Ten Days to Kamikaze is a series that explores the decision-making process and implementation of Japan’s use of suicidal crash dives during World War II. It provides an in-depth review of the critical ten-day period and examines the background leading up to those decisions. Events on the first of the ten days gave rise to a legend that has often been repeated and believed including down to today.
If you are just joining the series, you can read previous posts here:
Ten Days to Kamikaze – Part II
Now on to Part III – The Ten Days!
October 15 – the Legend of Admiral Arima. V-Adm. Teraoka’s 1st Air Fleet had a direct reporting relationship with the Combined Fleet in Tokyo but for operational matters it took direction from Mikawa’s Southwestern Force headquarters. In its operational guise 1st Air Fleet was known as Fifth Base Air Force (Koku Kichibutai). Unlike early 1944 when air flotillas reported to the air fleet and air groups reported to air flotillas, in October 1944 four air groups (153, 201, 761, and 1021) reported directly to the air fleet. Two of these air groups (201 and 761) were enlarged organizations almost as large as former air flotillas. Air flotillas such as the 26th at Clark and 61st at Davao were regional support organizations controlling operations of such units as were in their area at a given point in time. On October 12th when Fifth Base Air Force was tasked with attacking Task Force 38 then ravaging Formosa R-Adm. Masafumi Arima, commander 26th Air Flotilla, was charged with organizing the attack.
It was past noon on the 12th when orders to attack were received. It proved to be too late to assemble a striking force that day and no attack took place. The following day 170 navy and army planes were sent out. Their target was a group of aircraft carriers 280 miles north of Aparri. That was about 500 miles from Clark. Aircraft were to land at airfields in Formosa after their attacks. The planes, unable to find the carriers due to weather conditions encountered, landed at various bases on Formosa and Luzon with some apparently lost or damaged in the process. The scattered planes spent the following day returning to their home fields. Moreover, searchers found no carriers within range and there was no attack by Philippine based aircraft on the 14th. In contrast to the ineffectiveness of planes operating from Luzon, reports from Formosa indicated twenty-three American ships including nine aircraft carriers had been sunk or seriously damaged. Most attacks took place at night with the difficulties in accurate reporting that conditions of darkness impose. The fact that these reports were wildly exaggerated was not known to commanders in the Philippines at that time.
On the morning of October 15th R-Adm. Ralph Davison’s Task Group 38.4 was approaching Luzon. The previous day Davison’s ships had been missed by both Japanese long range and short range searchers. Davison’s plan was to strike Manila area airfields as well as shipping targets of opportunity in the morning followed by a later second strike against similar targets. TG 38.4 consisted of eighteen warships including fleet carriers Enterprise and Franklin plus light carriers Belleau Wood and San Jacinto. Early on the 15th the task force fighter director began to receive radar reports of bogeys in the vicinity. Eight combat air patrol (CAP) fighters were directed to intercept a bogey picked up on radar 45 miles from the carriers. Four of the CAP fighters sighted a Francis (Ginga land bomber). After a chase during which the Francis was described as “very fast” Lt. H.F. Weidman of VF-13 shot it down about 15 miles from the task force at 0830. At 0800 either this Frances or another bogey appearing on radar at 25 miles at 0740 which was not intercepted got off a sighting report.[1] By 0830 the Fifth Base Air Force headquarters received a report of a carrier force 66 degrees, 240 miles from Manila.
U.S.S. Franklin CV-13 target in several attacks in October 1944.
An hour and a half earlier a message from Fifth Base Air Force directed R-Adm. Arima to take command of forces at or which would be sent to Clark Field. Following the sighting report a strike force was hastily assembled at Clark. It was neither large nor powerful consisting only of Zero fighters seven of which carried bombs. Their bomb load was just two 60kg bombs each. Such small bombs would not sink a large warship, so the apparent aim was to disable the flight decks of carriers and reduce the defenses against a follow up strike. Lt. Masanobu Ibusuki of Air Group 201 led twenty-six Zero fighters and fighter bombers taking off by 0930. Apparently one fighter bomber failed to take off or aborted the mission leaving six to carry out the attack.[2] At 1022 another Ginga snooper which was probably a guide plane for the Zero strike from Clark was shot down and crashed within the task force screen where smoke from its burning wreckage was photographed.
