Page 645

Chapter 21: The All-Out B-29 Attack

THE preoccupation of XXI Bomber Command with urban incendiary attacks and support of the Okinawa campaign during the spring of 1945 did not mean that LeMay had abandoned the concept of precision strikes against priority targets. After the March fire blitz had proved so successful, he had attempted to perfect a technique for night attacks against individual targets. “This experimentation,” said a Washington spokesman for the Twentieth Air Force, “is primarily for the purpose of increasing our versatility, particularly during the bad weather periods that are fast approaching.”1 Weather was, indeed, the determining factor. It may seem strange, after the monotony of complaints about weather during the early months, to speak of bad periods approaching, but this observation was on the authority of meteorological statistics: normally, the summer monsoon was even more productive of clouds than the winter monsoon, and LeMay’s weather section had estimated that the main target areas in Honshu could be attacked visually on only three days in the months of April and May, once in June.2 The night precision technique was never perfected except for a special type of operations conducted by the 315th Wing, but the experimental mood continued and the command did achieve a versatility of attack that succeeded superbly in spite of Honshu’s clouds.

The key to this success lay in the flexibility of target selection: on clear days multiple forces were sent out to bomb various targets visually; in cloudy weather radar attacks were conducted against urban areas. The development of this program was delayed by the Okinawa campaign, which, with the incendiary attacks on the principal cities, has already been described.* Concurrently with those

* See above, Chapter 20.

Page 646

activities, and subsequently, XXI Bomber Command was engaged in other operations; the complex program may be described under the following rubrics: 1) attacks against individual industrial targets (24 March to 14 August); 2) incendiary attacks on secondary urban areas (17 June to 14 August); 3) the oil campaign (26 June to 14 August); and 4) VLR mining operations (27 March to 14 August). To treat these topics separately is to lose all sense of the growing intensity of the B-29 effort as it was split between these several tasks, but such an approach makes it easier to understand the command’s general objectives.

Industrial Targets

The experiment with night precision attacks was short-lived. It was suggested by Arnold who had been greatly impressed by night operations during the March blitz.3 LeMay was willing to try the experiment: if successful, the tactics contemplated would go far toward making precision bombing independent of weather, and night bombing was more efficient than day in tonnage lifted and less costly in planes lost.4 His first trial was a large one, a 251-plane mission against Mitsubishi-Nagoya on 24 March. In four daylight attacks this priority target had received only minor damage. For the night attacks, operational plans were based on a variation of the RAF pathfinder technique: 10 minutes before bombing time 10 B-29’s were to light the engine works area with M26 flares; 5 minutes later another 10 B-29’s would drop M17 incendiary clusters to start marker fires; then the main force was to attack with 500-pound GP’s. Unfortunately, the crews found Nagoya covered with a deep stratocumulus cloud. The several formations came over with excellent timing but thereafter the attack fell apart. Cloud complicated the bomb run and, combined with smoke from the incendiaries, obscured the lights from the flares; some of the incendiaries fell outside the factory complex and bombardiers sighting on them were off target. Thus, though the attack with its 1,533 tons was the heaviest ever sent by the command against a single aircraft industry target, the results were negligible: only 60 tons of bombs fell in the factory area, damaging a few minor buildings and causing no appreciable loss in production.5

Still interested in the possibilities of night tactics, LeMay’s operations officers realized that some more effective way of lighting the target would have to be devised. Because there was no agreement at

Page 647

the wing level as to the best procedure, LeMay decided to let O’Donnell and Storrie run single-wing missions, each trying his own technique. Storrie’s attack on 30 March, obviously on an experimental scale, involved only 14 of the 314th Wing’s planes – 3 loaded with flares, 7 with 5 tons each of 500-pound GP’s, and the last 4 with a mixed cargo of bombs and flares. In spite of the large ratio of flares carried, the twelve planes that got over Nagoya missed the Mitsubishi plant entirely.6 The 73rd’s effort was larger but hardly more successful. On 1 April, O’Donnell sent against Nakajima-Musashi 121 B-29’s, each topping off its bomb load with 4 flares. Of 1,019 tons dropped, only 4 hit in the area.7

Since none of the attempts had provided any positive evidence, LeMay ordered for the night of 3 April three one-wing attacks against Mitsubishi’s Shizuoka engine plant, Nakajima’s Koizumi assembly plant, and the Tachikawa aircraft engine plant. Command headquarters specified the mean point of impact, bombing altitude, and bomb load; each wing was allowed to choose its own method of lighting the target. Again, the results in each instance were negligible.8 The command just wasn’t equipped for night precision bombing; specifically, it needed target marker bombs, such as the 1,000-pounders used by the RAF, and reflex optic bombsights. Lacking the proper equipment, LeMay abandoned the experiment.9 For a while B-29’s were kept busy enough with airfield strikes in support of the battle for Okinawa,* but between trips to Kyushu they were sent occasionally against the aircraft industry in daylight missions.

The first break came on 7 April when LeMay split his forces between the two top-listed targets. The 313th and 314th Wings went to Nagoya to unload 610.7 tons of high explosives on the Mitsubishi engine works. Bombing from about 20,500 feet in CAVU conditions, the bombardiers put on a brilliant demonstration of precision technique. Post-strike photos showed 62 per cent of the roof area destroyed, and actual damage was worse than was realized at the time.10 After the war Japanese officials reported that this attack destroyed 90 per cent of the plant’s facilities. Dispersal, induced by earlier raids, had already slowed production to 129 engines in March; in April only 15 were built and only 44 during the rest of the war.11 After repeated failures at Nagoya, XXI Bomber Command had finally knocked out its second-listed target in a single mission.

* See above, pp. 627-35.

Page 648

The 73rd Wing hit pay dirt, too. It had been banging away at Nakajima-Musashi since the first Tokyo raid without much effect. Weather had made the target hard to hit, and the 500-pound GP’s used had been too light for the plant’s modern buildings of steel and reinforced concrete. On the 7th, however, Tokyo’s weather was as near perfect as Nagoya’s and the 101 planes that bombed were loaded with newly arrived 2,000-pounders. About 26 per cent of the 490 tons dropped hit in the target area, doing heavy damage in the machine shops and destroying about 10 per cent of the plant’s buildings. The same wing returned to Musashi on 12 April to finish up the job with 119 planes again loaded with 1-ton bombs. The aiming point was in the Tama section, untouched on the 7th. Because of a heavy haze, bombardiers had to make a radar run; they missed Tama but got sixty-four tons into the eastern section of the plant, causing heavy structural damage to 10 per cent of Musashi’s buildings. Eleven B-29’s, unable to bomb the primary, hit Mitsubishi’s Shizuoka engine works, a new plant just coming into production, and damaged approximately 86 per cent of the roof area.12

These raids about finished Musashi. The command was to make an abortive attempt against it in June and to return on 8 August when lucrative targets were scarce, but both missions were superfluous. By 12 April the B-29’s had staged eleven missions against this, the No. 1 target: four had failed entirely because of weather, and of the seven in which bombs were dropped, only the last two were more than moderately successful. It must have been somewhat embarrassing to the command that Task Force 58’s low-level strike on 17 February had done more damage than any single VHB attack, but it was the cumulative effect of repeated blows that ruined the plant. Nakajima officials had long since abandoned any efforts at repair, concentrating on removing machine tools and equipment to dispersal areas. Production had fallen to 425 engines in March and now, after the April attacks, ceased entirely.13

On the last Musashi show 102 P-51’s from VII Fighter Command had escorted the B-29’s and had found plenty to do when about an equal force of Japanese defenders came up, damaging 36 of the bombers.14 In hopes of drawing off some of Tokyo’s interceptors, LeMay had sent on the same day a moderate force from the 313th and 314th Wings to Koriyama, 120 miles north of the capital. The planes

Page 649

severely damaged two chemical plants, Hodogaya and Koriyama, and an adjacent aluminum plant. The primary targets had been chosen because they produced tetraethyl lead, an essential component of aviation gasoline; actually, the oil shortage was so severe that these plants were part of a surplus production capacity and hence the good bombing was of little strategic consequence. Nor had the mission succeeded as a diversion: the enemy fighters refused to leave Tokyo, and only ten individual attacks were reported over Koriyama.15

