A DIRECT HIT AT THE MUSASHINO AIRCRAFT ENGINE PLANT dug this crater and destroyed half of the industrial target, located on the outskirts of Tokyo.


Chapter IV
Planning the Strategic
Air War against Japan

Grand strategy for the Pacific began to receive formal reexamination at the Quebec (Quadrant) Conference of the Combined Chiefs of Staff, held at Quebec in August 1943. The conclusions of the conference made no mention of a specific bomber offensive against Japan. However, there was agreement on the overall strategic objectives for the prosecution of the war:
  1. In cooperation with Russia and other Allies to bring about at the earliest possible date the unconditional surrender of the Axis in Europe.

  2. Simultaneously, in cooperation with other Pacific Powers concerned, to maintain and extend unremitting pressure against Japan with the purpose of continually reducing her military power and attaining positions from which her ultimate surrender can be forced. The effect of any such extension on the over-all objectives to be given consideration by the Combined Chiefs of Staff before action is taken.

  3. Upon the defeat of the Axis in Europe, in cooperation with other Pacific Powers and, if possible, with Russia, to direct the full resources of the United States and Great Britain to bring about at the earliest possible date the unconditional surrender of Japan.


There were two specific agreements made at Quadrant that would affect the air operations in the Far East.

We are agreed that the re-orientation of forces from the European Theater to the Pacific and Far East be started as soon as the German situation, in our opinion, so allows.

General Stilwell will be Deputy Supreme Allied Commander South East Asia Theater and in that capacity will command the Chinese troops operating into Burma and all US air and ground forces committed to the South East Asia Theater.

The plan for defeat of Japan would be taken up at the SEXTANT Conference [held at Cairo, Egypt, in the late November -- early December 1943].

President Roosevelt headed the U.S. delegation at the Cairo (SEXTANT) Conference. With him were his personal military aide, Maj. Gen. Edwin M. "Pa" Watson; the Joint Chiefs of Staff (Gen. George C. Marshall, Adm. Ernest J. King, Gen. Henry H. Arnold); Adm. William D. Leahy; Maj. Gen. Muir Fairchild, AAF, from the Joint Strategic Survey Committee; Maj. Gen. Laurence S. Kuter, Deputy Chief of Staff, AAF; the Joint Plans Committee; and the Joint Logistics Committee, among others. The Joint Staff planners consisted of Brig. Gen. Frank N. Roberts, USA; Rear Adm. Bernhard H. Bieri, USN, formerly of the Joint Strategic Committee; and myself, recently returned from the Eighth Air Force. For this occasion Admiral Bieri, the senior member and chairman, chose to consider there were just two legitimate members of the Joint Plans Committee, one representing the War Department, the other the Navy Department. He considered me, if he noted my presence at all, as a sort of junior consultant to Frank Roberts on air matters.

Because Admiral Bieri would not bring himself to recognize my existence, he could not very well argue against the items I presented and supported. The chief air objectives I supported were: (1) Consolidating our strategic air forces under unified air command and control, both in Europe and in the Pacific; (2) Recognizing strategic air warfare as a principal, war-winning strategy, and its acceptance as such in the war against Japan; (3) Obtaining air base sites from which


strategic air warfare could be waged against Japan. General Roberts was cooperative, and I was able to get the Joint Plans Committee to agree to all of the important things affecting the air forces.

Long before the proposed invasion of Normandy, General Arnold had sought to strengthen the strategic air forces opposing the European Axis powers, through merger and establishment of a unified air command. The Eighth Air Force in England and the Northwest African Strategic Air Force should have been directed in a coordinated attack against the selected targets in Germany. But they were separated by command barriers. The strategic air forces in England operated under the direction of the Combined Chiefs of Staff, with Air Chief Marshal Sir Charles F. A. Portal as executive agent. The strategic air forces in the Mediterranean were under the theater commander in that area, an Army general. They were being used to support theater objectives. General Arnold endeavored to correct this fault by merging the command of the U.S. strategic air forces in both areas under a single U.S. strategic air commander, who would have authority to direct all European strategic air operations.

As the Army Air Forces' air planner, I succeeded with some difficulty in putting the issue through the Joint Plans Committee of the U.S. Joint Chiefs of Staff. As the American air member of the U.S.-British Joint and Combined Plans Committee, it was my job to put it through that committee at the Cairo Conference. It met stiff opposition from the British members. They pointed out that the strategic air forces in the Mediterranean were wholly dependent upon theater agencies for logistic support and administration. Our contention was that unity of command and concerted cooperation in the target area were more important than unity of command of logistics and administration in the base areas. The British, who had the overall command of all air forces in the Mediterranean, were quick to oppose a change that would rob their senior air commander in the Mediterranean of a large block of his air power. They stressed the complexities of logistic support and the fact that the U.S. strategic air forces in the Mediterranean were completely dependent upon the common logistic facilities. A separate operational command would still be at the mercy of the logistic allocations and capacity. Why not leave the command


chain as it was and direct the U.S. strategic air commander in the Mediterranean to cooperate and coordinate with his opposite number in England? This would, of course, leave the strategic air forces in the Mediterranean under the command of the local theater commander, who could use them as he deemed necessary and leave the arguments to be settled later.

In the final confrontation we prevailed, asserting that the RAF did in fact have unity of command of its own air forces through the Air Ministry, and that this practice bridged the boundaries between theaters. Why should not the American strategic air forces have a similar structure and unity of command?

The argument and agreement that unity of command and control over air operations to be exercised at the target areas was more important than that covering the base areas later served us in good stead when the Twentieth Air Force was created. The outcome of the issue at the Cairo Conference was the creation of the U.S. Strategic Air Forces in Europe and the organization of the Fifteenth (Strategic) Air Force in the Mediterranean as the second component (along with the Eighth) of that strategic force.

Even this consolidation could not prevent frequent diversion of strategic air forces from their primary mission to the support of local ones. Without this unified command, however, the diversions would have been far worse. Airmen grew distrustful of the powers of surface theater commanders. When the time for organization and command of air forces in the war against Japan came up for consideration, the fight was renewed on a broader scale. But command of strategic air forces in Europe was not the only air issue at Cairo.

I had returned from England to become the U.S. air member of the Joint Plans Committee just about four weeks before departing for the Cairo (SEXTANT) Conference in November 1943. During my preparation for the conference, I was surprised by one paper I came across. It was the outline of the proposed Joint War Plan for the conduct of the war in the Pacific. The opening statement of basic strategy, prepared by the Far East War Plans Group and endorsed by the Joint War Plans Review Board (which contained an Army Air Forces general officer), said in effect:


It has been clearly demonstrated in the war in Europe that strategic air forces are incapable of decisive action and hence the war against Japan must rely upon victory through surface forces, supported appropriately by air forces. Final victory must come through invasion of the Japanese home islands.

There was no dissenting voice from the air members of the committee and the review board. The draft plan had been sent to the various members of the Joint Plans Committee in October 1943. To be sure, the bombing offensive against Germany had not yet shown decisive capability. It had not yet been launched in strength and would not reach full power and application against primarily strategic targets in Germany for almost another year. And nothing had been demonstrated either for or against the potential of that strategic air offensive. But it was clearly evident the Far East War Plans Group and the Joint War Plans Review Board had written off the possibility of victory over Japan through final reliance on air power and were dedicated to victory through invasion. I knew that the Army and Navy members of the Joint Plans Committee would welcome this conclusion. With much difficulty I succeeded in amending that statement of basic strategy and establishing a provision for an initial potentially decisive strategic air offensive against the Japanese home islands. It was agreed preparations for invasion should proceed concurrently, in case such an air offensive should not be decisive. Strategic air power barely attained a reprieve, and strategic air forces gained a temporary stay against dismemberment and apportionment to various theaters for support of surface operations. But final reliance on surface invasion of the Japanese home islands was indelibly imprinted on Allied grand strategy, at least so far as the U.S. Army and Navy were concerned.

