Page 470

Chapter 16: Cutting the Enemy’s Lifeline

As the Philippines campaign progressed toward a successful conclusion, Kenney’s airmen, still heavily engaged in tactical support of ground and naval forces, assumed new responsibilities that were more nearly those of a strategic air force. The island of Formosa, from which the Japanese had launched their first attack on Clark Field in December 1941, was a key position in the Japanese Empire. Situated halfway between Japan and the southern extent of its military conquests, the island was the principal way-station along the routes leading from the homeland to the Netherlands East Indies and Malaya. In addition to guarding vital sea lanes, Formosa served as a staging and supply base for outlying garrisons and, from its own economic resources, contributed importantly to the maintenance of Japan’s home front. To protect U.S. forces in the Philippines and Ryukyus, it was necessary to neutralize the many Formosa airfields, but the sustained air attack launched in 1945 served an even larger purpose and was joined with efforts to interdict Japanese shipping in the South China Sea. In time, targets on the China coast also came under attack.

That purpose was to speed the enemy’s collapse by cutting his lifeline. In retrospect, the contribution to the achievement of that end made by the Fifth Air Force does not loom so large as it did in the eyes of its commanders at the time. It is now clear that submarines of the U.S. Navy already had gone far toward choking off the sustenance received by Japan from her southern conquests by the time the Fifth Air Force was in position to render major assistance, and the underwater blockade of Japan was to continue with increasing effectiveness.1 By the spring of 1945, moreover, the B-29’s of the Twentieth Air Force were adding their own significant contribution to the blockade of Japan through highly effective mine-laying operations

Page 471

in Japanese home waters. At the same time, the B-29’s were more heavily engaged in their primary mission, the destruction of Japanese industrial plants and cities. Blockade and bombardment were complementary but imperfectly coordinated methods of attacking the enemy’s war production: in some industries it was lack of raw materials, in others loss of factories, that led to a decline, and in some instances there was a duplication of effort.

Much of the postwar debate over the relative credit for Japan’s defeat that should be assigned to the several arms and services is beside the main point. That point, clearly, is that the United States and its allies, while destroying a major foe in Europe, had at the same time assembled and deployed in the Pacific air, land, and sea forces of such magnitude as to leave the Japanese no escape from an early defeat. And among those forces none was more experienced or battle-wise than Whitehead’s Fifth Air Force, which in assuming its new responsibilities demonstrated once again its own and air power’s extraordinary flexibility.

Formosa

With the single exception of Manchuria, Formosa was the most highly developed of the Japanese possessions. An island 249 miles long, it enjoyed the advantages of a subtropical climate for the production of sugar cane and its by-product alcohol. Many of its sugar refineries had been converted during the war to production of butanol, a hydrocarbon used in the manufacture of aviation gasoline. Electrical power plants in the mountainous backbone of the island were tied to the production of perhaps 10 per cent of Japan’s aluminum. Iron, copper, and salt, together with oil refining, rounded out the island’s industrial contribution to the Empire. Takao, Formosa’s leading city, boasted well-developed port facilities, as did Kiirun in the north. Two main railway lines, on either side of the central mountains, connected the north and south.2 Along the eastern and western coastal plains the Japanese had developed an airdrome system superior to anything found outside the home islands: photographic intelligence in the spring of 1945 showed a complex of some fifty strips, of which the chief were located at Heito, Tainan, Okayama, Matsuyama, and Takao.3

By that time, of course, Formosa had become an old target for U.S. planes. Though in December 1941 the AAF had been frustrated in its

Page 472

Formosa

Formosa

Balikpapan Attack by 
Thirteenth Air Force B-24’s

Balikpapan Attack by Thirteenth Air Force B-24’s

Balikpapan Damage to 
Cracking Plant

Balikpapan Damage to Cracking Plant

B-25’s of the 345th 
Bombardment Group Sink Frigate Near Amoy, 6 April 1945

B-25’s of the 345th Bombardment Group Sink Frigate Near Amoy, 6 April 1945

Page 473

plan to meet Japanese aggression with a counterattack on Formosa from Clark Field* and soon had been denied bases within reach of Formosa by the rapid development of Japan’s southward thrust, Chennault’s Fourteenth Air Force from its Chinese bases had begun photo-reconnaissance of the island early in 1943 and followed with a series of harassing attacks before the year was out. With the loss of the east China bases during the latter half of 1944, these attacks had to end, but in the autumn of that year both XX Bomber Command and Halsey’s Third Fleet struck hard at Formosa targets. In support of the landings on Luzon, the two forces had again joined in attacks on Formosa, and the Navy had carried its effort with devastating effect into the South China Sea.†

Effective as was Hakey’s neutralization of the enemy’s capacity to interfere with the Luzon landing, there remained the task of keeping Japan’s Formosa-based forces under control – a task best suited to the capabilities of land-based aviation. The assignment belonged naturally to the Fifth Air Force as the Sixth Army’s partner in the seizure of Luzon, an operation which had found part of its justification in the assumption that the Fifth Air Force would cover Formosa in conjunction with the April invasion of Okinawa. Fifth Air Force leaders, who had been restrained in their use of heavy bombers out of concern for the protection of the Filipino people, responded vigorously to the opportunity.4 Because of shortages of shipping, engineers, and materiel, GHQ refused to indorse AAF plans for the development of heavy bomber bases north of Clark Field,5 and only two of Whitehead’s four heavy groups would be based at Clark in time, but a successful campaign against Formosa did not depend upon acquiring the new bases.

The campaign began in a small way in January, when most of the bombers were still flying from Tacloban and a few from the new bases on Mindoro. On the night of I1 January the first mission was flown to Heito from Tacloban by two of the 63rd Squadron’s Seahawks and an H2 XB-24.‡ One plane turned back with an engine failure, but the other two bombed the storage, fuel dump, and administration areas, starting fires visible thirty miles away. Two more planes were back over Heito on the night of 12 January, and the

* See Vol. I, 203-12.

† See above, p. 415.

‡ Equipped with high-altitude radar bombsight devices, ten H2 XB-24’s had been attached to the 63rd Squadron on 7 January for traiaing in night missions.

Page 474

next night 3 planes dropped 500-pounders and incendiary clusters.6 Okayama airdrome was the target for the period from 14 to 21 January. On 16/17 January a trained crew in a specially equipped B-24 ran the first of a series of antiradar reconnaissance, or ferret, missions over Formosa.7 While on 21 January Fifth Air Force fighters opened up on Formosa with a fighter sweep from Luzon and the heavy bombers began daylight bombardment on 22 January, the night bombers switched their attention to the city of Takao and its Nippon Aluminum Company. Airdromes were again the primary target on the nights of the 27th and the 29th.8

Though the missions were still flown in small force, the hazards were great, as may be shown by the experience of one of the three 63rd Squadron planes over Takao on the night of 30 January. First Lt. Albert J. Goossens’ plane, second over the target, was picked up and held by searchlights in its bomb run over the tank farm. The bombs hit the target, starting four fires which soon merged into ane huge blaze, but no sooner had the load cleared the plane than an antiaircraft shell exploded in the open bomb bay. The doors had just been closed when a second shell tore through them, and another ripped the cowling off No. 3 engine. Then, just as the damaged plane began to pull away from the target, it was jumped by seven Japanese night fighters attacking from the rear and sides. The tail gunner, S/Sgt. Charles F. Trusty, exploded one plane before he was wounded by a 20-mm. shell. S/Sgt. Bruce H’. Willingham, one of the waist gunners, dragged Sergeant Trusty into the waist, took over the tail turret, and exploded a second Jap plane. The other waist gunner, S/Sgt. Willard W.’ Ogle, manned both waist guns until he’ was wounded in the arm. Sergeant Willingham came back and dressed the wound, then returned to the tail turret where he scored hits on a third Japanese night fighter. Meanwhile, Lieutenant Goossens had taken the plane down to the protection of a cloud layer at 1,000 feet, but just before entering it, a last Japanese burst hit the No. 4 engine. Although the engineer shut off the fuel supply to the engine, it was leaking oil and would not feather. The hydraulic system and automatic pilot had been shot out; there were holes in both wing flaps and in the left vertical stabilizer; the main gas tank in the right wing had been punctured; and although the crew did not yet know it, the left landing-wheel tire had been punctured and the brakes shot out. The plane was safe from fighters in the cloud layer, but the windmilling

Page 475

No. 4 engine, now without oil, began to heat up and fire broke out on the right wing. Fortunately, the engine heat melted the nose section and the entire propeller section spun off the shaft clear of the plane; after about twenty minutes the fire burned itself out. Contacting the newly captured Lingayen airfield, the pilot received permission to land but had to wait out a Japanese air raid. After cruising off the coast, the crew spent another thirty minutes manually lowering the damaged flaps and landing gear. Guided in by a searchlight beacon, they discovered the damage to the landing gear when the wheels touched and the plane slewed sharply to the left with one wing clipping the tails off two parked B-25’s. Lieutenant Goossens and 2nd Lt. Charles D. Phippen, the co-pilot, managed to strong-arm the plane back on the runway and into the sand at its end. After the two wounded gunners were sent to the hospital, the rest of the crew, as the squadron historian reported, sat down with hot coffee to congratulate their pilot and their ship, and to contemplate the solid earth beneath them.9