At 0848 TG 38.4 strike Able headed toward Manila. It consisted of 42 F6Fs, 20 SB2Cs and fourteen TBMs. In addition to the ordnance of the Helldivers (500lb. bombs) and Avengers (1,000lb. bombs) most of the Hellcats carried underwing rockets for ground attack. The Japanese radar alert system functioned well on this occasion and some fifty army fighters got into the air before the American planes appeared above Nichols and Nielsen airfields at about 1015. The Japanese had numerous radar stations in the Philippines and U.S. countermeasures became effective only later in the campaign. American reports are unanimous in describing Japanese attacks as “aggressive.” Some reports indicate the Japanese fighters attacked in sections of two or four aircraft. Others say Japanese formations were split up and individual attacks were encountered. It was mostly fighter versus fighter combat, but some Japanese fighters got close enough to the strike aircraft to be claimed as damaged or probable kills by gunners.
The combat included some interesting incidents. A Hellcat pilot broke up a formation of Oscars by firing his underwing rockets at them. High scorer of the combat was Ens. Robert Thomas of VF-21 with claims for two Tonys and two Oscars. Thomas commented that his Hellcat was faster and more maneuverable than the Tonys but that an Oscar was faster and could outclimb his Hellcat. No American pilot identified any of their opponents as a new model fighter, but Thomas’ comments suggest he had encountered a Type 4 fighter (Frank) of which several were involved in the combat. Total claims came to 22 Oscars, 13 Tonys, 8 Zekes and a Tojo destroyed with others probably destroyed or damaged. Additional claims were made for aircraft destroyed or damaged on the ground.
There is a conflict in Japanese sources, but it appears the Japanese lost ten fighters.[3] The 204th FR a veteran unit arriving just days before the action lost five pilots killed in combat with additional aircraft lost to ground attacks. American losses are also illusive. Enterprise’s VF-20 suffered no loss. There was apparently only a single outright loss in the combat area (a VF-13 pilot bailed out), but some planes returned with serious battle damage. These may have included planes jettisoned from the carriers because of the day’s action. These “deck losses” amounted to six Hellcats and two Helldivers. For example, Franklin had 31 Hellcats on board at the beginning of the 15th but only 25 at the start of the following day of which 20 were serviceable. As will be described below it suffered three losses in Japanese attacks on the ship. The cause of the other losses is hard to trace. One VF-21 Hellcat pilot returned near the carriers but bailed out. Another damaged Hellcat of the same squadron crashed on deck and was a write-off. In any event American losses were far less than the 27 sure and probable victories claimed by Japanese fighters. The top scoring Japanese unit the 19th FR equipped with Type 3 fighters claimed nine sure and four probable victories (“uncertain” in Japanese terminology).
SB2C Helldivers attacked manila area bases on October 15th.
After the midmorning shoot down of the Frances snooper American radar picked up the first wave of Japanese attackers. About 1045 Hellcats encountered a reported 8-10 enemy aircraft claiming four Zekes, three Oscars and two Frances. Ten minutes later an additional bogey of ten planes was encountered and all were claimed destroyed. Meanwhile four Zero fighter-bombers eluded the CAP fighters at high altitude and apparently went into approach runs from about 20,000 feet before beginning their steeper attack dives. They came out of cloud at 8 to 10,000 feet and were picked up visually. Two seemed to head for San Jacinto but according to American reports aborted their runs due to heavy AA fire from the carrier and destroyers. Their bombs fell nowhere near any U.S. ship but may have been the basis of a Japanese report of damage to a cruiser. Two others targeted task force command ship U.S.S. Franklin. Bombs from one fell close to the ship and caused fragment damage to the hull. A bomb from the other hit the aft corner of the deck-edge elevator causing damage that distorted but did not disable it. However, shrapnel from these bombs caused thirty casualties including three deaths. Fires were started both on the flight deck and hanger deck. Three planes were sufficiently damaged that they were jettisoned. There was other damage all of which was referred to as superficial in subsequent reports.
