Chapter 15: The Flame Thrower in the Pacific: Marianas to Okinawa
The Need for a Mechanized Flame Thrower
The stimulus for the development of the mechanized flame thrower, as for the portable, came from the war against Japan where the enemy’s excellent defenses and stubborn resistance called attention to the utility of flame. Although the portable flame thrower gradually proved to be an effective weapon against the Japanese, one of its inherent disadvantages, the vulnerability of the operator, suggested the portable’s installation in an armored vehicle. It is not to be inferred that CWS engineers in the United States had been unmindful of the possibilities of a mechanized flame thrower; the Chemical Warfare Technical Committee had advocated the development of such a weapon in May 1940. But for various reasons, work on the mechanized models in the zone of interior proceeded slowly and fitfully throughout the entire war.1
There were several attempts in the Pacific to mount portable flame throwers in some sort of armored vehicle, a combination made the more appealing by the lack thus far of cannon or other antitank weapons in the enemy bunkers. In the South Pacific Area, for example, the commander of a tank battalion in New Caledonia installed a flame gun in the pistol port of a tank, and a chemical officer on New Georgia modified the flame gun so that it could be fitted into the aperture for the tank’s bow machine gun. The 1st Marine Tank Battalion, serving in the Southwest Pacific, mounted several portable flame throwers on its tanks in preparation for the New Britain operation. None of these improvisions could have been called successful. The portable flame thrower was not constructed to withstand the vibrations and jarrings
of a moving tank, and its fuel capacity was much too limited.2 Word came from Washingon pointing out the disadvantages of range and fire hazards from such modifications and counseling patience until the arrival of perfected mechanized flame throwers from the zone of interior.3
After the bloody battle of Tarawa, which opened Allied offensive operations in the Central Pacific Area, an even greater clamor arose for a mechanized flame weapon. The portable flame thrower had done its part in that battle, but new weapons and techniques were urgently needed to help prevent the repetition of such staggering casualties. In preparation for the Marshall Islands operation scheduled for February 1944 both the 4th Marine Division and the 7th Infantry Division installed M1A1 portable flame throwers, modified by the chemical section of the 7th Division, in light tanks and LVTs, an amphibious tractor. Included in the 7th Division’s version were special fuel containers manufactured in Honolulu. But the attempts to waterproof the guns on the LVTs were unsuccessful, and the electrical systems of those flame weapons, drenched on landing, failed completely. The tank-mounted versions, plagued with the basic weaknesses of fragility and low fuel capacity, had but modest success.4
The poor results of flame thrower improvisation in the Pacific theaters was no cause for criticism of the responsible chemical officers and tank commanders; the portable flame thrower was basically unsuited for tank adaption. If nothing else, these efforts clearly indicated that improvisation was not the answer and underlined the real need for a mechanized flame thrower in the Pacific fighting.
The Marianas
After the experience at Tarawa, General Richardson, Commanding General, U.S. Army Forces in Central Pacific Area, asked the War Department if mechanized flame throwers were available in the zone of interior. Upon receiving a favorable reply, Richardson requisitioned
forty auxiliary bow gun flame throwers (E4-5) for use in the Marianas operation.5 When these failed to arrive, he utilized local resources to fashion substitutes.6 It was fortunate that this need arose in the Central Pacific, for Hawaii had factories and machine shops and an eager and able chemical officer—Colonel Unmacht, who approached the problem of the mechanized flame thrower with the same efficiency he had shown for the portable.
At this time the marines in the theater were perhaps even more interested in flame weapons than were the Army troops. Late in January 1944 the V Amphibious Corps, preparing for the Marianas, obtained twenty Ronson vehicular flame throwers from Canada. The development work involved in adapting these British-designed weapons for installation in M3A1 light tanks was to fall primarily upon Colonel Unmacht, who utilized CWS, Ordnance Department, Naval, and private facilities for this undertaking. The resulting main armament flame thrower, dubbed Satan, had a range of from 40 to 80 yards, a fuel capacity of 170 gallons, and a duration of fire of 2 minutes, enough, according to Unmacht, to reduce 40 or 50 pillboxes based on 2-second bursts.7 At a demonstration held for interested officers on 15 April 1944, the marines fully recognized the potential of the mechanized flame thrower. The V Amphibious Corps managed to get ten more Ronson units from Canada, and its commander, Lt. Gen. Holland M. Smith, asked the Army authorities in Hawaii to install the Ronson units in M3A1 light tanks in time for the Marianas operation.8 Although medium tanks would have had advantages of better protection and more space and mobility, none could be made available in time to meet the required deadline.9 The Chemical Section, CENPAC, with the cooperation of the 14th Naval District, the V Amphibious Corps, and the Seabees, equipped twenty-four light tanks with the flame
thrower units. As the Satans were being produced Colonel Unmacht conducted a series of 40-hour classes on flame tank operation which were attended by Marine Corps and Army officers and men. By mid-May the weapons were tested, waterproofed, and loaded on ships.
The invasion of the Marianas began early on 15 June 1944 when the V Amphibious Corps, known as the Northern Landing Force and consisting of the ad and 4th Marine Divisions, invaded Saipan.10 Each division had 12 Satan tanks. These 12 plus 3 conventional tanks formed a company of three platoons. Each platoon (4 Satans and light tank) was attached to a company of medium tanks—the organic armored support of a marine regiment. Landing on D plus 2 (17 June) the flame tanks saw infrequent use during their first day of battle and then only for the purpose of mopping up. Next day the tanks took part in front-line action and thereafter, as tankers and infantrymen alike quickly learned flame tank techniques, the Satan proved to be an effective weapon.