For the next two weeks airfields remained the primary target. Whenever the undercast was too thick over the airdromes, the night bombers would return for radar or ETA bomb runs on Takao, where the extensive development almost guaranteed damage to the Japanese. Kagi, Okayama, Tainan, Heito, and Reigaryo near Takao were hit repeatedly. On 12 February the 63rd Squadron was returned to its favorite target – Japanese shipping. Later its planes occasionally undertook nightly harassing attacks on Formosa airfields, chiefly during the invasion of Okinawa. The 90th Group’s H2XB-24’s were on their own after 3 February. They continued night missions against Formosa targets but were also increasingly used as pathfinders on daylight missions. Toward the end of February more H2XB-24’s arrived in the theater and were assigned to fill the pathfinder role in the 43rd, 22nd, and later the 380th Bombardment Groups. In addition to their pathfinding in bad weather and nightly harassing of the enemy, these bombers provided valuable weather information for the daylight attacks.10

The opening of the daylight attacks on Formosa had been beset with difficulty. Whitehead had originally planned to open the attack on 16 January with two groups of B-24’s and one group of B-25’s covered by P-38’s ,but bad weather and slow progress on the Mindoro airfields plus the demands of Luzon ground support forced cancellation

Page 476

of both this attack and one scheduled for 19 January. A similar attack was planned for 21 January with the 22nd Bombardment Group’s Liberators flying from their new base on Samar under cover of P-38’s from Mindoro. The Liberator part of the mission was scratched, however, when on take-off the plane of Col. Richard W. Robinson, the 26-year-old commanding officer of the 22nd, hit a parked Corsair and crashed at the end of the runway killing all the crew.11 This was a sore loss, for Robinson was one of the finest combat commanders developed in the Fifth Air Force. The 8th and 49th Fighter Groups sent their nearly eighty P-38’s on for a sweep of southern Formosa, where they met no opposition whatsoever. The Third Fleet’s carriers had struck earlier that morning, and the P-38 pilots brought back convincing testimony to the effectiveness of this attack.

Finally, on 22 January the first heavy bomber daylight attack on Formosa was carried out by the 22nd Bombardment Group. Covered by forty-nine P-38’s, again from the 8th and 49th Groups, the bombers dropped over one hundred 1,000-pound bombs on Heito air base with good effect. Although antiaircraft fire was heavy, only one plane was damaged and there was no interception.12 Still, the reduction of Corregidor* kept attacks on Formosa down to occasional strikes by one or two groups of the heavies. After the airborne landing on Corregidor on 17 February the way was clear for a sustained assault on Formosa. In the development of that assault the heavies were joined by mediums and fighters, with the heavies taking care of the better defended targets – the major airfields, towns, and industrial plants. Medium bombers and fighters were assigned the smaller and less well protected airfields and isolated industrial plants. Bad weather often protected the northern end of the island, but southern Formosa took a heavy beating.

Throughout the campaign Japanese airfields continued to receive the major share of the Fifth Air Force attack. The second daylight heavy bomber attack had been run on 29 January from McGuire Field on Mindoro† by the newly emplaced 90th Bombardment Group

* See above, pp. 430-34.

† The field was first named for Col. Gwen G. Atkinson of the 58th Fighter Group who was shot down over Luzon, but when guerrillas returned him alive and healthy, the newly constructed field on Mindoro was rededicated McGuire Field in honor of Maj. Thomas B. McGuire, Jr., of the 49th Fighter Group. McGuire, one of the best-liked and -respected pilots in the Fifth Air Force, was within two planes of Bong’s record of forty enemy planes destroyed before he himself was lost in action.

Page 477

which put eighteen of the “Jolly Roger” Liberators over Heito. Fifteen to twenty enemy fighters from Takao were driven off by the escorting P-38’s before any interception was attempted, but flak damaged six bombers, The 90th Group returned to Formosa on 31 January, on 1 february, and again on 7 February.13 While the heavies concentrated on Corregidor, the 38th Group’s B-25’s gave the Japanese at the Kagi airdrome a taste of low-level bombing and strafing on 13 February. With their commitments on Luzon eased, two heavy groups were slated for Heito on the 17th but bad weather diverted them to Takao. The 380th Bombardment Group had just completed its move from Darwin to Mindoro, and it joined the parade to Formosa on 18 February, when Brig. Gen. Jarred V. Crabb staged all four of V Bomber Command’s heavy bombardment groups and one medium group for a full-strength strike. Three heavy groups had Okayama as a primary target, but only one bombed there while the others hit Takao. Twenty-five B-25’s of the 38th Group made a very destructive low-level attack at Koshun. The next day clouds again protected Heito so that two of the three heavy groups hit Koshun and Takao instead. The 22nd Group was badly off target in its bombing of a new landing strip south of Heito.14

Missions to Formosa were run on a smaller scale for the rest of February while the main weight went to ground support in the Fort Stotsenburg, Ipo dam, and Balete Pass areas of Luzon. On 2 March, however, the Fifth Air Force returned in strength to attack Formosa airdromes. Twenty-four Liberators of the 90th and four of the 380th went to Matsuyama to unload 500-pound frags on dispersal and landing areas; 35 Mitchells of the 345th Bombardment Group dropped 23-pound parademos and strafed Toyohara airdrome; and 36 planes of the 38th gave the same treatment to Taichu. The A-20’s of the 312th Group, in their first appearance over Formosa, could not find Kagi airdrome but hit a small drome at Shirakawa and other targets of opportunity, including warehouses, locomotives and boxcars, two bridges, and one truck. Heavy weather interfered on 3 March: the 90th Group bombed Kiirun harbor by radar; the 22nd Group found Tainan airdrome cloud-covered but visible from the north, so that bombs had to be released on an estimated bomb run with unobserved results. Finding Kagi (the primary) and Hosan (the secondary) closed in, mediums of the 38th Group hit the tertiary, Basco drome on Batan Island.15

Page 478

Poor weather and ground support missions kept Formosa airdromes safe until 16 March when eighty-six B-24’s from all four of the heavy groups split up the job of plastering the enemy’s air bases. The next day, four groups put out 70 Liberators to blanket Taichu, Toyohara, Shinchiku, and Tainan airdromes with 500-pounders. The cloud cover made radar runs necessary for almost all of the bombing and forced a scratching of planned low-level attacks. Forty-six heavies followed up on 18 March at Tainan, Koshun, and Toko. Poor weather over other airdromes resulted in Tainan and its airdrome receiving the undivided attention of seventy-seven Liberators on 22 March. Tainan again on the 28th was bombed by twenty-four planes diverted from their primary, Okayama. On the last day of March twenty-three heavies of the 22nd Group blanketed Matsuyama with twenty-pound fragmentation bombs.16

The estimate of Japanese planes on Formosa airfields had dropped from 601 on 14 January to 375 by 1 April. But with the invasion of Okinawa scheduled for that date, the Fifth Air Force had continued to hit the fields regularly both day and night, and though most strikes after 1 April were made in smaller force than were the big raids of March, enemy air continued to be the target of over 50 per cent of the effort devoted to Formosa through April. On 1 April Giran airdrome was bombed by thirty Liberators of the 43rd and 22nd Groups; eighteen Mitchells of the 38th Group assisted the heavies by working over Karenko airdrome the same day. On the 3rd Kagi airdrome was attacked by one heavy and one medium group using frag bombs and another group hit Toyohara on 4 April. Three groups were out on the 7th and four on the 8th, despite bad weather. Tainan, Kagi, and Okayama were hit but many of the planes, closed off from their airdrome targets by the thick undercast, dropped on targets of opportunity. In repeat raids which must have been very discouraging to Japanese repair crews, three heavy groups were out on 11, 12, and 14 April bombing Tainan, Takao, Okayama, Kagi, Taichu, and Toyohara airdromes. The pattern was repeated on 15 April with Toyohara, Shinchiku, and Shinshoshi airdromes on the receiving end. Again on the 16th, 3 heavy groups attacked Matsuyama airdrome and aircraft parked in nearby Taihoku with 100-pound frag bombs. The same day the 380th Group hit Giran airdrome while the two medium groups, the 345th and 38th, each sent eighteen B-25’s to work over Nanseiho and Osono airdromes in northern Formosa, strafing targets

Page 479

of opportunity on the return trip down the length of the island. Four Liberator groups on 17 April bombed Taichu, Shinshoshi, Toyohara, and Shinchiku, with 18 “Sun-Setters” (38th Group) dropping 100-pounders on Taito airdrome.17

By 26 April the estimated plane count on Formosa had dropped to eighty-two, of which not more than 20 per cent were thought to be operational on any one day. For some time now there had been few enemy raids on Luzon air bases or on resupply shipping in Lingayen and Subic bays, and interception of U.S. planes over Formosa had become almost a thing of the past. Except for still potent antiaircraft defenses, Fifth Air Force planes enjoyed freedom of the air over Formosa. Consequently, toward the end of March, airfields had been replaced by industrial targets as first priority among Formosa targets.18