At Cairo I succeeded in making a very substantial change in the original statement of a grand strategy for the Pacific. The Combined Chiefs of Staff accepted and approved the "overall plan for the defeat of Japan" as submitted by the Combined Staff planners on December 2, 1943. The new description of grand strategy stated:

Our studies of the subject (of grand strategy) have taken account of the possibility that invasion of the principal Japanese islands may


GENERAL ARNOLD AND HIS AIR STAFF at the Cairo (SEXTANT) Conference of November 1943. The delegation was led by (front row, left to right) Brig. Gen. Hansell, Maj. Gen. Muir S. Fairchild, General Arnold, Brig. Gen. Laurence S. Kuter, and Col. Willard R. Wolfinbarger.

GEN. GEORGE C. MARSHALL (right) and General Arnold meet with the Combined Chiefs of Staff at Cairo's Mena House Hotel, December 4, 1943.


not be necessary and the defeat of Japan may be accomplished by sea and air blockade and intensive air bombardment from progressively advanced bases. The plan must, however, be capable of expansion to meet the contingency of an invasion.

At another point the Combined Chiefs agreed to this overall objective among others: "To obtain objectives from which we can conduct intensive air bombardment and establish a sea and air blockade against Japan and from which to invade Japan proper if this should be necessary."

I also succeeded in inserting a sentence in a paragraph of "Specific Operations for the Defeat of Japan, 1944," which was approved by the Combined Chiefs of Staff on December 3, 1943. The paragraph read:

Central, South, and Southwest Pacific. The advance along the New Guinea-Netherlands East Indies-Philippines axis will proceed concurrently with operations for the capture of the Mandated Islands (by then the Central Pacific). A strategic bombing force will be established in Guam, Tinian, and Saipan for strategic bombing of Japan proper. Air bombardment of targets in the Netherlands East Indies-Philippine Area and the aerial neutralization of Rabaul will be intensified.

The strategy underlying the bombing of Japan proper was similar to that applied against Germany: to defeat the enemy air force and to so weaken the Japanese capability and will to fight as to cause capitulation or permit occupation against disorganized resistance; failing this, to make invasion feasible at minimum cost.1

The position of air strategists regarding the air offensive against Japan was very weak in November 1943. B-29s were beginning to come off the line, but essential bases for their operation against the Japanese home islands had not yet been provided. General Arnold and his Air Staff were determined to employ B-29s against the Japanese homeland. We were extremely apprehensive lest they be apportioned to theater commanders for local operations. Once assigned to such


control, it would be hard to extricate them and concert their efforts against the prime targets in Japan.


An outline plan was prepared in August 1943 by the Air War Plans Division of the Air Staff for use of the B-29s from bases to be built by the forces of Chiang Kai-shek in China. It was the only way we could find to start using these aircraft (however ineffectively) against Japan proper, prior to the capture of the Marianas.

The idea of basing strategic bombers in China was not entirely new. At the Casablanca Conference in January 1943, President Roosevelt discussed with Prime Minister Churchill the possibility of air operations out of Chinese bases against the Japanese. General Marshall endorsed General Arnold's view that Japanese industry was very vulnerable to bombardment from the air. The President added that periodic bombing of Japan would have a tremendous effect upon the morale of the Chinese people. He suggested sending 200 to 300 planes to China, including heavy bombers (B-24s), and proposed that the bombers be based in India and staged through advanced bases in China.

The President had gone so far as to wire Chiang Kai-Shek that he was sending General Arnold to Chungking to discuss U.S. aid because he was "determined to increase General Chennault's air force2 in order that you may carry the offensive to the Japanese at once." However, this reference was evidently to the Japanese in China, not to Japan proper. In August 1943, a new outline plan for using B-29s based in India and staged through China to attack the Japanese home islands was first presented to the Combined Chiefs of Staff by General Arnold at the Quebec (Quadrant) Conference. The plan was tabled there for study by the Joint Logistics Committee and restudy at the Cairo (SEXTANT) Conference in November.

AWPD-42 had listed "iron and steel" as a primary target system


in the air offensive against the Japanese home islands. The Committee of Operations Analysts agreed that iron and steel were vital, both to the war-making capabilities of Japan and to the economic structure of the state. Steel production was in short supply and was running about 9.5 million tons per year. The consumption was divided equally between military and civilian usage. The following table shows the extent to which steel was vital in both categories:

Consumption Group Tons
Military and Naval
  Aircraft 190 2.02
  Armored fighting vehicles 142 1.51
  Ammunition 1,800 19.15
  Artillery & small arms 190 2.02
  Miscellaneous field equipment 998 10.62
  Shipbuilding 950 10.11
  Buildings & works 430 4.57
Subtotal 4,700 50.00
Industrial and Civilian
  Buildings & works 430 4.57
  Storage & transport 380 4.04
  Mining & quarrying 470 5.00
  Carbonization industry 190 2.02
  Agricultural machinery 95 1.01
  Machinery, equipment, tools 475 5.05
  Railways 1,140 12.13
  Motor vehicles 190 2.02
  Chemical & electrical industry 380 4.04
  Miscellaneous 950 10.11
Subtotal 4,700 49.99
Total 9,400 100.00

The Committee of Operations Analysts found that in Japan the steel production was uniquely vulnerable, because of the heavy concentration of coke ovens upon which steel production depended. Six coking plants (3 in the Japanese southern island of Kyushu, 2 near Mukden in Manchuria, and 1 in Korea) produced 73 percent of


Japanese coke. The Committee of Operations Analysts said the destruction of these 6 coking plants would deprive Japan of 66 percent of her total steel output. Coke ovens were susceptible to shock and their replacement would take years. The Air War Plans Division proposed that B-29s be based in India and operated from advanced bases in China, within range of some or all of these coking plants. The vicinity of Chengtu, China, was the preferred forward base. The location, relative importance, and approximate distance of these plants from Chengtu were:

Plant Location Percent of
Miles from
Anshan (near Mukden, Manchuria) 34.5 1,350
Penchihu (near Mukden, Manchuria) 11.2 1,300
Kenjiho, Korea 3.6 1,400
Yawata (Kyushu, Japan) (1) 16.2 1,500
Yawata (Kyushu, Japan) (2) 3.9 1,500
Omuta (Kyushu, Japan) 3.3 1,475

Chengtu, China, was situated about 1,150 miles from Calcutta, India.

General Arnold directed Brig. Gen. Kenneth B. Wolfe, who was to command the first combat unit of B-29 bombers -- the XX Bomber Command -- to prepare an outline operational plan to carry out attacks on these targets. General Wolfe's plan was expanded by the Air War Plans Division and became Project MATTERHORN.

The strategic purpose and concept of Project MATTERHORN were sound. But the logistic requirements were staggering and the logistic plan was horrendous. Based in India, the B-29s would stage through advanced bases in China. Even if the Chinese could be persuaded to build the air bases, it would be necessary to support B-29 operations from the advanced bases by air supply over the Himalayas. The B-29s themselves would have to ferry bombs and gasoline over "the Hump," and supply would have to be supplemented by B-24s converted into tankers. The effective rate of the operations would be very low indeed. Their primary virtue would be in striking an important blow against Japan proper and in preserving the command-and-control structure,



pending a time when other Pacific bases could be captured and prepared for other B-29 forces.

Chiang Kai-shek, who was present at the Cairo Conference in November-December 1943, agreed to the base construction at Chengtu. He was as good as his word. The bases were hand constructed by hundreds of thousands of workers.

Actually, Project MATTERHORN had a painful birth and a brief life span. Under General Kuter's supervision, the Air Plans Division of the Air Staff planned MATTERHORN. General Arnold presented the final plan to the Joint Chiefs of Staff and, after some discussion, they, at General Arnold's request, referred it to the Joint Staff planners (before my arrival in that committee). These planners and the Joint Intelligence Committee seriously questioned if the concept was desirable and feasible. Nevertheless, it was sent to the Combined Staff planners for comment or agreement.

The overall objective of MATTERHORN was to accelerate the destruction of selected systems of critical Japanese industry. This would reduce the industrial support of the Japanese war effort as a contribution to the overall strategic air plan. The plan, like that in Europe, contemplated an intermediate objective, the neutralization of the Japanese Air Force by combat and by destroying aircraft and engine factories. Then would come the destruction of primary targets whose paralysis would lead to fatal weakening (or collapse) of the Japanese will to resist and capability to continue the fight. Operations from Chinese bases would further this objective and reduce Japanese shipbuilding and naval resources. This would directly assist the later major air offensive from the Marianas. The plan called for 4 advanced bases near Chengtu to be furnished by the Chinese, and 4 main bases in the vicinity of Calcutta, India, to be provided by the British. Ten B-29 groups (28 aircraft per group) were to be ready by October 1944 for deployment to India and operation from China. Two thousand B-24s converted to transports were to support such operations over the Hump from supply bases around Calcutta. These 2,000 converted aircraft could be made available in the Calcutta area by October 1944.