Targets were varied—pillboxes, brush, canefields, buildings, and caves. Typical action against stiff opposition saw flame tanks neutralizing targets under cover of medium tanks. A tank commander, interviewed shortly after the end of the operation, told of one such action in which the Satan, supported by conventional tanks, came forward to flame a well-defended pillbox. As the target started burning, two Japanese sprang out, only to be cut down by rifle fire. Resistance ceased. An examination of the bunker revealed ten other Japanese, grotesquely dead at their firing positions. The tank commander added that the mechanized flame thrower proved to be the only effective weapon against caves.11
After the fall of Saipan the ad and 4th Marine Divisions immediately started preparations for the invasion of Tinian which was to begin on 24 July. In this attack the MIA’ light flame tanks, loaded aboard LCTs and LCMs, followed hard on the heels of the first assault wave. The composition and attachment of flame tank units during the
fighting were just as they had been on Saipan.12 The terrain on Tinian proved much more favorable to tank employment, and this, combined with the recently acquired combat experience, resulted in a profitable use of flame vehicles. The Satans again combined with the medium tanks against the more tenacious points of resistance. They also were successful against caves and when used to burn vegetation concealing enemy positions.13
The Southern Landing Force, as the III Amphibious Corps was designated for the Marianas campaign, assaulted Guam on 2r July with the 3rd Marine Division, the 1st Provisional Marine Brigade, and the 77th Infantry Division as its principal combat elements. The 3rd Tank Battalion, which landed on the first day, had among its armament six M4A2 medium tanks equipped with E4-5 auxiliary bow gun flame throwers. These weapons had arrived from the zone of interior for service tests shortly before the operation began. One of the flame tanks met the enemy at Assan Point on the second day of the battle. Supported by a conventional tank the flame vehicle approached an enemy cave and fired half of its 25-gallon charge into the mouth. Seventeen Japanese soldiers were incinerated. A similar attack near Chonito Cliff resulted in 30 enemy dead. During the next five days the auxiliary flame tanks continued to burn out resistance on Guam.14
A few logistical difficulties arose during these flame operations in the Marianas. Planners for the invasion had estimated that the daily expenditure of each flame tank would be one load of fuel; actually, two loads were required. A shortage of napalm meant that most of the flame fuel was either diesel oil mixed with Bunker C fuel obtained from the vessels, or, at times, straight Bunker C. A post-campaign recommendation called for adequate amounts of thickened fuel for future operations to insure a longer, more effective range for the mechanized flame thrower.15
Flame tank crews, contrary to some pessimistic predictions, suffered no casualties as a result of actual flame operations, although two men were injured when a vehicle struck an enemy land mine. Fatigue
became a problem because of the extremely cramped positions within the light tanks. While the marines were generally impressed by the flame tank performance, they were critical of the range of the flame gun, the limited visibility and light armor of the tanks, the lack of special tools and spare parts, and the manner in which the Ronson was installed in the tank. And experience in the Marianas substantiated the preinvasion opinion that flame throwers should be mounted in medium tanks.16
The successful debut of the mechanized flame thrower in the Marianas campaign generally overshadowed the accomplishments of the portable model. The truth was that the two types of flame throwers supplemented one another; the mechanized afforded greater protection to the operator and delivered larger amounts of flame for longer distances, while the portable, capable of quicker and more flexible employment, attacked targets inaccessible to the tank-mounted type. Portable flame throwers saw action on Saipan on D-day before the tanks were landed and proved invaluable during the street fighting in the village of Garapan. Later, these weapons helped overcome stubborn resistance of cave defenses located in Saipan’s cliff formations, defenses which the tank-mounted flame throwers could not reach. On Guam the portable flame throwers were committed against the cave defenses of Chonito Cliff within an hour of the actual landings, and they remained busy during the entire three weeks of fighting.17
On at least one occasion the two types of flame thrower were employed in the same action. During the battle for Tinian, marines first used the cannon of a medium tank to blast defended caves, then the light flame tanks to spray the openings, and finally assault teams with demolitions and portable flame weapons to actually reduce the positions.18
Of the two Army divisions in the operation the 27th continued to assign the weapons to the combat engineers while the 77th gave them to
the infantry. The marines, prompted by previous experience, assigned the portable flame throwers to the infantry.19
Although the portable weapon stood up well under hard usage in the Marianas, weather conditions caused innumerable problems. Many flame throwers were deadlined through battery and spark plug failure and because of inadequate waterproofing. The former difficulty was greatly relieved with the introduction of the new M2-2 flame thrower (flown from the zone of interior) and its improved ignition system.20 There was a good supply of flame throwers present, but no provision had been made for transporting them or the heavy service equipment needed for their maintenance.21
Peleliu
Tarawa had called the Navy’s attention to the possible use during the initial stages of an assault landing of a mechanized flame thrower mounted in some sort of landing craft. The National Defense Research Committee provided a speedy solution to this problem. The organization had recently developed the Q model flame thrower for light tanks only to find that vehicle out of favor. It now adapted the Q flame unit to fill the Navy’s needs. Newly christened the Navy Mark I, five units of the flame thrower reached Hawaii in April 1944. The Army inherited these units when Navy authorities turned down the model as unsuitable for amphibious operations because of its excessive weight. The 43rd Chemical Laboratory also found that this flame unit, with a fuel capacity of 200 gallons, a firing time of 74 seconds, and a maximum range of over 1oo yards, was too bulky and heavy for installation in tanks.22
Interest in the Mark I flame thrower then arose in another quarter. Early in June 1944 a Navy flame thrower detachment from the United States joined the 1st Marine Division on Guadalcanal which was preparing for the invasion of the Palau Islands. This detachment,
Flame thrower on an amphibious tractor blasts cave on Peleliu Island.
consisting of one officer and three enlisted men and attached to the division’s amphibian tractor battalion, mounted its three Mark I flame units on LVT(4)s, the armored amphibian tractor. Tests soon revealed that there were certain drawbacks to this combination. The severe vibrations of the vehicle operating on land shook loose the gunner’s protective shield and cracked the porcelain of the spark plugs, failures which were only partially remedied by improvisation and substitution. The naval detachment began to instruct the marines in the operation and maintenance of the flame tractors but, unfortunately, a shortage of napalm prevented adequate training of men and testing of weapons. Three flame units which arrived from Hawaii shortly before embarkation were held in reserve.23
The island of Peleliu was the principal objective of the 1st Marine Division attack in the Palaus. Each of the three assault regiments was to have a flame tractor in support. Crews of these flame vehicles received briefings on the over-all operation and detailed instructions in their own particular mission. The flame tractor with the 1st Marine Regiment would advance just behind the initial wave of landing craft. Upon reaching the beach it was to fire on targets of opportunity, following the infantry as it pushed inland. Any infantry officer could commandeer the flame tractor if an appropriate target appeared. The flame vehicle with the 5th Regiment would land and advance with the first wave of assault troops, and the 7th Regiment’s flame support would accompany a flanking group of landing craft.
The three flame tractors, plus two service tractors carrying an air compressor and extra fuel, were loaded in a landing craft, tank, which in turn made the 2,000-mile journey from Guadalcanal to Peleliu on the deck of a landing ship, tank.