During the last half of April FOA forces at Okinawa, beset by kamikaze attacks, demanded heavier attacks on Formosa airfields on the assumption that those fields were the principal source of this new and dangerous form of attack. Fifth Air Force intelligence officers disagreed, arguing that Japanese air power on Formosa had been reduced to such impotency as to require only occasional attention and that the kamikaze attacks against Okinawa shipping were flown from Kyushu. Postwar investigation has proved both parties to have been right. It was true that most of the kamikaze attacks were flown from Kyushu, with approaches which deceived U.S. naval commanders as to the point of origin. But it was also true that at least 20 per cent of them flew from Formosa.19

This confusion of U.S. leaders is explained in part by the enemy’s effective use of dispersal and camouflage on Formosa. By February 1945 it had become evident to Japanese air commanders that the Philippines were lost, and early aerial battles over Formosa had also shown them that an attempt to dispute control of the air would only result in their being quickly defeated. Consequently, they had decided on conservation of their aircraft for kamikaze attacks in order to repel an invasion of either Formosa or Okinawa. In February they had ordered flying discontinued between the hours of 0700 and 1600, when Allied aircraft were most likely to be over Formosa airfields. Although interceptors were used to pace bomber formations for the purpose of furnishing range and altitude information to the antiaircraft guns, aggressive interceptions were permitted only infrequently.20 Planes were widely dispersed, often towed miles from the

Page 480

airstrips and in one case ferried across a river; some planes were even partly dismantled and others, well camouflaged, were parked in scattered villages and towns. Many dummy airplanes and even entire dummy airfields were constructed. The frequent cloud cover over northern Formosa, which often prevented Allied observation, enabled the enemy to move operational planes from airfield to airfield. While Allied intelligence officers were aware of the Japanese dispersal program (a number of missions were run against Japanese planes parked between houses in towns and villages), they did not think it as thorough as it was. At a time when they estimated only 89 planes, the Japanese had approximately 700; the “twenty per cent operational on any one day” estimate was approximately correct.21

As a result of these measures, the Japanese 8th Air Division was able to send from Formosa against Okinawa shipping approximately 240 sorties, of which the Japanese estimated 140 to 170 were successful. The XXIX Naval Air Corps ( a consolidation of remnants of the First and Second Air Fleets) flew 135 suicide missions from Formosa of which 81 were reported as successful. To avoid detection by the naval radar screen and also because of the difficulty of getting more planes together at any one time, the usual number sent out on kamikaze missions was four to six planes accompanied by one or two escorts and reconnaissance planes. The escorts returned unless shot down by U.S. Navy fighters, but the kamikazes were committed to a one-way trip.22

GHQ had directed compliance with the Navy’s request for increased attention to enemy air on Formosa, and the Fifth Air Force itself found reason to retract its original opposition. During May and June photo reconnaissance revealed that planes previously listed as unserviceable had been repaired, that additional dispersal fields were under construction, and that operational planes frequently shifted from one field to another.23 Though the conclusion that some of the suicide attacks originated on Formosa was inescapable, the continuation of group strength missions against such widely dispersed targets, in accordance with Admiral Halsey’s wishes24 was regarded as an unnecessary waste of effort that could be profitably employed elsewhere. Wherever concentrations showed on Formosa airfields, the heavies responded in force, as when eighty-four Liberators on 15 June blanketed the airdrome at Taichu with frag bombs. Meanwhile, LAB and H2X bombers went out nightly in strengths of from four to ten

Page 481

planes to heckle and break up enemy preparations for suicide attacks – an effort which was probably more effective than any other form of attack employed. The great distance from Mindoro and the lack of airfields in northern Luzon, together with the persistently unfavorable weather over northern Formosa, made difficult the maintenance of any closer supervision of enemy activity. Attacks were begun, however, on enemy air bases on the China coast, which the Fifth Air Force believed to be a source of kamikaze attacks, and strikes against Formosa fields were continued. There was a sustained sequence of airdrome strikes, for instance, in the period from 5 through 11 July undertaken in response to a special request from the Navy. By the end of that month the plane count on Formosa had dropped to sixty-six.25

In the early phases of the Formosa campaign, when the main effort of the heavy bombers was directed against airfields, the night bombers regularly searched Formosa harbors for shipping or dropped incendiaries on docks and warehouses. Frequent antishipping sweeps were run along the east and west coasts of Formosa by flights of from six to twelve medium bombers, and fighters on escort duty often made strafing runs along the coast before heading for home. In a low-level attack on shipping in Mako harbor on 4 April twelve B-25’s of the 345th Group claimed destruction of or damage to six merchant vessels. The first daylight heavy bomber strike against harbor installations had been run by twenty-one Liberators of the 90th Group against Takao on 27 February.26 The oil storage tanks and installations of the large Japanese naval base at Mako were blasted and fired with 1,000-pound bombs in missions by B-24’s on 13, 14, and 15 March. On the night of 24/25 March the radar B-24’s stretched their night bombing to include the docks and shipping installations at Kiirun on the northern tip of Formosa. On the next night, while the night bombers heckled Kiirun, nineteen heavies hit the docks at Takao. Kiirun was hit in a daylight attack for the first time by eighteen Liberators on 29 March. Thirteen planes followed up on 31 March, 38 on 5 April, and a force of 102 Liberators on 19 May dropped loads varying from 2,000-pound bombs for buildings, docks, and storage areas to 100-pounders for small shipping in the harbor. That shipping received further attention on 16 June when the heavies dropped 260-pound fragmentation bombs fuzed to explode on contact with the water in a successful test of the theory that small

Page 482

vessels might thus be holed at the water line. By 1 July the damage accomplished at Kiirun seemed to justify shifting the night bombers back to the Mako naval base.27

An example of the many missions flown against smaller installations was that against Koshun on 18 February, when 24 Mitchells dropped 250-pound parademos on warehouses and barracks with excellent results. On 19 February twenty-two Liberators dropped sixty-three tons of frags on similar targets at Koshun, while fifteen heavies dropped frags on toxic-gas storage tanks at Hozan with excellent results. On 31 March twenty-six A-20’s of the 312th Group dropped parademos and napalms on barracks and tents of the Saiatau military camp and gave the area a thorough strafing. On 30 April 42 Liberators dropped 1,000-and 250-pounders on the oil storage tanks at Toshien with very satisfactory fires resulting. Through an undercast on 2 June the Hozan poison-gas storage was treated with 1,000-pounders by 45 Liberators. Twenty Liberators celebrated Independence Day by dropping frags on the barracks areas at Toshien, and on 12 July 26 others used 500-pounders against the oil storage at Toshien.28

Approximately 7 per cent of the Fifth Air Force’s attacks on Formosa was devoted to its railroad system-locomotives, freight and passenger cars, tracks, bridges, tunnels, marshalling yards, and repair shops. Only rarely did the heavies bother with these targets, but medium bombers weathered out of a primary target or fighters which had fulfilled their escort assignment worked to good effect.29 Marshalling and repair yards were frequently attacked by night bombers, and in June even the very heavy B-32 experimented with high-altitude attacks against bridges, unhappily without success.30 Twelve B-25’s in a low-level attack had shown how bridges should be destroyed on 15 February when they knocked out the north end of the Sobun River railroad bridge and the adjacent concrete highway span with 1,000-pounders. But the heaviest damage to the rail system resulted from strafing of rolling stock by fighters.31

The high paint in the attack on rail transportation was reached during the last half of May 1945. Of the approximately 240 low-level B-25 sorties sent against Formosa in that month, over 85 per cent was directed against marshalling yards, railway stations, and bridges. At the same time V Fighter Command kept its planes working over tracks and equipment between stations. The Marines got into the

Page 483

show on 26 May when four PV-1’s of Fleet Air Wing 17* scored rocket hits on the mouth of a railroad tunnel and damaged rails and freight cars south of Koryu. The 49th Fighter Group got 4 locomotives and 8 cars with an estimated 150 to 200 passengers on 27 May, and P-51’s of the 1st Air Commando Group from Laoag destroyed 2 engines and damaged boxcars on the following day.32 By June the Japanese were operating their trains only at night, and desertions of native laborers forced use of military personnel to move even essential civilian freight.33

Among the most damaging attacks on industrial targets were those against the island’s main source of electric power, two plants in the mountains of central Formosa. Four heavy groups were scheduled to hit these plants on 13 March, but heavy cloud cover forced resort to secondary targets by all save sixteen planes of the 90th Group; they bombed by radar with unobserved results except for an encouraging column of black smoke. Ten days later, however, twenty-three Liberators of the 22nd Group dropped ninety-two 1,000-pounders on the penstocks and transformer yard of one plant while fourteen B-24’s of the 43rd Group unloaded fifty-five 2,000-pound bombs on the other. The numerous direct hits on vital points cut off 60 per cent of Formosa’s power for the rest of the war.34 Fifth Air Force attacks had already damaged the power station at Mompaitan, burned out another at Keiko, and hurt steam power plants at Hokobu and Takao. Except for Taichu, none of the principal cities and towns on Formosa had power through the summer of 1945, and Taichu was the only city that did not suffer from loss of water supply through damage to its water and pipe systems. The significance of these attacks is indicated by the fact that three industries, the Japanese Aluminum Manufacturing Company of Takao, the Asahi Electro-Chemical Plant at Takao, and the Kiirun factory of the Taiwan Electrical and Chemical Manufacturing Company, had previously consumed 60 per cent of the electrical power on Formosa.35 It is true that shipping shortages already had cut seriously into the production of these plants by denying them necessary raw materials and that the plants themselves already had been damaged by bombing and were slated for further destruction. But by the late spring of 1945 it was becoming a not uncommon practice for U.S. forces simultaneously to deny the materials

* This unit, awaiting redeployment to the United States, had been assigned to V Bomber Command until movement orders came through.

