Crescendo on Okinawa

The men aboard the invasion fleet off the coast of Okinawa on Easter Sunday morning, 1 April 1945, had been told in shipboard briefings as they steamed north and west from Leyte, Guadalcanal, Hawaii, and other Pacific bases, that Okinawa was likely to be a bigger and tougher operation than Iwo Jima. The casualty reports from Iwo made this a gloomy reflection indeed. Aboard the Montauk, the command ship from Hawaii, Easter services were held at 0400, and Colonel Welch, Ordnance officer of Island Command (ISCOM), observed that about 75 percent of those who attended would not have gone to church back home.1

The beaches selected for the L-day landing were known as the Hagushi beaches, after a village at the mouth of the Bishi River about halfway down the five-mile-long landing area. North of the river two Marine divisions, the 1st and 6th under III Amphibious Corps, landing abreast, were ordered to capture Yontan airfield behind their sector and then advance north to the narrow neck of the island and east to the coast. South of the river, the 7th and 96th Infantry Divisions under XXIV Corps were to capture Kadena airfield and advance east and south. In addition to the 77th Division, which was to be available for special operations after Kerama Retto, Tenth Army also had the 2d Marine Division to make a feint against the east coast of Okinawa and the 27th Infantry Division to serve as floating reserve. As area reserve, the 81st Infantry Division, under the control of CINCPOA, was to stand by in New Caledonia. The plans for this big invasion force were based on the belief that the Japanese would put up a hard fight for the two airfields behind the Hagushi beaches.

The Landings on Hagushi

At sunrise, anxious faces at the rails of the transports were straining for the first look at Okinawa through gaps in the smoke of the heaviest concentration of naval gunfire ever to support a landing. It was a beautiful morning, just enough offshore breeze to blow the smoke back. The rising sun, in the words of Samuel Eliot Morison (who stood at the rail of the battleship Tennessee), cast "a peach-like glow" over the calm sea. But the air was not calm, filled as it was with the deep booming of the naval cannon, the rumble of exploding bombs, and the crack of rockets.2


To this familiar overture to an amphibious operation was added one rather novel contribution, the roar of 155-mm. guns based on an offshore island. Using a technique successfully employed at Kwajalein, Tenth Army had emplaced two battalions of Long Tom guns on Keise Shima, a group of coral islets that had been secured by the 77th Division on 31 March following the Kerama Retto operation. From Keise, about eight miles southwest of Hagushi and about eight miles west of the Okinawan coastal town of Naha, the artillerymen had the job of prohibiting enemy reinforcements from moving toward the landing beaches from the south.3

Before the sea, air, and land bombardment had lifted, the LVT's were issuing from the LST's and forming waves that followed the pattern set at Saipan (like Okinawa, a coral reef landing)—first, the LVT(A)'s (the amtanks), next the troop-carrying LVT's, all guided to their assigned beaches by control craft flying colored pennants to match the beach designations: from left to right, green or red for the 6th Marine Division; blue or yellow for the 1st Marine Division; purple or orange for the 7th Infantry Division; and white or brown for the 96th Infantry Division. Lined up behind the first assault waves were the landing craft—LCVP's, LCM's, LSM's, LST's. Of particular interest to the Ordnance Island Command men on the Montauk were the LSM's carrying the "swimming" Sherman tanks.4

The bombardment increased as the first waves approached the shore at 0830, and then suddenly stopped until nothing could be heard except the rumble of shells that were falling inland. In those first nervous waves, the tense men stumbling ashore had "steeled themselves to meet almost anything—except the Sabbath calm which greeted them."5 Crossing the narrow strip of sand and climbing the 10-foot sea wall, the infantrymen and marines to their utter surprise were able to go in standing up6 —no crouching, no foxholes, for there was no enemy resistance beyond some sporadic and ineffective sniper fire. They found themselves on farmlands peaceful and still in the brisk spring sunshine. Maj. Gen. Lemuel C. Shepherd, Jr., of the 1st Marine Division expressed the general feeling when he said with a smile, "There was a lot of glory on Iwo, but I'll take it this way."7