On 24 April, with weather forecasts unfavorable for Kyushu and favorable for Tokyo, LeMay interrupted the ICEBERG support operations to run a mission against the Tachikawa plant of the Hitachi Aircraft Corporation. Located at Yamato, a few miles north of Tachikawa and nineteen miles from the Imperial Palace in Tokyo, the factory built radial engines for army planes. Production had averaged 250 engines per month in 1944, but in the early months of 1945 had slumped, partly because of a dispersed program, partly because of a carrier strike on 17 February. A force of 131 B-29’s got over Tachikawa on the 24th to find the target obscured by haze, but 101 were able to bomb visually from 12,000 feet. The 473.5 tons of GP’s dropped completely wrecked the plant; no effort was made thereafter to repair the damage and production stopped altogether. Going in without escort and at an unusually low altitude, the B-29’s met stiff opposition from flak and fighters; four bombers were lost and sixty-eight damaged while their own gunners were registering claims of fourteen fighters destroyed and twenty-four probables.16

An attack against the Tachikawa Army Air Arsenal on 30 April was foiled by weather, though some of the planes were able to bomb the primary radar target, the Hamamatsu urban area.17 This mission, and one sent out on 5 May against the Hiro Naval Aircraft Factory at Kure, were coordinated with Kyushu airfield strikes. In the Kure attack, the 73rd Wing was joined by the 58th, which had recently moved from CBI and was now making its first Honshu strike from the Marianas. Attacking from about 20,000 feet, 148 B-29’s dropped 578 tons of 2,000- and 1,000-pound bombs with devastating effect; many buildings and more than 500 machine tools were destroyed or damaged and production was cut almost in half.18 On the 11th, LeMay sent a moderate force against the Konan plant of the Kawanishi Aircraft Company, an important manufacturer of airframes. Despite a

Page 650

4/10 cloud cover, the mission was highly successful: so heavy was the destruction that the company immediately removed almost all of the machine tools left.19

The Konan mission was run as a diversion for the last B-29 strikes against Kyushu airfields; on the same day, 11 May, Nimitz released XXI Bomber Command from support of ICEBERG. Arnold immediately reconfirmed current target directives, with the aircraft industry and the principal urban areas as the priority objectives.20 With the two most important engine factories stilled, LeMay chose to concentrate on the great urban areas; his highly successful campaign, already described,* absorbed most of the command’s energies from 14 May to 15 June, and except for an abortive mission against the Tachikawa Aircraft Company on 19 May,21 no precision attacks were scheduled until 9 June. The success of the recent strikes, however, had shown that daylight missions could play an important role in an articulated program. In part, the improvement in bombing had resulted from better weather and from better forecasting. Crews had gained confidence from the occasional fighter escort and greater skill with lead-crew training and with combat experience. Finally, there was the change in tactics which had lowered the mean bombing altitude from 30,000 to 20,000 feet and at times sent bomber formations in much lower. This increased the danger both from enemy fighters and flak, but like LeMay’s other calculated risks, it paid off in effectiveness without undue cost: of 1,433 B-29’s airborne against industrial targets between 24 March and 19 May, only 20, or 1.3 per cent, were lost to all causes.22

While great formations of B-29’swere burning out the six principal cities, LeMay’s staff was working overtime on a coordinated plan for the cloudy period of late spring and summer. When radar weather was predicted, the command would run incendiary missions against secondary industrial cities; when visible conditions were predicted, daylight precision missions would be dispatched against priority targets, most of them connected with the aircraft industry. Neither the industrial targets nor the urban areas were of a size to demand a maximum force; hence, the policy was to name for any strike day a number of targets with a separate force assigned to each. This system of multiple targets allowed the command to take full advantage of any good weather areas, but for success it required much study of routes,

* See above, pp. 635-44.

Page 651

forces required, bomb types, and altitudes and a careful coordination with weather services.23 This so-called “Empire Plan” was to govern most of the command’s bombing until the end of the war.

The plan went into operation on 9 June when, with good weather predicted, LeMay used three small formations. Two groups hit the Kawanishi Aircraft Company’s plant at Narao, an important source of navy planes; the 1,000-pound bombs did so much damage that virtually all surviving machine tools were dispersed to other locations. Two groups heavily damaged Aichi’s Atsuta factory, although only 4 bombs hit in the target area: they were 4,000-pound light-case projectiles and 1 touched off a devastating fire. A single group, sent against Kawasaki’s plant at Akashi, found the area covered with 9/10 cloud and, releasing by radar, put its 2-tonners into the village instead of the factory.24

Good weather was promised for the 10th also, so LeMay named six Empire targets. He sent the 73rd Wing back briefed to dump 2,000-pounders on ruined Nakajima-Musashi, and single-group formations against Nakajima plants at Ogikubu and Omiya, Japan Aircraft Company at Tomioka, Hitachi at Cbiba, and the Tachikawa Army Air Arsenal. Weather in the Tokyo Bay area, where all the targets were located, proved variable. Nakajima’s luck held: its three factories were cloud-covered and formations assigned to them bombed instead primary radar targets, doing heavy damage to the Hitachi engineering works at Kaigan and a seaplane base at Kasumigaura. Against the other targets, bombing was visual and quite effective in each case. VII Fighter Command provided an escort of 107 P-51’s.25

Again on 22 June there were six targets, this time in southern Honshu. Kure Naval Arsenal was assigned six groups and the other targets, all aircraft factories, forces ranging from one to four groups. In all, 446 B-29’swere airborne and 382 bombed, dropping 2,103 tons of bombs.26 Post-strike photos showed 72 per cent of the roof area at the Kure arsenal damaged.27 No analysis of the results at Mitsubishi’s Kagamigahara plant is available, but postwar investigations showed varying degrees of destruction elsewhere: only slight damage at Kawasaki’s works at the same town;28 at Kawanishi’s Himeji plant, great destruction among the buildings and total destruction of machine tools;29 at Mitsubishi-Mizushima, 135 of 235 machine tools and almost half the roof area destroyed, drastically curtailing production of Betty bombers and George fighters;30 at Kawasaki-Akashi, where

Page 652

the extent of damage done on 19 January was not fully appreciated, the empty buildings were completely destroyed.31

On 26 June LeMay sent out nine formations against southern Honshu and Shikoku targets, with a total force of 510 B-29’s and an escort of 148 P-51’s. Good weather was predicted but heavy clouds over much of the area made assembly difficult and many of the planes bombed targets of opportunity individually or in small flights.32 Damage for those targets where assessment is available (aircraft factories, light-metals industries, and arsenals) varied from light at Aichi’s Eitoku plant (already hit heavily) to unnecessarily severe at Kawasaki-Akashi, where well-placed 4,000-pound bombs served only to whip a dead dog.33 At Kawasaki’s Kagamigahara plant, however, the slight damage done on the 22nd was so greatly increased that every important building was knocked out.34

The weather had held up somewhat better than had been expected, allowing five daylight missions in April, three in May, four in June. Thereafter almost a month passed before visual conditions again obtained. On 24 July the command put up 625 planes, directed against 7 targets in the Nagoya and Osaka areas. The attacks were coordinated with a two-day carrier strike in the Inland Sea region. Targets for the B-29’s were chosen to give the several formations a wide choice according to local conditions, but in each case the force assigned was considered heavy enough to destroy its primary target.35 Weather turned out spotty; 26 aircraft dropped 166 tons on targets of opportunity and 573 dropped 3,539 tons on primary visual or radar targets.36 The Sumitoma Metal Company’s propeller factory, whence most of the machine tools had been removed, was completely wrecked.37 Kawanishi’s Takarazuka plant lost most of its buildings and no effort was made subsequently to repair them.38 The Osaka arsenal, though cloud-covered and attacked by only part of the assigned force, suffered additional damage amounting to 10 per cent of the original roof area.39 Aichi at Eitoku sustained its heaviest damage of the war, damage which was superfluous because of previous dispersal.40 Nakajima at Handa, struck for the first time, lost its principal assembly buildings, but the attack came too late in the war to have much direct effect on production.41

This fifth Empire attack was the last. Two weeks of cloudy weather followed, then the atomic bomb attacks against Hiroshima and Nagasaki ushered in a last spasm of precision strikes in an effort

Page 653

to end the war quickly and, one might suppose, to end it with conventional bombardment methods. XXI Bomber Command had brought its daylight tactics to a state of high efficiency,42 but the favorable opportunities had been rare and the individual targets were of diminishing importance; in the meantime, the B-29’s had carried to some fifty-odd of Japan’s medium-sized cities the same incendiary tactics that had ruined the greater industrial centers.

