks against the Japanese aircraft factories proved to be more effective than was realized. The U.S. Strategic Bombing Survey would discover that the damages caused by the B-29's were enough to convince the Japanese of a need to disperse their aircraft plants. The destruction inflicted, plus the confusion resulting from frantic dispersal efforts, reduced the pre-attack capacity of the aircraft engine plants by 75 percent, of airframe plants by 60 percent, and of electronic and communications equipment plants by 70 percent.

Japanese combat aircraft production peaked in 1944, before the B-29 attacks commenced in late November. Output for the year was 21,058, an increase of 662 percent from the 3,180 built in 1941. There were 9 producing companies. dominated by:

Nakajima 37.1
Mitsubishi 23.0
Kawasaki 14.9

followed by:

Aichi 6.9
Tachikawa 6.0
Total 87.9
Others 12.1
Grand Total 100.0

The primary and vital airframe and engine facilities were concentrated in the central manufacturing districts of these areas: Tokyo-Kawasaki-Yokohama; Nagoya; and Osaka-Kobe.

As shown in the chart, the drastic drop in actual aircraft



production began in the third quarter of 1944. This reflects the air attacks on the Japanese aircraft industry by the XXI Bomber Command.

Later Operations

When General LeMay assumed command of the XXI Bomber Command, he "stood down" the groups briefly. He retained my training program and improved the Lead Crew School. He focused on the maintenance effort, keeping on my A-4, Col. Clarence S. Irvine, who had worked wonders with this terribly difficult problem. After loss of the depot at Guam, Colonel Irvine needed to improvise a maintenance system based on the depot at Sacramento, 8,000 miles away. He set up an air transport system of his own, and I was careful not to ask where he got the transport aircraft or the authority to use them. In addition, he set up a streamlined maintenance organization in the 73d Wing by consolidating the resources of the combat and service groups.

General LeMay retained most of my tactical methods, including the airplane formations. This was hardly surprising since we had worked together in the Eighth Air Force. The capture of Iwo Jima by the Marines in February 1945 (an operation advocated by the Twentieth Air Force in May 1944) provided an advanced air base that was a boon to the bomber offensive. The air base was of enormous significance from the standpoint of morale, the recovery of crippled aircraft, and the improvement of operations in general. Damaged B-29s returning from raids could land at Iwo Jima, and it served as a fighter base from which escort fighters could support the Superfortresses or make strafing and bombing attacks on their own. Finally, the capture of Iwo Jima removed the Japanese early-warning station that had been giving 2- to 3-hour warning to Japanese defenders. From March 4, 1945 -- when the first crippled B-29 touched down there -- to the end of the war, 2,241 B-29s landed at Iwo Jima. Otherwise many of these would have been lost. Iwo Jima became the base of the VII Fighter Command.

For about six weeks, General LeMay carried forward the


operations I had started -- with almost identical results. It was apparent our preferred strategy (destruction of selected targets through precision optical bombing) could not be sustained in the face of the almost continuous cloud cover over Japan. Furthermore, the APQ-7 radar was still unavailable. General Norstad paid LeMay a visit at the end of six weeks, bringing the expected news that General Arnold was far from satisfied with performance. The factor of time was taking on a new insistence. The invasion of the Japanese home islands -- whose necessity had become an obsession with the Army planners -- had been agreed upon. If air power was to end the war without a massive bloodletting on the ground, its application could not be delayed. A drastic reappraisal was in order. LeMay made it.

The cities of Japan were vital to the ongoing war effort. Small factories were extremely vulnerable to incendiary attack. Although the first priority objective (destruction of Japan's aircraft industry) had not been fully attained, it had been approximated. Night attack of Japanese cities with incendiaries and radar bombing with the APQ-13 could be conducted on a consistent schedule, regardless of the weather. Japanese air defenses against night or all-weather assaults were minimal or nonexistent. Since it would not be necessary to operate in formation or at high altitude, bombloads could be much greater (up to 20,000 pounds per aircraft).

LeMay decided to switch from chief reliance on daylight precision bombing of selected targets to night incendiary attacks of Japanese cities. The first incendiary attacks against six of Japan's greatest cities were very effective and most impressive. The tonnage of bombs dispensed was extremely high, losses were very low, the rate and frequency of operations were unconstrained by poor weather, and the devastation of urban and industrial areas was startling. This new type of operations, attacking at low and medium altitudes at night, represented a superb tactical and strategic decision, and a most courageous and fitting one. Though it was suitable to the specific situation and circumstances, it was not necessarily appropriate to all requirements for the future. General LeMay recognized this. He went on using selected targeting whenever the weather and his equipment permitted.


These operations induced interesting reactions. The Joint Target Group of the Joint Chiefs of Staff seized upon the new tactic with enthusiasm. As depicted in Craven and Cate's The Army Air Forces in World War II:

The Joint Target Group, after studying reports of the blitz, concluded that there were no strategic bottlenecks in the Japanese industrial and economic systems except aircraft engine plants, but that the enemy's industry as a whole was vulnerable through incendiary attacks.

Set up in the Joint Chiefs of Staff organization to recommend Twentieth Air Force targets, the Joint Target Group began functioning in December 1944. The judgment on its part that "there were no strategic bottlenecks in the Japanese industrial and economic systems except aircraft engine plants" was wholly unwarranted, as intelligence studies had shown and the postwar Bombing Survey was later to prove. The Joint Target Group simply embraced a new tactic that was easier to perform and to measure.

