Abstract
The Japanese convoy HI-72, attacked by a wolf pack of American submarines after it departed Singapore in September 1944, has gained some notoriety because two of the ships sunk carried Allied prisoners of war. Beyond this, however, the convoy’s fate highlights some of the factors which made US submarines so effective while disclosing the shortcomings of Japanese escorts. An examination of the extant battle reports by the convoy ships reveals that the Japanese were aware of many of the factors which contributed to their shipping losses, but also unaware of how heavily the odds were stacked against them at this stage of the war.
Convoy HI-72 departed Singapore for Japan on 6 September 1944, transporting raw materials, Japanese troops, and passengers including Allied prisoners of war. Six days later the convoy was attacked by a three-boat wolf pack of US submarines which sank six ships in the convoy, including two transports carrying POWs. To the extent that historians have examined convoy HI-72, their attention has been on the fate of the prisoners, especially given the unique circumstances in which some of the POW survivors from the Rakuyo Maru were rescued by the same submarines which attacked the convoy. 1 The current study shifts the focus from prisoners of war to convoy HI-72 as a case study in the conflict at sea. Through this approach, the convoy’s fate highlights some of the factors which made US submarines so effective at this stage of the war, while also revealing some of the shortcomings in Japan’s efforts to protect merchant shipping. It is argued that while the Japanese appreciated many of these factors at the time, they were also unaware of how heavily the odds were stacked against them by late 1944.
I
HI convoys were a series of convoys between Singapore and the Japanese home islands. Convoy HI-72 initially consisted of six transport ships: two tankers, Zuiho Maru (5,135 tons) and Shincho Maru (5,136 tons); the passenger liner Kachidoki Maru (10, 509 tons) and the passenger-freighter Rakuyo Maru (9,418 tons), both carrying POWs; the freighters Asaka Maru (7,400 tons) and Nankai Maru (8,416 tons). In addition, there were four escort ships: three kaibokan (sea defence ships) Hirado, Kurahashi and Mikura, as well as the destroyer Shikinami. The convoy commander with authority over the transports was Captain Sukehiko Hosoya of the 16th Convoy Command Section who travelled aboard Kachidoki Maru. Rear Admiral Sadamichi Kajioka of the 6th Provisional Escort Convoy Command was in charge of the escorts and made Hirado his flagship.
As not uncommon for HI convoys, ships from another port joined HI-72 en route to Japan. In this instance, seven ships of convoy MAMO-73 from Manila made a rendezvous on 11 September about 100 miles northeast of the Paracel Islands. MAMO was the designation given to convoys travelling from Manila to Moji, Japan. This convoy included three transports: Kagu Maru (8,416 tons), Gokoku Maru (10,438 tons), and Kibitsu Maru (9,564 tons). The convoy was protected by four escort vessels, but three of these quickly returned to the Philippines after the convoys merged. The combined convoy proceeded north across the South China Sea at 10.5 knots, taking a zigzag course to discourage submarine attack. 2 As typical of Japanese convoys, the ships travelled in two columns with the escorts ranging up and down the flanks where it was believed merchant ships were most in danger of attack. Given the limited number of escorts, this arrangement left many gaps in the screen which could be exploited by submarines. Contrary to British convoys in the Atlantic, where sheer size and numerous columns offered the most valuable targets like oil tankers extra protection in the centre, the relatively small size of Japanese convoys meant that most ships were equally vulnerable. 3
In the early hours of 12 September the convoy suffered the first of a series of attacks by a wolf pack of American submarines including USS Growler, USS Sealion II, and USS Pampanito. As customary, the Americans nicknamed the wolf pack ‘Ben’s Busters’ in honour of the senior skipper, Thomas Benjamin ‘Ben’ Oakley, Jr, who commanded Growler. At about 2.00 a.m., Growler launched the first torpedo attack on the convoy as an escort bore down on it. The escort, Hirado, was observed by Growler’s skipper to explode ‘in a large ball of yellow flame and black smoke’. 4 The escort force’s commander, Admiral Kajioka, was lost with Hirado, probably contributing to the melee which followed. Before dawn USS Sealion fired six torpedoes at two overlapping ships in the convoy, hitting Nankai Maru and Rakuyo Maru. The submarine was then heavily depth charged. In the meantime, Growler returned to the fray in a submerged attack, firing six torpedoes at the Fubuki-class destroyer Shikinami. By daybreak the transports Nankai Maru and Rakuyo Maru, as well as the destroyer Shikinami, were sunk or sinking. 5
The diminished convoy continued sailing, but came under attack again at around 9 p.m. when approximately 200 miles east of Sanya, the southernmost city of Hainan Island. USS Pampanito sank two more convoy ships – the transport Kachidoki Maru and the tanker Zuiho Maru. Kachidoki Maru had begun its colourful history as an American passenger-cargo ship, SS President Harrison, before falling into Japanese hands in 1941. Captain Hosoya went down with Kachidoki Maru, depriving the convoy of its commander in addition to the already lost escort commander.