The Combined Logistic Committee concluded on September 14, 1943, that the plan was not feasible from a logistic viewpoint. General


Arnold received this negative opinion of the MATTERHORN proposal just before he departed for the Cairo (SEXTANT) Conference. Even so, he asked the Committee of Operations Analysts to give him a list of recommended targets in Japan for the final air offensive, and show which targets could be included in the initial operations from Chinese bases. The plan would embody the operations from other bases. Targets would be those suitable for air bombardment that would "knock Japan out of the War." Iron and steel were high on the list.

The operations analysts described seven industries which "now appear profitable aviation target systems," and listed thirteen others which "did not now appear to be profitable" but might become so. The seven preferred target systems were:

  1. Merchant shipping in harbors and at sea.

  2. Iron and steel production, to be reached through coke ovens (in Manchuria, Korea, and Kyushu, within range of Chengtu).

  3. Urban industrial areas vulnerable to incendiary attacks.

  4. Aircraft plants.

  5. The antifriction bearing industry, highly concentrated in six main factories.

  6. The electronic industry whose interruption would have immediate military effects.

  7. The petroleum industry.

The Committee of Operations Analysts contended, however, that the list was not in an order of desired priority.

In spite of all the criticism and opposition, Project MATTERHORN was approved at the Cairo (SEXTANT) Conference. Chiang Kai-shek agreed to build the advanced bases at Chengtu. The British, who would have to build the bases in the Calcutta area, went along. But this did not end the argument. As late as February 15, 1944, the Joint War Plans Committee still believed the best use of the B-29s before deployment in the Marianas would be from Australian bases against shipping and oil. In that committee's opinion, Chengtu, China, was a very poor choice. Nonetheless, MATTERHORN was approved at SEXTANT and proceeded as planned.


Allied leaders also agreed at the Cairo (SEXTANT) Conference upon two coordinated but semi-independent surface thrusts in the Pacific. One would be from the Southwest Pacific, under the command of Gen. Douglas MacArthur. The other would be across the Central Pacific, under the command of Adm. Chester W. Nimitz.

Conference in the Pacific Theaters

At the termination of the SEXTANT Conference at Cairo in December 1943, I was chosen to go with General Marshall to meet with our principal commanders in the theaters in the Pacific. Our flight plan took us to Karachi, India, then to Ceylon, and to Exmouth Gulf and Darwin in Australia.

Lt. Gen. George C. Kenney, General MacArthur's air commander, met us at Darwin. We next flew across the Arafura Sea to Port Moresby in New Guinea. The following day, General Marshall met with MacArthur at the latter's headquarters on Goodenough Island. Only four persons were present at the meeting: Marshall, MacArthur, Kenney, and myself. As a staff officer to Brig. Gen. Frank M. Andrews in the early days of GHQ Air Force, Kenney had made a major contribution toward creation of air power. He hit it off from the first with General MacArthur -- and little wonder. He did things with air forces that left airmen gasping. MacArthur owed much of his brilliant success in the Southwest Pacific to General Kenney's imaginative performance.

It was Kenney's idea to establish advanced air bases, bypass enemy strongholds, furnish air support for forward movements, cooperate with the Navy in cutting off sea movements by isolated Japanese garrisons, and even to supplement important supply movements by air. He proposed to General MacArthur that, instead of inching slowly over the formidable Owen Stanley mountains in New Guinea, an airborne force be dropped and landed on the other side behind Japanese strongholds, and supplied initially by air. This would leave a Japanese force behind the invading force, cut off by the jungle, the terrain, and the sea. When one of General MacArthur's staff officers asked sarcastically what the troops were supposed to do


without wheeled vehicles, Kenney said he would move these too. He proceeded to cut trucks and other vehicles in two with acetylene torches, stuff the parts into air transports, fly them over the mountains, and have them assembled and welded back together. It worked! He also promoted low-level skip bombing on ships, armed medium bombers with multiple batteries of .50-caliber guns for strafing, and sponsored the mounting of a 75-millimeter cannon on a light bomber.

No air strategist or tactician showed greater imagination and inventiveness than George Kenney. And it speaks volumes for mutual trust and confidence that General MacArthur embraced the daring proposals of his chief airman and pursued his audacious program of "island hopping."

The meeting place was a shack that had been fitted up as a private war room for General MacArthur. Maps and charts covered the walls. MacArthur described the situation and his plans. He stood at various maps, strode back and forth, and talked for about two hours without notes of any sort. He had at his fingertips all the dispositions and recent actions of his troops. He seemed equally well acquainted with his enemy. He named Japanese organizations and their commanders everywhere and seemed well informed of their competence. MacArthur revealed his plans, culminating in the recapture of the Philippines and preparations for the next campaign, whether it be Formosa or the coast of China. In minute detail he defined the force requirements (land, sea, and air), the timing, the objectives, and the logistic flow. Throughout the presentation he employed wit and charm with devastating persuasiveness. Although I had from the first been an advocate of a "Europe first" strategy, with attendant delay against Japan, I simply melted under the persuasive logic and the delightful charm of the great MacArthur. By the time he had finished, I was anxious to find some way to give him what he had asked for.

General Marshall was of far sterner stuff, though his position left him reason for sensitivity or even embarrassment. In their relationship years before, General MacArthur had been Chief of Staff of the U.S. Army when Marshall was still a colonel. In World War I, MacArthur had been a general officer who had achieved an aura of fame from


personal bravery on the battlefield, while Marshall was an able but little-known officer on General Pershing's staff. Now the tables were completely turned. General MacArthur, for all his great prestige, was really a supplicant for approval of the strategy which he had proposed with such elan. Those plans would have required the assignment to General MacArthur not only of the majority of the American forces arrayed against Japan, but also substantial diversion of forces destined for Europe.

This was seven months before the launching of the Normandy invasion. General Marshall had to tell MacArthur he could not have these forces and hence could not carry out the program he had described. This Marshall proceeded quietly to do. He reminded him that the basic grand strategy of defeating Hitler first, and of concentrating maximum forces to that purpose, was agreed upon and approved by the Combined Chiefs of Staff as well as by President Roosevelt and Prime Minister Churchill. Marshall stated quite calmly his own devotion to that scheme, of which in fact he had been one of the chief architects.

The meeting closed on the same level of punctilious courtesy on which it had commenced. If General MacArthur was chagrined and disappointed, he did not show it. We left for home via Hawaii where General Marshall met with Admiral Nimitz and his staff. There was a presentation of a plan for Central Pacific strategy that quite naturally advocated primary reliance on the U.S. Navy for progress toward Japan and for regaining base areas. General Marshall made no commitments that I know of, and we journeyed home.

Pacific Strategy

After our return from the SEXTANT Conference at Cairo, arguments on Pacific strategy rose. The next major strategic objective was depicted by General Marshall as the "Formosa-China Coast-Luzon" triangle, to be approached by General MacArthur from the Southwest Pacific and Admiral Nimitz from the Central Pacific. Proponents of the two thrusts presented their views to the Joint Chiefs of Staff, each


proposing that the preponderance of effort and of forces be assigned to his axis.

On March 7, 1944, Admiral Nimitz, supported by his Deputy, Rear Adm. Forrest P. Sherman, appeared before the Joint Chiefs of Staff. They suggested bypassing the Japanese base at Truk and seizing the southern Marianas. From there they would capture Ulithi Atoll (about 360 miles southwest of Guam) for use as a fleet base, together with nearby Yap where airfields could be constructed to support Pacific naval operations. Nimitz's schedule called for capture of the Marianas in mid-June, Ulithi-Yap by September 1, and the Palau Islands by November 1. It would then be possible to invade the Formosa-China Coast-Luzon area by early spring 1945. He later amended the plan to specify the capture of Palau before Ulithi-Yap.