Hellcats claimed nineteen victims; the Japanese reported six losses. The Japanese claimed seven victories; no Hellcats were lost in the encounters over the task force. Aside from the gains and losses in terms of physical damage an important effect of the Japanese attack was that it caused R-Adm. Davison to call off strike Baker the second Manila attack. Most of the strike group had been launched by the time of the Japanese attack. They were ordered to orbit away from the task force while fighters intercepted the incoming Japanese. After the attack it was deemed there was insufficient fuel in the planes to carry out a strike against Manila area targets. Moreover, despite many hours of daylight to recover, refuel and launch an attack, and plenty of time to reach his refueling point on the following morning Davison never carried out a second attack as had been ordered. He justified his action based on the activity of Japanese attackers in this and a subsequent attack. His rationale was not accepted as valid in Commodore Bates’ post-war Naval War College analysis. Bates opined that Davison may have been unduly influenced by the attack upon and damage to Franklin his flagship.
During the course of the morning Admiral Arima learned that his first strike had inflicted some damage on the carrier force northwest of Clark. He also learned that despite claims of several carriers sunk or damaged by Formosa based planes there were still several carriers operational east of Formosa. It was imperative that he execute his orders to strike Davison’s carriers. Arima was having trouble assembling an attack force. A partial solution, that of fighter escort was solved by early afternoon. Lt. Col. Tsune-enon Shindo commander of the 16th Flying Brigade (Hiko Dan) had been given operational control of Japanese army fighters a few days earlier and he organized the army fighter escort including fifteen Type 4 fighters of the recently arrived 51st FR. Per a previous army-navy agreement Shindo proceeded to take personal command of Japanese fighter operations assembling an army escort force of more than sixty fighters plus several navy Zeros.
Arima remained woefully short of strike aircraft. No army bombers had arrived from their airfields in Borneo and Malaya. All he had at Clark was three Type 1 land attack bombers of K704. At Tugegarao in northern Luzon were a dozen Tenzan carrier attack planes and a few Zero fighters. By afternoon it became clear that was all that could be organized in time to insure a daylight attack. Additional planes could be sent out later as and if they became available. Since the Tenzans when loaded with torpedoes had a relatively short range, Arima elected to have them fly from their base rather than transfer to Clark. Most likely he planned for the two forces to unite while on the way to the target so that the Tenzans would have the advantage of a strong fighter escort.
Before recounting the legend of Admiral Arima’s demise this might be a good place to explore his possible motivation in boarding one of the Type 1 bombers for the attack. First, Arima shouldered the burden of earlier unsuccessful efforts. Second, he needed to ensure that the two attack forces taking off from different bases carried out a coordinated attack. Sending just three Type 1 land attack bombers against a strong enemy task force in daytime might have seemed almost suicidal perhaps requiring a strong leadership example. Finally, if he did not join the attack for which the navy had the primary responsibility the senior officer involved would be an army officer. Various published renditions of the legend fail to mention these possible motivations.
Type 1 land attack bombers preparing to take off.
About 1400 hours the attack forces took off. Sixty-five army fighters and nine navy Zeros escorted three Rikko (Betty) torpedo bombers. Twelve Tenzan attack bombers and four Zeros sortied from Tugegarao. The most prosaic version of what was believed to have occurred appeared over twenty years ago in an unpublished English language manuscript titled Rikko and Ginga by Dr. Yashuo Izawa, edited here for grammar:
Fighters intercepted and all three Rikkos failed to return. An accompanying Army Ki 46 reported one carrier sunk from a torpedo hit by a Rikko. The Rikko was assumed to be the lead plane and a special announcement stated that the plane rammed a carrier. For an unknown reason Admiral Arima, O.C. 26th Air Flotilla was abroad the plane. It was propagandized as a suicide attack by the Admiral.