The surviving ships headed for sanctuary at Sanya, where they reorganized into a fast-ship group and a second subgroup of slower ships. Both groups sailed separately from Sanya on 16 September, but remained in harm’s way. In the early hours of 20 September, about 30 miles south of the Pescadores Islands, the fast-ship group was attacked by a formation of American B-24 bombers based on the Chinese mainland. Had the destroyer Shikinami survived the initial submarine assaults, it might have contributed to the ships’ defence since it had been fitted with additional anti-aircraft guns during maintenance in March–April and July 1944. 6 As it transpired, the transports Asaka Maru and Gokoku Maru were left dead in the water after suffering direct bomb hits and had to be towed to Mako in the Pescadores. Another transport, Kagu Maru, was also damaged as was the escort ship Mikura. Left without power and communications, Mikura drifted for three days until spotted by a Japanese search aircraft, and it too was eventually towed to Mako. In the meantime, the group of slower ships also came under attack from China-based B-24 bombers. At about 2.00 a.m. on 21 September, approximately 50 miles west of Takao, Formosa, the tanker Shincho Maru was hit by a bomb and lost all power. Fortuitously, convoy HI-74 was in the vicinity and one of its escorts, CD-21, came to the tanker’s aid and towed it to Takao.
The surviving fast ships reorganized again, this time at Takao and Keelung Harbour, Formosa. On 25 September the transports Asaka Maru and Kibitsu Maru, with three escorts, set sail from Keelung Harbour, but came under attack two days later by the submarine USS Plaice about 100 miles northwest of the island Amami Ōshima between Okinawa and Kyushu. The escort ship CD-10 was torpedoed and sunk. CD-10 had only been launched in January 1944, and in March was assigned to the Grand Escort Command’s First Surface Escort Division. 7 The remaining ships in the group dispersed and made their way independently to Japan, arriving at Moji on 28 September.
II
The fate of convoy HI-72 illustrates just how effective American submarines had become by this stage of the war, as well as the increasing role of aircraft in disrupting Japanese supply lines. Of the ten ships that departed Singapore, six were sunk on 12 September, while most of the remaining ships from the convoy subsequently sustained damage or were sunk. Among the nine transports and tankers in the combined convoy, four were sunk and another four damaged. In addition, two escorts and a destroyer were sunk, while another escort sustained damage. The only transport to survive the convoy unscathed was Kibitsu Maru. 8
The lack of extant naval records and ship manifests makes it impossible to calculate the total cargo lost as a result of the attacks. It is known that the ships carried rubber, mercury, drummed aviation fuel, at least 12,500 tons of bauxite, and over 12,000 tons of oil. 9 Nankai Maru had taken on 6,500 tons of bauxite, used for the production of aluminium, as well as drums of gasoline before departing Singapore. 10 According to some sources, in addition to bauxite the cargo included 4,000 drums of gasoline, 170 tons of oil, 525 passengers, 77 packages of mail, and the ashes of 18 war dead in urns. Kachidoki Maru carried another 6,000 tons of bauxite, as well as the ashes of 582 war dead. The tanker Zuiho Maru departed Singapore with 8,000 tons of oil. 11 The loss of oil was especially critical to the Japanese war effort, since it is claimed that from 1943 the lack of oil supplies proved the greatest limitation on Japanese naval operations. 12 From February 1944 tankers were designated priority targets for US submarines and they quadrupled the number sunk during that year. 13 In fact, the sinking of five Japanese tankers escorted by only one ship in February 1944 led to a proposal that all convoys between Singapore and Moji have at least three escorts in future. 14 Even so, it is claimed that only 13.5 per cent of the oil shipped from the East Indies during 1944 managed to reach Japan. 15
In addition to the problem of calculating the material loss from convoy HI-72, it is difficult to confidently calculate the loss of life. Of the 950 British prisoners of war on the Kachidoki Maru, 431 perished. The sinking also resulted in the loss of 45 passengers and 12 crewmen. There was even greater loss of life from Rakuyo Maru’s 1,317 prisoners. Days after their initial attack, US submarines recovered 159 British and Australian survivors, while a Japanese ship picked up another 136 prisoners from lifeboats. The majority of prisoners on Rakuyo Maru, over one thousand, perished at sea. A dozen crew and Army personnel from the ship were also lost. The Japanese casualties included 100 sailors lost with the destroyer Shikinami. It appears that 196 of the 525 passengers on the Nankai Maru were lost, along with three crewmen. Only eight men were rescued from the escort CD-10, while the remaining 148 personnel on board were lost. 16
There is little surviving evidence of the Japanese reaction to these losses. The naval records available for this period are scant, and because the convoy’s experience was highly classified there may have been little written in reference to it from the outset. Furthermore, since massive losses had become more or less normal, the convoy’s fate may not have been viewed as especially noteworthy. To take but one example, convoy HI-71, which departed Moji for Singapore on 8 August 1944, fell victim to US submarines when traversing the area between Mako and Manila on 17–18 August. Among the 20 transport ships in the convoy, four were sunk and another three damaged, while the loss of life included over 8,000 Army personnel. The submarines also took a heavy toll on the convoy’s escorts, sinking the sole escort carrier as well as three kaibokan. 17
It may be suggested that to some extent the Japanese were gradually numbed to their shipping losses. Early in the war, from December 1941 through September 1942, shipping losses to all causes amounted to less than 100,000 tons per month. Losses began to grow from October 1942, however, when losses of over 100,000 tons of shipping each month became common. From September 1943 losses continued to escalate, so that for the period through December 1944 Japan was losing on average 294,000 tons of shipping per month. Even though Japanese shipyards nearly doubled their production during 1944, Allied submarines and aircraft destroyed the equivalent of 18 months of shipbuilding between September and November of that year. 18 In the month of September 1944 alone, submarines claimed 49 Japanese ships totalling over 181,000 tons. 19 In this broader context, the destruction of convoy HI-72, while far from inconsequential, would not have appeared extraordinary either.