On the same day, Lt. Gen. Richard K. Sutherland, General MacArthur's Chief of Staff, tendered a plan for major support of an operation (Reno IV). This operation would be a push along the northern coast of New Guinea into Mindanao, Philippine Islands. In a covering letter, General MacArthur said: "The line of action presented in RENO IV will sever sea communications between Japan and the vital Borneo-Netherlands East Indies-Malaya region and will place our forces in the Luzon-Formosa-China Coast area at the earliest possible date under conditions that can be foreseen at this time."

Neither of these plans and presentations attached any importance to a strategic air offensive against Japan proper. When these proponents of rival strategies had reached the end of their presentations before the Joint Chiefs of Staff, General Arnold remarked that the Army Air Forces would like to present its views on Pacific strategy at the next meeting. As he was leaving the room, he turned to me and told me to prepare the presentation. I conferred with Generals Santy Fairchild and Larry Kuter and prepared an outline.

I went up to General Arnold's office to seek his approval or instructions and learned he had gone to the West Coast and would not be back for the next Chiefs meeting. Later I came to understand and appreciate this tactic, which General Arnold used several times. His position as a member of the Joint Chiefs of Staff was equivocal at best. The AAF was never accepted as an equal partner by the Navy. The


Navy Department did not openly try to quash the upstart air membership, but it worked quietly on the premise that there were by law just two recognized military departments -- the War Department and the Navy Department. General Marshall was a tower of strength in supporting General Arnold and the Army Air Forces. Still, Arnold was a subordinate Army officer and he could never afford to openly oppose Marshall. The Army Chief of Staff was universally fair and unbiased, but the Army people as a whole tended to support MacArthur just as Navy people tended to back Nimitz.

General Arnold knew that we airmen would lean toward Nimitz's position and the Navy, if we could be assured of the capture of the Mariana Islands as base areas for B-29 strikes against Japan. If in Arnold's absence from the next meeting, General Marshall should disagree with this approach, the AAF Chief on his return could dismiss us in good grace on the ground that we had not been instructed to take this line and had exceeded our authority. The status of the Army Air Forces and the support of Marshall could thus be preserved.

But in fact General Marshall did not take exception to the air position as I presented it on March 9, 1944. The presentation was favorably received. Subsequently, the Joint Plans Committee and the Joint Logistics Committee proposed this schedule:

Objective Command Date
Hollandia Southwest Pacific April 15, 1944
Marianas Central Pacific June 15, 1944
Palau Islands Central Pacific September 15, 1944
Mindanao Southwest Pacific November 15, 1944
Formosa Central Pacific February 15, 1945

The question of Pacific strategy had not been clearly resolved, and both the rival strategies were endorsed. Achieved, however, were the capture of the Marianas as air bases and support for the B-29 operations in their air offensive against the Japanese home islands. This precipitated the vital question of how the B-29 force would be organized, commanded, and controlled.


Roles and Missions for Strategic Air Forces

The command of strategic air forces was thornier than the technical problems posed by the new bomber airplane. Unity of command was a cherished military concept in both the Army and the Navy. The Army attained this unity by designating a single commander with authority over all units within specific geographical boundaries. The Navy achieved it by retaining control of major combat naval forces under the ultimate command of the top naval echelon of the nation. Fleet units were rarely assigned to territorial command areas. When they were, it was always with the proviso they could be withdrawn at any moment for use elsewhere if the naval situation should so require.

Strategic air forces did not fit either concept, but their command characteristics resembled more closely those of the Navy than those of the Army. Often the long-range air force straddled several land commands. Its bombers might be based in many areas, each under separate Army or Navy jurisdiction. But bombers of the strategic air forces demanded unity at the target area, and they needed continuity of application if they were to accomplish their strategic mission. The very flexibility that was the cardinal virtue of strategic bombers was also their greatest vulnerability. There was a constant temptation to divert them from their long-range strategic war objectives to targets critical only to local area commanders.

The problem of unity of command grew more acute as primary attention turned to Japan and the B-29 force started to emerge. To apply this very heavy bomber force against Japan proper -- its most important and potentially decisive role -- plans were made to set up a number of bases within action radius of Japan. These bases were to be in China, the Marianas, Alaska, the Philippines, and Formosa or Okinawa.

Project MATTERHORN called for bases in India and China. All U.S. forces there were under Gen. Joseph W. Stilwell, USA. He in turn was part of the Allied Southeast Asia Command headed by Adm. Lord Louis Mountbatten of the Royal Navy. (This command had been created at the August 1943 Quebec (Quadrant) Conference.) Although


Admiral Mountbatten was Supreme Allied commander, Southeast Asia, Chiang Kai-shek did not recognize any commander in China above himself. General Stilwell commanded all U.S. forces in the China-Burma-India Theater, which would include the B-29 forces. General Chennault commanded the U.S. Fourteenth Air Force and was at the same time Chiang Kai-shek's Chief of Staff for Air. Admiral Mountbatten had an Allied Air Commander in Chief, Southeast Asia Command, Air Chief Marshal Richard E. C. Peirse. Maj. Gen. George E. Stratemeyer was Commanding General, U.S. Army Air Forces in the India-Burma Sector, and Air Advisor to General Stilwell in China. Stratemeyer's command included the Tenth Air Force based in India. To further confuse an already complicated command arrangement, Stilwell gave Stratemeyer administrative command of the B-29 force (with its main bases in India) and also issued orders to the Fourteenth Air Force through him. General Stilwell proposed to exercise direct control of the B-29s, which he planned to use extensively in combined operations in China against Japanese ground troops. Admiral Mountbatten endorsed the initial operations entailed in MATTERHORN, but planned to use the B-29s later in support of Southeast Asia Command objectives. In his capacity as Chiang Kai-shek's Chief of Staff for Air, General Chennault appealed to President Roosevelt directly requesting that all B-29s operating out of China be placed under his control. He made a similar request to General Arnold, asking that the B-29s operating from Chinese bases be put under the U.S. Fourteenth Air Force.

The Joint Staff planners proposed that ultimately four groups of B-29s be based in the Philippines. Those islands, when recaptured, would be under the command of General MacArthur. His chief airman, General Kenney, was already making vigorous demands for B-29s to be used in the Southwest Pacific campaign. Kenney wanted B-29s stationed in Darwin, Australia, for strikes on targets in the Netherlands East Indies. In addition, plans were actively being prepared for positioning B-29s in the Central Pacific and in Alaska. The Marianas, due to be captured chiefly as the Central Pacific base for the B-29s, would under existing circumstances be under the command of Admiral Nimitz. Alaska was still another command area.



Every one of these base areas was under a separate theater commander, and these field commanders were powerful people. Each wanted to apply the B-29s to his own strategic theater purposes, and each resented any incursion into his area of control. Yet there was one area in which unity of air command and continuity of effort was imperative. That was the target area itself, Japan, which was under the control of none of them.

In March 1944 I presented to the Joint Chiefs of Staff the Army Air Forces concept of Pacific strategy. It envisioned a concerted bombing offensive against the Japanese home islands from the Marianas, to undermine the war-making capability of Japan. The plan also provided for the main B-29 force to be located in the Marianas. When the Philippines had been retaken, B-29 units were likewise to be situated on bases there within range of Japan. The B-29s in Chengtu, China, were to be moved forward when better base areas became available. A base was to be built in the Aleutian Islands as well.

Prior to the redeployment of the Eighth Air Force, the first plan for the final deployment of B-29s (and escort fighters when they became available) was as follows:

Chengtu, China 4 B-29 groups3
Mariana Islands 16 B-29 groups (3 squadrons each)
Ryukyu Islands 12 B-29 groups (3 squadrons each)
Philippine Islands 12 B-29 groups (3 squadrons each)
Aleutian Islands 4 B-29 groups (deployment questionable)
Iwo Jima Island 3 groups of long-range support fighters (type unspecified)
Ie Shima Island 2 groups of support fighters (type unspecified)
Kikai Island 2 groups of support fighters (type unspecified)
Okinawa Island 1 squadron of strategic reconnaissance aircraft (type unspecified)


The main thrust of the plan was a unified and concerted air bombardment concentrated on a single list of targets in the Japanese home islands and coordinated through a unified air command.