A few years after the war this version of the incident appeared in Japanese Monograph No. 84:
Rear Admiral Arima, Masafumi, who commanded the Clark base as commander of the 26th Air Flotilla, personally commanded torpedo attacks from a land based attack plane and was killed executing a suicide dive upon an enemy aircraft carrier. Thus, he gave an example of a special attack method, the so called “Kamikaze attack” which was later ordered to be executed.
Additional details were added in the book “Divine Wind” published in 1958. There it is said Arima suddenly announced he would lead the attack. His aides and staff officers unsuccessfully tried to dissuade him. Arima removed all rank insignia before entering the lead bomber. The account asserts that approaching the enemy Arima “ordered ‘all planes to attack’. His own plane led the way by crash diving into an enemy aircraft carrier.” A footnote inaccurately confuses the attack on the fifteenth with damage to the Franklin caused by a crashing Betty on the thirteenth. Even such a distinguished author as John Toland in his monumental “The Rising Sun” (1970) designates Arima as the first Japanese navy Kamikaze in a footnote while adding that he was shot down before accomplishing an actual crash.
Perhaps the most bizarre version of the legend first appeared in the French language. In 1971 Bernard Millot’s “Divine Thunder” was published in English translation devoting two pages of fiction to the legend. In this version Arima spent the weeks before the attack living in great austerity and in meditation. Instead of a Type 1 land attack bomber Millot places Arima in a Suisei dive bomber an aircraft type not involved in the attack.[4] Intercepting Hellcats decimate the Japanese dive bombers, but one gets through. Arima’s plane crashes Davison’s flagship Franklin from which huge orange-red flames erupt.
Fires rage through it and set off successive explosions of the powder magazines that worsened the already serious damage. Only after long hours of sometimes superhuman effort could [the crew] be sure it was not going to sink. Listing and unable to resume aerial operations, it had to withdraw to a repair base…Some of Admiral Arima’s subordinates were not surprised when they learned of his sublime gesture from the pilots who survived the mission.[5]
Millot’s version if not pure fiction is incredibly bad history. Franklin was damaged by Kamikazes in late October. Millot’s description roughly resembles bomb, not suicide crash, damage to Franklin occurring in March 1945.
A recent Wikipedia article on Arima summarizes his career in an essentially factual way but perpetuates confusion and uncertainty about the incident. Other internet websites contain various garbled versions of the story, some relying on Millot. The legend lives on. None of the accounts quoted including Dr. Izawa’s matter of fact version accurately describe what occurred.
So, a reader might ask what really happened. There was no torpedo hit and there was no crash dive. The Task Group 38.4 action report notes the shoot down of a Frances shortly after 1300 followed two and half hours later by the interception of a Dinah:
Two large bogies at this same time appeared on the screen, bearing 245 deg., distance 70 miles. These enemy planes appeared to be orbiting in this position and it is believed that they miscalculated the movement and position of this group which they were seeking [carriers headed northeast since mid-morning]. Nine divisions of fighters were launched to augment the four division combat air patrol then airborne. Three divisions of Enterprise fighters, later joined by two San Jacinto divisions, were vectored out and shot down twenty enemy planes about forty miles distant from the group. This raid was broken up completely and no planes got through to attack.
The only ramming incident, an accidental one, took place when VF-20 pilot Ens. Bruce Hanna collided with a Betty shearing off its right wing. Hanna bailed out and was rescued three days later. It is anyone’s guess what, if any, attack message about the sinking of a carrier was sent. In absence of any other explanation, we could speculate that an incomplete message about a collision between aircraft including a torpedo plane was misinterpreted as a crash involving an aircraft and an aircraft carrier. If there was such a message possibly the aircraft that signaled the message was under attack and that resulted in a garbled transmission.