Page 484

of life to the enemy’s industrial plant, to destroy the plant itself, and to cut off its power.

It was known that the Japanese had had large-scale projects for production of alcohol and butanol (for aviation gas) which depended in part upon the widely scattered sugar mills of Formosa.36 Consequently, air attacks against Formosa sugar mills and alcohol plants were the most persistent of any phase of the Formosa attack except those directed toward neutralization of enemy air power. Not too heavily defended by antiaircraft in the early phases of the campaign, these targets went chiefly to the medium bombers, after eighteen A-20’s of the 312th Group on 25 March 1945 had begun the raids by burning out, among other targets, an alcohol plant at Kyoshito. Four days later 18 of the Havocs flamed the entire area of the sugar refinery at Eiko with napalm and demolition bombs while 16 B-25’s dropped 250-pound parademolition bombs throughout the oil refinery and power plant at Byoritsu.37 A sweep by seventeen Mitchells on 30 March exploded and burned two small factories at Toyohara. Two locomotives and a string of freight cars received a treatment of 250-pound parademos as the formation went on to fire sugar refineries at Kori and Tenshi, as well as a refinery and barracks area at Taichu. A most destructive attack by the Havocs of the 312th Group against the sugar and alcohol plant at Shinei followed, and on 4 April the attack bombers fired the entire factory area at Suan Tau. A sweep on the 11th by thirteen of the 38th Group B-25’s got the Tsan-Bun plant. These targets had proved so explosive that the number of planes assigned to a particular mission was gradually cut down. The sugar and alcohol plant at Hokko was badly damaged on 23 April by nine B-25’s; the plant at Mizukami was considered knocked out of commission by seven Havocs the next day; and on the 25th five A-20’s of the 3rd Group did the same for the Taito sugar refinery. Eighteen B-25’s plastered the Heito sugar refinery on 26 April, and the same number of Mitchells worked over the Koshun alcohol plant and Koshun town with thirteen tons of 23-pound frags on the 28th. The alcohol plant at Taito was hit by five Mitchells on 5 May while six others hit the sugar refinery at Shoka. The Marines joined the attack with four PV-1’s rocketing the butanol plant at Kagi on 11 May. Flying now in two- or three-plane attacks, B-25’s hit the Byoritsu alcohol plant on 13 May, the Kizan plant on the 14th, the Shoka alcohol plant on 15 May, and on 16 May the Mataan, Ensui, and Shinei plants.

Page 485

Through the rest of May 19 plants were hit, and on 3 June two B-25’s hit the Sharoku alcohol plant while 24 Mitchells gave a full treatment to the Getsubi sugar refinery with 500-and 250-pounders.38

The increasing weight and accuracy of Japanese antiaircraft around these plants led to a shift in tactics in June, with the very heavy and heavy bombers participating for the first time. Two B-32’s dropped sixteen 2,000-pounders through clouds on the sugar mill at Taito on 15 June, On the 22nd one of two B-32’s scored with 500-pounders on the alcohol plant at Heito, but the other missed flak positions with 260-pound frags. The same day 34 B-24’s hit the oil refinery and flak positions at Toshien: 3 heavy gun positions were silenced by the frag bombs from 23 planes of the 22nd Group while 11 planes of the 43rd Group scored hits with 1,000-pounders on the cracking plant. From 26 to 30 June three heavy strikes damaged refineries and butanol plants at Tanchi, Keishu, and Heito. In July several single-plane attacks were made by the heavies with little damage observed, the final mission being run on 12 July by the new planes of the 3rd Group as nine A-26’s heavily damaged the Taiharo sugar refinery.39

In all, some thirty sugar refinery-alcohol-butanol plants were attacked – all the known plants in Formosa. The USSBS survey team credited the effort with destruction of at least 75 per cent of the island’s alcohol production. The report of the Japanese Governor-General’s Office listed seventeen plants completely destroyed, nine moderately damaged, and four slightly damaged. Still further reduction of the enemy’s potential supply was attributed to the disruption of rail transportation and a forced conversion from cane to rice crops in an effort to make Formosa self-sufficient in food supplies.40

So many of the significant targets on Formosa were situated in the island’s cities and towns that area bombing was frequently employed. The resulting destruction, it was assumed, not only would reach supplies of military importance and many small industrial units, but would impose upon the enemy, through destruction of housing and municipal services, a serious loss of labor. Such operations were looked upon also as preparatory to later attacks on the Japanese homeland. The missions were used to experiment with different types of bombs and fuzings and with the tactics best suited to a variety of objectives.41 The cities became a favored secondary target for planes weathered out of their primary target, as on 20 February when sixty-three B-25’s

Page 486

bombed and strafed Choshu town. On 24 and again on 26 February, Takao received the attention of a total of 67 heavy bombers and a mixed load of 1,000-pounders, 500-pound general purpose, and 500-pound incendiary bombs. Tainan was virtually written off the target list after a series of attacks in March. On 1 March, 44 Liberators dropped 387 x 500-pound incendiaries squarely on Tainan, which on 12 and 13 March was to receive a total of 84 x 1,000-pounders. On 20 March a formation of 35 B-24’s finished the job with a load of 260-pound frags, 100-pound napalms, and 100-pound incendiaries, burning out the military barracks and housing area in the northern half of the city. Tainan was hit occasionally as a target of opportunity thereafter, but it no longer offered targets for mass raids. Takao, a good or bad weather target, was on the receiving end of 1,000-pounders dropped by radar and carried by 22 B-24’s on 24 March and 24 on the 28th. Shinchiku city got identical treatment from sixteen heavies on the 17th. Koshun town received the unorthodox bomb load of 100-pound frag clusters on 10 April from 23 Liberators weathered out of Tainan airdrome. The Mitchells of the 38th Group began attacks against the smaller towns in April, hitting Kagi, Hokko, and Shoka.42

In these attacks the mediums were often joined by the heavies. On 24 and 26 April and 1 May, a total of 112 heavies weathered out of their primary targets hit 14 smaller towns in attacks of varying strength. Weather over Matsuyama airdrome on 6 May diverted nine heavies to Taihoku for a radar run while smaller numbers hit Koshun, Taito, and Kiirun. The same day twenty-three Liberators of the 22nd Group bombed Kiirun. The dock and warehouse area had been designated as the target, but with cloud cover over that area, the city proper was bombed on a radar run. Also on the 6th, the 2 medium groups sent 53 Mitchells loaded with 250-and 500-pounders to knock out Mato town. When they left almost the entire town was engulfed in flames.43 The 38th and 345th Groups teamed up to burn out the town and sugar refinery at Kari on 10 May, and the next day forty-eight mediums put the finishing touch on Kagi town, already partially destroyed. It was this mission which brought one of the unique experiences of the war to a plane of the 501st Squadron of the 345th Group piloted by Flight Officer William M. Mathews. On the approach over Kobi town and airdrome to Kagi the plane was hit in the nose and right engine by 40-mm. antiaircraft

Page 487

fire. Temporarily out of control, the plane skidded down and to the right onto the Kobi strip but it bounced back into the air, where Flight Officer Mathews regained control in time to join up, on single-engine operation, with planes of the 499th Squadron which had just pulled off the target. Mathews flew the damaged plane back over the Formosa Strait and landed at the emergency strip at Laoag with no serious injury to any of the crew. Examination showed 4 of the nose guns shot out and at least 118 holes in the right engine nacelle.44

On 11 May fifty-six Liberators left Toshien in flames. Twenty-seven B-24’s followed up on 14 May, their crews observing fires in the storage and warehouse areas as they pulled away. Next day eighty-two Liberators were over Shihchiku for excellent bombing, with hits on railroad yards, industrial plants, government buildings, and residential areas. On 22 May eighty-nine heavies divided their attention among Toshien, Okayama, and Koshun. Though the attack of 19 May on Kiirun harbor* had been an effective one, 98 planes unloaded 1,000-pounders on residential sections, warehouses, and dock areas. Memorial Day brought 117 B-24’s loaded with 260-pound fragmentation bombs, against Takao and its antiaircraft defenses, and Taihoku was the target for 114 heavies in another excellent mission on 31 May.45 For the fourth straight day, all four of the heavy groups were airborne with Takao city again the target on 1 June. On 2 June two groups were scheduled for Kiirun, but finding that target cloud-covered, seventeen of the planes hit Takao instead. Again on the 3rd two groups, weathered out of an attack on the Jitsugetsutan power plants, hit Takao, as did eleven other B-24’s originally scheduled for Hozan. Taito was the unlucky town on 5 June, when seventeen B-24’s, weathered out of both their primary and secondary targets, scored heavily on the center section. Meanwhile, the mediums had been busy with the smaller Formosa towns – an effort which ‘continued through 9 July.46 Commitments in support of the Balikpapan landings Slowed both heavy and medium operations against Formosa during June, and in July U.S. Navy requests for airdrome strikes cut into the tonnage available for urban destruction.