In order of priority, the Joint Target Group listed twenty-two of the most vital Japanese cities from the standpoint of the important industries they contained. Based on these recommendations, the Twentieth Air Force on April 3 issued a new target directive. The leading aircraft engine manufacturers, Nakajima-Musashi in Tokyo and the Mitsubishi at Nagoya, were given top priority. Both were selected targets. Then the directive listed six priority urban areas: Tokyo, Kawasaki, Nagoya, and Osaka Urban Areas 1, 2, and 3. Again quoting Craven and Cate:

The Joint Target Group based its recommendations on the assumption that the principal function of air attack was to pave the way for an invasion of the home islands. . . . But after studying the results of the March fire raids, LeMay came to the conclusion that with proper logistic support air power alone could force the Japanese to surrender -- a view shared privately by some members of Arnold's staff.

Thus, one outcome of the first urban incendiary attacks was an


THREE LEADERS OF THE XXI BOMBER COMMAND: Maj. Gen. Curtis E. LeMay (left), Brig. Gen. Hansell, and Brig. Gen. Roger M. Ramey, Deputy Commander. Saipan, January 1945.

ORDNANCE MEN ON SAIPAN ISLAND load B-29 with incendiary clusters. Fire raids on Japanese cities began in February 1945.


endorsement of this method to the near exclusion of selective targeting. There was also another reaction. Granting the obvious tactical advantages to single-plane night operations at relatively low altitudes, did it follow that selective targeting should be abandoned? Was it possible to apply these same tactics to selective precision targets?

General LeMay was ordered to find out, and he applied himself to the question with his customary zeal. There was every reason to believe he would have welcomed an effective tactic to destroy selective targets as well as urban areas. Even so, the bombing equipment on hand was of limited capability, and his crews were ill-trained in this technique. LeMay concluded that the APQ-13 radar was inadequate for precision bombing. This was almost certainly true, in the absence of good-quality radar maps and selected offset aiming points affording good radar returns. He chose to adopt and adapt the RAF night bombing technique that, late in the European war, had yielded surprisingly accurate bombing results. He ran four good-sized experimental missions.

On March 24, 1945, there were 251 planes of the 73d, 313th, and 314th Wings dispatched against the Mitsubishi plant at Nagoya. The RAF pathfinder technique was employed. Ten minutes before bombing time, 10 B-29s lighted the engine works area with M-26 flares. Five minutes later, another 10 Superfortresses dropped M-17 incendiary clusters to start marker fires. The main force then attacked with 500-pound, general purpose bombs-sighting visually with the optical sights on the fires started by the pathfinders. Nagoya, however, was obscured with clouds. Though 1,533 tons of bombs were dropped, the results were negligible.

On March 30 the 314th Wing sent a small force to bomb the same target, once more using pathfinder tactics. The bombers missed completely, again applying visual sighting with the Norden optical sight. On April 1 the 73d Wing dispatched 121 aircraft to strike Nakajima-Musashi. Of the 1,019 tons of bomb dispensed, there were just 4 hits. On the 3d of April, 3 attacks were conducted consisting of 1 wing each attacking Mitsubishi's Shizuoka engine plant, Nakajima's


Koizumi assembly plant, and the Tachikawa engine plant. Damage in each instance was slight.

The command was simply not equipped or sufficiently trained for night precision bombing. Specifically, it needed target marker bombs, such as the 1,000-pounders used by the RAF, and reflex optic bombsights. Lacking these, General LeMay abandoned the experiment at the time. However, with the arrival in May of the 315th Wing, equipped with the APQ-7 radar bombing equipment, he tried again with results that were satisfactory indeed, as described later.

When I commanded the XXI Bomber Command, I had hoped to use aircraft equipped with the APQ-7 as lead aircraft. This would have enabled the entire force to bomb in daylight in squadron formations through the undercast cloud cover. And it would have let individual aircraft fitted with the APQ-7 bomb at night. But arrival of the 315th Wing had been delayed, and there was slight chance to test this tactic. Interest in the continued application of selective targeting, and the directive to try it, may well have come from Maj. Gen. Laurence S. Kuter, General Arnold's Assistant Chief of Staff for Plans.

Methodically, General LeMay proceeded to destroy the urban industrial areas that had been prescribed for him. But he did not lose interest in selective targeting, and attacked Japanese aircraft and engine plants whenever the weather appeared favorable. He carried out such strikes on April 7, 12, 24, and 30; May 5 and 11; June 9, 10, 22, and 2C -- a total of 10 such missions in 3 months. Then he bombed selected targets again on July 24. From April 8 to May 11, 75 percent of the XXI Bomber Command's effort was diverted to tactical support of the invasion of Okinawa (Operation Iceberg), particularly to attacks on airfields in Kyushu to suppress kamikaze operations from there. After the B-29s were released from Iceberg, intensive incendiary attacks on Japanese cities were resumed at once.

Incendiary Strikes

During General LeMay's concentration on incendiary bombing of urban industrial areas, there were 17 maximum-effort attacks entailing


6,960 B-29 sorties and 41,592 tons of bombs. Losses were 136 B-29s or about 2 percent of the sorties. Thereafter, LeMay turned to the smaller cities on his list, eventually assaulting and devastating a total of 66 urban areas. The chronology of these incendiary strikes was:

May 14: Daylight incendiary attack on Nagoya, including the Mitsubishi plant area. 529 B-29s were dispatched, 472 dropping 2,515 tons of bombs from 12,000 to 20,500 feet. 3.15 square miles burned out.