On the day of attack, 15 September 1944, the elaborate plan for using the flame tractors completely broke down. The 1st Marine Regiment was stopped by stiff resistance just beyond the beach, and its flame tractor waited five hours for some kind of order. The flame vehicles with the other regiments were told to stand offshore out of danger. When the three flame tractors eventually landed they stood idle on the beaches, a result no doubt of extreme confusion and the unfamiliarity of the marines with the weapon. Inactivity on the second day was caused by the fact that the air compressor had not yet landed. The flame vehicles saw action on the third day, and from then on their commitment was regular.24
The troops used the flame tractor principally to neutralize caves, pillboxes, and dugouts and to burn the cover from the battleground. Japanese, hidden in defiladed positions, were often caught by arching rod-like streams of burning, thickened fuel which hit the reverse slopes. The marines controlled this type of fire by radio from observation posts. In the first phase of the fighting the flame tractors usually worked ahead of the infantry, even ahead of the tanks. As the attack slackened upon reaching the hills, the tanks and infantry provided support for the flame weapons. The latter could have been more effective, particularly from commanding positions on hills and cliffs, but for their
light armor plate and their lack of all around protection. After the first few days on Peleliu the employment of the six flame tractors was steady until about D plus 40 and then sporadic until D plus 75. The flame weapons attacked about 100 caves and twenty-five pillboxes and dugouts. They burned off some forty acres of cover, expending fifty loads of fuel in the process.
The over-all usefulness of this mechanized flame thrower suffered from the inadequacies of the tractors. Damage to the flame units themselves was negligible and usually repairable within several hours. But the amphibious tractors could not withstand the rugged terrain and were constantly out of action with engine troubles, torn tracks, and broken final drives. The 6 flame throwers were employed for 6s days (or a total of 366 flame thrower days) , and almost one-third of the time was spent on tractor maintenance. This unfortunate situation occurred despite the fact that the detachment had a total of 19 tractors in which to mount the 6 flame units. In addition to these frequent breakdowns, insufficient training of the crews hampered efficient operations. Gunners sometimes were unable to estimate the range of the target, and incomplete knowledge of the flame gun prevented the crew from taking care of small malfunctions in the field. All in all, experience on Peleliu demonstrated that the LVT was not a suitable vehicle for flame throwers; this was the conclusion of the Navy flame thrower detachment. But despite the failure of the mount, the marines were impressed by the Navy Mark I flame thrower. In some respects they considered it superior to the Ronson flame gun used in the Marianas, especially in length of range.25
An interesting sidelight to the Peleliu operation was the use of a high pressure hose to carry flame to targets beyond the range of mechanized flame weapons. During the fighting at Umurbrogol Pocket, engineers attached a 500-yard hose to a fuel tank and, with pressure provided by booster pumps, operators sprayed Japanese positions with flame much as firemen direct water on burning buildings.26 So impressed was General Richardson with this weapon that he asked Colonel Unmacht to continue the investigation of its combat potential. In subsequent tests the chemical section attached a 400-foot length of
standard 1½-inch rubber fire hose to a flame thrower tank and attained a range of 60 yards. Following a demonstration of the device in February 1945 Tenth Army ordered three sets of these hoses, all of which were to see service on Okinawa.27
The Philippines
Preparations for the Portable Flame Thrower
In May 1944 Sixth Army’s commander, General Krueger, expressed to his staff chemical officer some misgivings about the past performances of the portable flame thrower.28 As a consequence, a study was made under the aegis of Sixth Army’s G-3 to determine what revisions should be made in the Army’s official flame thrower doctrine.29 This study revealed several reasons why the flame thrower had been used with less than maximum effectiveness. Commanders had incomplete understanding of the capabilities and limitations of the weapon; as a result, it was of ten committed to action without any chance of success or, conversely, was not introduced into situations where it might have been helpful. A corollary to this was the failure of unit commanders to comply with Sixth Army doctrine requiring the employment of portable flame throwers as an integral part of the assault party.30 A commander, upon brief reconnaissance, would send forward a poorly protected flame thrower on a mission that would have been difficult for a much larger force. Or flame thrower operators would be accompanied by a security detachment which had been hastily formed for the emergency. These improvised groups seldom jelled into the efficient teams envisioned by the writers of Sixth Army’s training memo on flame thrower operations.
Commanders had difficulties even when they tried to adhere to the established doctrine. During training, an assault party working as a unit might attain a high standard of proficiency. But in the interval
between the completion of training and the test of battle, the group normally was so broken up by promotions, transfers, and casualties that the maintenance of organizational integrity was impossible. To complicate the situation still more, the flame thrower was officially classified as a secondary weapon. This meant that with the close of the training period flame throwers were returned to unit supply. Thereafter, whenever an attacking force unexpectedly encountered a fortified defensive installation, an assault party could not be sent forward at once. Instead, flame thrower operators first had to go back to the unit supply point to pick up and fill their weapons. Such delays were often costly.31
Chemical and other officers concerned with flame thrower operations recognized the inadequacies of Sixth Army flame thrower doctrine, especially as it applied to assault parties. As a remedy they suggested the establishment of permanent units whose principal mission would be attacks on fortifications and to whom the flame thrower would be assigned as a primary weapon. Specifically, some recommended the conversion of chemical processing, 4.2-inch mortar, or antitank units into chemical flame thrower companies or platoons attached to regimental headquarters.32 Although flame thrower platoons of this type had already been organized with success by regiments of the Americal Division on Bougainville, Sixth Army rejected this idea for its own units. It issued instead, on 22 June 1944, new instructions which emphasized the technical and tactical training of assault teams.33 Each infantry battalion and cavalry squadron was to form and maintain on a permanent basis at least one assault party, to include a leader, an assistant leader, 4 flame thrower operators, 2 demolitions men, 2 rocket launcher men, 2 BAR operators, and 4 riflemen. Trained to reduce fortified enemy positions, the assault party was to be held in reserve during combat until appropriate targets appeared. Company and battalion commanders and executive officers were expected to familiarize themselves
with the capabilities, limitations, and tactical employment of these assault parties, so that they would be committed with maximum effectiveness.34
Having resolved the question of tactical employment, Sixth Army turned to flame thrower servicing problems. It had always been difficult to refill fuel tanks, replace empty gas cylinders, and test and repair flame throwers rapidly and efficiently near the front lines. But as long as demands for the weapon were infrequent, speedy servicing had been a minor consideration. The difficulty reached critical proportions only on Biak, where for the first time flame throwers were in almost daily use.
