The Kamikaze concept was proposed by Vice Admiral Takijiro Onishi, commanding the 1st Air Fleet on Luzon, on 19 October 1944 as a means of attacking and ensuring the destruction of American carriers. Kamikaze means the 'Divine Wind'. A more formal Japanese designation was Tokko – 'Special Attack'. They were code-named Kikusui – 'floating chrysanthemums.' The plan was quickly implemented with the formation of four special attack units (Tokko-tai) and the first confirmed Kamikaze attack occurred on 25 October during the battle off Samar with an aircraft attack on the USS Suwanee (CVE27). The first aircraft struck the escort carrier, causing moderate damage, and others were downed attacking other ships of the task group. Kamikaze air attacks quickly became a major threat to Allied ships through the remainder of the war with 26 American ships sunk and 225 damaged at Okinawa alone.
The development of Kamikaze attack boats actually began earlier than that of Kamikaze aircraft. It was envisioned that swarms of small, high-speed, explosive-laden boats attacking American transports and warships off the landing beaches would play havoc with the invasion fleet. While Kamikaze aircraft generally attacked at dawn or through clouds during daylight hours, the boats would use the cover of darkness to close in slowly on the transports' anchorage and with a sudden burst of speed, strike unexpectedly. One other Kamikaze initiative was the Kaiten – 'Herculean task' manned torpedoes launched from mother submarines.
The concept for the development of IJA suicide boat units was conceived in April 1944. Lieutenant General Suzuki, commander of the Shipping Engineers at Ujina, proposed the use of what were essentially manned torpedoes launched from shore. They were to be employed to defend islands by executing surprise attacks on the enemy invasion fleet. Imperial General Headquarters approved the concept, but also urged its Technical Institute to study manned explosive-bearing assault boats for the same role. The Shipping Engineers went ahead with the development of their manned torpedo while the Technical Institute worked on the boat. The boat was tested in Tokyo Bay in June and it was determined that it would be the most effective of the two options. As usual the IJA and IJN further developed their own boat designs. In July, the Grand Directives on Future Military Operations were jointly issued by the Army and Navy Divisions of the Imperial General Headquarters. This document provided basic policies and procedures for the employment of special attack units. Construction of the boats began and the first IJA sea raiding regiments were organised on 1 September 1944.
A sea raiding regiment was actually a company-sized unit commanded by a captain. At full strength the 'regiment' consisted of 104 personnel and 100 boats, including spares. The regimental headquarters had 11 personnel. Each of the three companies had 31 personnel organised into a four-man headquarters and three nine-man platoons. The platoon was considered the minimum sized element to conduct tactical operations. Virtually all personnel assigned to the regiment were boat operators. Administration, supply, boat maintenance, security, and other support functions were provided by an approximately 900-man sea raiding base battalion bearing the same number as its associated sea raiding regiment. Prior to the 1 April 1945 American invasion of Okinawa, the sea raiding base battalions were converted to rifle battalions and employed in a defensive role on the main island or to defend some of the small islands of the Kerama Retto (archipelago) 15 miles to the west of Okinawa. The Kerama Retto became a principal special attack boat base. The sea raiding regiments were prepared to undertake their mission and would no longer require their attendant base battalion's support.
The 16 and 17-year-old volunteers were 'special' officer cadets of the Shipping Engineer Officer Academy. The cadets had been recruited in December 1943 from among third and fourth year high school students. If a boat pilot failed to return from his mission, he was presumed successful and posthumously promoted to lieutenant.
The boats themselves were referred to as 'Q-boats'. The IJA boats were 18 feet long and 5 feet wide. Their 85-horsepower, 6-cylinder Chevrolet engine gave them a 20 knot speed and a 3½ hour range at full speed, although it was unnecessary to maintain such high speed for a prolonged period. Maintaining a cruising speed of 6½ knots they could cover 70 nautical miles in less than 11 hours. This would allow them to move slowly and quietly to their battle zone, execute a full-speed attack, and possibly return to their base. The 20 knot speed was actually too slow to allow the 'whirlwind' attacks envisioned by the Japanese and certainly reduced their ability to survive the high volume of automatic weapons fire they would receive from American ships. A 30 to 40-knot speed would have been more desirable, and more survivable.