Through gaps in the sea wall blasted by naval gunfire rolled amtanks and DUKW's carrying 4.2-inch mortars. The swimming Sherman tanks did not go in with the leading waves, and when they did take to the water, it was discovered that reef conditions made their flotation equipment worse than useless. Standard Shermans, landed directly from LCM's at the reef edge, got ashore more easily. In any case, the firepower of the swimming tanks (greater than that of the amtanks) was not needed.8

Since there was no opposition, the supply


Photo:  Preparing Long Toms to fire on Okinawa from Keise Shima


DUKW's followed quickly in the wake of the assault troops. On the 96th Division beaches an advance party of officers and supply men of the 796th Ordnance Light Maintenance Company was ashore by 1030 in three DUKW's that it had combat-loaded at Leyte with emergency parts, cleaning and preserving materials, and weapons. The destination of these men was Sunabe, a little coastal town behind White Beach 2, but when they reached it they found that it had been leveled by the air and naval bombardment and that their designated site was a heap of smoldering rubble and debris. That afternoon they started out to find a better place for their supply dump and found it about a mile and a half east, across the road from a small deserted village.9

Rolling inland, past checkerboard fields of barley, sugarcane, sweet potatoes, and cabbages, and houses roofed with red tiles, they could hardly help thinking that Okinawa was a pleasanter place than the Philippines. The most striking feature of the inland scene was the circular coral tomb built into almost every hillside, said by some to represent a turtle's back, because of the low domed roof, and by others to represent a womb. For centuries the Okinawans had thus interred their dead,


first coffining the body in a seated position (some said in a fetus-like position), later removing the bones and placing them on shelves in ceramic pots. The Army had issued strict instructions against entering these tombs, but after sundown on L-day the air became chill and "more than one shivering doughboy," reported the 96th Division historian, "sought sanctuary among the 'ancestors' in their snug homes that first cold night." The top of a tomb was to serve Colonel Daniels, Tenth Army Ordnance officer, as the foundation of his tent when Tenth Army command post was set up next day near Yontan airfield.10

By nightfall of L-day most of the rest of the 796th Ordnance Light Maintenance Company was ashore, bivouacked at Sunabe, and the 632d Ordnance Ammunition Company was ashore also, setting up its first dump about 150 yards inland from the beach with ammunition brought in by DUKW and LVT. But here men of the 632d worked with heavy hearts for the company had suffered sadly in the landings. A lieutenant and eighteen enlisted men of the company had been drowned when their LCVP capsized at the coral reef.11

A few miles up the coast on the 7th Infantry Division's beaches, the main group of the 707th Ordnance Light Maintenance Company, consisting of sixty-seven men and the commanding officer, were ashore on Purple Beach 1 at 1400 on L-day. Moving to a point south of Kadena airfield to set up the bivouac, they were immediately busy with shore party work, returning to the beaches where the 644th Ordnance Ammunition Company was establishing its first ASP, and the army (196th) depot detachment was laying out a supply dump. The 707th had some trouble getting the division Ordnance supplies ashore, but this was less serious than it might have been because the XXIV Corps Ordnance companies, which began landing on L-day, were also located in the Kadena area. The division Ordnance company acknowledged that it received "splendid service throughout the campaign from the corps 183d Ordnance Depot Company.12

The rest of the 707th Ordnance Light Maintenance Company was split up into 11-man detachments, one behind each of 7th Division's regiments. Landing on the day after L-day, the detachments moved immediately down to Koza and set up a third echelon shop in support of the division's rapid advance. In the first few days after the invasion, spectacular gains were made in all sectors. By 4 April, in the north the 6th Marine Division had cut Okinawa in two at the line Nakodamari to Ishikawa (the L plus 15 line) and the 1st Marine Division had reached the eastern shore. In the south the 7th Infantry Division, having also reached the east coast on


the heights commanding Nakagusuku Bay, was ready to drive south to join the 96th Division, which had turned south and arrived at the narrow waist of southern Okinawa. The 96th had now reached the outposts of the Shuri defenses.13