Twentieth Air Force

Various schemes for centralizing control of the B-29s under the Joint Chiefs of Staff had been discussed ever since SEXTANT. Actually it was the similarity of this air problem to the traditional naval problem which finally was persuasive. At least, it was this resemblance which persuaded Adm. Ernest J. King to accept the idea of a strategic air force that would be assigned to none of the surface commands, but would report directly to the Joint Chiefs.

In retrospect, the way this significant agreement was reached seems almost trivial. I secured General Arnold's permission to discuss the subject with Admiral King. I found King and Arnold walking together down a corridor leading to the Joint Chiefs of Staff conference room. I asked Admiral King if I might have a word with him. I described briefly the problem of concerted command and control of the long-range bombers that would be attacking common targets in Japan, but would be operating from bases under the command of several separate theater commanders. I suggested a similarity with the problems attendant on control of the U.S. Fleet whose command was centralized under him as Commander in Chief, U.S. Fleet, as well as the Navy member of the Joint Chiefs of Staff. The U.S. fleets in the Pacific could be employed in concerted action against the Japanese naval forces under unified command, wherever the battle area might be and regardless of the geographical areas in which naval bases might lie. The B-29s had a like requirement. Would it not be sensible to concentrate the very long-range bombers arrayed against Japan in a strategic air force under General Arnold, Commanding General of the Army Air Forces? Unified operations against targets in Japan could be assured, notwithstanding the geographical areas in which the B-29 bases might be located. Under this


arrangement, the B-29s would in fact fall under the control of the Joint Chiefs of Staff, with General Arnold serving not only as Commanding General, but as executive agent for the Joint Chiefs. The Joint Chiefs would furnish unified strategic air objectives. As in the case of logistic support for fleet units of the Navy, such support could be provided the B-29s through directive to appropriate area and theater commanders. Admiral King reflected for a moment and said, "I could find such an arrangement acceptable."

I prepared a brief memorandum to that effect, discussed it with Generals Kuter and Fairchild, and took it over to the War Department Operations Division. The struggle was only half won because General Marshall was dedicated to the concept of theater unity of command. He had even forced through the first Allied unified command, the ill-fated organization under British Gen. Archibald P. Wavell. This command embraced all the forces in a specified area, and was formally called the Australian-British-Dutch-American Command. Set up in January 1942 at General Marshall's insistence, it operated until the fall of the Netherlands East Indies to the Japanese in February. I gave the paper to Maj. Gen. Thomas T. Handy, Marshall's Deputy for Plans and Operations. Tom Handy was one of the finest and most able officers with whom I have been associated. He had General Marshall's great integrity and intellectual grasp, coupled with a fine sense of humor. He accepted the paper, read it carefully, and looked at me. "I'll tell you the truth, Hansell," he said, "I don't like any part of this paper. It violates the principle of unity of command in a theater of war. It inserts operational forces into a commander's area of responsibility but gives him no control of those forces. At the same time, the theater commander is expected to defend and supply and support those forces in competition with his own requirements. I don't like it." Then he grinned and said, "But I don't have a better solution. I'll buy it." I said, "Do you think General Marshall will buy it?" He replied, "General Marshall isn't here. But I know how he'll react. If General Arnold and Admiral King are agreed on it, he'll go along. As a matter of fact, I'll approve it in his name." That memorandum served as the basis for a paper presented to the


Joint Chiefs of Staff by the Joint Plans Committee, of which I was the air member.

The Twentieth Air Force was born on April 4, 1944, with General Arnold its first commander. It was designated the Twentieth to distinguish it from the other numbered air forces. On April 10 the Joint Chiefs of Staff accepted the Joint Plans Committee paper, and it constituted the formal charter under which the Twentieth operated. I have always believed the wartime establishment of the Twentieth Air Force was one of the most important events in United States Air Force history. If it had not occurred, we might be still parceling out our big punch in penny packets to numerous theaters and lower commands. And there might have been no United States Air Force.

The Joint Chiefs, in approving the creation of the Twentieth Air Force, directed:

Commanders of the Theaters in which the Twentieth Air Force's XXth and XXIst Bomber Commands are based are directed to coordinate B-29 operations with other air operations in their Theaters, to construct and defend B-29 bases, and to provide logistical support and common administrative control of the B-29 forces. Should strategic or tactical emergencies arise requiring the use of the B-29 forces for purposes other than the missions assigned to them by the Joint Chiefs, Theater commanders are authorized to use the B-29 forces, immediately informing the Joint Chiefs of such action.

Admiral King's endorsement was vital because the bulk of the bombers would be in the Pacific Ocean Area, a naval command. General Marshall, with his typical breadth of vision, gave full support to the project.

The British Chiefs of Staff Committee plans called for participation of British bombers in the final air offensive against Japan. Also, the change in command relations would affect the Supreme Allied Commander, Southeast Asia, who operated under the Combined Chiefs of Staff with the British Chiefs of Staff Committee as executive agents.

The British Chiefs countered with the proposal that the air


offensive, including the Twentieth Air Force and later a British bomber contingent, be placed under the control of the Combined Chiefs of Staff rather than the U.S. Joint Chiefs of Staff. This would parallel the command relationship in the European Combined Bomber Offensive in which Air Chief Marshal Portal had been the executive agent of the Combined Chiefs. The Joint Chiefs demurred: The CBO in Europe was a British-American venture, while the air offensive against Japan was almost completely an American one. The British did not press the issue.

When General Arnold assumed command of the Twentieth Air Force and I became its first Chief of Staff, it was apparent we needed a staff for the new organization. But Arnold already had a staff -- a large one -- the Air Staff, or Headquarters Army Air Forces. He was loath to increase the "overhead" by creating yet another staff. I suggested to General Arnold that the Air Staff meet the needs of the Twentieth Air Force. He agreed somewhat reluctantly. The Air Staff was part of AAF Headquarters, the parent for all the numbered air forces. To single out the Twentieth as the special concern of the Air Staff as an operational headquarters for a combat air command -- would seem to slight all the others. But General Arnold did not want another headquarters staff in Washington, so he went along with my suggestion. Each of the Assistant Chiefs of Air Staff (Personnel; Intelligence; Training, Materiel, Maintenance, and Distribution; Operations, Commitments, and Requirements; and Plans) was told to wear two hats: one for Headquarters AAF and the other for Headquarters Twentieth Air Force. Each of these Assistant Chiefs selected one senior officer to represent him on Twentieth Air Force matters. General Order No. 1, Twentieth Air Force, assigned Col. Cecil E. Combs to the A-3 Division of the Twentieth as Chief of Combat Operations. I was designated a Deputy Chief of Air Staff as well as Chief of Staff of the Twentieth. On the whole, I thought the scheme worked reasonably well.

In many ways the Twentieth had unique features and problems. There was the need to draw up and approve tables of organization and equipment and to establish tactical doctrine and standing operating procedures. This would (1) permit the handling, control, and


coordination of many aircraft and units, and (2) provide a basis for uniform training.

Air Staff members who served as principal staff officers for the Twentieth Air Force while I was the Twentieth's Chief of Staff and Deputy Chief of Air Staff included:

Brig Gen. John H. McCormick A-1 (Personnel)
Col. Woodbury M. Burgess A-2 (Intelligence)
Col. Cecil E. Combs A-3 (Combat Operations)
Col. William F. McKee A-3 (Operations, Commitments, and Requirements)
Col. Llewellyn O. Ryan A-3 (Training)
Lt. Col. John W. Carpenter A-3 (Crew Training)
Col. Samuel R. Brentnall A-4 (Materiel, Maintenance, and Distribution)
Col. Sol Rosenblatt A-4 (Supply)
Maj. Gen. Laurence S. Kuter A-5 (Plans)
Maj. Gen. David N. W. Grant Surgeon
Brig. Gen. Harold M. McClelland Communications Officer
Col. Max F. Schneider Air Inspector
Col. Guido R. Perera Target Intelligence Officer and representative of the Committee of Operations Analysts

Arrangements had to be made for deployment to overseas bases and for logistic support. Personnel needed to be selected for key assignments. The top ones, of course, required General Arnold's approval, and he selected commanders at his own discretion. I anxiously watched the Materiel Command's progress in correcting a multitude of airplane and engine technical problems. I spent as much time as I could shepherding concepts and ideas through the Joint War Plans Committee, the Joint Plans Committee, and the Joint Chiefs of Staff and discouraging dismemberment of the force. At my request, the Committee of Operations Analysts was rendered responsive to the Chief of Staff of the Twentieth Air Force on matters relating to that command. I had to be ready at a moment's notice to answer General Arnold's questions, so he could be prepared to field those of the other


members of the Joint Chiefs of Staff, who were showing a lively interest in the new Twentieth Air Force.