Some of the cheaply constructed plywood boats carried a 551 lb (225 kg) explosive charge inside the bow. This was also standard for the IJN boats, which were intended to ram their target. Most IJA boats, however, had a rack on either side of the cockpit for a 264 lb (120 kg) Model 98 depth charge intended for dropping within five metres of a ship after making a sharp U-turn to allow an escape (see sidebar). The depth charges would detonate within seconds alongside the target ship and the hydrostatic shock wave of such a heavy close explosion would buckle the hull plates. It would also most likely destroy the attack boat. During an attack on the USS Terebinth (AN59), a net-laying ship, the depth charges were dropped from 12 to 18 metres away and the small ship suffered no damage. The fleeing boat was destroyed by 20 mm gun fire. Most boats were detected and engaged as they attempted to covertly approach their targets. The pilot had to possess the skill and will to execute an attack in such close proximity in order to be successful and his chances for survival were slim.
The tactics employed by special attack boat units and other special attack means were provided in Guidelines Pertaining to Military Operations Covering Small Islands. The boats would deploy at night in a combat unit of company groupings. The number of boats in a combat unit would vary greatly, from a half-dozen to 20 or more. The smaller units appear to have been more common. Selected pilots with proven skills were responsible for navigating the combat unit to the anticipated surface battle zone at an average speed of 6½ knots. This speed kept engine noise to a minimum, reduced the tell-tale wake, conserved fuel, and was less stressful for the pilots during the long, slow approach requiring many hours of transit time. During daylight, US amphibious ships and transports would lay close to shore, discharging cargo and receiving wounded for days after the initial landing. After dark they would retire further offshore to designated transport areas making it difficult for air attacks, submarines, and special attack boats to locate them. At Okinawa the Japanese identified three areas off the island's west and south coasts where they thought the night transport areas would be. One was off the lower west coast, north of Naha, the island's capital and main seaport. This proved to be the main transport area as it was opposite the Hagushi Beaches near the Yontan and Kadena Airfields, the landing force's initial objectives. For the 1st and 2nd Special Attack Boat Regiments based in the Kerama Retto, this meant a 69-mile round trip requiring approximately 10½ hours, if they returned. Another pre-planned route from the Kerama Retto took the boats through two anticipated surface battle zones. The first one was due west of Naha and south of Keise Shima, four islets north-west of Naha. The other was off the south-east end of Okinawa, near the Minatogawa Beaches (Minatoga in US documents). The Japanese considered it to be a principal landing beach but no transport areas were located in these areas after the first two nights. The south-east area had been used by a diversionary force attempting to lead the Japanese into believing that the landing would be there. Special attack boats transiting this route from the Kerama Retto found themselves running a gauntlet of patrol craft and destroyers screening the southern approaches to the actual transport area to the north. This 56 mile two-way route required approximately 8½ hours to transit.
When the US ships were detected the boats would separate into groups of two to four, with each group heading for a specific ship. If a boat was separated from its group and the pilot disoriented he was instructed to attack the nearest enemy ship alone. Once the enemy had detected the special attack boats, either when they attacked or were discovered by sharp-eyed lookouts, it was essential for all other boats to press home their attacks immediately, as the enemy was alerted. Ideally, several boats would approach a target ship nearly simultaneously from different directions to prevent the enemy gunners from concentrating their fire on single boats and to achieve the maximum surprise and confusion. The reality was that the boats were too easily detected and far too slow to manoeuvre amongst dispersed ships, and the patrol craft among them, and to execute co-ordinated attacks from multiple directions. The only way to begin to ensure success was for all boats in a group to immediately attack one ship as soon as it was detected.