May 16: Nagoya urban area assault at night. Of the 522 B-29s taking part, 457 dispensed 3,609 tons of bombs. 3.82 square miles burned out.

May 23: Night bombing of urban Tokyo. 520 of 562 B-29s reached the target, dropping 3,646 tons of bombs from 7,800 to 15,000 feet. 17 Superfortresses were lost. 5.3 square miles burned out.

May 25: Strike against Tokyo urban area at night. 501 B-29s were sent, 26 being lost to flak. 3,262 tons of bombs dropped. 56.3 square miles destroyed. 50.8 percent of city burned out.

May 29: High-altitude, daylight attack on Yokohama urban area. 517 B-29s were escorted by 101 P-51s. 2,570 tons of bombs dispensed. 6.9 square miles burned out.

June 1: Osaka urban area struck in daylight by 458 of 521 B-29s dispatched. Escort of 148 P-51s suffered heavily from violent weather. 2,788 tons of bombs released from 18,000 to 28,500 feet. 3.15 square miles burned out.

June 5: Kobe hit in daytime by 473 of the 531 B-29s airborne. 3,077 tons of bombs dropped from 13,650 to 18,000 feet. 11 Superfortresses lost. 4.35 square miles burned out.

June 7: Day assault on Osaka by 458 B-29s with an escort of 138 P-51s. Radar bombing was from 17,900 to 23,150 feet. 2,540 tons of bombs dispensed. 2.21 square miles burned out.

June 15: Attack on Osaka at night. 444 of the 516 B-29s dropped 3,157 tons of bombs. 1.9 square miles burned out.

The bulk of XXI Bomber Command's operations was devoted to urban industrial area incendiary attacks. In the entire period of its operations, the Twentieth Air Force applied its capacity as follows:


Sorties Tons of Bombs
Precision bombing attacks:
    Aircraft and engine targets 2,838 14,152
    Petroleum targets 1,437 10,600
    Assorted industrial targets 1,459 8,093
Total 5,734 32,845
Urban industrial area attacks 21,671 138,215
Aerial mining 1,750 _______
GRAND TOTAL 29,155 171,060

Thus, just 19 percent of the total effort in terms of both sorties and bomb tonnage was directed against selective targets; 80 percent went to urban incendiary attacks; and less than 1 percent to mining.

The somewhat precipitate decision of the Joint Chiefs of Staff to move the 58th Wing from China-Burma-India Theater to the Marianas had at least one ill effect. It postponed the deployment of the 315th Wing (equipped with the new AN/APQ-7 radar bombing system) from April to June 1945. So upon the 315th's arrival in the Marianas, it carried out fewer than two months of operations before the end of the war. The deployment delay, however, was put to good use by intensifying training, particularly to perfect radar bombing accuracy. Consequently, the 315th's performance with the APQ-7 was spectacular. It clearly showed that selected targets could be hit at night or when obscured from visual bombing.

The first group of the 315th touched down at Northwest Field, Guam, late in June. Only one runway was available, although the other was nearing completion. Many of the base facilities were not yet installed, for construction of the field had been seriously delayed. The decision of the Navy Commander in Chief, Pacific Ocean Area, to move his advanced headquarters to Guam had caused critical changes in construction priorities there. Roads and naval facilities enjoyed a higher priority than B-29 bases. As a matter of fact, Northwest Field was slipped to Priority 95 on the Island of Guam. It taxed General Harmon's great persuasive powers to get the project moving again.

By the time the 315th Wing arrived, Lt. Gen. Barney M. Giles was the Deputy Commander, Twentieth Air Force. General Giles


INCENDIARY BOMBS SHOWER on the dock area of Kobe, Japan, on June 5, 1945.

SMOKE BILLOWS FROM AN INDUSTRIAL SECTION OF YOKOHAMA, JAPAN, as B-29s continue to dump fire bombs during a daylight raid on May 29, 1945.


SCENES OF WIDESPREAD DESTRUCTION greeted the first Americans arriving a Yokohama harbor three days after the Japanese surrender on September 2, 1945.

A JAPANESE SURVIVOR AMONG THE RUINS of Yokohoma, now occupied by American forces.



AMONG THE BURNED-OUT RUINS OF TOKYO, a survivor drinks from a broken water pipe.


established his headquarters on Guam. He endorsed and supported General LeMay's decision to direct the 315th operations against a set of selected targets comprising the Japanese oil industry. These targets had been given in the 1942 Air War Plan (AWPD-42), and more recently strongly recommended as a consequence of the analysis contained in the recently completed Strategic Bombing Survey of the European Theater. General Spaatz, who became Commander in Chief, U.S. Army Strategic Forces in the Pacific, in mid-July 1945, strongly supported the decision. Because the destruction of oil refining was not specified as a top-priority objective in the current assigned target list, LeMay described the initial attacks as shakedown training operations.

The selection of strategic bombing objectives was being argued back and forth in the Joint Target Group in Washington. But more and more the picking of such objectives was being evaluated in terms of influence upon the proposed invasion of Japan. By April, however, the Strategic Air Intelligence Section in Washington was contending that the state of the Japanese petroleum industry was so critical that the destruction of facilities and storage in Japan would instantly influence the tactical situation. So the position of Generals Giles and LeMay had considerable backing. This fine decision by General LeMay afforded the opportunity to test again the feasibility of all-weather attack on selected targets by radar bombing, and at the same time to contribute substantially to the conduct of the war.