Supporting the Assault on the Shuri Defenses

Following closely behind the combat troops, an advance party of the 796th Ordnance Light Maintenance Company made a long jump forward on 5 April to a point less than a mile west of Futema and about a mile from the 96th Division's front lines. Next day, after the rest of the company had arrived, enemy shellfire began to fall in the company's area. This was but a taste of what was happening in the combat zone. The Japanese were reacting strongly with machine gun, mortar, antitank, and artillery fire of great volume and intensity. At noon on 8 April three 674-pound projectiles—the biggest the men had ever seen—came screaming into the 96th Division lines. They came from a 320-mm. spigot mortar emplaced on one of the ridges. The "boxcars," as the shells were nicknamed, traveled so slowly and were so large that the men swore they could see them coming. Unless a direct hit was scored the boxcars buried themselves in the earth, digging craters as big as swimming pools, and did little damage (beyond that caused by the rocks thrown into the air) except to the nerves of the soldiers already under strain from shelling by the more effective 150-mm. pieces.14

As a result of the intensity of the enemy resistance, which stopped the 96th Division cold at Kakazu Ridge the second week in April and was also being felt strongly on the 7th Division front, the 27th Infantry Division was landed on the southern Hagushi beaches and sent down the west coast. But as serious as the need for more combat troops was the need for more artillery ammunition before the attack could be resumed.15

At the same time the Japanese began to resist on land at the outposts of the Shuri defenses, they began to resist by kamikaze attacks. On 6-7 April they inflicted heavy damage on the shipping off Okinawa. In Kerama roadstead kamikazes crashed into two Victory ships converted to ammunition carriers, the Logan and the Hobbs. Abandoned by their merchant marine crews, the vessels drifted, burning and exploding, for over a day, until the U.S. Navy sank them.16 The loss of these two ships contributed to the ammunition crisis but by no means caused it. The main cause was inability to unload the ships off the XXIV Corps beaches. Getting the ammunition ashore was in the opinion of the ISCOM Ordnance officer the biggest Ordnance problem on the beachhead.17 There were not enough men to do it, nor enough equipment of the right kind. Until the 61st


Ordnance Ammunition Company came in with the 27th Division, the XXIV Corps beachhead area had only two ammunition companies, the 644th with the 7th Division and the 632d with the 96th Division. The 632d arrived crippled from the accident at the reef. The 644th for the first two days was operating most of the time on the beach at less than half its strength. It had been obliged to leave two platoons aboard three LST's to act as ship's platoons, and on L plus 1 had to send one platoon inland to set up an advance dump. At first the 644th had to do all unloading by hand, for it had no cranes. All the ammunition companies keenly felt the need for heavy equipment such as dump trucks and bulldozers, which were lacking either because they were not provided in the tables of equipment or because they had been left behind as a result of limitations on assault shipping.18

Native labor, depended upon for ammunition operations in other campaigns, did not exist on Okinawa. The Okinawans who emerged from the caves and tombs to which they had fled during the preliminary bombardment were mostly children or old people—"under six or over sixty," one officer noted; there were no young men and very few young women. All looked miserable and undernourished. The most wretched were the bearded old men in their black kimonos or filthy blue smocks. One 70-year-old man seemed to a Time correspondent "one of the most pitiable of God's creatures"; while being treated by a medical corpsman for a bayonet wound inflicted by the Japanese before they evacuated his village, "he tried repeatedly to rise from the stretcher, in an unknowing, fearful manner."19

Lacking native labor, the 1st Engineer Special Brigade had used tactical units when it took over shore party operations on 9 April. Three antiaircraft battalions and an armored group furnished more than 700 men for labor in the ammunition dumps for periods from four days to two weeks. A Quartermaster company was also pressed into service. And when three ammunition companies designated to set up army ammunition depots began to arrive on 16 April—the 61st, the 68th, and the 693d (diverted, except for a detachment, from the 77th Division's Ie Shima operation) —they were used as handlers on the beaches for varying periods until 6 May.20