As an interim measure, Lt. Col. Frank M. Arthur, 41st Division chemical officer, devised a plan for mounting two charging sets, each consisting of two hydrogen and five nitrogen cylinders and a fuel mixing kit, on jeeps and trailers. Teams of technicians from the 94th Chemical Composite Company operated these sets and serviced flame throwers down to the regimental level. Experience proved this system inadequate, since it was at the battalion and the company level that servicing was found to be most desirable.35
Although Sixth Army agreed that the assignment of CWS personnel to lower combat units was the most likely solution to the servicing problem, the limited number of chemical troops prevented such dispersal. In planning for the Leyte operation Sixth Army therefore decided that battalion and company personnel would service the flame throwers and perform first echelon maintenance. The task of second and third echelon maintenance and the preparation of proper fuel mixtures was assigned to chemical service platoons attached to corps and divisions. This arrangement was to work even more successfully on Luzon than it did on Leyte, primarily because of the presence on Luzon of a greater number of chemical service platoons.36
Employment of the Portable Flame Thrower
The return to the Philippines took place on 20 October with the assault on Leyte by Sixth Army’s X and XXIV Corps. The portable
flame thrower saw action the first day and continued to play a significant role throughout the operation. In all, 229 weapons were on hand and, except for the 11th Airborne, each of the divisions on Leyte used the flame thrower in combat.37
The campaign demonstrated the validity of the new Sixth Army policy on tactical doctrine and servicing. Flame thrower operator casualties were light, malfunctions rare, and assault teams successful in the large majority of their missions.38 Japanese bunker, cave, and dugout defenses on Leyte were elaborate and often ingenious. The success of the flame thrower pointed up the merits of the weapon and the training and skill of the operators.
On 29 October a battalion of the 17th Infantry, 7th Division, encountered the strangest defensive position of the campaign. During the fight for Dagami, Company L came upon a cemetery south of the town which was overgrown with weeds and filled with stone crypts built above the ground. Encountering no resistance the leading elements of the company passed through the graveyard. The support platoon followed. When the platoon was halfway through, a headstone tilted back revealing four Japanese in the grave, armed with rifles and an American BAR. The enemy troops could not be dislodged until a flame thrower came forward and burned them out. The platoon broke up into small details and pushed its way through the rest of the cemetery, eliminating enemy fighters as they were located. Company K, following Company L, also received fire from the stone crypts. It became evident that the enemy had removed the bodies from their tombs, punched holes through the stone, and had thus established a series of small pillboxes. The company commander withdrew to a path in the cemetery and, lining up his men shoulder to shoulder, sent them through the cemetery behind a battery of six flame throwers. This effective, albeit unorthodox, formation burned its way through the macabre defenses, destroying about 30 of the enemy in the process.39
Flame thrower assaults by other units on Leyte, though less bizarre, were no less successful. Regiments of the 96th Division, equipped with the new M2-2 flame thrower, found it useful in burning off kunai grass which hid Japanese emplacements. On one occasion a flame
thrower was used at close quarters to repel an enemy attack on American tanks.40 The division reported that Japanese troops were reluctant to stay in positions attacked by flame throwers and that the weapon was “a very important factor in overcoming the enemy’s inherent will to resist.”41
Reports from the 1st Cavalry Division and the 32nd Infantry Division substantiated this estimate of the portable flame thrower. In mid-December an element of the 1st Cavalry Division encountered an enemy emplacement on a narrow ridge in the Mount Minoro sector which commanded the approach along the mountain trail. Repulsed in several frontal attacks cavalrymen, supported by an assault team, circled wide and approached the position from the rear. The flame thrower sprayed the emplacement, and the defenders were cut down by small arms fire as they fled from their position.42
On 15 December troops of the 126th Infantry, 32nd Division, were halted in the Ormoc corridor by strong resistance from a pillbox and riflemen situated about twenty yards below the crest of a hill. After one attempt to overcome these defenses had failed, the regiment formed an assault party of fifteen riflemen, a bazooka team, and two flame thrower operators. The party crept around the enemy and opened fire simultaneously with grenades, bazookas, and flame throwers. One flame thrower operator directed his fire on the pillbox, while the other sprayed the rifle positions. The attacking riflemen found that the badly burned and demoralized Japanese offered little resistance.43
Units on Leyte for the most part adhered closely to Sixth Army flame thrower doctrine. Those who did not soon came to grief. Some unit commanders, for example, continued to use but one flame thrower in an assault party. By thus failing to allow for possible misfires or casualties they endangered the success of their mission. Moreover, despite repeated warnings, weapons were occasionally committed to action without sufficient servicing. The 32nd Division, which arrived on Leyte on 14 November 1944, unloaded its flame throwers and sent them directly to the front line without a check. Misfires and malfunctions
resulted. Not until the weapons were called back and properly serviced did they function satisfactorily.44
The Leyte Campaign was followed in January 1945 by the invasion of Luzon, the largest and most populous island in the Philippines.45 Although suitable flame thrower targets did not develop during the 9 January landing at Lingayen Gulf, they began to appear a week later in I Corps sectors north and east of the beachhead. Three units of I Corps had particular success with the weapon. On the 26th a company of the 158th RCT, operating in mountainous terrain, encountered a series of long, curved, and defended tunnels which ran about thirty feet into a slope. An assault team moved toward the entrances under cover of smoke from white phosphorus grenades. The flame thrower operator fired a long burst of fuel into the opening of the first tunnel and flushed a number of burning Japanese, who were killed by small arms fire as they fled from the entrance. The other openings were treated in a like manner. Japanese bodies, victims of burns and suffocation, were found along the smouldering corridors in the most remote parts of the tunnels.46
Although other units reported similar success with the portable flame thrower, the weapon was not ideally suited for combat in many parts of Luzon. The 43rd Division, for example, reported that in the open terrain which featured much of the island, flame thrower operators found their approach much more hazardous than in the undergrowth of the jungle. As a consequence, weapons with longer ranges had to be used in the reduction of enemy fortifications.47
The flame thrower proved more effective against urban targets. The bitter battle for Manila marked the first extensive city fighting in the Southwest Pacific, and flame throwers, although untested in this kind of combat, saw much use in the street-to-street struggle. The weapon proved particularly valuable in routing the enemy from the intricate positions of the Intramuros, the old walled city of Manila.48 The portable flame thrower also played a part in the fighting in the Philippine
General Hospital, the Legislative Building, and the General Post Office—fighting which often turned into a series of fierce room-to-room and floor-to-floor battles. American troops blasted holes in walls with explosives and then used flame throwers against the enemy in the next room. The 37th Division reported that it used over three hundred fillings during the Manila operation, or an average of more than four fillings for every flame weapon in the division.49
The capture of Manila did not end the fighting on Luzon. There remained the task of clearing out the more remote mountainous regions of the island. In this type of action flame throwers were infrequently used. Units reported that the flame thrower was too heavy, making it difficult for one man to carry for prolonged periods, and, since few extra men were available to haul the weapons, they often were left behind. On several occasions I Corps resorted to Filipino carriers to transport the weapons from the service point to the line of departure, a distance of several thousand yards and at times over steep mountain trails. Under such conditions flame thrower employment was limited to reasonably accessible targets.50
The policy of using flame thrower assault parties worked out successfully as long as infantry casualties were low. But when casualty rates mounted and every available man was wanted for the immediate needs of combat, it became impossible to hold these organized teams in permanent reserve. Commanders broke up these groups and sent their personnel to line companies. Missions for which assault parties had been organized fell to the rifle squad or platoon closest to the action.51
Sixth Army, disturbed by this situation and desirous of avoiding similar difficulties in its forthcoming campaign on the Japanese mainland, changed its training doctrine somewhat. Representing the accumulated experience of several years of combat, a new directive was issued which differed in only one significant respect from its predecessors—it rejected the concept of “organized assault parties in permanent reserve.” Instead, it stipulated that at least three squads from each rifle
company be trained to use special assault weapons and that four men in each of these squads be flame thrower operators. These trained groups were to operate as normal rifle squads until a situation arose in which they were needed to perform their specialty.52 This change did not alter the basic concept of using flame throwers solely as component elements of an assault team. What was attempted was a further integration of the flexible assault team within the organizational framework of the normal infantry unit. The war ended before this new system could be tested in combat.