Seven sea raiding regiments were deployed to Okinawa Gunto by April 1945 under the 5th Sea Raiding Base Headquarters. The first three were organised in September 1944 at Yukinoura on Eta Jima in the Home Islands. The other four were organised in October. The units were shipped from Kyushu, the southernmost of the Home Islands, between September 1944 and February 1945. The Japanese assessed the most likely landing beaches, with three on the lower west coast, one on the south-east end, and one on the lower east coast in Nakafusuku (later Buckner) Bay. The regiments were deployed accordingly to be in position to attack the invasion fleet in these areas.
The 1st Sea Raiding Regiment (Akatsuki Regiment No. 16777) under Captain Hiroshi Umezawa was based on Zamami Jima, the second largest island in the Kerama Retto. The unit struggled to assemble its boats, hampered by heavy rains and equipment shortages. Less than half the boats were completed and little work was accomplished on the island's defences after most of the supporting 1st Sea Raiding Base Battalion was withdrawn to Okinawa. After three days of bombardment the 27th Infantry Division landed on 26 March ahead of the 1 April landing on Okinawa. Unable to launch, the commander ordered the remaining boats destroyed and then his men withdrew into the hills. Most of the regiment and attached engineer troops on the island were killed or ultimately committed suicide.
The 2nd Sea Raiding Regiment (Akatsuki Regiment No. 16778) was commanded by Captain Yoshihiko Noda. Most of the regiment was based on Aka Jima, the third largest island in the Kerama Retto. Off its south shore is small Keruma Jima (also known as Geruma Jima) on which the 1st Company was based. Plagued with technical problems and bombardment, the regiment was unable to launch its remaining undamaged boats. The 1st Company did launch four boats and claimed to have sunk two enemy ships and damaged two. Most of the regiment withdrew on to Aka's high ground and were contained there until after the war.
The 3rd Sea Raiding Regiment (Akatsuki Regiment No. 16779) under Captain Yoshitsugu Akamatsu was based on Tokashiki Jima, the largest island in the Kerama Retto. The regiment was operationally ready even though most of the 3rd Sea Raiding Base Battalion had moved to Okinawa in February. When the bombardment began on 23 March the situation began to deteriorate drastically with the realisation that the regiment would have little effect on the enemy as the Kerama Retto was being overrun, it was ordered to evacuate to Itoman near the south end of Okinawa using its own boats. At dawn on the 26th it was clear that it would be virtually impossible for the boats to get through the US fleet. Most of the boats were burned and American troops began to land on the 27th. Fighting as infantry, most of the regiment was wiped out after conducting nightly Banzai attacks on the Americans.
Over 250 boats were destroyed or captured in the Kerama Retto, about half of the suicide boat force in the Okinawa Gunto. Virtually none were successfully committed to combat.
The 26th Sea Raiding Regiment (Akatsuki Regiment No. 19765) was initially based on Tokashiki Jima, but soon moved to Itoman on Okinawa. Under the command of Captain Mutsuo Adachi, the unit prepared its boats after the main American landing on 1 April. On the night of 8/9 April virtually the entire regiment was ordered to proceed north and attack the transports. Japanese troops in defensive positions reported 20 American ships destroyed or damaged. Few boats returned. Many had been destroyed attempting to infiltrate through American patrol boats and US records show only a destroyer, USS Charles J. Badger (DD657), damaged by a suicide boat on 9 April. The reported burning ships were probably suicide boats. Additional attacks were launched on the nights of 14 and 20 April with claims of seven American ships sunk and four damaged. No US ships were reported sunk or damaged by boat attacks on those dates. The few survivors of the regiment were absorbed into the 32nd Infantry Regiment and died in June attempting to retake the Maesato Heights.