Siege Warfare With a Difference: The Cave Positions

By such efforts, enough artillery ammunition was rushed to the front to enable XXIV Corps to launch a full-scale attack on the first Shuri defenses on 19 April. In planning the attack, Pacific Ocean Areas placed great confidence on massed artillery fire, assisted by naval bombardment and air bombing, to blast the Japanese from their hill positions. These were the tactics— fairly standard for mountain warfare—that had been employed in Italy. It is significant that when they failed on Okinawa,


as fail they did. General Buckner referred to a proposed second landing as "another Anzio."21

The reason for the failure was the enemy's intelligent and effective use of the deep caves that honeycombed the limestone hills of southern Okinawa. This became clear only after the hard fighting and heavy casualties of the assault on the first Shuri defenses, 19-24 April. During the action a newspaper correspondent explored a typical cave position:

In the face of a rocky hill was a narrow horizontal slit five feet long and eighteen inches high, with pine log for header and low growing bushes for camouflage. At a distance of a hundred feet it could not be seen.

We had to wriggle through on our bellies to reach the interior of the cave. There in a room about 15 feet square was a Japanese antitank gun mounted on pneumatic-tired wheels. Piled high around three walls were cases of ammunition, some high explosive, some of armor-piercing variety.

From that room a drift or tunnel ran 125 feet through the hill, opening on another side. At intervals along its length were other rooms in each of which large quantities of ammunition were stored.

From the slit through which we entered the gun crew commanded the field of fire covering the broad, level plain across which American tanks had to advance to reach Kakazu Ridge.

The cave was finally spotted and artillery was trained on it, but "Our gunners might just as well have tried to level the Rock of Gibraltar. Their high trajectory fire from howitzers landed on the round top and steep sides of the hill but accomplished nothing more useful than showering the cave mouth with dust. Aerial strikes were equally ineffective."22

Against positions like these a great deal of the artillery ammunition put ashore with such strenuous efforts was undoubtedly wasted. The Tenth Army Ordnance officer estimated that between 1 April and 2 May it took 1.65 tons of ground ammunition (not counting air and Navy) to kill one Japanese. Direct fire by self-propelled 155-mm. guns or 8-inch howitzers might have been effective if Tenth Army had had any such weapons (the only self-propelled artillery consisted of three battalions of 105-mm. tank destroyers); on the other hand, self-propelled guns were too vulnerable to be brought up very close to their targets.23

Heavy Tank Losses

The tanks were having a hard enough time against the withering Japanese fire, especially fire from the 47-mm. antitank gun, which was accurate and deadly. The gun was small and easily concealed and its high muzzle velocity would send a projectile through any part of a medium tank except the glacis plate. At Kakazu on the morning of 19 April in the 27th Division's sector, Company A of the 193d Tank Battalion lost four tanks from a single 47-mm. piece firing only sixteen shots and later in the day had many more cut down by artillery fire, some of them the scarce and


valued flame throwers. The company returned with only eight out of thirty tanks— the greatest loss suffered by American armor in any single engagement on Okinawa.24

Six of the tanks lost in this action at Kakazu were destroyed by satchel charges placed by Japanese suicide squads that seemed to spring out of the grass beside the tank, sometimes forcing turret lids open and throwing in grenades that killed the tankers. These squads of three to nine men did not cause as many tank casualties as the 47-mm. gun, but they continued to be a constant source of danger. Each man of the squad had his own job: one would blind the tank with smoke grenades, another would force it to button up by hurling fragmentation grenades, another would immobilize it with a mine under the track. The final act was destruction of the tank and crew by a satchel charge.25

Against such tactics, Tenth Army Ordnance in Hawaii had devised the backscratcher but just what it accomplished was hard to determine. Only a few were actually mounted on Army tanks, none on Marine tanks. In one case the device wounded rather than killed the attacker; in another, rain shorted the circuit so that the mine did not detonate. The few times the backscratcher appeared to be effective, friendly fire took equal credit for destroying the enemy.26