Incendiary Attacks on the Smaller Cities

The incendiary campaign begun by XXI Bomber Command in mid-May was based on a Joint Target Group study of 28 March, which listed thirty-three urban areas concentrated in eight of Japan’s largest cities.* One, Yawata, was not hit; as the others were reduced to cinders in successive raids, it became increasingly obvious that the same tactics should be applied to the smaller cities: the efficiency and light cost of the night raids and the weather outlook were convincing arguments. LeMay’s A-2, Col. James D. Garcia, stressing the importance of cumulative effects of raids compressed within a short period of time, recommended a systematic attack on medium-sized cities now that there was “a possibility of achieving a decisive effect with air power.”43 His choice of preferred targets was based on the following factors: 1) congestion and inflammability; 2) incidence of war industry; 3) incidence of transportation facilities; 4) size and population; and 5) adaptability to radar bombing. The list of 25 cities, with populations ranging from 323,200 (Fukuoka) to 62,280 (Hachioji) was merely a tentative one but it served well enough to get the campaign under way: of the first 15 targets struck, 13 were from Garcia’s selection and eventually all but 5 were hit. His estimate of forces required – “an educated guess” in advance of photo reconnaissance – was about double what was actually used, and as the original targets were quickly scratched, others were added until by 14 August fifty-eight towns had been fire-bombed.

On 16 June, the day after the last incendiary attack against the major cities, LeMay alerted his wing commanders: the new program would begin next night with four one-wing attacks according to the following assignments – Omuta, 58th Wing; Hamamatsu, 73rd; Yokkaichi, 313th; Kagoshima, 314th.44

The cities, all in the 100,000-to-200,000-population bracket, were relatively congested; all had war

* See above, pp. 624-25.

Page 654

industries and each was a transportation center: Omuta and Kagoshima were ports on Kyushu, Yokkaichi on Honshu, and Hamamatsu was a rail point on the main Tokyo-Nagoya line.45

Except for the use of multiple targets the mission of 17 June was run off pretty much like those against the major cities. The planes were loaded with the familiar combination of M47 and M69 incendiaries and attacked by radar at altitudes between 7,000 and 9,200 feet, with pathfinders marking the targets. Enemy opposition, expected to be weak, was almost nil – the only loss suffered was chalked up to unknown causes. The ratio of effective sorties, 456 out of 477, was high and the total weight of bombs dropped, 3,058 tons, was heavy. Omuta received the heaviest attack but suffered least, with a destroyed area of .t 17 square miles, only 4.1 per cent of the city’s area. The results elsewhere were much more satisfactory: at Kagoshima, 2.15 square miles, or 44.1 per cent, destroyed; at Hamamatsu, 2.44 square miles, or 70 per cent; at Yokkaichi (where only three groups were sent), 1.23 square miles, or 60 per cent. The total area burned out, 6.037 square miles, was considerably better than the average results achieved by +wing missions against a single city, and as usual, some numbered industrial targets were damaged in the general fires.46

The success of the first multiple-target mission insured the continuance of the program, and the operational pattern established on 17 June became standard during the remaining weeks of the war. Whenever a force of B-29’s was ready to go and radar conditions were predicted, a night incendiary mission was scheduled. Targets for a particular night were based on operational considerations – weather, radar, and relative position of the several towns – as well as on data furnished by intelligence reports. As the campaign progressed and the available targets became smaller and of less significance, it became increasingly difficult to calculate accurately the proper type and weight of bombs required for each, which put an additional strain on intelligence and operations personnel in command, wing, and group headquarters.47

On strike nights XXI Bomber Command usually attacked four cities, with one wing assigned to each. Occasionally, target cities were considered large enough to require two wings (Fukuoka, 19 June, and Omuta, 26 July), in which case only three cities would be named.48 Conversely, as the choice of worth-while targets narrowed, smaller

Page 655

forces were sent against more towns.49 Sometimes the 315th Wing’s night strikes against the oil industry” were integrated with the incendiary attacks and, near the end of the war, day and night missions were coordinated in a furious round-the-clock effort to bring the enemy to surrender.50

In all, multiple incendiary attacks were sent out on sixteen occasions, an average of two a week. During that period, the main weight of B-29 attacks was directed against urban areas, as it had been, except during the Okinawa campaign, since the March fire blitz: between 9 March and 14 August about 70 per cent of the bombs loaded were expended in incendiary raids, about 22 per cent in precision attacks. After 17 June, 8,014sorties, with a total of 54,184 tons of incendiaries, were sent against 58 secondary cities.51 Because of the similarity of the methods used and the large number of attacks, it is more convenient to summarize the results of the missions in tabular formt than to describe them individually. One urban area attack, however, differed sharply from the others in technique.

Yawata, center of Japan’s steel industry, had been the target of two B-29 attacks from Chengtu, neither successful, and had long been carried by XXI Bomber Command on its priority list.‡ Because of its layout the city offered a poor target for radar bombing and had been marked for a daylight mission. Yawata was probably the most important industrial city left when, after the atom-bomb attack on Hiroshima, a strike was scheduled for 8 August. Eleven groups drawn from three wings were dispatched, carrying a mixed load of incendiaries. Although visual conditions had been forecast, the crews found heavy clouds over Yawata; smoke from fires started by the first-comers further obscured the target and 136 of the 221 planes bombing did so by radar. Results were considered only fair: 1.22 square miles were burned out, 21 per cent of the urban area and 33 per cent of the planned target area.52 Because fighter defense in the Yawata area was considered strong, the mission was escorted by three groups of P-47’s from the 301st Fighter Wing on Ie Shima. The Japanese were up in force and though the P-47’s knocked off a dozen or so, the enemy destroyed one B-29 and five U.S. fighters.53

The Yawata mission, in spite of the force involved, was less successful

* See below, p 658-661

† See below, Table 1, p. 674-75.

‡ See above, pp. 624-25.

Page 656

than the average incendiary attack and more costly. The Japanese air forces never devised an effective defense against night attacks; they had no first-rate night fighter and no efficient means of vectoring in the interceptors they sent up. Though B-29 crews could expect to meet some enemy planes on any night mission, they were never jumped by a large force, and such passes as were made were neither aggressive nor well coordinated. The big cities were defended by heavy AA guns working with searchlights but not the smaller towns,54 so that combat losses in the night raids against the latter were unbelievably low. Only one B-29 was known to have been destroyed by enemy action in the whole campaign, a Superfort attacked by three Japanese fighters over Omuta on 26 July.55 Flak and fighters damaged sixty-six others, but the remaining eighteen losses were chalked up to operational or unknown causes.56

The feebleness of enemy opposition led LeMay to try another tactical innovation that was less dangerous in reality than in appearance. To increase the psychological effect of his wide-ranging B-29’s, he decided literally to call his shots, warning about a dozen cities of an impending attack and then actually hitting four of them. Such a warning, delivered by leaflets, would be a grand gesture of confidence and might lessen the stigma attached to area bombing.57 Nimitz’ psychological warfare section approved the project and preparations were made to drop leaflets before an incendiary mission scheduled for 28 July. Three Japanese officers, prisoners of war, volunteered to translate the text, which developed the theme that “in accordance with America’s well-known humanitarian principles, the American Air Force, which does not wish to injure innocent people, now gives you warning to evacuate the cities named and save your lives.” OWI’s printing presses on Saipan ran off the 660,000 copies required.58 They were crammed into M26 bomb cases, 10,000 to the case, and on the night of the 27th, 6 B-29’s dumped them over 11 cities: Aomori, Tsu, Ichinomiya, Uji-Yamada, Ogaki, Uwajima, Nishinomiya, Kurume, Nagaoka, Koriyama, and Hakodate. Next night the command bombed the first six cities in that list, and although the enemy made some show of opposition, the forty or fifty fighters that rose shot down no B-29; they damaged only six and flak, five more.59 Even with advance warning the Japanese opposition was feeble, and LeMay used leaflets with the same impunity on 1 and 4 August; radio broadcasts from Saipan carried a similar warning. The stratagem came too late

Page 657

to reach the whole of the civilian population, but it came at a time when every additional pressure was important, and according to morale surveys made soon after the surrender the announcements achieved a wide circulation and a considerable effect, both in causing residents of the proscribed towns to move and in persuading them of the good intention of the Americans. A high Japanese official later characterized the warnings as “a very clever piece of psychological warfare, as people in the affected regions got extremely nervous and lost what faith they still had in the Army’s ability to defend the mainland.60