Between June 26 and the end of the war on August 14, the 315th Wing flew fifteen night missions against oil refineries or synthetic plants in Japan. These missions are listed below:

Date Mission
June 26/27 1 Utsube Oil Refinery at Yokkaichi
June 29/30 2 Nippon Oil Company at Kudamatsu
July 2/3 3 Maruzen Oil Company at Shimotsu
July 6/7 4 Maruzen Oil Company at Shimotsu (Repeat)
July 9/10 5 Utsube Oil Refinery at Yokkaichi (Repeat)
July 12/13 6 Mitsubishi Oil Company at Kawasaki


Date Mission
July 15/16 7 Nippon Oil Company at Kudamatsu (Repeat)
July 19/20 8 Nippon Oil Company at Kansai
July 22/23 9 Imperial Fuel Industry Company at Ube
July 25/26 10 Mitsubishi Oil Company at Kawasaki
July 28/29 11 Toa Fuel Company at Smimotsu
August 1/2 12 Mitsubishi Oil Company at Kawasaki (Repeat)
August 5/6 13 Imperial Fuel Industry Company at Ube (Repeat)
August 9/10 14 Nippon Oil Company at Kansai (Repeat)
August 14/15 15 Nippon Oil Company at Tsuchizaki (near Akita)

The missions were conducted by streams of single aircraft at night, bombing from 15,000 feet. The initial bombloads averaged 14,631 pounds per airplane but, with experience, this grew to 20,684 pounds. Only 4 planes were lost and 66 damaged in the entire campaign. The 315th Wing launched its first mission on the night of June 26/27, under the command of Brig. Gen. Frank A. Armstrong, Jr., who had been one of my wing commanders in the 1st Bombardment Division of the Eighth Air Force. Two groups attacked the Utsube Oil Refinery at Yokkaichi which was producing aviation gasoline. The mission was only partly effective.

The second mission was flown against the Nippon Oil Company at Kudamatsu on the night of June 29/30. This oil refinery was on the coast, west-southwest of Kure. Of the thirty-six aircraft airborne, thirty-two bombed the primary target. No B-29s were lost. The attack was but moderately successful, so the target was hit again on Mission No. 7 during the night of July 15/16, using seventy-one aircraft of which fifty-nine bombed the primary target. Again there were no losses. The cumulative results of both missions were damaging indeed.

On the night of July 2/3, Mission No. 3 struck the Maruzen Oil Company at Shimotsu (located on the coast, south-southwest of


THIS OIL REFINERY IN THE TOKYO AREA was a key target during the final five months of the war against Japan.

ALL THAT REMAINS OF THE POWER AND GENERATOR PLANT at the Imperial Fuel Industry Company at Ube, after the bombing attacks of July and August 1945.


Osaka). Because the attack was not a complete success, the target was hit again on the night of July 6/7 by Mission No. 4. This time the outcome was superb as the mission reports attested:

Target: MARUZEN Refinery

Located immediately north of Shimotsu and 7 mi. South-Southwest of Wakayama. The Plant produced aviation gasoline, lube oil, ordinary gasoline and fuel oil. It had extensive storage facilities. Crude capacity was 5000 barrels per day.

On Mission No. 3, 40 aircraft were airborne. 30 dropped 297 tons of general purpose 500 lb. bombs on the primary target (95.7 percent of the bombs which were airborne).

On Mission No. 4, 60 aircraft were airborne. 59 bombed the primary with 441 tons of 500 lb. general purpose bombs (98.2 percent of bombs which were airborne).

Damage from Mission No. 3 was just moderate, but photographs from Mission No. 4 disclosed that ninety-five percent of the installation was damaged. Only five large tanks and several small ones were left standing.

General LeMay, who never extended unearned praise, sent this message after photo reconnaissance and interpretation of pictures confirmed the mission report:

I have just reviewed the post-strike photography of your strike on target 1764, the MARUZEN Oil Refinery at Shimotsu, the night of 6/7 July. With a half-Wing effort you achieved ninety-five percent destruction, definitely establishing the ability of your crews with the APQ-7 to hit and destroy precision targets, operating individually at night. This performance is the most successful radar bombing of the Command to date. Congratulations to you and your men.

On the night of July 28/29, Mission No. 11 demonstrated the high degree of accuracy obtainable with the AN/APQ-7 Eagle radar bombing equipment. The target was the Toa Fuel Company at Shimotsu. Extracts from the mission report revealed:

An important refinery of crude petroleum with large and modern facilities and good shipping and rail connections, the target



also had a tank capacity of about 600,000 barrels. It was roughly the shape of an equilateral triangle about 2,500 feet along each side. The storage area extended 1,000 feet beyond the northern top of the refinery area.

84 aircraft were airborne and 78 bombed the primary target, dropping 658.3 tons of 500 pound GP bombs. Opposition was light and there were no aircraft losses or crew member casualties.

Photos showed it was unnecessary to return to the refinery for in this one mission the target was almost completely destroyed. 927,000 barrels of the 1,245,000 barrel storage capacity was damaged while the 1,274,000 cubic foot gasometer capacity was almost completely destroyed. 69 percent of the 210,254 square foot group area was destroyed. The target was thoroughly saturated with bombs and obliterated beyond repair.