One of the first challenges facing the Twentieth was communications -- both command-and-control and administrative. The Joint Chiefs had approved our request to set up our own separate communications system. This was a sizable order seeing that we would have units and headquarters scattered over half the world. I briefed General Arnold on our communications requirements and watched with amusement as he applied his famous technique for attacking the impossible. He called in Brig. Gen. Harold M. McClelland, head of communications for the Army Air Forces. General Arnold said with caustic emphasis that he wanted:

A net that would include Washington, Hawaii, the Marianas (which had not yet been captured), Calcutta, India, and Chengtu, China, with provision for extension to somewhere in the Philippines (when they were captured). He wanted TOP SECRET security with instantaneous coding and read-out by teleprinters. He wanted the net in operation twenty-four hours a day.

Fantastic, I thought! It couldn't be done. General McClelland didn't bat an eye. He said "Yes, sir," saluted and departed. It left Arnold a little flat and completely deflated me.

General McClelland had the last chuckle. He produced. Shortly after I had set up headquarters in Guam, (about six months after this conversation) the machinery was working. Within six weeks, I was sick of it. The machine worked twenty-four hours a day all right, without stopping. Most of the messages seemed to consist of questions I could not answer. I began to understand the meaning of the remark ascribed to the English statesman, Lord Palmerston, that the disintegration of the British Empire had begun with the invention of the telegraph.

During one of our daily staff meetings at Headquarters Twentieth Air Force, I got rather upset because supplies were not being provided for at a rate I thought satisfactory. The staff representative for materiel was Sol Rosenblatt, a temporary wartime colonel. I delivered myself of a somewhat intemperate diatribe. I mentioned that the U.S.


Navy always got the best of everything, with plenty to spare, while good fighting people in the Army -- and the Army Air Forces in particular -- made do on a song and a shoestring. The Twentieth Air Force was on its way to becoming the most powerful fighting force in the world, and it deserved the best and we were going to provide it.

I felt that I had expressed myself well and that the point was clearly understood -- and I dismissed the subject from my mind for a while. But intemperance often breeds surprising results. It did so in this case. It was some time before I found out what a fire I had started. Colonel Rosenblatt took me seriously. He used not only my meager name and authority to churn the brew, but he used General Arnold's name and authority as well. I do not suppose we will ever find out the true level of supplies that were ordered for the Twentieth. I think it likely supplies were still being shipped to the Marianas long after the war was over and the troops had come home. Through his efforts I also acquired for a brief time my own personal "fleet" of cargo vessels. But that is another story.

During the period I served as Chief of Staff of the Twentieth Air Force, I had one particular experience which I well remember. The B-29s were coming off the production line, and there was increasing interest and speculation in the aviation press. We tried to keep a tight rein on security. The B-29s were destined for the major air assault on Japan. They would be operating at high altitude, unescorted. If the Japanese learned this and also discovered the salient elements of B-29 performance and defensive firepower, they would try by every means at their disposal to provide defenses against them. Security could be directly equated in terms of mission success or failure and in terms of the lives of American crewmen. We were deeply worried about news leaks. Of course there was wartime censorship, but skilled aviation writers who indulged in speculation could, and often did, hit upon the truth.

We had an exceptionally fine public relations officer in Twentieth Air Force Headquarters. His name was Rex Smith, and he was a wartime colonel in the Army Air Forces. He was a veteran newsman and had been at one time a foreign editor of Newsweek magazine. He came to me with a suggestion and recommendation. He said:


I know my people and my associates. They are as loyal and patriotic as any Americans you can find anywhere. They will respond to a gesture of faith, if they understand the issues and if they are treated fairly and equitably. I suggest we have a general meeting of all the professional aviation writers, tell them the truth, put them on their honor not to divulge, and assure them that, when the news can be released, they will all be told so at exactly the same time so that there will be no "scoops." Let them write their stories and file them with us. We will release them simultaneously at the earliest time that will not jeopardize our mission.

I was somewhat shaken by this bold suggestion. But I had a lot of faith in Rex Smith, and I realized I knew practically nothing about the press and news media. I bought the idea and then obtained General Arnold's agreement.

We had the meeting in a midwestern city. There were several hundred people present. The security arrangements were carefully prepared and carried out. We "spilled the beans" to a degree that left me quaking. But it worked like a charm. Reporters and writers filed their stories. When the first bombs released over Tokyo were still in the air, a message was flashed back to Guam and was automatically relayed to Washington. The President was the first recipient. But almost simultaneously, all the stories and reports Rex Smith had been holding were released to the press and the other news media. Whether this approach would work again, I do not know. But it worked once to perfection and every attendee at the conference proved completely trustworthy.

Original plans called for the Twentieth Air Force to eventually have three or four bomber commands: the XX Bomber Command in China-India; the XXI in the Marianas; the XXII in the Philippines or Formosa or Okinawa; and perhaps the XXIII in Alaska. The Twentieth's total aircraft would be 1,000 to 1,500 operational B-29s and such escort fighters as could be developed or modified.

The decision to concentrate the B-29s under Joint Chiefs of Staff control made possible the development of the concerted bomber offensive against Japan. However, it did not mark the close of the argument from the theater field commanders. They continued their


efforts to gain control of the B-29 units in their areas. Requests from General MacArthur's headquarters were especially insistent. They were coupled with personal letters from General Kenney to General Arnold contending that B-29 operations out of the Marianas against the Japanese home islands were militarily and technically unfeasible.

The XX Bomber Command

The initial operations of the XX Bomber Command in India and China did not go well. Logistic problems had been expected, but the operational tactics were not yielding results even when the B-29s had sufficient gas and bombs to attack their targets. Brig. Gen. Kenneth B. Wolfe was using night operations exclusively. The coke oven targets (prescribed as first priority) did not present good radar images and were not easily seen at night. In consequence, the bombs were not being placed on their targets. As Chief of Staff of the Twentieth Air Force, I prodded General Wolfe to improve bombing results. I requested daylight bombing of the coke ovens in the Mukden area in Manchuria, where Japanese fighter defenses were not very effective. The available B-29 force was deemed by some to be too small to penetrate the air defenses of the Japanese islands themselves. Others of us believed it could be done. As a matter of fact, the XXI did pierce the air defenses of Tokyo in raids from the Marianas later in the year, with only one wing of B-29s -- the same strength available to the XX. General Wolfe vigorously denied that his B-29s could fly in formations in daylight to these targets. He also categorically said B-29s could not reach their targets in daylight in formation from the Marianas. This assessment dealt a real body blow to the operational plans of the XXI. Wolfe was the only air commander having actual experience with the airplane, and he was the real expert and final authority on the technical aspects of the B-29 itself.

I directed Colonel Combs, Chief of Combat Operations for the Twentieth, to conduct practice tests to confirm or refute this contention that the B-29 had insufficient range to operate in formation as required. He went to Eglin Field, Florida, and set up a test run over the Gulf of Mexico simulating the flight from the


Marianas to Tokyo and back. Simulated bombloads of 8,000 pounds were carried as well as full loads of ammunition. Combs could muster but 3 B-29s for the test, but it was run with wartime combat tactics imitated as closely as possible, including the nature of the formation itself. That is to say, the test entailed initial assembly, loose formation en route, climb to 30,000 feet for the bomb run, tight defensive formation in the areas of potential fighter interception, and retention of that formation until beyond the range of enemy fighters, then return to base in loose formation.

The aircraft all returned successfully to the original base, Eglin Field, bat gasoline reserves were admittedly too low. Though the operation was feasible, much remained to be learned about fuel consumption and daylight tactics if large formations were to be flown over those distances. Upon receiving the report of the test, General Wolfe still did not move from the stand he had taken. The B-29 was a magnificent engineering achievement, but it was new and different and it had new engines that we did not fully understand.