American forces began the Philippine campaign only partially equipped with the new M2-2 flame thrower. As these weapons became available, and as stocks of the M1A1 model were exhausted, the new weapons were issued to combat troops.53 Supplies of the M2-2 flame thrower never were sufficient to enable the M1A1s to be completely withdrawn from service, but by V-J Day the newer type was in the hands of the majority of units operating in the Philippines.
Those who were obliged to carry the M1A1 had to maintain a vigilant servicing policy, especially during the rainy season, for despite waterproofing the ignition system of the weapon was still unreliable. Those equipped with the M2-2 also had problems. The pressure regulator proved to be entirely unsatisfactory and before the end of the Philippine operations had been replaced by an entirely new type.54 Flame thrower operators complained that the M2-2 could not maintain pressure long enough, a result of inherent deficiencies in the weapon as well as of poor maintenance procedures. Troops who filled the flame thrower forgot that heat was generated when a cylinder was charged and the subsequent cooling-off process could result in a drop of as much as 200 pounds in pressure. Conversely, pressure built up in the fuel tank when the tank was exposed to the direct rays of the sun, blowing out the safety discs in extreme cases.55
The Mechanized Flame Thrower
Not counting a limited use of improvised models, the first employment of the mechanized flame thrower in the Southwest Pacific took place in the Philippines. The XXIV Corps had been slated for the assault on Yap, in the Central Pacific Area, when the Joint Chiefs of Staff canceled the Yap operation and ordered General MacArthur and Admiral Nimitz to invade Leyte on 20 October 1944, two months to the day before the planned date of the operation. The XXIV Corps, which was now to take part in the assault on Leyte, had already embarked for Yap, when this change took place. A terrain study of that island had recommended the utility of the mechanized flame thrower particularly during the assault, and ten Ronson flame throwers, nine on light tanks and one in an amphibious tractor, were in the convoy.56
Not only was Leyte’s terrain far less suitable for tracked vehicles than that of Yap, but the invasion of Leyte took place during the rainy season, a circumstance which all but immobilized the mechanized vehicles. During the landing one resisting enemy bunker was quickly neutralized by the lone flame tractor, and during the first few days on Leyte the light tank flame throwers burned off the foliage from a number of concealed enemy positions. But the fighting soon reached the mountains, which put the tanks at further disadvantage.57
The Luzon fighting saw the arrival of both auxiliary and main armament flame throwers from the zone of interior. The XIV Corps included a tank battalion, equipped with bow gun flame throwers, that was employed in the house-to-house fighting in Manila. In one instance a flame tank neutralized in a few minutes a barricaded building which had defied the infantry for two days. Twenty-five more bow gun flame throwers reached the Sixth Army in mid-April.58
On 3 April 1945 the first and only main armament flame throwers from the United States to see action with Army troops overseas arrived on Luzon. These were four NDRC Q model (E7-7) flame throwers mounted in M5A1 light tanks and scheduled for service testing. The flame throwers and a service truck were assigned to the 13th Armored
Group.59 By this time the fighting on Luzon had reached its final stages, and the combat activity at Ipo Dam, Balete Pass, Villa Verde Trail, and Baguio took place in mountainous terrain, which made even the maneuvering of light tanks difficult. Moreover, the enemy artillery and mortar fire limited the activity of the thin-skinned M5A1s. In mid-April these main armament tanks were attached to I Corps for 25th Division operations at Balete Pass. Following a demonstration before officers of the division, the tanks, which were assigned to the 27th Regiment, saw action almost at once. The objective was an enemy position located on the reverse slope of a ridge. One of the flame tanks moved along the narrow ridge which provided the only avenue of approach and, when even with the target, turned down the slope and attempted to flame the Japanese. Thwarted by the limited maximum depression of the gun, the tankmen fired two quick bursts and withdrew.60
Mechanized flame throwers experienced the usual difficulties of tank-infantry communications. Close contact between the two was particularly important because of the limited visibility in the mountains where infantrymen often had to guide the tank to its objective. A power telephone on the rear deck of the tank was rigged to the turret, allowing the turret operator to communicate with the infantry but only at the cost of cutting off turret traversing power whenever he spoke.61
The four main armament flame tanks advanced to the front lines several times during the next twelve days, but remained uncommitted. On 27 April the advance of Company K, 27th Infantry, was stalled for more than twenty-four hours by deadly machine gun and 47-mm. fire from enemy emplacements on the ridge’s forward slope and the valley beyond. Thick underbrush, which made friendly grenades ineffective, covered the entire defensive position, while fire from U.S. tanks and artillery was unable to reach the enemy’s defiladed position.