In general the incendiary attacks on the smaller cities were highly successful. On 12 July, because of operational difficulties, Uwajima and Ichinomiya were only slightly damaged and the command had to stage repeat visits on the 28th;61 for similar reasons Omuta was attacked twice, on 17 June and 26 July.62 But in all other cases one strike was enough, if not to destroy the town at least to scratch it as a profitable target. The burnt-out areas ran to 43 per cent of the total built-up area of the cities and in the case of Toyama to the fantastic figure of 99.5 per cent.63 In area destroyed per 1,000 tons of bombs expended, the attacks from 17 June to 14 August were not as destructive as those against the great cities, but that was to be expected because of the physical layout of the medium-sized urban centers.64

The economic effects of the incendiary attacks are harder to evaluate than the physical, and it is impossible sharply to differentiate in the over-all picture between the results of the raids against the greater and the smaller cities. The best guide to the problem is a study made by the U.S. Strategic Bombing Survey after the war. In spite of an increasing shortage of raw materials, Japanese war production increased during 1944, reaching a peak in such important items as aircraft, metals, and ordnance in October, just before the first B-29 raids from the Marianas. The decline thereafter was not wholly attributable directly to air attacks: the blockade and inefficient dispersal – itself brought on by fear of air attacks – would have reduced the production rate without bomb damage. But sampling processes and the existence of a few unbombed cities which could be used as controls made possible some quantitative judgments. In urban areas not bombed, production decreased only slightly after October 1944: at Hiroshima it stood in July 1945 at 83 per cent of the peak, and in six un-bombed cities on Hokkaido at 93 per cent in June. Yet by July production in

Page 658

the bombed cities surveyed had fallen to 33 per cent of the peak; even in factories not themselves hit but located in those cities the output had declined to 51 per cent of the October rate.65

In the small cities, as in the large, the area attacks worked vast hardships upon the Japanese people. Statistics of dead, wounded, and homeless tell little about the personal sufferings, nor about the dislocation which occurred as thousands fled the towns. The effect of this dispersal on industry, as had been expected by air planners, was great. So also was the effect on morale, as the terror which had earlier been confined to a few great cities was spread throughout the country. In a report on the effects of urban area bombing submitted in December 1945, a group from the faculty of the Imperial University of Tokyo wrote that “with the shifting of the attacks from cities to local districts, the people became concerned over the future of the war. In consequence, their fighting morale was weakened.”66

Attacks on the Oil Industry

In the program of coordinated strikes begun in June, ultimate choice between Empire Plan or urban industrial targets was determined by weather. As the program got under way, the 315th Bombardment Wing (VH) became available for combat, and its operations, in some measure independent of those of the other wings, were dictated largely by the special equipment of its units.

The 315th Wing, authorized in December 1944 for deployment in the Pacific, settled at Northwest Field, Guam, during May and June; its commander, Brig. Gen. Frank A. Armstrong, Jr., was a veteran of the strategic air offensive against Germany.67 The wing’s B-29’s differed in two important respects from those of other units. They were equipped with the AN/APQ-7 (Eagle) radar instead of the conventional AN/APQ-13. The latter had been designed primarily as a navigational aid, and though crews had improved with experience in their use of it for night or bad-weather bombing, it was not an instrument of precision. The Eagle radar, developed for bombing, possessed a much greater degree of definition, and though it required a long bomb run (average, seventy miles), that was no serious handicap in the present tactical situation in Japan. The Superforts had been stripped of all armament except the tail gun; this modification and the Eagle radar marked the 315th as a night-bombing outfit.68 There had been several proposals for use of the specially equipped B-29’s:

Page 659

for very high-altitude bombing, for area bombing, and for aerial mining.69 Before the 315th Wing was combat-ready, however, the 313th had become proficient in mining, and all the wings in area bombing, with the AN/APQ-r3. Unit training for the 315th had stressed night radar tactics to the neglect of visual bombing and daylight formation flights, and it was obvious that if the Eagle radar was to be given a scientific test, it should be against a discrete set of targets – preferably large in size and located along the coast line.70 In the estimation of XXI Bomber Command, the oil industry met these requirements.

As a target system, that industry differed sharply from its counterpart in Europe, though there was one fundamental similarity: Japan, like Germany, was dependent on imports for its petroleum. Homeland wells had produced only 2,470,000 barrels in the peak year, 1937, and only 1,941,000 (less than 0.1 per cent of the world’s total) in 1941. During the 19~0’s, the war lords had built up a backlog of 55,000,000 barrels by extensive importation and severe restrictions on civilian use, but by Pearl Harbor heavy consumption and the U.S. embargo had reduced the stock to 43,000,000 barrels. The need for oil had been the main incentive for Japan’s drive southward, and her quick success in the Netherlands East Indies had gained for a while a ready access to petroleum and its refined products, just as Germany’s drive into southeast Europe had given Hitler control of fields in Rumania and Hungary. But whereas Germany could depend for imports on a complex transportation system, rail and barge, which was long proof against Allied attack, Japan had to depend on shipping and had begun the war with inadequate tonnage in tankers. The immediate demands of the war consumed much of her newly gained production, and the Allies took an ever increasing toll of shipping in attacks by submarines and carrier- and land-based planes. By August of 1943 oil shipments from the south had begun to decline, and as the Allied forces moved north-westward into the Marianas and Philippines, the flow decreased sharply. There had been no opportunity to build up sufficient reserves at home, nor had Japan the raw materials, plant capacity, or technological skill to build a synthetic oil industry equal to that which had served Germany after the loss of her stolen Balkan wells. The Japanese made desperate efforts early in 1945 to improvise a synthetic industry – including a fanciful pine-root oil project – but it was too late and too little.71

Page 660

Much of this decline in the oil industry and stock in the home islands had preceded the first VHB missions; the COA in its reports of November 1943 and October 1944* preferred attacks on shipping to direct strikes against oil installations as a method of increasing the stringency in Japan, and the JTG had never listed the industry as a primary target objective. When the flow of oil into Japan ceased entirely in April 1945, the only B-29 attacks on the industry had been by XX Bomber Command: a mission against Palembang in the previous August and some strikes against storage dumps in the Singapore region early in 1945.†

By April, however, AAF intelligence had come to the opinion that the petroleum industry in Japan was in so critical a state that the destruction of facilities and stores would react immediately upon the tactical situation.72 Consequently, LeMay and Lt. Gen. Barney M. Giles, who came to Guam as the deputy commander of the Twentieth Air Force, eventually decided that during its combat-testing period the 315th Wing would devote its efforts exclusively to oil targets.73 This decision had the enthusiastic indorsement of Gen. Carl Spaatz, slated to command all B-29’s under USASTAF, who had been an ardent advocate of the oil campaign in the ETO.74

Actually, the attacks had begun before the 315th Wing was ready for combat. The first came on 10 May and was considered, by Nimitz as well as LeMay, as a blow in direct support of the Okinawa battle.75 It was a three-pronged strike, with the 73rd Wing hitting the Third Naval Depot at Toltuyama; the 314th, the Iwakuni Army Fuel Depot; and the 58th, oil storage installations at Oshima. Most of the planes bombed visually and each attack was successful; damage ranged from 20 per cent at Tokuyama to 90 per cent at Oshima and would have been heavier had it not been for the lack of oil in many tanks and pipes.76 On 22 June, six B-29’s of the 313th Wing, unable to find their primary target, dropped seventeen 4,000-pound bombs on the Second Naval Fuel Depot at Yokkaichi, damaging about 15 per cent of the plant.77 Besides these storage areas, several refineries were hit by one or another of these wings, incidentally in area attacks or as targets of opportunity.

The 315th Wing opened its specialized campaign on 26 June with

* See above, pp. 26-28, 132-34.

† See above, pp. 107-10, 163-64.