The target photo and interpretation showed the following distribution of bombing effort and impact of aircraft salvos:

Group Number of Aircraft
Bombing Primary
Percentage of Aircraft
Dispatched Bombing
Primary Target
16th 20 95
331st 13 93
501st 16 100
502d 15 91
Total 64 96

Centers of impact of 80 percent of salvos were identified in the photographs. Of these, 78 percent were in the target circle of 1,000-foot radius; 7 percent were over the target circle of 1,000-foot radius; and 15 percent were short of the target circle of 1,000-foot radius. Thus, 96 percent of aircraft dispatched bombed the primary target; 80 percent of the salvo centers were identified; and 78 percent of those identified were within 1,000 feet of the aiming point. This meant that 60 percent of the bombers dispatched placed their salvo centers within 1,000 feet of the aiming point, giving an average circular error probable of about 850 feet for salvo centers. This was an astonishing degree of accuracy for bombing at night from 15,000 feet through an undercast. It was actually much better than the average


CEP bombing distribution of XXI Bomber Command for visual daylight formation bombing (1,250 feet), though this was conducted at much higher altitude. Of course, one mission does not establish a CEP that can be taken as a reliable basis for forecasting and planning. Unfortunately, the other mission reports of the 315th Wing did not contain bomb plots.

The final mission of the 315th, flown on the night of August 14/15, was also remarkable. The mission report stated:


This target was attacked on the 15th and last mission flown by the Wing, flown on the night of 14-15 August 1945 with bombs released only a few hours before the announcement by President Truman that the Japanese had accepted the United States terms.

The mission was the longest nonstop combat flight ever made, a distance of 3,740 statute miles from base at Guam to the target on the northern coast of Honshu island and return.

Postponed for several days by the peace negotiations, the mission took off, led by the Wing Commander (Brig Gen Frank Armstrong) at 1637 hours on 14 August. 143 aircraft were airborne and 134 dropped 953.9 tons of 100 pound and 250 pound GP bombs on the primary.

Results of photo-interpretation brought now familiar words: "Almost completely destroyed or damaged." Photographs disclosed that no portion of the target was untouched. The three refining units were a tangled mass of wreckage, the main power plant still standing but seriously hit. More than 66 percent of the tank capacity was destroyed. Lesser installations, including the worker's barracks, were destroyed.

Note: This mission was conducted by the Twentieth Air Force, after redesignation of the XXI Bomber Command, under the command of Lt. Gen. Nathan F. Twining.

[The figure for bomb size in the report is probably in error; it should have been 1,000-pound GP bombs rather than 100-pound GP bombs as reported.]

The operations of the 315th Wing showed conclusively that it was feasible to destroy selected targets by radar bombing when the target


location was well known and the radar returns of the target itself were clear or its location relative to a prominent radar feature was well known. As suggested earlier, B-29s with AN/APQ-7 radar systems might have been used as lead aircraft for daylight selective bombing by formations of the other B-29s. This technique would have permitted employment of the entire force for daylight attack of selected targets even if those targets were obscured by clouds.

Effects of the Air Offensive

The U.S. Strategic Bombing Survey was able to report the effects of strategic bombing on the Japanese aircraft and engine target system, the aerial mining campaign, and attacks against selected targets in the iron, steel, and petroleum industries. On the other hand, it had trouble reporting the results of strategic bombing on Japanese war production and upon Japan's war economy, because no related system of targets was set up by the Joint Chiefs of Staff and the Twentieth Air Force as selective priority strategic targets. Nevertheless, the Bombing Survey submitted reports on a number of war production industries.

The Report on Japanese War Production Industries set forth the objectives of our strategic air force:

All four of these objectives were met to some degree. Strategic bombing did hasten and intensify the decline in war production, and prevented the Japanese from saving the production of high-priority items from the general decline. The report did not say from whence the objectives were derived; probably they were deduced from various


statements by military leaders and from mission directives and orders. It chose to consider Japanese "war production" as comprising six categories: aircraft industry; army ordnance; naval ordnance; naval shipbuilding; merchant shipbuilding; and the motor vehicle industry. Of these, only the aircraft industry was subjected to selective air attack.

The effects of the strategic air offensive were catastrophic in the aggregate. They stemmed from a series of interacting results of air bombardment: direct damage from bombing; indirect results of bombing reflected in frantic efforts to disperse industry; loss of basic raw materials through blockade, including aerial mining; and absenteeism of workers whose homes had been destroyed and who had to forage for food and the essentials of life for themselves and their families.

The report gave this analysis of the selective bombing of the aircraft industry:

No figures are available for loss of production due to physical destruction of plant, machinery, and equipment. Loss of production capacity through unsuccessful attempts at dispersal (which resulted from fear induced by the early attack on airframe and engine plants) was:

Airframes 33 percent
Engines 57 percent
Propellers 42 percent

About 55 percent of the whole aircraft industry's facilities were out of production due to dispersal alone.