The test of the sanded paint applied to the tanks of the 713th Armored Flame Thrower Battalion for protection against magnetic mines was also inconclusive, for the battalion encountered no mines of this type.27 Antitank mines did, however, account for about 31 percent of all tank losses on Okinawa. Luckily there were few antipersonnel mines among them to interfere with tank recovery. The problem in bringing back the tanks was the lack of enough tank recovery vehicles to do an adequate job. Often after a tank was immobilized by a mine, satchel charge, or antitank projectile, or because it had simply thrown a track or bogged down in bad terrain, it was abandoned by its crew, and if not recovered by nightfall it would be demolished at leisure by the returning Japanese.28

By the end of May, the four Army tank battalions and the one armored flame thrower battalion had suffered 221 tank casualties, not counting Marine losses. Among the 221 tanks put out of action, 94 (including at least 12 of the irreplaceable flame throwers), or 43 percent, had been completely destroyed. The effort by Tenth Army Ordnance to make up these losses was painfully unproductive. To begin with, losses on such a scale had not been anticipated. In the planning, a small reserve stock of 13 medium tanks was to be placed on Saipan for fast emergency shipment to Okinawa. These were sent for on 25 April. But even this small reserve had not been established on Saipan. The tanks had to be ordered from Oahu and did not arrive at Okinawa until 10 June—


some ten days before the campaign was virtually over. An additional shipment of 65 Shermans and 25 tank recovery vehicles was requested from Oahu on 28 April. They were shipped on 20 May and did not arrive until 15 July—almost two weeks after the campaign had been declared officially at an end.29

Because of the delay, all serviceable medium tanks were stripped from the 193d Tank Battalion (the unit crippled at Kakazu) and distributed to other tank battalions. By early May the tankers were asking for a tank heavier than the Sherman, with a weapon of higher muzzle velocity and thicker armor plate. The M26 Pershings seemed to be the answer. In mid-May (after V-E Day) twelve were requested from the United States, but they were not received until August. In the meantime, to provide better protection against the 47-mm. gun, the Ordnance tank maintenance companies welded steel track sections to the side sponsors, turret sides, and glacis plate of the Shermans, and also used armor plate from wrecked tanks to reinforce the sponson and shield the suspension system.30

In the end, however, the best defense for the Sherman turned out to be the infantrymen who accompanied it as part of the tank-infantry team, a truth forcibly brought home to commanders in the catastrophe at Kakazu where the tanks had been operating alone. In the early assault on the Shuri defenses, the Japanese, who knew the value of the infantryman, attempted to pin down the infantry with artillery and mortar fire, often successfully. But as the long bloody battle for Shuri dragged on through April and May, the Tenth Army commanders learned about cave warfare and how to win it. They directed artillery fire on cave entrances, forcing the Japanese gunners back into their tunnels; then infantry and tanks closed in, the infantry protecting the tanks from the suicide squads. Having gained a foothold, the troops could move down on cave openings from above, sealing them with flame or explosives, a method that General Buckner called "blowtorch and corkscrew."31

Shortages in Mortars and Mortar Ammunition

This type of warfare demonstrated again the usefulness of mortars, which could drop projectiles into Japanese positions on top of the hills and on the reverse slopes, and, moreover, could be hand-carried. The Chemical Warfare Service's 4.2-inch rifled mortar was extensively used and was a great favorite with the marines. Observers believed that the 155-mm. mortar developed by Ordnance in its jungle warfare program would have been most effective; however, this remained speculation, for none was available. Since the 155 was lacking, the cry arose for more 81-mm. and 60-mm. mortars. This demand was more easily met by Tenth Army Ordnance than that for medium tanks because these light


weapons could be supplied by air. On 9 May a hundred 81-mm. and thirty 60-mm. mortars were requested from Oahu and the entire reserve stock of twenty-seven 81-mm. mortars on Saipan was called forward, all shipments to be made by air. All of the Saipan mortars were on Okinawa four days later, those from Oahu in about ten days. For this quick reaction two infantry commanders expressed "great satisfaction and thanks."32