Page 661

a strike against the Utsube Oil Refinery at Yokkaichi, the top-priority target. By 14 August the wing had run fifteen missions against ten targets, petroleum refineries or synthetic plants, which included besides the first: Maruzen Oil Company at Wakayama; Mitsubishi Oil Company at Kawasaki; Nippon Oil Company plants at Akita, Kansai, Kudamatsu, and Amagasaki; Imperial Fuel Industry Company at Ube; and Toa Fuel Industry at Wakayama.78 In all, the 315th Wing dispatched 1,200 planes, of which 1,095 bombed primary targets, dropping 9,084 tons of 500-pound GP’s, the bomb considered most effective against the scattered installations attacked. The very heavy bomb load lifted was possible because the planes were stripped and bombed individually at night; with experience, the crews were able to increase the weight carried from an average of 14,631 pounds on the first mission to 20,684 on 9 August. Removing most of the guns did not prove too dangerous: only four planes were lost and sixty-six damaged during the entire campaign.79

On the whole, the experiment was markedly successful. The formations were able to attack the primary target on every mission, and while the results varied they were generally good. On most missions, General Armstrong sent a two-group force, and releasing at low or medium levels, bombardiers were able to get enough bombs into the target area to do substantial damage. In some cases it was necessary to return a second or even a third time, but by the end of the war most of the plants were completely or largely inoperable. USSBS statisticians calculated that 315th Wing bombardiers had achieved an accuracy rate of 13.5 per cent, as compared with 5.4 per cent achieved, under more difficult tactical conditions, with the Eagle radar in Europe.80

If more accurate, however, XXI Bomber Command’s campaign was much less important than that in Europe. The Twentieth Air Force estimated that B-29 attacks had destroyed about 6,000,000 barrels of tank-storage capacity, USSBS that they had reduced Japan’s refining capacity from 90,000 barrels a day (in December 1941) to about 17,000 barrels.81 The strategic effects were more apparent than real, however, because many of the storage tanks were empty and refinery production had fallen to only 4 per cent of capacity before the VHB campaign began.82 The lack of precise intelligence on the state of Japan’s economy justified the effort spent on the oil program, a sort

Page 662

of reinsurance policy, but the blockade had dried up the nation’s oil resources so that tankers lay idle at the docks. Other B-29’s, however, had contributed importantly if tardily to that blockade.

Mining Operations

At the beginning of World War II, neither the Navy nor the AAF was keenly interested in the use of the mine as a strategic offensive weapon and consequently there was a serious lag in the mining program, both in the development of new weapons and in their employment. Indifference was gradually overcome, partly through the missionary work of enthusiasts in the Navy’s Mine Warfare Section, partly through the influence of the British, who had been engaged in mining and countermeasures since 1939. Japan’s island position, with her dependence upon sea transportation both for vital imports and for supplying military outposts, made the nation particularly vulnerable to mining operations; distances in the Pacific made it most efficient to deliver the mines by aircraft. There was thus during 1943 and 1944 an increasing use of aerial mines in the Pacific Ocean Areas and in CBI, either in connection with specific amphibious operations or as a campaign of attrition. As the westward drive of U.S. forces emplaced them within VLR striking distance of Japan, it was natural that the B-29 should be considered as a mining instrument.83

Most of the initiative in this respect came from the Navy, which recognized the superiority of the heavy land-based bomber over carrier or amphibious aircraft for a sustained mining campaign but which lacked the proper equipment. Navy agencies in Washington had argued early in 1944 for the use of XX Bomber Command B-29’s for mining and CINCPOA was even more insistent. On 6 July, when the advanced echelon of XXI Bomber Command was passing through Pearl Harbor en route to Saipan, Nimitz’ staff tried to sell the idea of a VHB mining campaign against Japan’s home waters, with the command furnishing the B-29’s and crews, CINCPOA the mines and technicians.84 The advanced echelon officers, of course, were unable to act on the suggestion, but Navy staffs continued to urge the campaign, preparing detailed operational and logistical plans for cooperation with the Twentieth Air Force.85 These were referred to the COA, currently engaged in revising its report on strategic targets in the Pacific.86 In both its over-all report (10 October) and a special subcommittee report on shipping (20 October), the COA reviewed

Page 663

the possibility of a mining campaign.* In the latter paper particularly the committee agreed with the Navy’s point of view: the Allied blockade had already forced the Japanese to funnel most of its shipping through the Inland Sea and the spout of the funnel, the Shimonoseki Strait, could be plugged with aerial mines. Operations on the scale suggested by the Mine Warfare Section, the COA agreed, would force the Japanese to abandon the outer zone by August 1945 and would weaken the defense of the home islands. The proposed schedule, based on estimates of availability of mines and capabilities of the B-29 force, was as follows:87

Period Sorties Per Month Mines Per Month Type of Mines
Phase I Dec. 1944 405 1,500 Magnetic
Phase II Jan.–Mar. 1945 590 2,100 80% Acoustic
20% Magnetic
Phase III Apr.–Aug. 1945 540 1,400 60% “Unsweepable”
20% Acoustic
20% Magnetic

Arnold’s staff greeted the proposal without enthusiasm. Kuter, his top planner, pointed out that only ten days earlier the COA’s revised report had listed as priority targets for the B-29’s the aircraft industry, urban areas, and shipping-in that order. To divert a sizable part of the VHB effort to mining – 405 of a total of 500 projected sorties in December – would make it impossible for the Twentieth to perform its primary mission, and hence mining on a large scale must await the build-up of forces in the Marianas.88 Actually, the COA report of 10 October had contained two alternative sets of recommendations; in one, predicated on the supposition that Japan could be defeated without an invasion, mining had been listed in top priority for the B-29’s. The strategy currently accepted, however, included an invasion, and within Norstad’s staff the COA proposal was characterized as “another hope for a relatively painless method of winning the war,” a slow process that might require two years. Extensive mining should be delayed until after the aircraft industry was knocked out and should be tried then only if a complete blockade seemed feasible and preferable to any other target system.89

Nimitz nevertheless continued to press for an early commitment of B-29’s, asking Arnold to assign to mining 150 sorties per month

* See above, pp. 133, 552.

Page 664

beginning in January and a heavier effort beginning in April.90 This request was more modest than the earlier proposal, but it still would have handicapped XXI Bomber Command’s program of precision bombing during the “good weather” months of winter, and hence Arnold demurred in a letter of 28 November, promising aid when the command’s forces were larger and the weather was less suitable for daylight missions.91 This was not mere temporizing; if not enthusiastic about the project, Arnold was determined to give it a try. Harmon, apparently more sympathetic to Nimitz’ plan, learned of Arnold’s delayed acceptance and on 13 December alerted Hansell.92 Hansell, busy with his precision attacks, protested vainly: on 22 December he received from Arnold an order to initiate planning for mining operations as requested by Nimitz, but beginning on 1 April rather than 1 January.93

The AAF thus embarked on a B-29 mining program, grudgingly, because the Navy lacked aircraft suitable for a Navy job; the situation was not wholly unlike that which had taken the AAF into antisubmarine work in 1941.* Evidently the air planners did not envisage the extraordinary success that was to follow, but it is questionable whether they could have acted differently if they had. Like the Navy advocates of a blockade, they hoped to win the war without an invasion; their weapon and their doctrine had been conceived in terms of an attack on industrial targets, however, and in their view any sustained air operations over Japan, bombing or mining, demanded first the destruction of the sources of Japan’s air power. This had been the experience in Germany, and the opposition to Nimitz came, it must be realized, when the overworked 73rd Bombardment Wing was taking heavy losses. To have inaugurated mining in January would have disrupted the bombardment program, as yet unsuccessful but going through a necessary period of adjustment, and in objecting to this Hansell and Arnold’s office were not unreasonable. Yet in light of the spectacular results of B-29 mining operations later, it is ironical that the decision to cooperate with Nimitz came not from any great liking in the AAF for mining but rather from the sort of logic that often colored inter-service comity during the war-the fear that otherwise the AAF might allow “a possible major usage of long-range aircraft to develop, by default, into a matter of special interest to the Navy.”94

* See Vol. I, Chap.15.