For other categories of war production, the overall drop in production capacity through physical destruction of plant, machinery, and equipment was:

Army ordnance 26 percent
Naval ordnance 28 percent
Merchant and naval ships 10-15 percent
Motor vehicles negligible


The decline in production capacity due to unsuccessful attempts at dispersal (induced by the threat of bombing) was:

Army ordnance 12 percent
Naval ordnance 12 percent
Merchant and naval ships small
Motor vehicles sufficient to bring a complete collapse of production

The loss of production capacity by bombing was brought about by a combination of heavy urban area attacks and a relatively small amount of bombing (24,000 tons or 17 percent of total bomb tonnage) directed at selected targets. Strategic bombing alone did not reduce Japanese production. Loss of raw materials from shipping losses and blockade had an impact as well, especially in the case of steel. The report pointed out that:

The loss through ships sunk of 17 percent of all Army supplies shipped overseas (including food, clothing, fuel, and construction materials as well as ordnance) in 1943; 30 percent in 1944; and 50 percent in 1945 shows that increased production alone would not have been sufficient to provide adequate supplies for the Japanese Army overseas.

The report summed up its findings on the effect of strategic bombing of Japanese war industry production in these terms:

There was a 53 percent decline in war production between September 1944 (just prior to the launching of the air offensive) and July 1945 -- the last full month of production before the end of the war. The magnitude of the decline in output of each of the major categories of war production from peak levels to the July 1945 level is shown below:

Category Percentage drop from
peak production
Aircraft 57
Army Ordnance 54


Naval Ordnance 56
Merchant ships 82
Naval ships 53
Motor vehicles 96

There can be no doubt that the air offensive crippled Japanese war industries, even though only one of them (aircraft) was selected for direct attack. And the urban area incendiary attacks indirectly crippled other industry.

The Bombing Survey recognized another basic Army Air Forces strategic objective -- one that had been clearly defined in AWPD-1 and AWPD-42, though it had been submerged by other considerations in later plans. This objective was not only to destroy the war-supporting structure, but the economic framework on which the Japanese state depended. The combination was meant to bring about surrender, when it became apparent to the Japanese they could no longer supply the basic needs upon which the population relied for its life and social survival.

The U.S. Strategic Bombing Survey issued a report on the effects of strategic bombing on Japan's war economy. It concluded:

By July 1945 Japan's economic system had been shattered. Production of civilian goods was below the level of subsistence. Munitions output had been curtailed to less than half the war-time peak, a level that could not support sustained military operations against our opposing forces. The economic basis of Japanese resistance had been destroyed. This economic decay resulted from the sea-air blockade of the Japanese home islands and direct bombing attacks on industrial and urban-area targets.

The urban-area incendiary raids had profound repercussions on civilian morale and Japan's will to stay in the war. Sixty-six cities, virtually all those of economic significance, were subjected to bombing raids and suffered destruction ranging from 25 to 90 percent. Almost 50 percent of the area of these cities were leveled. The area raids interrupted the normal processes of city life to an extent that interfered seriously with such production as the shrinking raw material base still permitted.

The bombing offensive was the major factor which secured agreement to unconditional surrender without an invasion of the


home islands -- an invasion that would have cost hundreds of thousands of American lives. The demonstrated strength of the United States in the B-29 attacks contrasted with Japan's lack of adequate defense made clear to the Japanese people and to the government the futility of further resistance. This was reinforced by the evident deterioration of the Japanese economy and the impact it was having on a large segment of the population. The atomic bomb and Russia's entry into the war speeded the process of surrender already realized as the only possible outcome.

The effectiveness of strategic air attack was limited by the concepts of its mission. Had the purpose of strategic air attack been primarily to force an independent decision rather than to support a ground-force invasion in November 1945, there would have been no occasion to attack oil, tetraethyl lead, arsenals, or, after March, aircraft. Efforts could have been concentrated against food and fuel supply by attacks on internal transportation and against urban areas, thus striking solely at the main elements upon which continued Japanese resistance was based. Moreover, a part of the bombing effort merely duplicated results already achieved by blockade. Attack on the rail transportation system would have secured full coordination with the blockade program. The railroads were overburdened, defenseless, and had only limited ability to replace rolling stock or major installations.

The testimony was overwhelming that the air offensive against Japan -- essentially an anti-Japanese Air Force operation followed by an urban-area strategy -- was a magnificent success. The conclusion that the bombing effort should have been concentrated upon aircraft and engine production, transportation, and urban areas alone is interesting, significant, and worthy of further evaluation.

Debate over Grand Strategy in the Pacific

As the time for the Potsdam Conference drew near, President Harry S. Truman asked that the Joint Chiefs of Staff and the Service Secretaries meet with him to discuss Pacific strategy before his meeting with Prime Minister Churchill and Marshal Stalin. The Joint Chiefs immediately asked theater commanders for their views on strategy to defeat Japan.

General MacArthur had previously advocated invasion of Honshu


at the plain of Tokyo. He had stipulated that Russia must be induced to enter the war, so as to tie down the million-and-a-half Japanese soldiers believed to be in Manchuria. Without this provision, MacArthur advised against direct invasion of the Japanese home islands. In a staff report of March 8, 1945, he was quoted as saying he was in complete agreement with the Army that the sole means of defeating Japan was by invading the industrial heart of Japan. (There is a striking parallel here. Gen. Dwight D. Eisenhower had stated with regard to Germany that it was necessary to conquer the Ruhr, the industrial heart of Germany. Neither Eisenhower nor MacArthur seemed to understand that the heart of a great nation can be stilled by strategic air operations, as well as by occupation by troops.) General MacArthur believed Russia would demand and get Manchuria, but thought she should pay for it by joining in the fight against Japan. He was quoted as saying that he understood the Navy still favored a plan whereby Japan would be surrounded with air and naval bases, and eventually blockaded and bombed into submission. MacArthur contended that this would never succeed.