Even so, the job had been done thoroughly enough. Out of eleven principal cities, the Governor – General’s Office later reported five almost completely destroyed (Kiirun, Shinchiku, Kagi, Tainan, and Takao), four 50 per cent destroyed (Shoka, Heito, Giran, and Karcnko),

* See above, p. 481.

Page 488

a third of Taihoku lucked out, and only Taichu relatively undamaged. Targets in or around the key cities of Takao, Tainan, Kiirun, Shinchiku, Taichu, and Taihoku had received 3,214 sorties and 8,435 tons of bombs, figures which represented, respectively, 42 and 53 per cent of the total air effort expended by the Fifth Air Force on Formosa. In these and lesser centers 6,100 persons were listed by the Japanese as killed, 435 missing, 3,902 severely wounded, and 5,335 slightly wounded. The Japanese also listed the total number of “sufferers” – presumably those individuals who lost their housing or whose lives were in some other way directly affected – at 277,383. In addition to government buildings and industrial plants, 10,820 buildings were totally destroyed by bombing and 15,965 half destroyed, while 18,371 were completely and 1,162 half burned out.47

At Takao the city was almost obliterated and the wreckage of ships sunk in the harbor’s mouth blocked the channel so effectively that only vessels of less than 100 tons could put into the harbor. At Kiirun the channel was clear but the damage to town and facilities hardly less extensive than that at Takao. Karenko’s harbor was blocked, Mako’s heavily damaged, and many of the smaller harbors seriously hurt. No exact figures exist, but estimates of small boats sunk or badly damaged ran as high as 600, and the loss clearly had cut heavily into coastwise shipping and fishing, this last being a main source of the island’s food.48 Up to January 1945 the Japanese consistently had sent some sixty to seventy ocean-going ships per month into Formosa’s ports; thereafter a rapidly climbing rate of loss sustained by vessels en route to or from Formosa* combined with the destruction of the island’s harbor facilities to cut the enemy’s sea communications with this major outpost. One ship from Japan got into Kiirun during May 1945, and it was the last one until after the war. At Takao, the headquarters charged with unloading, loading, and routing of ocean vessels was dissolved by summer.49

Brig. Gen. Jarred V. Crabb’s V Bomber Command had carried the main burden of attack, flying 87 per cent of the total sorties and dropping over 98 per cent of the bomb tonnage. Among the command’s planes the B-24’s, with over 5,000 sorties, contributed most heavily. The B-25’s flew more than 1,400 sorties; the remainder of just under 200 sorties was divided among the A-20’s, A-26’s, and B-32’s which

* Sixty-three ships of a total tonnage of 24,869 were recorded as sunk in or near Formosa waters. For the AAF’s role in this blockade, see the following pages.

Page 489

were credited with 14 sorties in all. Grand totals showed 7,709 sorties (including fighter sorties), 15,804 tons of bombs dropped, and 62,445 gallons of napalm.50

Although the Japanese had chosen to conserve their aircraft for suicide raids rather than to challenge the U.S. planes in the air, antiaircraft defenses had been the most formidable yet encountered by the Fifth Air Force, except possibly earlier at Rabaul. In overcoming these defenses, which repeatedly proved intense, accurate, and skillfully varied, U.S. aircrews depended heavily upon the experience of flak intelligence officers assigned to command and air force headquarters. The B-25’s of the 91st Photo Reconnaissance Wing flew special antiradar missions on the basis of information supplied by radar ferret missions undertaken by night bombers of the 63rd Squadron. Aerial photography revealed to trained eyes many antiaircraft positions, and it became a practice to assign certain planes on each mission to take out these defenses. Carefully planned approaches and evasive tactics also helped, but flak damages remained relatively high until summer.51

Except for a few fighter missions run in August, the Formosa campaign was completed in July. Beginning on a small scale in January, the campaign reached its peak in May and slacked off somewhat in June because of commitments on the China coast and at Balikpapan. The continuance of these commitments, plus the beginning of Fifth Air Force displacement northward in preparation for the assault on Japan, brought a further decline in July.

The South China Sea

Only in 1945, as Allied air units moved into Philippine bases, did the Far East Air Forces reach positions permitting a sustained attack on Japanese shipping in the South China Sea. Even then other targets claimed priority. The support of various operations designed to round out the victory won in the northward thrust of SWPA forces and the effort to take out Formosa kept FEAF busy well into the summer. But the time an the means were also found for work which gave to the air forces a significant share in closing off Japan’s most vital lifeline.

The Allied Air Forces brought to the new task a variety of experience. After the celebrated Allied air victory of March 1943 in the Bismarck Sea the enemy had followed a policy of keeping large ships out of range of low-level attacks by B-25’s, but the famous strafers had kept in practice on the lesser barges and luggers with which the Japanese

Page 490

transshipped in tactical areas. LAB-24’s with special radar equipment had begun to cover southern Philippine waters in the middle of 1944. Based on Morotai from October 1944, they had joined Navy “Black Cats” in night operations, with PB4Y’s operating by day, to extend the coverage into the central Philippines and to the northeastern coast of Borneo.52 RAAF Catalinas meanwhile had developed great skill in mining enemy waters.

Though the first significant use of aerial mining against the Japanese seems to have come in Tenth Air Force operations against the port of Rangoon in February 1943,* the RAAF Catalinas had been engaged in an increasingly successful mining campaign since April of the same year. The Catalina, a long-range flying boat, could carry 2,000 pounds of mines a distance of 1,000 miles or 4,000 pounds for 750 miles. It was a dependable plane, required no extensive base facilities, and could be refueled through use of naval equipment far in advance of its own base. The RAAF received its general directive for mining operations from the Allied Naval Commander through Kenney; the selection of particular targets and the scale of the operations themselves were left to the discretion of RAAF Command. In addition to occasional operations in tactical support of other Allied forces, the RAAF sought to reach ports of general importance to enemy shipping. Usually, there was no attempt actually to close a port but rather to effect a maximum disruption of shipping in the port and to impose as heavy a burden of minesweeping on the enemy as was possible at the lowest cost. Missions lasting for as long as twenty-four hours were timed to reach the target at night. After an approach under 1,000 feet to avoid radar detection, 2 to 6 planes would make their runs, some of them for the sole purpose of forcing the enemy to sweep 2 or 3 times the area actually mined.53 Having learned their first lessons in operations against Kavieng, the Catalina crews in July 1943 launched from Darwin a sustained program against NEI targets.

Reaching as far out as Soerabaja with the aid of refueling by a Seventh Fleet tender, the Catalinas added Balikpapan to their list in February 1944. In April of that year a third squadron was added to the original two at Darwin in an action fully justified by later enemy testimony to the effectiveness of their operations. According to that testimony,

* Such operations were continued on a small scale and in 1944 expanded with the assistance of XX Bomber Command. See above, pp. 158-59.

Page 491

“the destruction of tankers and delay in oil shipments was particularly serious” from early in 1944.54 Some of the Catalinas followed SWPA forces northward to Morotai and then to the Philippines, From more northern bases it was now possible to mine the ports of Formosa, China, Hainan, and Indo-China. By 1 August 1945 RAAF Command had registered, since April 1943, a total of 1,215 mining sorties and put on target 2,498 mines at a cost of only eleven Catalinas.55

Enemy shipping along the Asiatic coast had first fallen under air attack by Chennault’s Fourteenth Air Force in the early days of the war. He had repeatedly used the opportunity for such an attack as a main point in his arguments for greatly increased U.S. air power in China.* But his force had remained small and dependent on air supply for its logistics. Until well into 1943, moreover, it had been impossible from the Fourteenths west China bases to reach more than a few shipping targets outside the Gulf of Tonkin and the upper Yangtze River. By the fall of 1943 new forward bases and an increase in the force available brought targets from Formosa southward under bombing and mining attack.56 This assault was short-lived, however, for within a year of its inauguration the Japanese army had overrun the forward bases upon which continuation of the offensive depended. The Fourteenth Air Force claim of 596,620 tons of enemy shipping sunk in the course of its limited effort has been correctly considered as an exaggeration,57 but the campaign, and especially the mining operations, was not without effect.58 Had Chennault been able to keep his forward bases, he would have been in position to strengthen the final attempt to cut the enemy’s lifeline.