A Fifth Base Air Force message report summarized the mission as follows:
Three land based bombers, nine Type Zero fighter planes and 65 army fighters left MLB, and twelve Tenzan airplanes and four Type Zero fighters left MPH. They were intercepted by G fighters. Four Tenzan that returned had no engagement with the enemy and did not carry out the attack. Results of the attack by other Tenzan aircraft unknown. Sank one aircraft carrier. Shot down ten G fighters (three probables). Twenty G fighters were scored by the Army airplanes. Our losses (including missing): Three land based attack planes. Eight Tenzan airplanes. Three Type Zero fighter planes. Nine Army airplanes. [6]
Since no results are reported for the Tenzans and a carrier was claimed sunk, this report implies that the carrier sank as a result of the Type 1 land attackers which were armed with torpedoes. This in turn supports Dr. Izawa’s description of events. Exactly how data in this report was received is unclear. Dr. Izawa associates the report with an Army Type 100 HQ reconnaissance plane (Ki 46), Dinah. The Americans, Cdr. Fred Bakutis of VF-20 and his wingman, shot down a Dinah near the task group during the time the attack occurred (Bakutis claimed additional kills among the attack formation). The 2nd FR operated from Clark and part of its mission was to act as guide unit for fighter planes. A 2nd FR Dinah captained by 1Lt. Genyo Nakagawa was lost on this date. Evidence suggests that the report of a carrier being sunk stems from a message from the army’s reconnaissance plane. Whether this was wishful thinking, a garbled transmission or something else cannot be explained. The rest of the elements of the Arima legend including a crash dive and intentional suicide attack are as Dr. Izawa states, false. They stem from a point in time close to the events and seem to have been used to promote and sell the suicide attack concept.
The Imperial General Headquarters announcement of the two actions which appeared in the Japanese press the following day stated in part:
Our air forces in the Philippines immediately intercepted the enemy and pursuing the enemy aircraft which withdrew towards the sea, repeatedly pounded the enemy task force in the waters east of the Philippines. The war results attained during the attack on the enemy task force are as follows: One enemy aircraft carrier was sunk and three aircraft carriers and either a battleship or a cruiser were heavily damaged. In addition our air force accounted for more than twenty enemy aircraft shot down. Some of our aircraft have failed to return to their base.
Admiral Arima’s loss in action was not announced from Tokyo for more than a week. Possibly there might have been hope he had survived and would be recovered. It seems quite possible the Admiral Arima myth began with a garbled message and wishful thinking. It may or may not have been concocted as part of the recruitment of the initial cadre of suicide pilots. Later the myth served an agenda when a much broader net was cast for recruits and as part of an apologia for the decision to adopt such tactics.
The rest of the fifteenth was anticlimactic. Admiral Davison withdrew without launching a second attack. Late in the day the Japanese sent out twenty planes in two additional attack waves which returned due to weather without finding Davison’s carriers.
[1] San Jacinto (CVL-30), War Diary (WD), October 1944; Air Combat Action report, Commander Air Group 13, October 1944.
[2] Translations of captured Fifth Base Air Force orders and messages for this period are found in ADVATIS Translation No. 99. Additional details are in Inoguchi (note 2, Part II). See also Japanese monographs Nos. 12 and 84 (note 1, Part I).
[3] Per Urgent Intelligence Report No. 32, ATIS Bulletin No. 1602, known losses up to 1500 hours were one failed to return and seven “damaged.” An incomplete 3rd Air Army message dated 17 October 1944 indicates nine were lost.
Secondary sources, Hata et al, Japanese Army Fighter Units and Their Aces, London: Grub Street (2002) p. 286; Ichimura, Ki-43 ‘Oscar’ Aces of World War 2, Oxford: Osprey Publishing (2009) p. 84, place air combat losses at ten or nine (Ichimura), respectively. Several Japanese fighters were destroyed or damaged on the ground.
[4] V-Adm. Matome Ugaki who survived the incident in which Adm. Yamamoto was shot down was commander of the 5th Air Fleet in 1945 when the Emperor’s surrender message was broadcast. Before a formal cease fire message was received he ordered, participated in, and died in a suicide mission on 15 August 1945. He flew in a Suisei dive bomber along with other aircraft and crews. No Allied ships were hit that day. Several Japanese aircraft were claimed shot down approaching Allied ships.
[5] Millot, Bernard Divine Thunder, New York: Pinnacle Books (1971), p. 45.
[6] 5 FGB [Fifth Base Air Force] Urgent Combat Report (15 Oct), para (b), ADVATIS translation No. 99 (Bulletin 50-B- 99).