M26 Mine Dropped by B-29

M26 Mine Dropped by B-29

Okinawa: Motobu Airfield

Okinawa: Motobu Airfield

Page 665

Hansell and Arnold’s staff in Washington hoped to fulfill their obligation with a single B-29 group, perhaps even a reinforced squadron.95 Harmon was more realistic in his estimate,96 and LeMay, who replaced Hansell in mid-January, soon agreed that a larger force would be needed. His plan, submitted on 26 January, called for 1,500 mines to be laid in April and for an entire wing to be used-though not exclusively – in the project.97 LeMay selected the 313th Bombardment Wing (VH), then establishing itself at North Field, Tinian. Its planes were equipped with the AN/APQ-13 radar, suitable for the task at hand, and minor modifications to provide anchorage for parachute static lines could be done locally. Training, which began in February, consisted of indoctrination in the theory of aerial mining and a series of four to eight practice flights involving five radar approaches each, with a couple of mine drops on the last flight.98

Tactics, as they were worked out in command and wing headquarters, were influenced by considerations similar to those which had shaped incendiary tactics and by the experience of XX Bomber Command in CBI. Daylight missions were rejected because they would entail high-altitude formation flights, expensive in fuel and with parachuted mines generally inaccurate even on clear days. The decision was for individual approaches at night with a radar release; this technique would be safer, more accurate, less dependent upon weather, and far more efficient in terms of useful load. To increase the lift potential, LeMay removed from the B-29’s .50-caliber ammunition and two crewmen.99

The Navy, as it had promised, provided technical assistance and logistical support, gearing its production and shipping programs to meet the 313th Wing’s requirements.100 Channels were fairly complicated: the wing’s requests went up through XXI Bomber Command and the Navy’s Commander Forward Area, who was in charge of ammunition storage and allocation of shipping, to the higher Navy echelons. Though complex, this system worked fairly well in routine logistical support and in emergency calls for materiel, personnel, and technical assistance; such shortages in mine types as occurred were attributable to slow production rather than to theater red tape. Mines were prepared, tested, and stored on Tinian by the Navy’s Mine Assembly Depot No. 4. The 313th Wing transported the mines to its own area, installed parachutes and other flight gear, and through a specialized unit designed modifications to meet immediate tactical

Page 666

needs. The command was fortunate in having as mining officers Capt. G. A. Grossman, an early convert who had been connected with the course in aerial mining at the tactical school at Orlando, and Cdre. Ellis A. Johnson, USN.101

LeMay’s operational plan, as he described it in a directive to the 313th Wing on 23 January, was conceived as a four-phase program: (1) 15 February to 15 March, training; (2) 16 to 31 March, partial blockade with standard mines; (3) 1 April to 31 May, complete blockade with new-type mines; and 4) 1 June and after, “further mining.”102 His target list, based on Navy studies, was in the following order of priorities: Shimonoseki, Bisan Seto, Kobe-Osaka, Hiroshima-Kure, Sasebo, Nagasaki, Nagoya, Tokyo-Yokohama, Yokosuka, Tokuyama, and Shimizu.103 This program Arnold approved, but Norstad made it clear that the commitment to mining was an experimental one which should not be allowed to interfere with established bombardment policies, as some in Washington feared might happen.104 In spite of these misgivings in Arnold’s headquarters, LeMay went on with his preparations and on 11 March ordered the 313th Wing to execute its first two mining missions, coded appropriately STARVATION I and II, between 22 and 27 March; later the dates were postponed to between 27 March and 1 April.105 The target was the Shimonoseki Strait, always an important thoroughfare but at the end of March, for reasons that may be described briefly, certainly the most important shipping center in the Empire.

Japan had entered the war with about 6,000,000 tons of shipping, to which 823,000 were added by seizures during the early conquests. This sizable merchant marine was divided about equally between the army, navy, and civilian pools; the lack of a common control made for inefficient employment, and the failure of a plan to return needed tonnage to civilian use put a continuing burden on Japanese industry, Long-range shipbuilding programs and facilities were grossly inadequate for a major war, and no provision had been made for a convoy system; consequently, Japan was wholly unprepared for the Allied attacks on shipping which began immediately after Pearl Harbor. Even in 1942, sinkings exceeded replacements and thereafter the net losses increased in spite of redoubled efforts in the shipyards (which produced by V-J Day 4,100,000 tons of ships) and in spite of the establishment of convoy routes. Until late in the war, and for the whole of the war, the submarine was the chief killer, but it was ably

Page 667

seconded and made more effective by Navy, AAF, and Marine planes. The steady war of attrition was punctuated by especially heavy losses inflicted during the amphibious campaigns and the great carrier strikes which began in 1944. As early as December 1943 the Japanese started closing down convoy routes; by the following September they had abandoned regular contact with the South and Southwest Pacific and the mandated islands. The Philippines campaign produced a crisis, destroying 1,300,000 tons of shipping and threatening the routes southward to the Indies. The capture of Iwo Jima and the imminent assault upon Okinawa completed the stoppage of regular traffic to the south: harbors on Tokyo and Ise bays became less active and the convoy routes southward from Kyushu to Formosa to Singapore were given up. By March 1945, according to Japanese sources, thirty-five out of forty-seven regular convoy routes had been closed down; an additional burden had been put, where possible, on Japan’s inadequate rail system, and traffic between the home islands and the Outer Zone was confined to the Yellow Sea, the Tsushima Strait, and the Sea of Japan. This situation enhanced the importance of ports on the Asiatic side of Kyushu and Honshu and in the Inland Sea, a sheltered natural canal which had long been the vital central link in Japan’s transportation system. The southern entrances into the Pacific at either end of Shikoku were no longer used, and the great bulk of Japanese shipping passed through the eastern narrows, Shimonoseki Strait.106

This was the strategic situation that had determined priorities in LeMay’s target list. At the end of March, Shimonoseki had also a tactical importance. The assault on Okinawa was set for 1 April, and mining the strait would block the flow of reinforcements and supplies and would restrict the movements of the remnants of the Japanese fleet. To close the strait, LeMay expected to use 1,500 mines. Brig. Gen. John H. Davies, commanding the 313th Wing, set up the first attack for the night of 27 March with 3 formations totaling 105 B-29’s; planes were to go in singly, with only enough time between formations to prevent mix-ups. With release altitudes set at from 5,000 to 8,000 feet, most of the B-29’s were able to carry 12,000-pound loads, a mixture of 1,000- and 2,000-pound acoustic and magnetic mines.107 The planes got off as scheduled and ninety-two dropped mines in the primary areas. Enemy air opposition was light, but at low altitudes the B-29’s ran into a lot of antiaircraft fire, including some from ships in the Wakamatsu area, and three planes were

Page 668

lost. Minefield M (the western approach to the strait) was laid pretty much as planned though slightly south of the main shipping lane; in Minefield L (the eastern approach) a heavy concentration was laid in the main channel, but the field thinned out on either side and a rack failure in one B-29 left a three-mile gap.108

In the second mission, flown on the night of 30 March, eighty-five B-29’s mined primary targets, completing the closure of Minefield L and blocking the approach to Sasebo and the southern approach to Kure and Hiroshima; the northern approach, where the attack was weakened by the aborting of four planes, was not entirely closed but was rendered dangerous for traffic.109

Although the Japanese were caught flatfooted by the mining attack, they immediately reacted with minesweeping activities, and to keep the fields replenished LeMay ordered the 313th Wing to sow a minimum of 2,000 mines in April.110 Between the 1st and the 12th, Davies ran five small missions in which a total of forty-five B-29’s, operating without a single loss, mined a number of areas: the two Shimonoseki fields, the approaches to Hiroshima-Kure, and Kure harbor itself.111 But the 313th, like the command’s other wings, was involved in support of the Okinawa battle and in strategic strikes and was behind schedule in mining; by 18 April it had planted only 367 of the 2,000 mines, so LeMay canceled the remaining small-force missions in favor of 2 full-wing efforts lifting 1,500 mines.112 Other demands continued to interfere, and the wing, although dropping 1,070 tons of mines in 2 nights in March, expended only 288 tons in April.113

LeMay’s directive of 18 April called for repeat missions to Shimonoseki, for blockading completely the approaches to Kobe and Osaka, and for laying only attrition fields at Tokyo, Yokohama, and Nagoya harbors. The latter objectives reflected an appreciation of current traffic patterns which has been validated by Japanese statistics made available after the war. Kobe and Osaka, at the eastern end of the Inland Sea, constituted together a shipping area second only to Shimonoseki Strait; the other ports, opening on the Pacific, had declined to where they were not worth an intensive campaign. Expressed in terms of the monthly average in the peak year 1942, tonnage handled in March 1945 was as follows: Kobe, 71.7 per cent; Osaka, 48.1; Yokohama, 11.6; Nagoya, 4.7; Tokyo, 2.3.114

The missions were not run until early May when the strikes in support of ICEBERG were tapering off. On the 3rd, 88 B-29’s sowed