As it turned out, SWPA planes took over chief responsibility for the job that had been started by Chennault. While Whitehead’s Fifth Air Force directed most of the aerial operations, the total operation was inter-service, inter-theater, and inter-Allied. Seventh Fleet’s PB4Y’s from their Philippine bases flew daylight search missions as far north as Shanghai and covered almost all of the South China Sea. In the more southern reaches of that sea they were supplemented by planes of the Thirteenth Air Force, flying first from Morotai and later from southern Philippine bases. LAB-24’s, belonging to the Fifth, Thirteenth, and Fourteenth Air Forces, searched the seas by night, when RAAF Catalinas also went out to mine the coastal waters of

* See, for example, Vol. IV, 435-36, 442.

Page 492

Formosa, China, and Indo-China. No small part of the air force task was to assemble the intelligence which might guide attacking planes, submarines, or surface forces to the target. Coast-watchers along the China coast and friendly observers in widely scattered ports supplemented the information of enemy movements supplied by submarine or aerial observation and photography.

The LAB-24’s, because of commitments to Formosa targets, did not begin sea searches until 12 February 1945.59 As procedures were perfected, these night bombers usually searched in triangular vectors from their home field and return-going out along one leg of the vector, searching along its base, and then returning along the other leg. One sector had its base in the Formosa Strait, another in the area from Swatow to Hong Kong, another from Hong Kong to Hainan, another Hainan Island, and still others Tonkin Gulf and the Indo-China coast line down to Cape St. Jacques near Saigon. Not all sectors were covered each night. The missions were usually flown in a strength of from three to six planes. Sometimes lucrative convoys were shadowed and attacked successively by bombers taking off from two to four hours apart, but this was permitted only in blind-bombing zones where there were known to be no U.S. submarines. Where friendly submarines might be surfaced, the planes’ task was to shadow the target until morning, when other planes could be directed to a daylight attack. The PBY’s and PB4Y’s of the Seventh Fleet shared the work with the LAB-24’s and carried the main burden of search by day.

The method is well illustrated by attempts to bring to bay an enemy naval force of two battleships, a cruiser, and three destroyers in February 1945, even though the effort failed. These Japanese units had gambled on a long-range forecast of bad weather in undertaking the run from Singapore to Japan. They were picked up between the Anambas and Great Natoena islands at 1340 on 11 February by a submarine, which gave the signal for a coordinated effort by all services in accordance with plans (CRUSADE)agreed upon in anticipation of the attempt. SWPA search planes made contact on 12 February, and thereafter relays of Army and Navy radar-equipped planes tracked the vessels almost continuously. A strike was planned for 1100 on 13 February, when the force would come within range of heavies on Leyte and the mediums and fighters at Mindoro. With forty-eight P-51’s on assignment as fighter escort, the B-24’s of the 90th, 43rd, and 22nd Bombardment Groups and forty B-25’s of the

Page 493

Allied Air Forces SWPA 
– Search sectors April 1945

Allied Air Forces SWPA – Search sectors April 1945

Page 494

345th Group made a perfect rendezvous, but the clouds, virtually unbroken at all altitudes of attack, covered the targets, and to protect submarines also in the hunt, blind bombing had been forbidden. As fuel tanks drained, the planes returned to base.60 Search planes continued to track the enemy ships and another strike was set for the 14th. By this time the target had got beyond the reach of Leyte-based groups and the strike was limited to the 90th Group’s Liberators, the 345th Group’s Mitchells, and a fighter escort. Again, H2X bombing was forbidden and the planes failed to make visual contact because of the weather. And that was the last chance. One submarine had attacked on 13 February, reporting damage to one battleship and one cruiser, but the claim went unconfirmed. AAF planes had shot down one Topsy over the target area on the 13th, and search planes accounted for additional enemy fighters during the period.61 But the Japanese had made good their escape.

Ten Mitchells of the 38th Group had better luck against a convoy of four escort and four merchant vessels on 22 February. Attacking in 2-plane elements, the B-25’s claimed a destroyer sunk and an 8,000-ton freighter left smoking heavily, but these claims have not been officially credited.* In response to a call from one of the nightly LAB-24 search missions, a force of nine B-25’s on 23 February hit a seven-ship convoy in Phanrang Bay. Direct hits were claimed on two escort ships and one freighter; official credit has been given for one submarine chaser sunk. One Mitchell was lost to flak.62

Sailing junks were assumed to be Chinese fishing vessels and were not attacked, but power-driven junks in the open seas were considered to be operating for the Japanese and were attacked whenever bigger game was not found. Ten to fifteen of them were sunk near

* The official credit cited on this and following pages is the listing found in “Japanese Naval and Merchant Losses During World War II,” February 1947: prepared by the Joint Army-Navy Assessment Committee (cited as JANAC). This listing does not show merchant vessels smaller than 500 tons. Ships of larger tonnage and all naval vessels are listed by date of sinking, name, type, tonnage, position of sinking, and agent of sinking. These listings are used as official confirmation of claims by Army, Navy, and Air Force agencies, but in the opinion of the author the committee findings are to some extent prejudiced toward naval claims in doubtful cases. Findings of the Anti-Submarine Warfare Assessment Committee (Office Chief of Naval Operations), for instance, were accepted by the joint Army-Navy committee without further evaluation. Numerous instances occur of submarine sinkings in an area where previous air force attacks had made claims a day or two before. That the JANAC is not above error is indicated on page 82 by a listing of a 6,500-ton tanker sunk on 1 February 1945 at 1° 20′ N, 109° 5′ E and credited to Army aircraft. As near as can be determined this position is twenty-five to fifty miles inland on the western cape of Borneo.

Page 495

Hong Kong on 27 February by a B-25 shipping sweep, and twenty-one B-25’s on a hunt between Hong Kong and Swatow on 1 March sank a 1,500-ton cargo ship. Night bombers got a tanker in the Hainan Strait on the night of 2/3 March; two nights later they sank a submarine chaser in the same area. The B-25’s sank an 887-ton cargo ship in the China Sea and claimed another ship sunk and still others damaged on 5 March. Five days later 12 of the “Air Apaches” found and sank a 5,239-ton tanker in Tourane Bay; they also claimed another smaller cargo vessel sunk and 1 damaged. Night bombers found and sank a Japanese frigate off the China coast on the night of 12/13 March. A combined strike by 22 Mitchells from the Fifth Air Force’s 38th Group and the Thirteenth Air Force’s 42nd Bombardment Group swept the coastal waters from Swatow to Hainan Island on 13 March with confirmed scores of a frigate and a 2,742-ton cargo vessel sunk. On 15 March, 13 B-25’s from the 38th and 345th Groups, sweeping the Hong Kong-Swatow area, scored a direct hit with a 500-pounder amidships on a 4,500-ton cargo vessel. According to the mission report, a secondary explosion broke the ship in two, but the official listing fails to credit the claim. Other planes claimed a direct hit on the stern of a destroyer. Twenty-two planes of the 38th Group, overtaking four freighters and four escorts off Quemoy Island on 20 March, claimed a cargo ship and a destroyer escort sunk, with two more freighters and one escort damaged.* Two of the B-25’s fell to flak and two others headed inland into China for crash landings. Next day, off the Indo-China coast, the B-25’s found a seven-ship convoy covered by eight to ten enemy fighters. One 779-ton cargo ship, an 834-ton tanker, a submarine chaser, and a 2,000-ton repair ship were sunk; in addition, 4 of the enemy fighters were shot down with 2 more listed as probables. One B-25 was lost and enemy fire damaged five others. Also on the 21st, the 38th claimed the sinking of two cargo ships which have not been officially credited to the group.63 One of the night bombers is credited with a 2,857-ton cargo vessel sunk off the Luichow Peninsula on the night of 27/28 March. That same night an unarmed search plane reported a large convoy off the Indo-China coast. A B-24 of the 63rd Squadron, piloted by 2nd Lt. William H. Williams, responded to the call, reaching the target at about

* JANAC credited the mission with one 500-ton cargo vessel and another 1,577-ton ship.