The XXI Bomber Command

The Twentieth Air Force was under extreme pressure to perform. One major slip and the critics would have their way -- the Twentieth would have been dismembered and parceled out to the various theaters. An understanding of this tension and pressure is vital to an understanding of the XXI Bomber Command's early struggle to meet its commitments. We had pledged to launch an air offensive against Japan in November 1944. This proposed assault was tied into the carefully prepared plans for the Pacific campaigns of Admiral Nimitz and General MacArthur. The target date had to be met and the success of a highly controversial operation had to be shown, if strategic air power was to reach fruition in the Pacific.

The XXI Bomber Command was activated at Smoky Hill Army Air Field, Salina, Kansas, on March 1, 1944. At that time, the XX Bomber Command was stationed at Kharagpur, India, in the China Burma-India Theater. The 73d Bombardment Wing, originally scheduled for the XX Bomber Command, had been transferred to the XXI


when the XX was reduced from two wings (eight groups) to one wing, the 58th. The XXI Bomber Command was trained and staffed by the Second Air Force. Headquarters of the XXI was later moved from Salina, Kansas, to Peterson Field, Colorado Springs, Colorado. The XXI was to consist of 1,000 B-29s and it had to be given the necessary training.

In late spring 1944, General Arnold told me I was to have command of the XXI Bomber Command destined for the Marianas. My replacement as Chief of Staff of the Twentieth Air Force was Brig. Gen. Lauris "Larry" Norstad. His arrival was delayed because he felt he should "visit the troops" before becoming Chief of Staff, and he insisted on going to the India-China Theater. This took time so I could not get away until August to get a look at training and organize Headquarters XXI Bomber Command. When I finally assumed command of the XXI on August 28, 1944, the units of the 73d Wing were training for night radar bombing, along the pattern of the XX Bomber Command, of which it was to have been a part.

Due to its location, logistic troubles, and relationship to the chief target areas, the XX had been given target priorities different than those of the XXI. The force was thought to be too small to fight its way through the defenses of the Japanese homeland in daylight. And it could reach solely the southern portion of Japan from bases around Chengtu, China. The coke oven targets had proved unsuitable for night radar bombing. Other targets needed to be suited to radar bombing or situated in lightly defended areas. Aside from the coke ovens, this left little of real importance as targets for the XX.

The advice of the Committee of Operations Analysts was sought on the strategic targets of the Twentieth Air Force regardless of basing locations. The committee recommended using the B-29s against merchant shipping, steel production (through coke ovens), urban industrial areas, aircraft plants, the antifriction bearing industry, the electronics industry, and belatedly the petroleum industry. The committee repeated its conviction that the coke oven plants in Manchuria were highly vulnerable to bombing and were vital to Japanese steel production. It further pointed out the extreme vulnerability of Japanese urban areas to incendiary attack.


Memory of the Luftwaffe still fresh in its mind, the Air Staff advocated destruction or neutralization of the Japanese Air Force as an overriding priority for the XXI Bomber Command. The Joint Chiefs of Staff agreed. The aircraft and engine plants assigned as top priority targets to the XXI (based in the Marianas) were precision targets. Thirteen aircraft and engine plants were known to exist in Japan. It was estimated that eight of them turned out seventy percent of Japanese aircraft engines. The towns hosting these factories were known. Even so, the actual plants had not been pinpointed -- a major task for the reconnaissance squadron of the XXI.

We had some general knowledge of the industry. Right after World War I, the Japanese had canvassed European and American aircraft and engine builders and had obtained production licenses. Three major Japanese producers emerged at that time: Nakajima, Mitsubishi, and Kawasaki. They had continued to dominate the Japanese airframe, engine, and propeller business. As the U. S. Strategic Bombing Survey later reported:

While waves of Japanese technicians were studying American factories, America's top engineering schools were training men who, on their return to Japan, were to design the Zero fighter, Betty bomber, and other planes on which the Japanese bid for Pacific domination was to be based.

By 1930, the Japanese Army and Navy had decided the industry should stand on its own feet, and established a policy of self-sufficiency, whereby only aircraft and engines of Japanese designs were to be considered. No more foreign engineers were to be hired. This was intended mainly as a sop to Japanese nationalistic pride, however, and did not prevent their technical missions from continuing to buy the best foreign models as starting points for Japanese designs. In 1935 Nakajima purchased licenses on the early Corsair from Chance Vought Corp., and it acquired designs of the Whirlwind and Cyclone engines from Wright Aeronautical Corp. in 1937. Mitsubishi purchased a French radial engine, which became the basis for their famous Kinsei series, and secured plans for a Curtiss fighter in 1937. Sumitomo Metals bought rights to the American Hamilton Standard and German VDM propellers. Kawasaki secured rights on the German Daimler-Benz engine, from which came the only Japanese liquid-cooled engine of the war. . . .


We knew that Japan had embarked upon a vast and hurried expansion of her military aircraft industry. We knew, for example, that the Japanese government had directed a near-doubling of the aircraft plants in 1941. Japanese newspapers bragged to the world that a great new airframe and assembly plant had been built at Musashino, near Tokyo, and another close to Nagoya was heralded as the second largest in the world. Kawasaki set up immense modern ones near Nagoya. However, the precise location and description of these plants was a mystery to us in the fall of 1944. We recognized that those concentrated in the vicinity of Tokyo, Nagoya, and Kobe would be extremely vital precision targets -- if and when we discovered their precise locations and descriptions.

The aircraft targets could not be found, hit, and destroyed with the radar bombing equipment and the meager information we had. So the units of the XXI Bomber Command required crash retraining to do high-altitude, daylight precision bombing and to fly in formations not yet selected. We had to plan on reconnaissance after we had created a base on Saipan. The airplane and engine factory targets were at the extreme limit of the B-29 radius of action as it was then understood. Formations flying always reduces range, and it made completion of our missions (marginal at best) even more of a problem. In fact, it took several months of actual operation to master the techniques of fuel control that would give the B-29 its design capability.

There was spirited dispute at the time over this change in bombing tactics. The dispute persists, but the reasoning is not hard to trace. Our only real experience in massive bombing operations was over Europe. Had we not learned a painful lesson there? In Europe the whole concept of American air power -- the selection of vital targets on the ground and their destruction through precision bombing -- had faced the possibility of disastrous failure. The ability of massive bomber formations to fight their way through enemy defenses and reach remote targets, without intolerable losses, came dangerously close to being disproved. If the German fighter forces had been left free to expand, the price might have been too high. And if it had been, the air offensive would have failed and with it any hope of surface invasion.


In Europe the bombers of the Eighth and Fifteenth Air Forces were directed against the sources of German fighter development and strength -- aircraft and engine factories, air bases, and sources of aviation fuel. These comprised the targets of the "intermediate objective:" the enemy air force. As soon as possible, the penetration capability of the bomber formations was supplemented by escort fighters. This experience in Europe obviously weighed heavily in establishing target systems in Japan. The aircraft and engine factories and to a lesser degree the oil resources were designated the intermediate objective in the war against Japan. They were to receive first priority in point of time. Dangerous as unescorted daylight missions might be, they had to be undertaken against Japanese aircraft and engine plants lest the Japanese air force grow strong enough to make our missions too costly to sustain.

The other lesson of European air combat could not be applied initially to the Twentieth Air Force. The range of the B-29 prevented escort fighters from accompanying the formations from the Marianas, though steps were taken to secure a very-long-range escort fighter. Until Iwo Jima could be captured and a fighter base set up there, the bombers would be completely on their own. This was really the most controversial point of all. Seasoned experts on every hand assured us the B-29s would simply be shot out of the air. But it was a risk that had to be taken if the strategic purposes were to be achieved. And the B-29s had some factors working for them -- greatly improved defensive firepower and high-altitude performance.