In this situation the commanding officer of the flame detachment suggested the use of his tanks. The heavy underbrush made the approach exceedingly difficult and a medium tank and then an armored
bulldozer were pressed into service to clear a way. Two flame tanks were at the assembly point 200 to 300 yards from the principal target—a pillbox across a ravine at the foot of a large tree. Because of the narrow approach, only one flame tank could advance at a time and, as there was no room for armored support by the medium tanks, two rifle platoons from Company K furnished protection. An infantryman, walking behind the tank, guided it over the crest of the ridge. When the enemy resisted with hand grenades, the bow gunner sprayed the suspected areas with machine gun fire. Approximately fifty yards from the target the tank ran on a log twelve inches in diameter. The driver carefully maneuvered the vehicle until it balanced just forward, thus permitting direct aim down toward the target. The first burst of flame flushed eight Japanese who were killed by supporting infantrymen. Several short bursts, fired across the target, caused the entire area to burn briskly for about five minutes. When the smoke cleared, another dugout became visible and received what was left of the flame fuel. The infantry advanced promptly, and by dusk Company K had taken its objective of Lone Tree Ridge, 200 yards beyond. There were 53 enemy dead in the area, 6 dying in the flame attack and the rest killed by the infantry as they ran from cover. Friendly forces suffered no casualties.62
The flame tanks had varying success in the subsequent fighting on Luzon. Sometimes they burned out the enemy; other times they were stymied by their vulnerability to the heavier Japanese weapons. There were few mechanical difficulties and, in spite of heavy rains, no ignition failures. Although infantrymen and tankers were enthusiastic about the flame tanks’ ability to rout the enemy from strong positions, these combat tests indicated that the thinly armored light tank was not a satisfactory mount for the flame thrower.63
Iwo Jima
The V Amphibious Corps invaded Iwo Jima in February 1945. It was a costly campaign waged against an enemy entrenched in a superlative defensive system. But as one Marine Corps historian has stated:
“Never before in the Pacific War had troops engaged in amphibious assault been able to see so clearly the immediate importance of the objective.”64 The tiny island would provide an intermediate base for fighter escorts for the bombers headed toward Japan. It would also be a haven for crippled B-29’s unable to make it to airfields in the Marianas.
The preinvasion preparations of the three Marine divisions earmarked for Iwo included work with both the portable and the mechanized flame throwers. For the smaller weapon this preparation consisted mostly of integrating the flame throwers with the assault teams. It was 3rd Marine Division policy to have in each of its battalions an assault platoon made up of 6 men from each rifle company and 2 men from the battalion headquarters company. These 20 men, specially trained in the use of flame throwers, rocket launchers, and demolitions, were placed under the immediate control of the battalion commander. The commander in turn could attach the whole platoon, or a part of it, to his assault companies as the situation required.65 Other Marine divisions devised similar platoons. Regiments of the 4th Marine Division organized 39-man platoons which were attached to each battalion landing team. Squads from these platoons were then assigned to companies for specific missions.66
A plentiful supply of M2-2 portable flame throwers made possible these elaborate preparations—the battalion had 27 weapons or 243 for the division. A battalion usually assigned I flame thrower to each of its 9 rifle platoons, 9 to the battalion supply section, and the remaining 9 to the regimental service platoon. This arrangement insured the rapid replacement of lost or damaged weapons. It also made possible a reserve flame thrower for every one in use, a justifiable ratio because of the weapon’s vulnerability to enemy fire and the fact that it was difficult to service.67
The 4th and 5th Marine Divisions each received four Hawaiian-made main armament mechanized flame throwers. The development and procurement of these flame throwers was undertaken by the
Neutralizing a cave, Iwo Jima
Chemical Section, Pacific Base Command, with help from the Navy, the Marine Corps, other Army elements, and civilians. Briefly, the flame unit, designated the POA CWS Hi, consisted of a Ronson flame thrower installed in a salvaged 75-mm. gun tube and mounted in a medium tank. It had a maximum effective range of about 100 yards (with thickened fuel) , a capacity of nearly 300 gallons, and a firing time of 150 seconds.68
Meanwhile, the Japanese defenders of Iwo Jima had been busy strengthening their defenses. Engineers well versed in fortifications came from Japan to supervise the work. They added a series of trenches and pillboxes to the natural caves which formed the backbone of the island’s defenses. Whenever possible these features were integrated
into a single network permitting troops to move from one place to another without exposure. Some of the strongpoints had multiple entrances and apertures which were invulnerable to artillery fire.
The Japanese did their work well. Not only did it take a month to capture this tiny island, but during the course of the operation marines actually saw very few of the enemy. The Japanese commander expressly forbade banzai charges and other mass heroics, ordering his troops to stick to their positions to the last. Moreover, once a position had been silenced, it was by no means safe to assume that it would give no further trouble. The Japanese infiltrated at night, both above and below ground, and re-occupied defenses which previously had been neutralized.69
On 19 February 1945 the 4th and 5th Marine Divisions landed on the south coast of Iwo Jima with the 3rd Marine Division in reserve. From the beginning, portable flame throwers were a big help in cleaning out caves and pillboxes. A battalion in the 3rd Division described the usual employment of the portable weapon. Covered by rifle and BAR fire, the flame operator moved to a firing position on the flank of a target. He then fired across the face of the cave or embrasure, advancing quickly under his own fire until he could shoot directly into the opening. This tactic consistently silenced the position, which was then destroyed by demolition charges.70
At the end of the campaign some units commented on the superiority of the M2-2 model over its M1A1 predecessor, particularly in regard to its ignition system. The familiar cry for greater range continued to be heard. As far as maintenance was concerned, experience varied—one battalion reported no difficulty or malfunctions, while others lamented the complete absence of spare parts.71
Because enemy emplacements were difficult to approach, flame thrower operator casualties shot upward. The casualty rate on Iwo Jima was probably higher than in any other campaign in the Pacific.