Page 496

1100 the next morning, after a seven-hour flight. In the face of concentrated antiaircraft fire, Lieutenant Williams made his run at 300-foot altitude over a ship he took to be a tanker. Three bombs, strung nicely across the vessel, failed to explode; so Williams turned his now damaged plane back for another run and this time scored two hits which flamed the ship and caused it to list sharply. He got back home to claim a sinking, later credited as a 6,925-ton cargo vessel, but the way home was hard. After pulling off target, the injured B-24 was jumped by two aggressive Oscars. The Japanese fighters killed the co-pilot with a 20-mm. shell on the first pass, got the radar operator on a second, and knocked out the No. 3 engine on their third pass. The top turret gunner sent one of the enemy smoking into the clouds, but the B-24, its electrical and hydraulic systems also gone, was so nose-heavy that it took two men to hold the control wheels back. Over its base at last, the plane, unbraked by its flaps, came in fast for a crash landing at the end of the strip.64 With the aid of continued tracking, thirty-one B-25’s of the 345th Group attacked the same convoy in a position farther north along the Indo-China coast on 29 March. They claimed the sinking of 3 large merchant vessels, 1 small one, 4 destroyers or frigates, and 1 patrol craft; they are credited with 2 frigates and a merchant vessel of 956 tons.* Eleven of the Mitchells were holed by the heavy concentration of ship’s fire. A 2,860-ton tanker (credited to Army aircraft on 29 March as sunk in a position farther north) may have been in this convoy.65 The following night (29/30 March) marked the extension of night bomber operations northward to Shanghai and inland along the Yangtze River in search of shipping. Whitehead had asked permission to extend LAB searches to the Yangtze on 22 March, and CBI’s approval came through promptly. Since the overwater route passed through POA, prior notification of missions had to be sent to the commander of the Fifth Fleet (Admiral Spruance) as well as to Chennault.66 Of the 3 bombers sent on this first mission only 1 found shipping, but it claimed a 1,900-ton merchantman.67

V Fighter Command regularly sent escorts to cover B-25 operations and rescue missions flown by Catalinas. Sometimes the fighters covered B-24’s on search during daylight hours. Scoring regularly on Japanese fighters, the escorts helped cut down Japanese strength to

* Submarines have been credited with a 5542-ton tanker and a frigate sunk along the convoy’s route.

Page 497

such an extent that by April 1945 few enemy convoys had protective covering. As one Japanese convoy commander put it, “When we requested air cover, only American planes showed up.”68 By that time the mounting air attack on Formosa had forced the Japanese First Escort Fleet’s air squadron to withdraw most of its remaining planes to bases along the China coast, and their withdrawal still further to Shanghai and Kyushu was already near completion. This left the convoys largely unprotected from the air.69

Through the Sulu Sea, the Celebes Sea, Makassar Strait, the Java Sea, and off the coast of Borneo, the LAB-24’s of the Thirteenth Air Force were also active. Few large ships operated in these areas, but there was still considerable traffic of small vessels. Thirteenth Air Force planes are officially credited with sinking 3 submarine chasers in March, 1 in April, 1 in June, a small merchant ship in March, a 6,863-ton converted seaplane tender on 30 April, and 2 merchant vessels on 7 May. Though heavily committed to support of southern Philippine operations and Australian landings on Borneo, the Thirteenth Air Force carried out attacks against Japanese airfields throughout Borneo and even mounted extremely long-range B-24 attacks against shipping and airfields at Soerabaja – the first of these being led by Maj. Baylis E. Harris on 19 April.70

Land-based Navy patrol planes, which carried their full share of the job, normally were content to signal the Fifth Air Force on the location of defended convoys and concentrated their own attacks on single ships trying to run the blockade. To discourage Japanese shipping from holing up during the day in harbors too strongly protected for B-25 attack, General Crabb sent his heavy bombers out in a series of missions against harbors. On 31 March 13 Liberators used 500-pounders against shipping in Yulin Harbor on Hainan Island; 2 large merchant vessels were claimed sunk, but there is no official confirmation. Forty-three B-24’s from the 43rd and 22nd Groups carried 1,000-pounders to Hong Kong on 3 April;* three large merchant vessels were claimed sunk as well as many smaller ones, but the official listing is only one 2,750-ton cargo vessel and one 2,172-ton cargo vessel. Twelve direct hits were scored on oil storage tanks. Eleven planes were hit by flak, and both groups were intercepted. Forty-one planes from the same groups repeated on 4 April when 2 hits were claimed on a 50,000-ton vessel being repaired in the Tai Koo dry dock. Bombs

* See below, p. 502.

Page 498

were strung through concentrations of small shipping in the Victoria harbor area and hits on oil storage and the power plant there were claimed. While the heavies worked over Hong Kong, twelve B-25’s attacked the Mako naval base in the Pescadores, where two small tankers were caught tied up on opposite sides of the fueling pier. A direct hit on one tanker exploded it and caused burning oil to be thrown across the pier to the other tanker, which was soon engulfed in flames. Both tankers (658 and 834 tons respectively) are officially credited; 1 smaller ship and 5 barges were also claimed and hits were scored on residences, administrative buildings, and barracks. Accurate AA fire got one plane and damaged four others.71

On 5 April three A-20’s of the 3rd Group joined in with a unique antishipping mission. Col. Richard E. Ellis, group commander, had rigged up extra wing tanks for his short-range A-20’s. At 25 he was a veteran with over 200 combat missions in medium bombers and was the youngest colonel and group commander in FEAF. Tired of routine ground support missions on Luzon, he requested permission to test his long-range A-20’s against Japanese shipping. His suggestion met with a cool reception from Generals Whitehead and Kenney, but unknown to them he had made the same proposition to Col. D. W. Hutchison of the 308th Bombardment Wing at Lingayen, Hutchison agreed to a trial on the next convoy with the understanding that B-25’s precede him and work over the convoy before the A-20’s attacked. When a convoy was reported off Hong Kong on 5 April, the B-25’s took off from Lingayen and were followed after a thirty-minute interval by three A-20’s, led by Colonel Ellis. His two wingmen were the group deputy commander and operations officer. The B-25’s missed the convoy, but Colonel Ellis found it-a fat cargo ship with two escorts. Each wingman attacked one escort while Colonel Ellis attacked the cargo ship, which was hit and sunk in shallow water.* One of the destroyer escorts was left dead in the water, and the other damaged as the A-20’s went home. When he heard of the mission, General Kenney seems hardly to have known whether to reprimand Ellis or pin a medal on him. The dilemma was solved by forbidding him further combat flying and moving him to the job of Assistant Deputy Chief for Operations, FEAF.72

The next day, 6 April, saw one of the most vicious shipping strikes of the entire Fifth Air Force campaign against China Sea shipping.

* A 2,193-ton ship is officially listed as sunk on 5 April.

Page 499

Twenty-four Mitchells of the 345th’s “Air Apaches” went out to attack a convoy reported off Amoy-two frigates and a destroyer of the newest type with a jury-rigged bow. This destroyer by now was familiar and the Air Apaches had a score to settle with it. Photographs had first shown the vessel under repair at Singapore on 10 February. In a sweep of Yulin Harbor on 30 March by the 345th, the ship was again photographed, at which time it had shot down one B-25. Reconnaissance photographs of Hong Kong harbor on 2 and 4 April had shown it there. It had made half the distance to Japan but it got no farther. The first two squadrons attacked the convoy’s two frigates, one of which quickly sank, but the other was still afloat and firing as the third squadron came in. One plane was hit and barely made it back to base on one engine, but three more hits were scored, sinking the second frigate. The fourth squadron went on to the destroyer; two planes of the leading element were hit on the approach but continued their run and scored one direct hit. One went into the sea just beyond the destroyer, but the other made it back to base, though the pilot, co-pilot, and navigator had been wounded by an explosive shell in the cockpit. The third squadron, having finished off the second frigate, now attacked the destroyer. Again the first plane was hit by ack-ack and ditched, but one or more direct hits were scored, and as the group headed for home, the destroyer was burning fiercely. All three claims are officially credited.73

The next target was a twelve- to fifteen-ship convoy found on the night of 5/6 April. The 63rd Squadron, sending out a total of ten planes, shadowed it for three days and nights and claimed one 7,000-ton transport, one destroyer, and damage to a light cruiser. Seventeen Mitchells of the 38th Group failed to locate the convoy on 7 April because of poor weather, and when the convoy was finally out of range, the night bombers returned to their regular searches, concentrating on the Shanghai and Yangtze River area. They are credited with a 901-ton cargo ship off Shanghai on 14 April but had very poor hunting the rest of the month. By this time, however, permission had been granted to bomb certain targets on the Chinese mainland, so that secondary targets (chiefly airdromes) began to be regularly visited at the end of a fruitless sea search.74

With few ship sightings in open water, the heavy and medium bombers concentrated on harbors in China and Indo-China. Twenty Liberators bombed shipping at Saigon with 1,000-pound bombs on

Page 500

both 20 and 21 April,* some direct hits being scored on large merchant ships. Intercepting fighters were driven off on both missions with one claimed destroyed on 20 April. Eight B-24’s were holed. Again on 23 April, twenty B-24’s went to Yulin Harbor on Hainan claiming a number of barges and two freighters. Twenty-five Mitchells, after shipping in the Canton River, turned back on 26 April because of weather, but two days later fifteen planes over the Saigon River claimed four large merchant vessels as probably sunk along with a number of smaller craft. Antiaircraft fire and an enemy fighter shot down three B-25’s and five were holed. Except for searches, few shipping missions were run in May. Night bombers made many attacks in the Shanghai-Yangtze River area during the month, but none of their numerous claims are confirmed. On 13 June an experimental mission, run by 62 Liberators loaded with 55-gallon drums of napalm for the smaller wooden ships clustered in Hong Kong harbor, left the bay a sea of flames.75

The Japanese had tried every stratagem in the book by summer 1945, but the Allied air-sea blockade had cut the enemy’s lifeline. In the long and grueling test chief honors belong to U.S. submarine forces, with the airplane, both U.S. and Australian, Army and Navy, finally giving to the interdiction of shipping through the South China Sea a truly tight effectiveness. By 9 April 1945 Whitehead could report to Kenney: “As of this date the Japanese sea lane to its captured empire from Hong Kong south, is cut. ... While there is some clean-up work remaining to be done, namely small shipping around Hainan Island and along the China Coast, not many targets remain.”76 Statistics on subsequent ship sinkings fully substantiate the prediction. The sea searches continued, but virtually all missions now were briefed for a secondary target in China or Indo-China.