Early in September 1944, I issued orders for converting the 73d Bombardment Wing to daylight tactics, and I established tactical doctrine for daylight operations including a standard formation. Opposition to this change was severe, especially from the 73d Wing. Training was intensive. But training missions from Kansas to Cuba, simulating the mission from Saipan to Japan, left bombers down all over the Gulf States. Meanwhile, the pressure to commit the command to combat rose. Final practice missions were flown. Groups of the 73d flew two long-range missions that stressed takeoff, assembly, rendezvous, formation flying, and simulated frontal weather penetration. Still, it was simply impossible to train bombardiers to an acceptable


precision accuracy in the time remaining. Training would have to be completed in the Pacific.

Capture of the Marianas as a base for B-29 operations stemmed from the Army Air Forces' initiative and insistence. However, the decision was arrived at before the crews had enough flying experience with the B-29 to know what its performance truly was. Early experience in the training areas revealed that the round trip from the Marianas to Tokyo was marginal for the B-29, even on paper and without opposition. Depending on the location of the base and the target, the distance could be as much as 1,550 miles one way. Clearly there would be no land-based escort fighters for the first part of the campaign, prior to the capture of Iwo Jima. Apart from the marginal range of the B-29, nearly 1,500 miles of hostile water separated the Marianas from Tokyo.

When the first units deployed to Saipan six weeks later, the crews had averaged less than a hundred hours of total flaying time in the B-29. The average high-altitude formation flying experience was under twelve hours. Moreover, the B-29's engines developed a mean tendency to swallow valves and catch fire. The magnesium crankcases burned with a fury defying all extinguishing. Besides, gunsighting blisters were either blowing out at high altitude or frosting up so badly that the gunners could not see through them. But there was not time to fix them properly.

The burning out of exhaust valves was finally solved by fitting a goosenecked pipe that sprayed cool air directly on the valve housing, and by putting cuffs on the props to pump more air through the engine cowling: Oil flow through the exhaust valve housing was also improved. The other problems -- frosting of panes in the cockpit and of plastic bubbles at gunners' scanning stations -- were solved by running hot air hoses to the affected areas. With the cockpit blanked out, it obviously would have been impossible to keep formation. And with the scanning bubbles clouded, the gunners could not see to shoot. These problems were solved at literally the eleventh hour.

I requested that the units of the XXI Bomber Command be flown in squadron formation to Saipan under Air Transport Command control. This would let them get needed experience flying in formation


over considerable distances. The request was denied on the ground that the B-29 had not the range to fly in formation the 2,400 miles from Sacramento to Hawaii. The flight would have been without a bombload, in the face of no opposition, and with excellent communications, weather reporting, and base facilities. These same units, on arrival in Saipan, were faced with a round trip of about 3,000 miles, with bombloads, in the face of expected enemy opposition, and with no weather data or communications.

Two bases, each with two 8500-foot paved runways and 80 hardstands, necessary shops, housing, fueling facilities, and other essentials were supposed to be ready on Saipan. The bases were to have been built by the Central Pacific Area Command, but stubborn interference by the Japanese garrisons in the Pacific and competition from U.S. Navy construction work had set the schedule back by several months.

I paid a departing visit to General Arnold and General Marshall in Washington in mid-September. In response to his inquiry, I assured General Marshall we would carry out our pledge to attack Japan in November. Departing on October 5, I took the first B-29 to the Marianas and started the flow which ultimately grew massive. I flew with the crew from the 73d Wing, the aircraft commander being a bright and capable young major named Jack J. Catton. Catton and I alternated in the pilot position; I took it from Sacramento to Hawaii; he took it to Kwajalein; and I flew the last lap to Saipan. We took off from Mather Field near Sacramento. The original design gross weight of the B-29 was 120,000 pounds. Wright Field reluctantly permitted an overload weight to 128,000 pounds. With our spare engine in the bomb bay and the various kits we carried, we weighed in at about 130,000 pounds.

When we reached Hickam Field in Honolulu, Admiral Nimitz, Commander in Chief, Pacific Ocean Area, greeted us as did Lt. Gen. Millard F. "Miff" Harmon, Deputy Commanding General for Administration and Logistics, Twentieth Air Force. General Arnold retained direct control for operations as Commanding General of the Twentieth as well as executive agent for the Joint Chiefs of Staff. I asked Admiral Nimitz and General Harmon for an opportunity to discuss


my mission and my requirements. Meeting with Admiral Nimitz the next morning, I explained my mission and its peculiar command relationships. That is, I would be completely dependent on him for construction of bases in the Marianas, for movement and delivery of all supplies by surface transportation, and for defense of the bases. At the same time, I would be independent of his authority in operational matters, except for serious emergencies.

This command relationship had apparently not been clearly spelled out or explained to Admiral Nimitz, because he expressed surprise on some counts. I, for my part, was surprised too. General Kuter had made a trip to Hawaii in late March or April for the express purpose of describing these relationships. On his return, a staff unit, headed by Maj. Gen. Walter H. "Tony" Frank, in a followup visit, had spent a week reaching accord on the principal details. Colonel Combs represented Headquarters Twentieth Air Force on this mission. The results were favorable indeed.

Thinking the Joint Chiefs-approved command relationship had been explained, perhaps I was undiplomatic in presenting my understanding of it. Fortunately I had the forethought to bring a copy of the Joint Chiefs of Staff agreement on the subject, which I produced. Admiral Nimitz studied it intently and said:

I must say to you that I am in strong disagreement with these arrangements. If I had been aware of their extent I would have expressed this disagreement to Admiral King and the; Joint Chiefs. I command all of US forces in the Pacific Ocean Area. This is an abrogation of the chain of command. However this is the decision of the Joint Chiefs of Staff, and I say to you again, I will give you all the help and cooperation in my capability. You have my very best wishes for success.

He was a good as his word. I had good reason to be grateful for his continued support. But he ended on an ominous note. He said: "You are probably in for a rough time. You are going out to the Forward Area where my commander, Vice Admiral John Hoover, breaks my admirals and throws them overboard without the slightest compunction. God knows what he is going to do to you." When I got


to know Admiral Nimitz better, I recognized the vein of merry humor that flowed beneath a sometimes stern visage. But at the time I was somewhat shaken.

When I finally arrived at Saipan, I called upon Vice Adm. John H. Hoover, Commander of the Forward Area. He could not have been more cordial and helpful. I kept him constantly advised of my problems and progress, and I enjoyed his strong support. I made it a point to follow his advice and suggestions whenever I felt I could.

My movement to Saipan had been conducted in supposed secrecy. When I left my wife in San Antonio to return overseas, I did not even tell her which ocean I was going over -- Atlantic or Pacific. The night of our arrival in Saipan, "Tokyo Rose" broadcast a welcome to Saipan for "General 'Possum' Hansell" over the Japanese radio network. This may have seemed amusing to many, but to one it was almost tragic. Col. Richard H. Carmichael, commander of a B-29 group operating from Chengtu, was shot down over Japan and captured. He was hauled before a Japanese investigator who demanded to know why I was called "Possum." When he professed ignorance, he was beaten unmercifully. This went on for days, until the Japanese finally concluded he really didn't know -- which was all too painfully true. Years later at a cocktail party, he asked my wife, Dotta, why I was called "Possum." She said she had found an old prep-school annual bearing my likeness at age thirteen with the explanation, "He is called 'Possum' because he looks like one." I have been steadfast in a minority dissent on this report ever since, to no avail.

A survey of conditions on Saipan caused dismay. Of the two bases under construction, one could not be used at all by B-29s. The other had one runway 7,000 feet long (5,000 feet of it paved), a taxiway at one end only, about 40 hardstands, and no other facilities whatever except for a bomb dump and a vehicle park with gasoline trucktrailers. It was hardly ready to receive the 12,000 men and 180 aircraft of the 73d Wing. Ground crews put up borrowed tents in what was surely one of the most disorderly military encampments of the war, but they worked day and night to meet the demands for the first strike.


Table of Contents ** Previous Chapter (3) * Next Chapter (5)


1. The expression "occupation," as distinct from "invasion," was deliberately inserted by the airmen in SEXTANT discussions.

2. Brig. Gen. Claire L. Chennault was commanding the newly created AAF Fourteenth Air Force, stationed at various airfields in China. At the same time, he was air advisor to Chiang Kai-shek and Chief of Staff for Air of the Chinese Air Force.

3. These groups, making up the 58th Bombardment Wing of the XX Bomber Command, were later transferred to the Marianas.

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