One battalion reported 92-percent losses.72 As a result, there was a scarcity of competent and experienced operators, particularly as the battle reached its final stages. By way of correction, commanders recommended greater emphasis on flame thrower training and suggested that all enlisted men be given at least a working knowledge of the weapon.73
The eight main armament flame throwers were lightly used during the first few days of the operation, largely because of the rough terrain. But within the week the marines became dependent upon flame tank support and constantly requested its use.74 Typical tactics were as follows: the Marine infantrymen advanced, drawing fire from enemy positions; armored dozers then pushed ahead and, protected by tank and small arms fire, prepared a road to points within range of enemy positions; flame-throwing tanks moved forward and neutralized the enemy positions with flame and machine gun fire so that the infantry could advance; engineers followed closely behind with demolitions and destroyed the enemy caves and pillboxes to prevent their further use. This was a slow and laborious process, but any attempt on the part of the marines to continue the advance before bypassed positions had been destroyed resulted in heavy casualties, and no real gain could be made as long as the Japanese could fire on American troops from the rear.75
The northern area of the island was featured by deep gullies, saw-toothed ridges, sheer cliffs, and eroded, boulder-strewn plateaus, and here the Japanese holed up for their last desperate stand. A liberal use of portable and mechanized flame throwers helped to eradicate this kind of resistance, although the formidable terrain made it hard to get the flame tanks within effective range of enemy targets. Generally, a tank dozer nudged out a few yards of roadway and moved aside. The flame tanks then advanced, sometimes two or three in rapid succession, and fired while regular tanks stood back and provided cover. This operation was repeated so often that on some days the 5th Marine Tank
Battalion used between 5,000 and 10,000 gallons of thickened gasoline in its three or four flame tanks. The limited supply of flame tanks and tank dozers was a big handicap. And despite the intricate procedure outlined above the rugged terrain often defied all attempts at penetration.76
The fixed assignment of flame tanks to tank companies proved to be impracticable on Iwo. Placed in a battalion pool and stationed at the refueling area, flame tanks were made available by the battalion wherever needed. Tank companies or platoons sent requests through battalion, the infantry companies through liaison personnel. As soon as a flame thrower exhausted its fuel, it returned to the refill station for a new load and possibly a new assignment. In spite of all these measures, troop advances were frequently held up because all flame tanks were engaged elsewhere.77
The 4th and 5th Marine Divisions emphasized the greater effectiveness of the main armament type as opposed to the auxiliary flame thrower, of which the 4th Division had twenty-four. These divisions recommended more large capacity flame throwers and their incorporation as organic equipment in all tank battalions.78 Although the 3rd Marine Division entered the Iwo Jima operation without main armament flame throwers, it was later on able to borrow some from the other divisions. In its opinion this type was better against enemy defenses than the auxiliary bow gun flame throwers with their shorter range and limited traverse.79
After-action reports for the Iwo Jima operation attested to the value of the flame weapons. Of the two types, the mechanized flame thrower seemed to have come out on top. One battalion commander called it the “best single weapon of the operation.”80 The V Amphibious Corps report referred to the mechanized weapon as the “only effective means”
of reducing many of the concrete pillboxes and blockhouses encountered on the island.81
Okinawa
Okinawa was the last battle of the war and one of the hardest. Conducted by Tenth Army on the very doorstep of Japan, it was to involve more CWS equipment than any other Pacific campaign. Each of the four Army divisions of XXIV Corps had 141 portable flame throwers; each of the three Marine divisions of III Amphibious Corps had 243, the same number as had been taken to Iwo Jima. The 713th Tank Battalion, converted to a provisional flame thrower unit for the operation, received a complement of 54 POA main armament flame throwers.82
All seven of the divisions to see action on Okinawa were combat tested and experienced in portable flame thrower operations. Employment of the portable weapon closely paralleled that of earlier actions. Especially significant was the fighting in the southern portion of the island where enemy positions most closely resembled those which had been found on Iwo Jima.83 Despite these successes the portable flame thrower was overshadowed on Okinawa by its mechanized counterpart.
The 713th Tank Battalion changed over to a flame battalion in Hawaii, and its own troops had assisted in the installation of the POA flame throwers in the medium tanks. Tank crews test fired and adjusted their flame weapons, while others of the battalion received instruction in mixing the fuel. A tentative table of organization and equipment was drawn up to reflect the differences inherent in a flame tank unit.84
In general, each of the three companies of the 713th was attached to a standard tank battalion in support of a division. In turn, flame platoons joined tank companies, and sections joined standard platoons.
Flame-throwing tank attacks enemy infantrymen dug in along an old road on Okinawa.
A tank platoon, including a section of three flame tanks, was placed in support of an infantry battalion. Flame tank deployment differed within these tank-infantry teams. Sometimes they operated alone with the infantry; sometimes they formed an integral part of the tank platoon; most frequently they waited at forward positions until called upon for particular missions by the tank platoon. When the latter tactic was used the flame tanks, supported by conventional tanks, operated ahead of the infantry until an area had been cleared for the latter’s advance. The troops followed closely, protecting the tanks and occupying the ground before the enemy could recover. This support was extended to both Army and Marine divisions.85
Japanese troops defending Okinawa first faced mechanized flame throwers on D plus 18 (r 9 April 1945) , when thirty-one flame tanks expended 3,500 gallons of napalm in the attack on a hill known as Rocky Crags. Conventional tanks led the assault, with flame tanks and infantry from the 7th Infantry Division close behind. The flame throwers continued in support throughout the 6-day battle for the Crags, which extended from the northwest and covered the approaches to Hill 178 and the high ground west to Tanabaru, the division objective.
From 20 to 29 April 1945 other main armament flame throwers, in support of various units of the 7th, 27th, and 96th Infantry Divisions, continued to fire on rocky ridges, fortified caves, dugouts, pillboxes, machine gun nests, and villages. When the Japanese withdrew, they did so in many cases only because of the devastating effect of the flame. American forces appreciated the psychological value of flame; the commanding officer of one tank battalion reported that the infantry in many cases preferred to advance behind a flame tank rather than behind a standard one:86
During May the fighting took place along the defensive position in the high ground north of Shuri, the ancient capital of Okinawa. For the first ten days flame tanks supported divisions in the battles for Kochi Ridge, Flattop, Dick Hills, Hill 60, and Nan Hill. Meanwhile, other flame tanks neutralized the network of caves on the reverse slopes of the Maeda escarpment. Still others supported the marines in the savage fire fight for Shuri Heights. The last ten days of May saw torrential rains turn the ground into a quagmire of soft clay that immobilized tanks and other vehicles, while the Japanese 32nd Army executed a successful withdrawal to the high ground on the southern tip of Okinawa. Until the rains came, the flame tanks had used about 75,000 gallons of flame fuel. Only 28 of the 54 original main armaments flame tanks were still in operating condition.
The weather cleared in June and on its ninth day a battle began in which mechanized flame throwers played a significant role. Facing the XXIV Corps on its 6,000-yard front was a strong rocky line running from Yuza-Dake on the west to its eastern anchor, Hill 95. On this end vertical cliffs 250 to 300 feet high rose to a plateau topped by knobs, coral buttresses, and pinnacles. The 7th and 96th Infantry
Divisions made little progress in their first attack against the 11,000 troops of the 32nd Army which defended this line.
On 10 June naval gunfire, artillery, and tanks pounded the cliffs, but the 7th Infantry Division advancing upon Hill 95 met accurate machine gun and sniper fire from at least 500 Japanese deployed in depth in that section of the escarpment. Five flame tanks then maneuvered into position and burned off the cover to the approaches to Hill 95. Finally, two skeleton companies of infantry struggled to the top only to be pinned down by enemy fire from farther up the ridge. The commanding officer and two men of Company C, 713th Tank Battalion, scaled the ridge to within thirty yards of the enemy position, pulling a 200-foot fire hose after them. From this point they burned out the defenders who were then slain by the infantry.87
The next day the attack was resumed. At one point a section of tanks and flame tanks attacked a Soo-yard frontage of the escarpment, driving the Japanese from their position and cutting them down with machine gun fire. Several days later two flame tanks came to the front lines and an extension hose was hauled up a sxo-foot high section of the escarpment by means of a rope. The flame, hurled over the far edge, was blown by the wind into the caves on the reverse slope. By moving the hose from one flame tank to another the men destroyed ammunition dumps, fortified positions, and a large number of Japanese troops. That same afternoon five flame throwers lined up at the base of the cliff and fired on caves in its face. The infantry pushed forward slowly and by 1600 hours Hill 95 was wholly in American hands. On the right flank the 17th Infantry advanced with support from flame tanks to capture the village of Azato before nightfall on 13 June 1945 and to tie in with units of the 3zd Infantry. Although some mopping up remained to be done, the battle for Hill 95 was over, a typical struggle on Okinawa in which flame tanks helped turn the tide of battle.88
Hill 89, the last major strongpoint of southern Okinawa and headquarters of the Japanese 32nd Army, was attacked by the 7th Division on 20 June 1945. Here again was almost unbelievably difficult terrain cut up by natural caves, tunnels, and crevices. For weeks the area had been systematically pounded by naval gunfire, long-range artillery, rockets, and aerial bombing. A number of Japanese surrendered but many others fought to the bitter end.