China and Indo-China

In January 1945, General Chamberlin had expressed GHQ’s view that SWPA air units had no commitment for attacks against the Chinese mainland. It soon became evident to Whitehead, however, that it would be difficult to block the China Sea effectively unless enemy air bases along the China coast could be attacked. Not only did these bases shelter the remaining Japanese potential for convoy cover, but ports and harbors heavily defended by antiaircraft guns were being

* See below, p. 502.

Page 501

used for haven, repair, and refueling of convoys in their painful progress along the China and Indo-China coasts. Kenney, at Whitehead’s request, petitioned GHQ for permission to attack these targets in early February. At that time GHQ felt that the pressure of current operations in the Philippines and the requirement for support of POA at Okinawa by attacks on Formosa would strain SWPA logistics to the limit. But a promise was given to request approval from the China Theater at a later date, and by 20 March, after conferences with representatives of the Fourteenth Air Force, the permission was granted for attacks on the Chinese mainland after other commitments had been met.77

FEAF outlined particular target areas agreed upon with China Theater representatives. These areas included hostile air forces, air installations, and air bases along the China coast between Minhow (inclusive) and the Indo-China boundary (inclusive) and along the Indo-China coast between Tourane (exclusive) and Saigon (inclusive). Also listed was the railroad and its bridges between Saigon and Tourane (exclusive). Water-front areas and supply bases at Hong Kong, Saigon, and Canton, as well as other targets specifically designated by Chennault, were listed. It should be noted that many important and valuable targets within these areas were withheld for national interests. The Fourteenth Air Force was to continue attacks on Japanese air north of Foochow and between Tourane and the Indo-China-China border; it also covered the rail lines from Tourane north. This railroad was of particular significance because it was the only remaining means for the Japanese to shift their Malaya-Siam-Indo-China forces north to oppose an Allied counteroffensive in China planned for the summer of 1945. Allied convoy attacks had virtually stopped seaborne movements north of Saigon, but the Japanese were capable of moving men and supplies by sea to Saigon, then by rail to China.78

Whitehead had anticipated this plan: his B-24’s had hit the Canton airdromes White Cloud and Tien-Ho as secondary targets on the nights of 17/18, 18/19, and 19/20 March. On 21 March, two groups sent thirty-seven Liberators with fighter cover to work over Samah airdrome on Hainan Island (this portion of Hainan lay within the SWPA . Seven planes and two hangars were claimed destroyed, with other damage; ten B-24’s were holed by intense and accurate flak. A series of attacks designed to cripple Japanese ship repair facilities began on 31 March when thirteen B-24’s hit shipping and docks in Yulin

Page 502

Harbor. Thirty-seven Liberators went to Hong Kong on 2 April dropping 1,000-pounders on the Kowloon and Tai Koo docks with good results; two Tojos attempted interception but the P-38 cover drove them off. The next day’s mission to Hong Kong was primarily directed to shipping, but warehouses and an oil pipeline on Stonecutters Island were also hit. On 4 April, 41 more Liberators were back over Hong Kong with 1,000-pounders scoring in the Victoria city harbor area on the power plant, oil storage tanks, and the Royal Navy Yard. For the fourth straight day, on 5 April, twenty B-24’s bombed the Kowloon docks and the Kai Tak airdrome at Hong Kong. On 6 April twenty-three Liberators covered the Yulin Harbor docks and oil storage, giving Hong Kong a rest.79

The “take-out” of Saigon started on 19 April when eight Liberators of the Thirteenth Air Force’s 307th Group staged through Palawan and bombed the harbor through clouds. On 22 April, twenty 380th Group B-24’s hit the naval yards and shipping with 1,000-pounders, followed the next day by twenty-four 90th Group planes which scored on the dry docks, warehouses, and oil storage tanks. The two groups joined forces on 25 and 26 April to put forty-six and forty-seven planes, respectively, over Saigon. Hits were scored on dry docks, warehouses, ships at dock, barracks, an alcohol plant, and other installations. The same groups shifted their sights to the Texaco, Standard, Shell, and Socony-Vacuum oil installations at Saigon on 3 May, putting 47 Liberators loaded with 100-pounders over the targets and the same number loaded with 250-pounders on 4 May.80

Meanwhile, the LAB-24’s, the medium bombers, and fighters on shipping sweeps, running short of primaries, hit their secondary land targets. The night-flying Liberators hit airdromes, oil storage, supply depots, or arms plants-any vital target suitable for radar bombing from the Hong Kong-Canton area north to Shanghai. The medium bombers hit targets of opportunity on Hainan Island and along the Asiatic coast until 7 May, after which they were assigned Indo-China railroad targets, often as a primary. After being released from their bomber escort duties, fighters would go down on strafing runs over targets of opportunity. Favorite locations for independent fighter sweeps were the tributaries of the Canton and Saigon rivers as well as the coastal areas from Saigon to Tourane and Hong Kong to Swatow. While there were still Japanese fighters on Asiatic coastal bases and unescorted bombers might be intercepted, the last fighter encounter

Page 503

occurred on 2 April over Hong Kong, where escorting P-51’s claimed one destroyed and two probables.81

On 2 April V Bomber Command’s estimate of enemy air strength (based on photo interpretation) in the Hong Kong-Canton area was seventy-eight serviceable planes (fifty-nine fighters). The number went down to 62 planes on 27 April, but rose to 131 planes (119 fighters) on 7 May. This was interpreted as indicative of a decision to defend the area, but postwar testimony indicates that the increase was transitory – resulting from the shift of available air strength from Malayan and Indo-China bases northward toward Shanghai and the homeland. Maximum-strength heavy bomber missions were sent out on 9 May. Two groups, totaling 41 B-24’s, loaded with 100-pounders and 20-pound fragmentation bombs struck the White Cloud dispersal and revetment areas through the cloud cover; 44 planes loaded with 20-and 260-pound frags bombed Tien-Ho airdrome. Four groups again were sent out to the Canton airdromes on 10 May, and when the plane count the following day dropped again to sixty-eight planes, the airdrome targets were turned over to night bombers for occasional harassing strikes.82

Attacks on Indo-China rail lines started on 7 May, when fourteen B-25’s swept along the coastal railroad bombing and strafing three stations, two bridges, and rolling stock. Next day, forty-eight Liberators in six-plane formations bombed railroad bridges, with hits scored at Phu Khe, Nhatrang, Phanrang, and Tuyhoa while the Bong Son, Thoa River, and Ve River bridges were missed. The same day, sixteen Mitchells swept the railroad from Phanrang to Binh Dinh. Again on 13 May, 43 heavy bombers using 1,000-pound bombs attacked bridges along the same coastal stretch. Railroad yards were the targets for group formations on 27 May. Twenty-four of the 90th Group’s “Jolly Rogers” made a damaging attack on the Muongman railroad yards while twenty 380th Group B-24’s scored equally well on the Phanrang yards. Next day, the 90th Group again had twenty-four planes over the targets, twelve dropping on Muongman, six on the Gia Ray rail yards, and the other six hitting rolling stock between Saigon and Phanrang. On this mission the heavies made strafing passes after expending their bombs. Twenty-three 380th Group planes repeated their Phanrang mission. On 12 June, the two groups combined to send forty-four Liberators against the railroad yards at Saigon.83

Two special missions were run by the 90th Group after permission

Page 504

had been obtained from Chennault. On 12 July 23 planes dropped 500-pound bombs on the Canton supply depot scoring on warehouses and leaving several fires. On a similar mission, fourteen planes were sent to finish off the Canton small arms plant, often a secondary target for previous LAB-24 missions. Over 50 per cent of the bombs were on the target, leaving good fires as the planes pulled off.84

The technique employed on missions to the Asiatic mainland was usually that of the mass strike on a key point. The intervals between strikes are explained chiefly by other commitments, especially those at Formosa. Like other American forces which earlier had anticipated that the China coast would hold for them targets of major importance, the Fifth Air Force by July was moving forward to Okinawa with the focus of its plans now placed on Japan itself.

Harmon Field, Guam

Harmon Field, Guam

Isley Field, Saipan

Isley Field, Saipan

B-29 Maintenance – 
R-3350 Engines

B-29 Maintenance – R-3350 Engines

B-29 Maintenance – 
Night Work during the March fire blitz at Guam

B-29 Maintenance – Night Work during the March fire blitz at Guam