On April 20, 1945, General MacArthur analyzed the strategic problem under three possibilities:

Course 1. Encircle Japan by further expansion to the westward, deploying maximum air forces preparatory to attacks on Kyushu and Honshu in succession or directly against Honshu.

Course 2. Encircle Japan by further expansion to the westward with a view to its complete isolation, and endeavor to bomb Japan into submission without effecting landings in the homeland.

Course 3. Attack Kyushu and install air forces to cover a decisive assault on Honshu.

He dismissed Course 1 as time-consuming and diversionary away from the decisive area -- the plain of Tokyo. Turning down Course 2 as time-consuming and ineffective, he said:

It assumes success of air power alone to conquer a people in spite of its demonstrated failure in Europe, where Germany was subjected to more intensive bombardment than can be brought to bear against Japan, and where all the available resources in ground troops of the


United States, the United Kingdom and Russia had to be committed in order to force a decision.

He recommended Course 3.

In calling his conference with the Joint Chiefs for June 18, the President stressed the cost of invasion in terms of casualties. He wanted to know the time needed and the losses that would ensue from an effort to defeat Japan by isolation, blockade, and bombardment by sea and air forces. He asked General Marshall for his views. Marshall said he personally believed the operation against Kyushu was the only course to pursue. He felt air power alone was insufficient to put the Japanese out of the war, adding that it was unable alone to put the Germans out. General Eaker was present, representing General Arnold who was in Guam at XXI Bomber Command Headquarters. Arnold had been reached, and he asserted that bombing could end the war. Still, he instructed Eaker to support General Marshall's position.

General Arnold later divulged his reasons for favoring the invasion of Kyushu. He held, with General LeMay, that Japan was already tottering and air power would complete the collapse. But capture of Kyushu would afford certain benefits -- areas for basing forty groups in an additional air force. These groups, chiefly equipped with B-17s, would be close to targets in Honshu. The air units were actually available for transfer to the Pacific. And the capture of Kyushu would make it unnecessary to invade Honshu. Besides, this policy position would be an expression of loyalty to General Marshall. The latter had stood "in loco parentis" behind the birth and growth of the Army Air Forces, and had given evidence of supporting a separate Air Force, coequal with the Army and Navy after the war. General Arnold knew, as did the other Chiefs, that Japan had already started peace negotiations through the Russians in Moscow, who deliberately failed to forward the peace feelers. Admiral King concurred with General Marshall.

Admiral Leahy, however, voiced considerable concern over casualties and seemed to favor blockade and bombardment. He asked General Marshall what the casualty rates had been in the other Pacific invasions and how many troops would it take to invade Kyushu.


Marshall said the plan for invading Kyushu called for 766,700 men. Admiral Leahy said the Okinawa casualties (34,000 Army, 7,000 Navy) constituted about 35 percent of the force. If this yardstick was applied to Kyushu, the casualties would be numerous indeed (268,000). Admiral King thought the casualty rate would be somewhere between that on Luzon and that on Okinawa. Admiral Leahy went on to question the insistence on unconditional surrender, asserting that lesser terms would still allow our absolute control of Japan. The President finally accepted General Marshall's views, and the target date for invading Kyushu was set at November 1.

On his trip to the Pacific, General Arnold had viewed the scene firsthand from the Command Headquarters in Guam. He was impressed with the devastation already visited upon Japan and with the immense air power that was in place. He also saw the mounting strength of the Eighth Air Force, one wing of which was established in Okinawa and the rest in transit from Europe and the United States. Then, too, Arnold received a preliminary report of the Bombing Survey's findings on the effect of the strategic war against Germany. It was a staggering testimonial of the impact of strategic air warfare on a modern state at war. General Arnold was quoted at the time: "If we could win the war by bombing, it would be unnecessary for the ground troops to make a landing on the shores of Japan. Personally I was convinced it would be done. I did not believe Japan could stand the punishment from the air that Germany had taken." Arnold sent General LeMay back to Washington to brief the Joint Chiefs and, if they concurred, the Secretary of War and the President. LeMay had arrived too late. The President had already agreed to the policy of invasion and the machinery had been set in motion, not only for the invasion of Kyushu but for the subsequent invasions of Honshu.

At the Potsdam Conference, President Truman learned of the successful test of the "atomic device." He queried his advisors and top commanders about using it. They agreed to its use, with one exception -- General Arnold, the man whose aircraft would deliver it. Arnold, just back from the Pacific, questioned the need to drop the atomic bomb to assure Japan's defeat without an invasion. Japan had already been weakened by blockade and beaten to her knees by air


bombardment. In August, September, and October the Twentieth and Eighth Air Forces could double the total tonnage dropped on Japan to date. That should be sufficient to force surrender, since Japan was on the verge of collapse. However, if it became a question of dropping the atomic bomb or launching an invasion, he favored the bomb. Other considerations clinched his conviction that the bomb should be dropped. He later strongly supported the decision to use atomic weapons.

In anticipation of the President's decision to use the atomic bomb, potential targets had been selected. On orders from General Arnold, Col. Cecil E. Combs, Twentieth Air Force Deputy for Operations, had set aside four cities not yet bombed and passed the word to LeMay that they were not to be attacked. These were Hiroshima, Nagasaki, Kokura, and Niigata.