III
Despite a relative dearth of official military documents, the battle reports of Nankai Maru and Asaka Maru provide some insights into Japanese thinking at the time. Both reports included a section on ‘lessons learned’ which is highly instructive on Japan’s views of the submarine war being waged against them. For this reason, their assessment of US submarine strengths and the weaknesses of Japanese protection for shipping is worth considering in some detail. Given the toll taken by aircraft on the remnants of convoy HI-72, the battle reports in part addressed countering the success of low-altitude bombing attacks. It was recommended that convoys should regulate their speed more carefully at night, in order to avoid leaving telltale wakes. Also at night, it was recommended that ships should begin defensive anti-aircraft fire as soon as enemy planes were suspected to be within range, rather than waiting until enemy aircraft were definitely identified. 20 For the most part, however, the reports focused on the threat posed by submarines.
The battle reports attributed much of the American submarine success to their use of electronic equipment, specifically radar and wireless telephones. In the case of radar, the Japanese were certainly correct in identifying one of the Americans’ principal assets at this stage of the war. SJ surface search radar was arguably the most important technical development of the submarine war. 21 The Japanese had entered the war with night-vision superiority due to the quality of their binoculars, and early on the Japanese Navy gained a high reputation for their night actions in the Solomon Islands. 22 SJ radar allowed the Americans to turn the tables, providing them with a clear advantage in low-visibility conditions and night actions.
The first SJ radar sets were trialled on USS Haddock in August 1942, although initially there were teething problems. 23 For example, following the USS Snook’s first patrol using the equipment, skipper Charles O. ‘Chuck’ Triebel described it as ‘temperamental and delicate’. 24 The technology steadily improved, however, with submarines able to make contact on the enemy from impressive ranges. Having received a radio message on the track of convoy HI-72, the submarines Sealion, Growler, and Pampanito formed a scouting line covering their radar horizon, roughly eight miles apart from one another. 25 USS Growler, the first submarine to attack convoy HI-72, initially made radar contact with the convoy ships from almost 30,000 yards (15 miles). 26 Before Pampanito sank Zuiho Maru and Kachidoki Maru, it was also able to make radar contact with the ships from 15 miles. 27 Without radar, it is highly unlikely that Pampanito would have sunk the ships, at least on 12 September. Prior to making its later attack on 27 September, USS Plaice also located and tracked remnants of the convoy with its SJ radar, initially picking it up from 20,450 yards. 28 Given the often small size of Japanese convoys, consisting of five or six ships including escorts, SJ radar proved invaluable in locating them. 29
For American submariners, the use of radar proved a huge advantage in making night surface attacks. The equipment included a Planned Position Indicator (PPI) display which enabled submarine crews to not only track a prospective target, but to monitor the position of its escorts. This in turn facilitated both making an attack and planning an escape. 30 The use of radar in mounting night attacks was further enhanced by the introduction of the Target Bearing Transmitter (TBT), essentially a device that allowed submariners to place binoculars in a bracket on the bridge and easily transmit a target’s bearings to the Torpedo Data Computer. By 1944, well over half of US submarine attacks were being carried out on the surface at night. 31 Growler, for example, carried out its night surface attack on Hirado using radar ranges and TBT bearings. 32
Japanese development of radar failed to keep pace with the Americans. The battle reports of Nankai Maru and Asaka Maru criticized an over-reliance on visual efforts to spot enemy submarines at night. 33 The Japanese military was slow to recognize the potential of radar, and its development was handicapped by inadequate resources as well as an insistence that army and navy electronics research be carried out separately. 34 In late 1942 the Japanese Army began research on radar sets that could be installed on ships for protection against submarines, but when tested in February 1943 they lacked sufficient range. 35 The same month, when Mochitsura Hashimoto assumed command of the submarine I-158, his submarine was fitted with experimental surface radar. The radar proved disappointing, however, and was only able to pick up a surfaced submarine within about 2,000 yards. According to Hashimoto, Japanese submariners longed for radar like farmers longed for rain during a long drought. 36
In early 1944 radar began to be installed on escort ships, and most escorts had radar by the end of the year. 37 As late as August 1944, however, Admiral Matome Ugaki described Japanese radar as ‘far behind’ the Allies, lacking in ‘capacity and accuracy’. 38 The US Naval Technical Mission, which investigated Japan’s military technology immediately after the war, concluded that a lack of confidence in Japanese radar meant it was rarely used for searches. 39 Japanese radar also lacked the PPI display that would have allowed ships to deal with an array of moving vessels. 40 Pampanito reported that when the convoy was attacked, there was no evidence of the enemy using radar. 41
In contrast to the Asaka Maru and Nankai Maru battle reports’ assessment of American radar, it appears that they overestimated the benefits of wireless telephones on board US submarines. By mid-1944, short-distance VHF radio telephones allowed submariners to use voice contact with less danger of being detected by enemy direction finders. 42 The battle reports suggested that this innovation allowed US submarines to communicate and coordinate with one another very effectively. In reality, the submarines which attacked convoy HI-72 largely eschewed this technology. When the VHF radio was tested by the US wolf pack after it departed Midway, Pampanito reported that it was ‘not very satisfactory’, while Growler complained that communications were only good within 7,500 yards. 43 Many American skippers feared that their transmissions might be intercepted or make their submarines vulnerable to enemy direction finding. Before attacking convoy HI-72, Sealion, Growler, and Pampanito made a rendezvous at sea on the evening of 11 September with the boats coming together within about 25 yards of one another. For ‘security’, a discussion was then conducted by voice through megaphones rather than by radio. 44 Once the submarines formed a scouting line to search for the convoy ships, radar was used to keep track of their respective locations; again there was no use of radio. 45 After USS Growler departed for Fremantle on 14 September, Sealion and Pampanito held another rendezvous on the morning of 15 September, again communicating through megaphones. 46