Page 669

668 mines, with good patterns in the Kobe-Osaka area and Shimonoseki’s Minefield L; on the other side of the strait, where only 17 of 30 planes assigned were able to attack, the field was not a tight one. On this mission the new A-6 pressure mechanism was used for the first time, and when a string of mines equipped with it fell into shallow water, the device was considered compromised.115 On 5 May, eighty-six B-29’s planted mines in eight fields. At Tokuyama, Aki-nada, Bingo-nada, and Shodo-shima-Bisan Seto, the patterns were good, but at four port areas – Nagoya, Kobe-Osaka, Hiroshima-Kure, and Tokyo – results were less satisfactory.116

In spite of occasional faulty drops, however, and the slowdown in April, the mining campaign had got off to a fine start. The enemy’s first reaction had been to freeze his shipping in harbors affected by an attack until a channel could be cleared. By sweeping, by bombing shallow fields, and by using small suicide vessels, he was able to open some channels, but his countermeasures, never very effective, were made difficult as the 313th Wing varied its tactics and its mines and increased the number of its targets. The shortage of foodstuffs, aggravated by the B-29 fire blitz which had destroyed 25 per cent of the emergency rice stocks, was so stringent that ships had to keep moving; it became customary to allow the individual ship captain to decide whether to attempt to run a ruined channel or not.117 Some ships were able to bull it through, others were hit. LeMay’s A-2 estimated that by 27 April more than thirty ships had been sunk or damaged at Shimonoseki.118 Postwar investigations scaled this down to 18 ships of 30,917 tons sunk or permanently disabled,119 but the exact figure is of less importance than the fact that much of Japan’s shipping was immobilized after each attack: the real aim was blockade rather than attrition. Tactical results of the mining had also been gratifying. No large warship passed through the Shimonoseki Strait after 27 March; some destroyers used it during the Okinawa campaign, but according to Japanese reports, at least four were sunk. When the task force headed by the Yamato sortied from the Inland Sea on 6 April, it was forced to use the Bungo Strait between Kyushu and Shikoku, a course which led to its detection and defeat. This futile gesture, and other movements of the remnants of the fleet, diverted minesweepers from their more important task of keeping the merchant fleet afloat.120

The missions of 3 and 5 May were actually a part of the planned April effort; before they were run, on 1 May, LeMay issued a target

Page 670

directive for the new month calling for 1,500 more mines to be divided between Shimonoseki and the harbors along the west shore of Kyushu and the northwest coast of Honshu.121 The latter targets had been listed by the COA back in October, and now with the Inland Sea bottled up they had assumed a greater importance. Between 13 and 27 May the 313th Wing sent out 209 planes in 8 missions to lay 1,313 effective mines. Most of these went to Shimonoseki, the others being divided among a number of ports on the Sea of Japan: Miyazu, Maizuru, Tsuruga, Fushiki, Nanao, Niigata.122

During May the Japanese made some improvements in sweeping methods, particularly against the A-3 acoustic mine, but as they admitted later, the weight, spread, and variety of the 313th Wing’s attack made countermeasures extremely difficult. The small western ports, where minesweepers were scarce, remained closed three to five days and traffic was dangerous long afterward. Even at Shimonoseki Strait the Japanese were able to sweep only a narrow channel, 200 to 500 meters wide, and that imperfectly. The strait was closed completely for four days in May and partially on other days; even with these precautions about a third of the ships attempting the passage were put out of service. Through traffic shrank to 404,000 tons, less than half the April figure, and there was a daily average of 80 ships tied up by the mine blockade. During May mines for the first time took a heavier toll of shipping than submarines, sinking or permanently disabling 85 ships of 213,000 tons, about 9 per cent of the existing merchant marine. Ship repair yards, some suffering from the direct or indirect effects of B-29 area attacks, were made difficult of access by the blockade just as they were needed most; shipbuilding could not keep up with sinkings, and the pooling of all cargo vessels under the new War Power Council came too late to be of any service.123

LeMay’s original plan had been indefinite in respect to mining operations after 1 June; there had been some hope in his headquarters that large-scale mining might taper off into a policing job, but in the face of the evident success of the campaign, he was bound to continue it. Actually, in spite of a full docket of bombing, LeMay stepped up the pace of the mining program. He gave first priority to Empire strikes, even for the 313th Wing, but ordered Davies to use I group, when not otherwise employed, to sow 4,050 mines in small increments.124Davies assigned the task to Lt. Col. Charles M. Eisenhart’s 505th Bombardment Group (VH),which flew fourteen missions between 7 June and 3 July. In 404 B-29 sorties the group planted 3,542

Page 671

effective mines in 10 areas in the Inland Sea and Sea of Japan: the two Shimonoseki fields, Kobe-Osaka, Fukuoka-Karatsu, Fushiki-Nanao, Niigata-Sakata, Funakawa, Nagi-Senzaki-Yuya-wan, Sakai, and Maizuru-Miyazu-Obama-Tsuruga. Mines were adjusted to sink ships of various sizes, and the proportion set for smaller ships (55 per cent for vessels of 2,000 tons or less), like the unimportance of many of the targets in normal times, was an index of the desperate condition of Japanese shipping.125

The task of modifying and loading the various types of mines to be used for each mission was an exacting one, done usually under pressure of time, for mines were rarely available more than a day or two before needed. The supply situation, always tight, was further strained by the increased weight of attacks. Sometimes the designated mine types were not available in proper quantities; sometimes missions had to be postponed until a shipment of mines arrived – the slowdown in April was caused in part by lack of mines and mine assemblies.126 The shortage in any mine type worked against the system of mixing the weapons used on each mission and aggravated the concern lest the Japanese develop effective countermeasures against all.This concern, a healthy sign in a combat outfit, proved to be superfluous. Interviews with Japanese mining experts after the war indicate that they had little chance to break the blockade. After being caught by surprise, they tried desperately to meet the new form of attack: in all, 349 vessels and 20,000 men were assigned to minesweeping; staffs engaged in research on countermeasures were doubled and given highest priority. But the scientists were not given full cooperation by the military; measures for detection remained inefficient, making small use of radar and depending chiefly on visual spotting by watchers on shore or afloat. On the basis of these methods the Japanese estimated that the B-29’s dropped only 3,690 mines during the campaign instead of the more than 12,000 actually expended. They developed a fairly effective method of sweeping acoustic mines and tried, unsuccessfully, to sweep magnetic mines with airplanes. They were able to explode subsonic mines with bombs (many others went off prematurely without enemy help) and apparently brought out late in June a new floating electric loop switch for use against magnetic mines. But the A-6 pressure mechanism, very effective on the MK25 Model 2 mine, was apparently not picked up by the Japanese until 27 May and no sure defense against it was discovered.127

Whatever the merit of the enemy’s countermeasures, the weight of

Page 672

the combined attack against his shipping was too heavy to combat, and by July his shipping situation was hopeless. Submarines, which had moved into the Sea of Japan in April, sank 92,000 tons in June; aircraft, operating in Japanese and Korean waters, accounted for 56,000 tons, including 18,000 destroyed by B-29’s in an incendiary raid on Osaka on 1 June. Mines dropped by the 313th Wing sank or disabled 83 ships of 163,000 tons, some in the Sea of Japan, but more than half of them in the Shimonoseki Strait. The decline in tonnage sunk as compared with May results was indicative of the shrinkage in the enemy’s merchant marine rather than of any improvement in defensive measures. Ports opening onto the Pacific had closed down for the rest of the war: Nagoya, on 27 April; Shimizu, 14 May; Yokohama, 23 May; Tokyo, 27 May; Shiogama, 29 June. During June, Shimonoseki Strait was again closed for five days; by the end of the month Moji, the principal port, and the anchorages at Matsue and He-saki appeared to be completely abandoned-ships were anchoring in the swift current of the strait or in small unmined harbors outside. Small wonder that a member of the War Power Council excused its failure to relieve the situation by remarking that “all members knew the circumstances, and knew that Japan was hopeless.”128

During the last weeks of the war, the 313th Wing continued to cooperate with other aircraft and with submarines in a crescendo of attacks on the enemy’s dwindling merchant fleet. Army and Marine planes from Okinawa struck at shipping along a broad arc from Kyushu to the Asiatic continent.* Carriers from the Third Fleet dealt a devastating blow to shipping and the railroad ferries in the Tsugaru Strait areas, disrupting traffic between Hokkaido and Honshu; later, with the aid of British carriers, they hit various harbors in the Inland Sea. The B-29’s continued their blockade of the Shimonoseki Strait and northwest Honshu ports and extended their coverage to include Korea. Because southern Korea was within range of planes based on Okinawa, the 313th Wing concentrated on ports farther up the peninsula – Wonsan, Hungnam, Chongjin, and Najin. For these targets, correctly assumed to be weak in countermeasures, the wing used magnetic and acoustic mines, types against which the enemy had developed a partial defense; this saved for Shimonoseki the more lethal pressure mines, and when other types were used there they were made more dangerous by the use of arming delays and ship-counting mechanisms.129

* See below, pp. 695-99.