Second Change of Command

With Victory in Europe in May 1945, the second phase of global grand strategy was put in motion. Forces released from combat in Europe were transferred to the Pacific and preparations were made for the final offensive against Japan. On July 16, 1945, a major reorganization of the air forces in the Pacific took place. General Arnold turned to the strategic air team that had been so successful in the air war against the Third Reich. Gen. Carl Spaatz, Commander in Chief, U.S. Strategic Air Forces in Europe, was chosen to command the strategic air assault against Japan. His new designation was Commander in Chief, U.S. Army Strategic Air Forces in the Pacific, headquartered on Guam. He reported directly to General Arnold, who would continue to command the strategic air forces from Washington, as executive agent of the Joint Chiefs of Staff.

The Fifteenth Air Force in the Mediterranean was demobilized, but its Commanding General, Lt. Gen. Nathan F. Twining, was sent to the Pacific as a member of Spaatz's winning team. He would take command of the Twentieth Air Force with headquarters in Guam. The veteran Eighth Air Force, under the other member of the command team, Lt. Gen. James H. Doolittle, would be returned to the


United States, reequipped and trained with B-29s, and moved to Okinawa. The old XX Bomber Command Headquarters would be absorbed in the new Eighth Air Force Headquarters. The XXI Bomber Command would be reconstituted as the Twentieth Air Force. Again there was an awkward problem. General Arnold sent for General Twining and told him there was going to be a reorganization of the entire command structure in the Pacific: "I want Spaatz and Doolittle and you to take over right away. Now get on with it."

General LeMay was relieved as Commanding General, XXI Bomber Command, and General Twining assumed command of Twentieth Air Force. (Through oversight or neglect, General Arnold failed to acquaint LeMay with the decision he had reached, so it remained for Twining to answer LeMay's query on his arrival, "What are you doing here?" Perhaps Arnold had expected Spaatz to notify LeMay.)

General Twining's comment upon heading what had been the XXI Bomber Command and was now to be the Twentieth Air Force was typical -- and appropriate. He said, "Taking over this outfit from Curt LeMay is about like taking over the Notre Dame football team from Knute Rockne." Fortunately, General LeMay's broad experience and proven talents were saved for continued application in the strategic air war. General Spaatz made him his Chief of Staff -- a role that would keep him active in the final phase of the strategic air war against Japan. Lt. Gen. Barney M. Giles became General Spaatz's deputy.

General Twining had barely settled into his new command when he received orders to deliver the atomic bombs. Twining literally possessed no knowledge of the atomic bomb, for as Fifteenth Air Force Commander he had been deeply engrossed in mounting maximum-effort combat missions in Europe. After a supersecret briefing, he saw no need to waste time in questioning the judgment of his superiors.

The Finale

During the course of the war, the Twentieth Air Force flew



31,387 bomber sorties; 3,058 of which were flown by the XX Bomber Command, 28,329 by the XXI. The war had taken 414 B-29s, 80 from the XX Bomber Command and 334 from the XXI. Losses on combat missions averaged 1.3 percent of sorties airborne, with a total of 147 bombers lost. Of these combat losses:

50 percent were caused by enemy fighters.
36 percent were caused by enemy antiaircraft.
13 percent were caused by a combination of both.
1 percent were self-inflicted by accident.

Fighter losses were 80. By the end of the war the Twentieth Air Force comprised:

Bombers B-29 1,042
Fighters P-47 733
P-51 349
Night 18
Reconnaissance F-7 26
F-13 52
Staff and Transports   93

The total inventory of B-29s on hand in the Army Air Forces was about 3,700.

On the basis of photo coverage, intelligence estimated that 175 square miles of urban area in 66 cities were wiped out. Total civilian casualties stemming directly from the urban attacks were estimated at 330,000 killed, 476,000 injured, and 9,200,000 rendered homeless. There were 2,2 10,000 houses demolished or burned down and another 90,000 were partially damaged. This bombing "dehoused" 50.3 percent of the 1940 population of these cities. A total of 159,862 tons of bombs was dropped. Japanese casualties resulting from the strategic air attack, from all causes, was estimated at 900,000 deaths and 1,300,000 injured. Following are a few quotes from Japanese sources on the effectiveness of American air power:


"If I were to give you one factor as the leading one that led to your victory, I would give you the Air Force."

Admiral Asami Nayano
Imperial Japanese Navy
Chief of Naval Staff and Supreme
Naval Advisory to the Emperor

"If I were to give you the decisive factors in the war in the order of their importance, I would place first the Air Force."

Vice Admiral Shigeru Fukudome
Imperial Japanese Navy
Chief of Staff, Combined Fleet

"The determination to make peace was the prolonged bombing by the B-29s."

Prince Fumimaro Konoye
Premier of Japan

Japan accepted defeat while still possessing over 2.5 million combat-equipped troops and 9,000 airplanes capable of being equipped and flown as kamikazes. No Allied troops were present on the soil of the Japanese home islands when Japan surrendered.


Table of Contents ** Previous Chapter (5) * Next Chapter (7)


1. General Harmon respected but never really approved this arrangement. He was on his way back to Washington to try to have it changed when his airplane was lost without a trace in late February 1945.

2. Three times more by December 3; an initial rate of four missions in ten days.

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