The 27th Sea Raiding Regiment (Akatsuki Regiment No. 19766) was based at Yonabaru Town on the south side of Nakafusuku Bay. The regiment did not launch an attack until 25 April and then with only 15 boats. Most were lost, but a destroyer was claimed to have been sunk. No ships were reported sunk or damaged on that date. The regiment's 1st Company dispatched 14 boats on the 27th and less than half were able to infiltrate through American patrols. Regardless, they claimed a destroyer and transport were sunk. A destroyer, the USS Hutchins (DD476), was in fact damaged that night. Ashore, the Japanese attempted to conduct a major counteroffensive on 3 May. This was supported by attempted counterlandings in the American rear on both coasts. The 27th Sea Raiding Regiment and 23rd Shipping Engineer Regiment were tasked with this mission on the East Coast. Twenty boats attacked a transport area. Most were interdicted by patrol craft, but they claimed that a destroyer, two Landing Ships, a Tank (LST), and three transports were sunk. No US ships were reported damaged or sunk. The commanders of both the 27th and 28th Sea Raiding Regiments had been lost on 17 May and a company commander assumed command of the remaining personnel, consolidated as the Matsumoto Regiment. Withdrawing to the Chinen Peninsula they trained with canoes as no powerboats remained. On 27 May they paddled out to sea carrying small explosive charges only to be destroyed by patrol craft. Remnants of the unit fought as guerrillas on the peninsula until wiped out in early June.
The 28th Sea Raiding Regiment (Akatsuki Regiment No. 19767) was initially split between Naha on the lower west coast and Minatogawa on the south-east coast. It was commanded by Captain Toshio Honma. The 1st and 2nd Companies at the latter base suffered extensive boat and personnel losses when depth charges accidentally detonated as they prepared to attack the pre-invasion bombardment force. This explosion identified the unit's location and the survivors relocated to villages on the west coast below Naha. On 27 April two platoons finally attacked American ships, but none returned to report their results. No US ships were damaged that night. On 3 May the regiment supported the ill-fated counteroffensive as the 26th Shipping Engineer Regiment attempted a counterlanding on the West Coast. Two additional attacks were launched in May, but again no boats returned. The last reported American ship damaged by suicide boats was on 3 May, the cargo ship USS Carina (AK75). By June little of the regiment remained and the troops were assigned to the 44th Infantry Regiment and most were killed by the end of the month.
One of the transports carrying 1st and 2nd Companies of the 29th Sea Raiding Regiment (Akatsuki Regiment No. 19768) sank en route to Okinawa, but most of the unit was landed on Okinawa. The regimental headquarters under Captain Hisanori Yamamoto (no relation to the admiral), the 3rd Company, and most of the mechanics were forced to land on Amami-Oshima north of Okinawa due to heavy air attacks in mid-February. There they sat out the war. The 1st and 2nd Companies were deployed in the vicinity of Chatan just to the south of the landing beaches where they were attached to the 26th Sea Raiding Regiment. These companies were the first to launch attacks in the early morning hours of 30 March against the invasion force off the Kerama Retto. Most were destroyed, although the survivors reported sinking six destroyers and three transports. No US ships were reported sunk or damaged on that date. Upon landing the assault troops discovered blasted suicide boats and small craft choking the Bishi Gawa, the boundary between III Amphibious Corps and XXIV Corps. These boats probably belonged to the 1st and 2nd Companies. The few remaining boats and troops were assigned to the 23rd Shipping Engineer Regiment to support its 3 May counterlanding on the West Coast. The survivors of that failed attempt died fighting with the 266th Shipping Engineer Battalion in late June.
The boats were hidden in caves and concealed with camouflage nets and local foliage in ravines, gullies, and streams. In the early days of the Okinawa landing the weather conditions were favourable for the special attack boats. The night temperature was in the mid-60°F range, but the water was a debilitating 70°F. Winds were moderate from the east-north-east and the sea state was light swells. There was usually a light surface haze at night and rain storms providing a concealed if not comfortable approach. Rains increased as the campaign wore on.