IV
In contrast to the putative advantage of US submarine communications, the battle reports of Asaka Maru and Nankai Maru considered their own ship-to-ship communications as deficient. They relied heavily on signal blinkers, which not only increased the likelihood of detection by the enemy, but were often difficult for smaller ships to see in rough weather. Generally, the poor quality of intra-convoy signals handicapped the manoeuvrability of convoys, especially in response to an attack. While there was some belated success in introducing intra-convoy radio telephones for communication between escort ships, merchant ships remained without the equipment. For convoys fortunate enough to have air support, neither merchant ships nor escort ships were able to communicate directly with the airmen. 47
As recognized by the battle reports, the success of American submarines in making night surface attacks was greatly aided by their high-speed performance. In contrast, the reports pointed out that convoy HI-72 was handicapped by combining fast and slow ships, with the slower ships essentially acting as a drag on their faster counterparts. When USS Growler initially made contact with the convoy, it estimated the convoy’s speed at 9 knots, while USS Sealion later estimated the convoy’s speed at 10.5 knots. 48 The kaibokan used as escorts were typically capable of top speeds ranging from 16 to 19.5 knots. 49 They were faster than the ships they protected, but US submarines could make close to 20 knots on the surface. At one stage following its predawn approach on the convoy, Sealion was pursued on the surface by two escorts firing their guns. Sealion worked up to 19.2 knots in response; the patrol report conceded that ‘It was close and, for a while, too hot for comfort’, but the escorts gave up the chase. Sealion was then able to overtake the convoy again and make another attack torpedoing Nankai Maru and Rakuyo Maru. 50 The battle reports of Asaka Maru and Nankai Maru recommended that fast and slow ships should sail in separate convoys, essentially the ad hoc arrangement made by the remnants of convoy HI-72.
Overall, the anti-submarine warfare capabilities of Japanese escorts were considered insufficient. The fact that Japan entered the war with few escort ships specifically designed for anti-submarine warfare was often misunderstood at the time and since overlooked by historians. As noted by some contemporary commentators, the Japanese Navy’s preoccupation with a decisive fleet battle meant that inadequate resources were provided to effectively protect transport ships and conduct anti-submarine warfare. The battle fleet received first priority for the construction and allocation of ships, as well as the distribution of personnel and equipment. 51 At the end of the war, Atsushi Oi, a staff officer with the Grand Escort Command Headquarters, put the matter succinctly: ‘In a nutshell, Japan failed in ASW largely because her navy disregarded the importance of the problem.’ 52
At the beginning of the war, the only Imperial Japanese Navy ships specifically designed for anti-submarine warfare were about two dozen small sub-chasers (kusentei). 53 In practice, however, these lacked the deep-water sailing capabilities that might have made them effective for protecting merchant ships crossing the Pacific. On the positive side, the kusentei were armed with 36 depth charges, and at least in areas close to land they could effectively defend transports. 54 Between 1940 and 1944, 49 sub-chasers of the CH-13 class were completed for service. These were better armed than their predecessors, but they were also slower with a top speed of 16 knots. 55 The ships remained in short supply; until mid-1943, for example, only three sub-chasers were based at Balikpapan for the protection of tankers passing through the Makassar Straits. 56
The ships which were potentially best suited for anti-submarine warfare, destroyers, were also in short supply for escort duty. Those destroyers employed as escorts tended to be the older and least efficient ships. 57 At the formation of the Grand Escort Command in November 1943, the better escort ships included 15 destroyers built between 1920 and 1925. 58 The heavy consumption of fuel by destroyers proved a growing liability as petroleum grew scarcer. 59 Shikinami, which accompanied convoy HI-72, was launched in June 1929 and had played an active role from early in the Pacific War; the destroyer contributed to the sinking of USS Houston at the Battle of Sunda Strait and later participated in the Guadalcanal campaign. 60 A month before joining convoy HI-72, however, Shikinami ran aground on an uncharted shoal, causing ‘fairly heavy damage’. 61 It is unclear how far this affected the ship’s performance; presumably Shikinami remained formidable, but it was on loan from the Combined Fleet rather than part of a trained ASW team.
In practice, convoys relied heavily on the protection of kaibokan. The Japanese Navy kaibokan, literally translated as sea defence ship, was often described by the Allies as a ‘frigate’. USS Growler’s patrol report mistakenly identified the 860-ton Hirado as a ‘destroyer’ of 2,300 tons. 62 In reality, these ships were originally designed to protect fishing craft in the Kurile Islands and, at least initially, their anti-submarine capabilities were limited. Each kaibokan carried only 12 depth charges until the fall of 1943 when the number of allocated depth charges was increased to 60. 63
Until the second half of the war, the numbers of these ships used by the Navy were few in number. When the war began, the Imperial Japanese Navy included four Shimushu-class kaibokan, with only one attached to the Combined Fleet while the other three were assigned to the Kurile Islands. The Navy did not receive a fifth kaibokan until mid-1943. 64 Between March 1943 and May 1944, construction began on an additional 26 kaibokan of the Etorofu and Mikura classes. 65 By the end of the war Japan had produced a total of 169 kaibokan. 66 It was also relatively late in the war when the Japanese Navy began building lighter destroyers of the Matsu and Tachibana classes specifically for anti-submarine warfare, with the first of the Matsu class commissioned in April 1944. By the end of the war, a total of 32 of these smaller destroyers were built, fitted with sonar, radar and up to 60 depth charges. 67 As with other aspects of protecting merchant shipping, however, this proved too little too late.