Page 673

Davies used in the last series of mining missions two successive groups-the 6th Bombardment Group (VH) from 8 to 20 July and the 504th from 24 July to 14 August. In 15 missions, 445 B-29’s out of 474 airborne planted 3,578 mines in 17 fields. About thirty planes went out on each mission, divided among several targets. Because of the distance involved, B-29’s mining Korean ports staged through Iwo Jima. Losses were somewhat heavier than in previous months, amounting to six B-29’s, of which three were destroyed by the enemy. The Japanese had moved more heavy AA guns and searchlights into the Shimonoseki area, forcing the B-29’s to raise the altitude of attack to over 12,000 feet on occasion. Parachuted mines dropped from that height sometimes drifted widely, but even without the strays there were more than enough to tighten the blockade.130

Kobe-Osaka, which had cleared 320,000 tons in March, handled only 44,000 in July. Between 1 July and 14 August the Shimonoselti Strait was closed completely on 16 days and on many others only a ship or two got through: the total for the first 15 days of August was only 30 ships of 29,954 tons, about 7 per cent of the March traffic. In spite of the increasing stringency of the blockade, the enemy was forced to send his ships out without thorough minesweeping; in spite of the diminished size of the merchant fleet, he lost 478,000 tons of shipping in July, of which 198,000 were attributable to mines.131

During July the 313th Wing dropped 4,500,000 propaganda leaflets urging the Japanese to surrender before they suffered starvation. This was no idle threat. By mid-August the merchant marine had been reduced to about 1,500,000 tons afloat (exclusive of the useless tankers) and sea-borne traffic had almost ceased. Dependent upon imports for much of their food supply, the Japanese had cut off all shipments of other raw materials and were using the scanty traffic entirely for foodstuffs from the mainland, but blockade runners could do little toward supplying the demand and the caloric content of the average man’s fare had shrunk dangerously.

The 313th Wing’s B-29’s were not, of course, solely or mainly responsible for those conditions; blockade and attrition had been the coordinated task of the several services as the previous pages have shown. In its final report, the U.S. Strategic Bombing Survey credits all agencies with sinking 8,900,000 tons of Japanese shipping, and divides the credit according to the following list of percentages: submarines, 54.7; carrier-based planes, 16.3; AAF planes, 10.2; mines (largely dropped by B-29’s), 9.3; Navy and Marine land-based planes,

Page 674

4.3; surface ship gunfire, less than 1; and the remainder, about 4 per cent, to marine accidents. The 313th Wing got into the game late, operating with mines for only four and one-half months and at a period when the enemy’s merchant fleet had contracted in size and in scope of its activities. During that short period mines planted by the wing were more destructive than any other weapon, accounting for about half of the total tonnage disposed of. To accomplish this task, the 313th sent out 1,528 sorties and planted 12,053 mines, a much heavier effort than had been suggested by the Navy in the negotiations of 1944 and, indeed, the heaviest aerial mining campaign ever waged. That this could be accomplished in the midst of a rigorous bombing program and at a loss of only sixteen planes (only nine to enemy action) speaks well of the efficiency of the 313th Wing and of mine warfare.132

In the early weeks of the campaign, mining operations carried a top-secret classification, and LeMay was concerned lest the absence of any public recognition hurt the morale of B-29 crews engaged in mining – an unsatisfactory type of operations at best since the crewman never sees the results of his strike.133 Recognition came eventually however: the British, old hands at the game, said the B-29 mining was “very much like a dream come true,” and Nimitz’ messages of congratulations, including the final one in which he spoke of “phenomenal results;” went beyond the demands of interservice protoco.134l More restrained, but no less satisfying, was the postwar remark of a Japanese mine expert who had been in charge of minesweeping in the Inland Sea: “Surely B-29’s as a minelaying weapon were quite a hit in this war.”135 More eloquent than any encomium, however, were the bare statistics.

Table I: Incendiary Missions Against Secondary Cities

Mission No. Date Target Population Square Miles Destroyed Per Cent of Total
206 17 June 45 Kagoshima 190,250 2.11 44.1
207 17 June 45 Omuta 177,000 0.217 4.1
208 17 June 45 Hamamatsu 165,000 2.44 70.
209 17 June 45 Yokkaichi 102,000 1.23 60.
210 19 June 45 Toyohashi 142,700 1.7 52.
211 19 June 45 Fukuoka 323400 1.37 21.5
212 19 June 45 Shizuoka 212,200 2.25 66.
234 28 June 45 Okayama 163,560 2.13 63.
235 28 June 45 Sasebo 206,000 0.97 48.
236 28 June 45 Moji 139,000 0.302 26.9
237 28 June 45 Nobeoka 79,426 0.52 36.
240 1 July 45 Kure 277,000 1.3 40.
241 1 July 45 Kumamato 211,000 1.0 20.
242 1 July 45 Ube 100,600 0.42 23.
243 1 July 45 Shimonoseki 196,000 0.51 36.
247 3 July 45 Takamatsu 111,200 1.4 78.
248 3 July 45 Kochi 106,650 0.92 48.
249 3 July 45 Himeji 104,250 1.216 63.3
250 3 July 45 Tokushima 119,600 1.7 74.
251 6 July 45 Chiba 92,000 0.86 43.4
252 6 July 45 Akashi 90,000 0.81 57.
253 6 July 45 Shimizu 68,600 0.71 50.
254 6 July 45 Kofu 102,400 1.3 65.
257 9 July 45 Sendai 233,630 1.22 27.
258 9 July 45 Sakai 182,150 1.02 44.
259 9 July 45 Wakayama 195,260 2.1 52.5
260 9 July 45 Gifu 172,340 1.93 74.
263 12 July 45 Utsunomiya 87,868 0.94 34.2
264 12 July 45 Ichinomiya 70,800 0.01 0.8
265 12 July 45 Tsuruga 31,350 0.77 68.
266 12 July 45 Uwajima 52,100 0.14 14.
271 16 July 45 Namazu 53,165 1.4 89.5
272 16 July 45 Oita 61,000 0.555 25.2
273 16 July 45 Kuwana 41,850 0.63 77.
274 16 July 45 Hiratsuka 42,150 1.04 44.2
277 19 July 45 Fukui 98,000 1.6 84.8
278 19 July 45 Hitachi 82,700 0.88 64.5
279 19 July 45 Choshi 61,200 0.379 33.8
280 19 July 45 Okazaki 84,070 0.65 68.
293 26 July 45 Matsuyama 66,300 1.22 73.
294 26 July 45 Tokuyama 38,400 0.47 37.
295 26 July 45 Omuta 177,000 2.05 38.
297 28 July 45 Tsu 68,625 0.84 57.
298 28 July 45 Aomori 100,000 1.06 64.
299 28 July 45 Ichinomiya 70,800 0.99 75.
300 28 July 45 Uji-Yamada 52,555 0.36 39.
301 28 July 45 Ogaki 56,100 0.48 40.
302 28 July 45 Uwajima 51,100 0.53 52.
306 1 Aug. 45 Hachioji 62,280 1.12 80.
307 1 Aug. 45 Toyama 127,860 1.87 99.5
308 1 Aug. 45 Nagaoka 67,000 1.33 65.
309 1 Aug. 45 Mito 261,300 1.7 65.
312 5 Aug. 45 Saga 50,400 0.02 1.5
313 5 Aug. 45 Maebashi 87,000 1.0 42.5
314 5 Aug. 45 Nishinomiya-Mikage 111,800 2.8 29.6
316 5 Aug. 45 Imabari 60,000 0.73 76.
319 8 Aug. 45 Yawata 261,300 1.22 21.
321 8 Aug. 45 Fukuyama 56,653 0.88 73.3
329 14 Aug. 45 Kumagaya 49,000 0.27 45.
330 14 Aug. 45 Isezaki 40,0000 166 17.

Page 675

Remainder of table merged with previous page