Already alerted to the possibility of such attacks from their use in the Philippines and their detection in aerial photographs, the ships of Task Force 51 were prepared to meet suicide boat attacks. A confidential pamphlet was issued on 28 May, Suicide Weapons and Tactics, relating lessons learned countering the various threats. Night lookouts were posted on all quarters of ships. Machine guns, 20 mm and 40 mm anti-aircraft guns, and searchlights were manned at all times. Precautions were imposed to prevent gunfire directed at surface targets from hitting nearby friendly ships. Men armed with rifles and Browning automatic rifles were on standby to be turned out when boats were detected approaching. Radar of the period was generally ineffective for detecting boats. It was designed to detect aircraft and large surface ships. Also, the boats were low in the water and their pips difficult to distinguish from surface chop and swell. Sonar was sometimes able to detect the engine noises. Ships were warned that if they detected small, scattered and intermittent radar pips, coupled with the sound of high-speed screws by sonar, that suicide boats were approaching. The most successful means of defeating suicide boats was by extensive small craft patrols around the transport areas. Any and everything was pressed into service as 'flycatchers.' This included destroyer escorts, Patrol Torpedo (PT) boats, sub chasers, minesweepers; Landing Ships, Medium (rocket) (LSM[R]); and Landing Craft, Infantry (LCI). These small craft possessed the manoeuvrability and firepower to effectively engage the boats. From 9 April a cruiser and two destroyers were assigned to continuously illuminate Naha Harbour with star shells each night to detect suicide boats attempting to sortie.
Almost 700 suicide boats were deployed to Okinawa Gunto to be operated and supported by some 7,000 troops. They managed to sink one ship, damage five, and inflict 29 casualties. Besides the general ineffectiveness of the boats and their tactics, the effectiveness of the depth charges detonated close abeam to a ship was questionable. The 2,924 ton Fletcher-class destroyer USS Hutchins was lifted several feet by two depth charges, the port engine and shaft were damaged, and the crew knocked off their feet, but there were no casualties. Granted, it was deemed too costly to repair the ship and she was scrapped after the war.
Some two to three hundred boat pilots died in these futile attacks. While many of the pilots and over 6,000 base troops went on to die as infantry replacements, the effort was essentially an awful waste of resources and personnel. Organised and trained as an infantry brigade those 7,000 troops would have made the conquest of Okinawa much more costly for the Americans.
A small number of the boats were captured intact in the Kerama Retto and returned to Hawaii for examination. A few were overhauled and used as ship's boats by the US Navy. One such boat, serving as the captain's gig by the USS Laffey (DD724), was inadvertently sunk in Settle Harbour in 1945 and probably still rests on the bottom. A fully restored IJN Shinyo boat is displayed with the USS Massachusetts (BB59) at Falls River, Massachusetts. One or two others are reportedly displayed in the United States, but the author has been unable to locate them.
by Gordon Rottman
Further reading
Headquarters, Sixth Army, Japanese Navy Suicide Crash Boats, G-2 Weekly Report, No. 79, 14 March 1945
Kauai, Mashiro (Col.), The Operations of the Suicide-Boat Regiments in Okinawa: Their Battle Results and the Countermeasures Taken by the US Forces, National Institute for Defence Studies, undated Military Intelligence Service, War Department, Tactical and Technical Trends, 'Japanese Suicide Boats', No. 58, May 1945
Rottman, Gordon L., Campaign 96: Okinawa 1945 (Osprey, 2002)
Rottman, Gordon L., Campaign 137: Saipan & Tinian 1944 (Osprey, 2004)
Rottman, Gordon L., Battle Orders 7: US Marine Corps Pacific Theater of Operations 1943-44 (Osprey, 2004)
Rottman, Gordon L., Fortress 1: Japanese Pacific Island Defenses 1941-45 (Osprey, 2003)
Rottman, Gordon L., Battle Orders 12: US Special Warfare Units in the Pacific Theater 1941-45 (Osprey, 2005)
Samuel E. Morison, History of United States Naval Operations in World War II, Victory in the Pacific, 1945, Vol. XIV (Little, Brown and Company, 1960)
Wright, Derrick, Campaign 77: Tarawa 1943 (Osprey, 2000)