The material deficiencies of Japanese escorts were exacerbated by organizational and administrative problems. At the time America entered the war, there was no Japanese unit exclusively dedicated to the protection of merchant shipping. In April 1942 the Japanese Navy established two escort groups, including one devoted to escorting ships from Singapore to Japan, but it was thinly resourced. 68 Responsibility for maritime escort was handled by one subsection within the Naval General Staff, without anyone at a higher command level to make a case for enhanced escort services. 69 The situation improved from 15 November 1943 with the establishment of the Grand Escort Command Headquarters when, at least on paper, this command functioned at a level equal to the Combined Fleet Headquarters, with both commands reporting to the Chief of the Naval General Staff. In practice, however, the maritime escort services remained dependent on the Combined Fleet to release its ships for escort duty. 70 In any case, the creation of an independent command came much too late to turn the tide of the shipping war. Admiral Osami Nagano, Chief of the Naval General Staff, admitted as much at the time, noting that ‘the establishment of the Grand Escort Command at this time may be like calling a doctor only when the patient’s condition has turned critical’. 71
The command and organizational problems at fleet level were replicated in the case of individual convoys. A ‘Convoy Controller’ for each convoy was only appointed immediately prior to its formation and sailing. This position was later renamed ‘Convoy Commander’ with somewhat more authority, but those assuming this role remained unable to sufficiently train the forces under their command as a unit. 72 Escort groups, too, lacked unified training and doctrine, but were temporarily assembled as needed. Most of the escorts were skippered by graduates of merchant marine academies with little knowledge of naval tactics or drawn from reserve officers. 73 It was only in March 1944 that the Imperial Japanese Navy renamed its ‘Mine School’ the ‘Anti-submarine Warfare School’ and directed that the curricula include ASW tactics in addition to mining and minesweeping tactics. Even then the school lacked an actual kaibokan for instruction. 74 The Nankai Maru and Asaka Maru battle reports referred specifically to the lack of unit training for escorts as a reason for their shortcomings in anti-submarine warfare capabilities. They recommended that escorts should be organized into fixed rather than ad hoc formations under a regular commander, and that they should undergo unit training in anti-submarine tactics and convoy escort. 75
Both the material deficiencies of escorts and the lack of specialized training meant that the Japanese anti-submarine tactics were mainly defensive and reactive rather than offensive. Without radar, and often without a radio-direction finding capability, Japanese escorts relied mainly on visual contacts. In the absence of a visual sighting, the typical response to submarine attacks was to drop depth charges indiscriminately in the hope of scaring the attackers off. If a submarine was sighted and submerged, escorts would attempt to establish sonar contact, but there was a blind spot as escorts approached a target. Once a depth-charge attack was initiated, the escorts more often than not lacked persistence, partly because of the necessity of keeping continued watch over the convoy. 76
In the Atlantic, the British developed training, weapons, and tactics to counter the sonar blind spot and make anti-submarine warfare more effective. In a notable example, Commander F.J. Walker developed the ‘creeping attack’ in which one escort used its sonar to drive a submarine into an ambush, along with ‘carpet bombing’ by multiple escorts. Especially effective in eliminating the sonar blind spot ahead of warships was the development of forward-firing and multi-headed mortars. The Japanese lack of a comparable forward-throwing weapon to the Allied ‘Hedgehog’ proved a critical limitation. 77
In light of the decimation of convoy HI-72, the battle reports of Nankai Maru and Asaka Maru advocated the discontinuation of forced passages through submarine-infested waters. Instead, it was recommended that efforts should be made to suppress as many submarines as possible. 78 Presumably what the battle reports had in mind were offensive sweeps by hunter-killer groups. But for all of the reasons discussed above, there was little prospect of countering the danger posed by submarines. Meanwhile, the number of American submarines in the Pacific was increasing along with forward bases closer to Japanese shipping lanes. In the 15 months from February 1943, the number of US submarines patrolling the Pacific more than doubled, from 47 to 104. 79 At the time of the attack on convoy HI-72, the Americans had also recently established a forward submarine base at Saipan. Before hunting and attacking the convoy, Sealion had been able to replenish its torpedoes there rather than being forced to terminate its patrol at Pearl Harbor or Fremantle, Australia. 80
V
Just as important as the observations made by the Asaka Maru and Nankai Maru battle reports were the omissions, for these disclose that the Japanese were still not fully aware of how heavily the odds were stacked against them. Most importantly, the Japanese remained oblivious to the breaking of their codes and Allied signals intelligence collectively known as ‘Ultra’. From early 1943 the Allies were able to read the Japanese Maru Code and thus track shipping routes and schedules, a huge advantage in the vast expanses of the Pacific. 81 By late 1944, Pearl Harbor could despatch intelligence information to US submarines on patrol within 30 minutes of receiving it. 82 The American submarines which attacked convoy HI-72 had been specifically directed to it, when at 3.00 a.m. on 11 September they received an Ultra alerting them that a convoy from Singapore was headed northeast through the Paracel Islands to the Formosa Straits. 83 Not only did intelligence facilitate the interception of Japanese shipping, but Allied code-breakers often had a better understanding of the results from attacks than the submariners making them. Within six hours of the first HI-72 convoy ship being attacked, for example, code-breakers knew that one hundred survivors had been rescued from Hirado and that the ship’s captain had been killed. 84