Before the infantry attack the flame thrower and conventional tanks moved out to burn off the remainder of the foliage from Hill 89 and its approaches, which they left an ugly blackened mass of jagged coral. On the second day the flame-throwing tanks advanced in increasing numbers to pour thousands of gallons of burning fuel into caves and crevices. Then the infantry, supported by flame tanks, medium tanks, and artillery, successfully stormed the hill.
Between 22 and 24 June the infantry continued to mop up the top of Hill 89. Flame thrower tanks, used in conjunction with loudspeaker tanks, encouraged soldiers and civilians to surrender. As a team, infantry and flame tanks killed over 100 Japanese and persuaded three times as many to surrender. From the top of the cliff flame tanks repeated the technique of firing toward the sea so that the wind could blow the flaming napalm back into caves too high to be reached from the beach. Concealed ammunition dumps blew up in loud explosions. Load after load of fuel was fired into a huge cavern in which over 1,000 Japanese soldiers were thought to be hiding. Thus the remnants of the Japanese 32nd Army were destroyed, although some mopping up remained in the rest of the island. The battle for Okinawa was over.89
During the Ryukyus Campaign the 713th Tank Battalion saw almost continuous action for seventy days and officially received credit for the death of 4,788 Japanese and the capture of 49. At the same time the battalion lost 7 men killed and 1ro men wounded, injured, or missing. No one was killed inside a main armament flame tank as a result of enemy action, although 41 tanks were knocked out during the 630 flame tank sorties.90
The XXIV Corps reported that the main armament flame tank was one of the most effective weapons used on Okinawa and that the Japanese fear of flame was greatly exploited by the flame tanks. The 7th Infantry Division spoke of the outstanding success of the flame tanks.91 The periscope mounted flame thrower found little use in the fighting on Okinawa as the large-capacity flame tanks were available.
Tank crews developed a fear of the periscope type after the fuel container of one burst and ignited upon being struck by an enemy shell, burning the tank crew to death. This type of auxiliary flame thrower, therefore, was never fully tested in battle.92 Nor was the bow gun type, with which the III Amphibious Corps was equipped, put to much use, for main armament flame tanks of the 713th Tank Battalion were preferred by Marine as well as Army units.93
The 713th Tank Battalion suggested that, instead of having separate armored flame thrower battalions, commanders should replace the light tank company in each standard tank battalion with a flame thrower tank company or else that the latter be added to each battalion. This system, it was thought, would simplify administration and supply, since a flame thrower tank company was normally attached to standard tank battalions for operations. Some general mishandling of the weapon brought forth the recommendation that infantry commanders familiarize themselves with the use and limitations of armored flame throwers and learn how to coordinate these weapons with other supporting weapons and infantry.94
The Tenth Army and the 713th Tank Battalion agreed that the 75-mm. gun should be retained in flame tanks and mounted coaxially with the flame thrower, but that there should be no reduction in fuel capacity and no decrease in the effectiveness of the flame gun.95
Preparations for the Invasion of Japan
During the summer of 1945 United States planners were concerned with the problem of holding down casualties in the invasion of the main Japanese islands. Such battles as Tarawa, Iwo Jima, and Okinawa had illustrated how costly victory in the Pacific could be. The assault on Japan could be expected to exact a heavy toll of American lives unless some means were found for reducing casualties. One of these
was the full utilization of the main armament mechanized flame thrower.
According to U.S. intelligence reports the terrain of the next objective, the island of Kyushu, was well suited to the type of defense which the Japanese had used so advantageously on Iwo Jima and Okinawa—caves, crevices, fortified reverse slopes, pillboxes, and interlocking fields of fire from mutually supporting positions. Information gleaned from enemy prisoners taken on Okinawa revealed that many positions on Kyushu would be difficult or impossible to reduce with conventional weapons. But flame throwers, particularly mechanized flame throwers, could rout the Japanese from these otherwise impregnable fortifications with a minimum of casualties. Experimental use of extension hoses showed that positions hitherto inaccessible to armor could now be attacked. A large number of flame tanks might burn the Japanese from their well-prepared positions and keep the battle from dragging on interminably.
On I July 1945 the Sixth Army, relieved of all combat operations in the Philippines, received directions to prepare four corps, consisting of twelve divisions, and two additional divisions as army troops, for an assault on southern Kyushu on I November 1945.
For the Kyushu operation (OLYMPIC) over twice as many main armament flame tanks per division were to be used as had been employed on Okinawa. A tank battalion was to be attached to each assault division and Sixth Army requested sufficient main armament flame throwers to equip one company in each of these battalions.96 At this time the Sixth Army understood that some 600 main armament flame throwers were being manufactured in the zone of interior, but that only 40 of these would be available in time for OLYMPIC.97 Therefore, the plan called for the distribution of armored flame throwers in equal lots to each tank battalion as soon as the weapons arrived in the theater. Some 56 additional main armament flame throwers of the POA model were to be obtained from Hawaii.98 Thus a total of 96 main armament
flame throwers would be used in the initial assault. Since this number was considered insufficient to meet the expected needs for flame, the Sixth Army also decided to equip each medium tank platoon with two auxiliary flame throwers of the bow gun type. The conclusion of hostilities on 15 August 1945 negated the need for these plans.
In the trial of battle the Army and Marine Corps in the Pacific discovered that the value of the flame thrower lay in its ability to reach the enemy in his prepared positions, which were for the most part impervious to conventional weapons. The excessive infantry casualties ordinarily incurred in rooting the Japanese from their defenses were thereby reduced. It appears certain that had the war not come to a sudden close the mechanized flame thrower would have demonstrated its ability to cut these losses to an even greater degree—not to mention the casualties which would be inflicted upon the enemy.