While the Japanese battle reports appreciated the surface speed of US submarines, they underappreciated their diving capabilities. The Japanese often tended to set the fuses on their depth charges to explode at too shallow a depth. 85 As a result, many American submariners survived extensive depth charging for the simple reason that the depth charges exploded well above them. After Sealion’s attack on the convoy on 12 September, although the crew could hear the sonar echo ranging of Japanese escorts and the dropping of 31 depth charges, the submarine remained safely submerged, at times over 500 feet below the surface. 86 The diving capabilities of American submarines helps to explain the gross disparity between Japanese claims to have destroyed hundreds of submarines and the reality in which the USA lost 52 boats to all causes during the war. 87
In 1944 American submarines reached the zenith of their effectiveness. During that year US submarines sank over 600 ships totalling some 2.7 million tons, surpassing the totals for 1942 and 1943 combined. 88 The battle reports of Asaka Maru and Nankai Maru indicate an appreciation of some of the key elements which contributed to this success. While the reports overestimated the importance of short-distance communications between submarines, they were certainly correct in highlighting the significance of radar. The attacks on convoy HI-72 reinforce the claim of historian Robert Dienesch that, more than any other single factor, effective surface search radar improved the performance of American submarines in the latter stages of the war. 89 The absence of comparable radar for the use of escorts was a major reason for the vulnerability of Japan’s convoys. By a subtle irony, one of the convoy HI-72 ships, Asaka Maru, had carried a naval delegation to Germany in 1941 on a visit which reinforced just how far behind Japan was in developing radar. 90
Only two weeks before convoy HI-72 departed Singapore, Admiral Matome Ugaki lamented in his diary the great losses to submarines, acknowledging that ‘our antisubmarine policy isn’t perfect’. 91 The convoy’s fate demonstrated that this was a considerable understatement. Nevertheless, there is at least some evidence that Japanese anti-submarine warfare was improving by this stage of the war. With the Americans’ loss of 19 submarines during 1944, including five in the month of October alone, the silent service suffered its worst year of the war. 92 But ironically, just as Japan was devoting more resources to counter American submarines, the battlefield shifted as Allied aircraft increasingly took the initiative in destroying Japan’s merchant marine. The bomber attacks on the remnants of convoy HI-72 presaged the transition; by the end of 1944 American air supremacy extended to the Japanese home islands, so that aircraft and air-dropped mines displaced submarines as the main threat to shipping. 93 With the benefit of hindsight, convoy HI-72 may be viewed as a marker in the changing war at sea. Wedged between two decisive naval battles, the battle of the Philippine Sea and the battle of Leyte Gulf, the convoy’s fate signified a comparable disaster for Japan’s merchant marine.
Footnotes
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the Australia-Japan Foundation (grant number 00431).
1
Joan Blair and Clay Blair, Return from the River Kwai (New York, 1979); Gregory F. Michno, USS Pampanito: Angel Killer (Norman, OK, 2000); Gregory F. Michno, Death on the Hellships: Prisoners at Sea in the Pacific War (Annapolis, 2001); Don Wall, Heroes at Sea (Mona Vale, NSW, 1991); Aldona Sendzikas, Lucky 73: USS Pampanito’s Unlikely Rescue of Allied POWs in WWII (Gainesville, 2010).
2
Details of HI-72 and MAMO-03 were compiled from the following sources: ‘Dai-ichi Kaijo Goei-tai Senji Nisshi Showa 19-nen 9-gatsu Tsuitachi Showa 19-nen 9-gatsu 30-nichi’ (Wartime Diary, First Maritime Protection Force, 1 September 1944 to 30 September 1944), Library of the National Institute for Defense Studies, Tokyo, Japan (NIDS Library); ‘Showa 19-nen 9-gatsu 27-nichi HI-72 Sendan Hibakugeki Sento Shoho Unso-sen Asaka-Maru’ (27 September 1944, Detailed Battle Report of Bombing of HI-72, Transport Asaka-Maru), NIDS Library; ‘Showa 19-nen 10-gatsu 12-nichi Asaka-Maru Hi-rai-baku Chinbotsu Sento Shoho Unsosen Asaka-Maru’ (12 October 1944, Detailed Battle Report of the Torpedoing, Bombing, and Sinking of the Asaka-Maru, Transport Asaka-Maru), NIDS Library; ‘Showa 19-nen 10-gatsu Toka Tokusetu Unsosen Nankai-Maru Sento Shoho Showa 19-nen 9-gatsu 12-nichi Minami-shina-kai ni okeru Tai-sensuikan-sen Tokusetu Unsosen Nankai-Maru’ (10 October 1944, Detailed Battle Report of Converted Transport Nankai-Maru: Anti-Submarine Battle on 12 September 1944, in the South China Sea, Converted Transport Nankai-Maru), NIDS Library; Shinshichiro Komamiya, Senji Yuso-sendan-shi (History of Wartime Transport Convoys), Kyodo Shuppansha, 1987, pp. 246–9.
3
Gordon Williamson, U-Boat Tactics in World War II (Oxford, 2010), p. 22.
4
USS Growler Tenth War Patrol Report, 12 September 1944, Disc 10, Submarine Memorabilia (henceforth cited as SM).
5
USS Sealion Second War Patrol Report, 12 September 1944, Attack Data, Disc 22, SM; USS Growler Tenth War Patrol Report, 12 September 1944; John D. Alden and Craig R. McDonald, United States and Allied Submarine Successes in the Pacific and Far East during World War II (Jefferson, NC, 2009), p. 203; Eli T. Reich, The Reminiscences of Vice Adm. Eli T. Reich (Annapolis, 1982), vol. 1, p. 201.
7
Robert J. Cressman, The Official Chronology of the U.S. Navy in World War II (Annapolis, 2000), p. 258; Combined Fleet, CD-10.
8
Komamiya, Senji Yuso-sendan-shi, pp. 246–9.
9
Komamiya, Senji Yuso-sendan-shi, pp. 247–8.
10
Hiroyuki Shindo, ‘The IJA, IJN and Japanese Strategy for the South Pacific / SWAPA’, in Peter Dennis (ed.), Armies and Maritime Strategy: 2013 Chief of Army History Conference (Canberra, 2014), p. 164.
11
12
Evan Graham, Japan’s Sea Lane Security, 1940–2004: A Matter of Life and Death? (London, 2006), p. 88.
13
J. Rohwer and G. Hummelchen, Chronology of War at Sea 1939–1945: The Naval History of World War II (Annapolis, 1992), p. 305; Joel Holwitt, ‘Unrestricted Submarine Victory: The U.S. Submarine Campaign against Japan’, in Bruce A. Elleman and S.C.M. Paine (eds), Commerce Raiding: Historical Case Studies, 1755–2009 (Newport, RI, 2013), p. 234.
14
Atsushi Oi, ‘Why Japan’s Anti-Submarine Warfare Failed’, United States Naval Institute Proceedings 78.6 (1952), p. 598. Also published in David C. Evans (ed.), The Japanese Navy in World War II: In the Words of Former Japanese Naval Officers (Annapolis, 1986), p. 407.
15
Robert Gannon, Hellions of the Deep: The Development of American Torpedoes in World War II (University Park, 1996), p. 195.
16
Komamiya, Senji Yuso-sendan-shi, pp. 247–9; ‘Senpaku Yuso-kan ni okeru Sonan (Kaibotsu) (Yukuefumei) Butai (Gunijin Gunzoku) Sieri Shiryo’ (records to Organize Losses (Lost at Sea) (MIA) of Units (Military Personnel and Auxiliaries) during Transport of Ships), NIDS Library; Combined Fleet, Nankai Maru, CD-10; Alden and McDonald, Submarine Successes, p. 203.
17
Boeicho Boeikenshusho Senshishitsu (War History Office, Defence Agency), ed., Kaijo Goei-sen (Maritime Protection War), Asagumo Shinbunsha, 1971, pp. 384–5. Komamiya, Senji Yuso-sendan-shi, pp. 225–8.
18
H.P. Willmott, The Battle of Leyte Gulf: The Last Fleet Action (Bloomington, 2005), pp. 11, 32.
19
Atsushi Oi, Kaijo Goei-sen (Maritime Protection War), Asahi Sonorama, pp. 392–3.
20
Detailed Battle Reports of Asaka Maru and Nankai Maru.
21
Robert Dienesch, ‘Radar and the American Submarine War, 1941–1945: A Reinterpretation’, The Northern Mariner 14.3 (2004), pp. 29, 31–2, 39.
22
Mochitsura Hashimoto, Sunk: The Story of the Japanese Submarine Fleet 1942–1945, trans. E.H.M. Colegrave (London, 1955), p. 56; William Bruch Johnson, The Pacific Campaign in World War II: From Pearl Harbor to Guadalcanal (London, 2006), p. 181; Ian Pfennigwerth, ‘A Novel Experience: The RAN in 1942 Defending Australian Waters’, in Peter J. Dean (ed.), Australia 1942: In the Shadow of War (Cambridge, 2013), p. 184.
23
Peter Padfield, War Beneath the Sea: Submarine Conflict 1939–1945 (London, 1997), p. 337.
24
Quoted in Rick Cline, Final Dive: The Gallant and Tragic Career of the WWII Submarine, USS Snook (Placentia, CA, 2001), p. 35.
25
Reich, Reminiscences, p. 200.
26
USS Growler Tenth War Patrol Report, 12 September 1944.
27
USS Pampanito Third War Patrol Report, 12 September 1944, Disc 25, SM.
28
USS Plaice Second War Patrol Report, 26 September 1944, Disc 26, SM.
29
See Shindo, ‘Japanese Strategy for the South Pacific’, p. 165; Edward Beach, Run Silent, Run Deep (New York, 1955), pp. 148, 150.
30
Dienesch, ‘Radar and the American Submarine War’, pp. 31, 33; Kenneth Poolman, The Winning Edge: Naval Technology in Action, 1939–1945 (Annapolis, 1997), p. 196.
31
Dienesch, ‘Radar and the American Submarine War’, p. 36.
32
USS Growler Tenth War Patrol Report, Torpedo Attack Report.
33
Detailed Battle Reports of Asaka Maru and Nankai Maru.
34
Roger I. Wilkinson, ‘Short Survey of Japanese Radar’, Electrical Engineering 65 (August–September 1946), pp. 372, 460; Louis Brown, A Radar History of World War II: Technical and Military Imperatives (Bristol, 1999), pp. 140, 373.
35
Wilkinson, ‘Japanese Radar’, p. 373.
36
Hashimoto, Sunk, pp. 102, 119.
37
Oi, ‘Why Japan’s Anti-Submarine Warfare Failed’, pp. 596–7; Norman Polmar and Edward Whitman, Hunters and Killers: Volume 2 Anti-Submarine Warfare from 1943 (Annapolis, 2016), p. 51.
38
Matome Ugaki, Fading Victory: The Diary of Admiral Matome Ugaki 1941–1945, trans. Masataka Chihaya (1991, reprint ed., Annapolis, 2008), p. 439.
39
US Naval Technical Mission, Japanese Anti-Submarine Warfare, pp. 24, 41.
40
Brown, A Radar History of World War II, p. 363.
41
USS Pampanito Third War Patrol Report, 12 September 1944.
42
See Jonathan J. McCullough, A Tale of Two Subs: An Untold Story of World War II, Two Sister Ships, and Extraordinary Heroism (New York, 2008), p. 172; Carl Boyd, American Command of the Sea: Through Carriers, Codes and the Silent Service (Newport News, VA, 1995), p. 36.
43
USS Pampanito Third War Patrol Report, 17 August 1944; USS Growler Tenth War Patrol Report, 17 August 1944.
44
Reich, Reminiscences, p. 200; USS Sealion Second War Patrol Report, 11 September 1944; USS Growler Tenth War Patrol Report, 11 September 1944; USS Pampanito Third War Patrol Report, 11 September 1944.
45
Reich, Reminiscences, pp. 200–1.
46
USS Sealion Second War Patrol Report, 15 September 1944; USS Pampanito Third War Patrol Report, 15 September 1944.
47
Oi, ‘Why Japan’s Anti-Submarine Warfare Failed’, pp. 596, 598.
48
USS Growler Tenth War Patrol Report, 12 September 1944; USS Sealion Second War Patrol Report, Attack Data.
49
Mark P. Parillo, The Japanese Merchant Marine in World War II (Annapolis, 1993), p. 103.
50
USS Sealion Second War Patrol Report, 12 September 1944.
51
Oi, Kaijo Goei-sen, p. 24; Shindo, ‘Japanese Strategy for the South Pacific’, p. 155.
52
Oi, ‘Why Japan’s Anti-Submarine Warfare Failed’, p. 587.
53
Polmar and Whitman, Hunters and Killers, pp. 45–6.
54
Oi, Kaijo Goei-sen, p. 96.
55
Polmar and Whitman, Hunters and Killers, p. 50.
56
Parillo, Japanese Merchant Marine, p. 96.
57
Parillo, Japanese Merchant Marine, p. 104.
58
Oi, ‘Why Japan’s Anti-Submarine Warfare Failed’, p. 595.
59
Y. Horie, ‘The Failure of the Japanese Convoy Escort’, U.S. Naval Institute Proceedings, October 1956, p. 1078.
60
Combined Fleet, Shikinami.
61
Ugaki, Fading Victory, pp. 438–9.
62
USS Growler Tenth War Patrol Report, 12 September 1944.
63
Oi, Kaijo Goei-sen, pp. 95–6.
64
Polmar and Whitford, Hunters and Killers, p. 45.
65
Graham, Japan’s Sea Lane Security, p. 86.
66
Parillo, Japanese Merchant Marine, p. 103.
67
Polmar and Whitman, Hunters and Killers, p. 50.
68
Oi, ‘Why Japan’s Anti-Submarine Warfare Failed’, pp. 589, 592.
69
Oi, Kaijo Goei-sen, pp. 144–5.
70
Oi, Kaijo Goei-sen, pp. 146–7, 154–6; Polmar and Whitman, Hunters and Killers, p. 61.
71
Boeikenshusho Senshishitsu, Kaijo Goei-Sen, p. 306.
72
Oi, Kaijo Goei-sen, pp. 215–16.
73
Oi, ‘Why Japan’s Anti-Submarine Warfare Failed’, p. 596; Horie, ‘Failure of the Japanese Convoy Escort’, p. 1075.
74
Boeikenshusho Senshishitsu, Kaijo Goei-sen, pp. 332–5.
75
Asaka Maru and Nankai Maru Detailed Battle Reports.
76
Mark Stille, Imperial Japanese Navy Antisubmarine Escorts 1941–45 (Oxford, 2017), p. 6.
77
Paul Kennedy, Engineers of Victory: The Problem Solvers who Turned the Tide in the Second World War (London, 2014), pp. 41, 55, 57, 67; Williamson, U-Boat Tactics, p. 50; Stille, Antisubmarine Escorts, p. 10.
78
Asaka Maru and Nankai Maru Detailed Battle Reports.
79
Parillo, Japanese Merchant Marine, p. 98.
80
Reich, Reminiscences, p. 199.
81
Edwin T. Layton, ‘And I was There’: Pearl Harbor and Midway – Breaking the Secrets (Annapolis, 2006), pp. 471–3; McCullough, A Tale of Two Subs, pp. 162–3; Padfield, War Beneath the Sea, p. 391.
82
Reich, Reminiscences, p. 224.
83
Eugene Fluckey, Thunder Below! The USS Barb Revolutionizes Submarine Warfare in World War II (Urbana, 1992), p. 106.
84
Combined Fleet, Hirado.
85
US Naval Technical Mission, Japanese Anti-Submarine Warfare, p. 11.
86
USS Sealion Second War Patrol Report, 12 September 1944.
87
See Tom Paine, The Transpacific Voyage of His Imperial Japanese Majesty’s Submarine I-400 (Tom Paine’s Journal, July–Dec. 1945) (Los Angeles, 1984), p. 8; Parillo, Japanese Merchant Marine, pp. 122–3; Polmar and Whitman, Hunters and Killers, p. 61.
88
Max Hastings, Retribution: The Battle for Japan, 1944–45 (New York, 2008), p. 269.
89
Dienesch, ‘Radar and the American Submarine War’, p. 29.
90
Brown, A Radar History of World War II, pp. 135–7.
91
Ugaki, Fading Victory, pp. 444–5.
92
Rick Cline, Submarine Grayback: The Life and Death of the WWII Sub, USS Grayback (Placentia, CA, 1999), p. 219; Parillo, Japanese Merchant Ships, p. 122.
93
Oi, ‘Why Japan’s Anti-Submarine Warfare Failed’, p. 601.