Tremendous quantities of mortar as well as artillery ammunition were expended in the assault on Shuri. By mid-May, XXIV Corps had fired 268,904 rounds of 81-mm. mortar ammunition—six times as much as it had expended during the first six weeks of the Leyte Campaign—and stocks were running alarmingly low. On 21 May Tenth Army requested Oahu to ship 100,000 rounds by air. Nothing like that amount was available in Hawaii, but the Central Pacific Base Command responded so promptly with what it had, aided by the 962d Ordnance Ammunition Renovation Company, that it received a radio of appreciation from General Buckner. A total of 36,000 rounds came in by Air Transport Command and Naval Air Transport Service planes in two shipments, the first arriving 28 May, the second on 9 June. It was reported that combat troops cheered loudly when told of their arrival.33

Certain types of mortar shells continued to be the most critically short ammunition items of the campaign, not only the 81-mm. light round with the superquick M52 fuze, but the 60-mm. mortar illuminating shell. Illuminants of all kinds, including trip flares, became crucially important because the Japanese on Okinawa made skillful use of their customary tactics of night counterattacks and infiltration. These tactics had been stressed in the Roberts report of July 1944. But the hope that the 60-mm. mortar illuminating shell would be available in sufficient quantity was not fulfilled. The first large shipment of 20,480 rounds did not arrive at Okinawa until 24 June. In this dilemma the Navy came to the rescue, providing XXIV Corps with about 75 percent of its illumination. Most effective was the Navy's 5-inch star shell, usually fired by destroyers.34

The Ordnance Build-up in the Mud

By the end of the third week in May, Tenth Army had reached the inner ring of the Shuri defenses at the narrow waist of the island, the line known as Naha-Shuri-Yonabaru. This had been accomplished by bringing the entire Marine III Amphibious Corps (the 1st and 6th Marine Divisions) down from the north early in May (sending the 27th Division up to take over the marines' garrison duties in the north-


Photo:  Signal Corps men in a Weasel, Okinawa


ern area); by committing the 77th Infantry Division after its capture of Ie Shima at the end of April; and by the hard, costly tactics of digging the Japanese out of their caves. To support the two Marine and three Army divisions now in the line, Ordnance companies had been landing almost every day. In addition to the 209th Ordnance Battalion, which now controlled three Marine ammunition companies as well as six Army, the 1st Engineer Special Brigade had a second battalion headquarters, the 192d, to control depot operations, and had established a second Army depot on the east coast. Island Command, which was to release the 1st Engineer Special Brigade on 31 May, had received an Ordnance battalion headquarters, the 214th, and its first Ordnance group headquarters, the 61st, and was in the process of establishing two Ordnance service centers, the northern and the southern, under the group.35

Then the rains came. Until 21 May the weather had been good, but that day


strong gusts of wind and an overcast sky heralded the beginning of the season that the Okinawans call the "plum rains." The downpour began during the night and continued until 5 June, bringing a total of twelve inches of rainfall. Rivers flooded, fields became lakes, and roads were washed out. At one ammunition supply point ammunition stacks located near a small stream were caught in swift-flowing torrents and carried several yards downstream; when the flood subsided, some of the crates were found in the branches of trees five feet from the ground. Tanks were mired, and trucks could not get through to bring up ammunition and food. The little tracked Weasel (cargo carrier M29C), which up to that time had been of little use, assumed new importance, though there were places where the mud was so heavy that neither the Weasel nor the LVT could get through. The M5 tractor performed better, and was used by the contact parties of the divisional light maintenance companies and also to bring up ammunition. But at many front positions supplies had to be carried in, and casualties carried out, by men up to their knees in water or mud. During the last week in May, Tenth Army was virtually bogged down.36

This seemed "awfully tough luck" to General Buckner,37 because just at this time intelligence reports indicated that the Japanese, severely crippled by the American attacks during May, were withdrawing south from Shuri. Despite the mud, which was still so deep that supply trucks had to be dragged through quagmires by winches or bulldozers, Buckner ordered his two corps to drive rapidly forward. On the morning of 29 May, a Marine company had already penetrated to Shuri Castle. Two days later the III Amphibious Corps and the XXIV Corps joined lines south of Shuri and at dawn on 1 June began the pursuit.

Three weeks of hard fighting were ahead. On the left of the advance, the 7th Division, having captured the little port of Yonabaru on the east coast on 22 May, cleared Chinen Peninsula. Wading over green, rain-soaked hills and meeting slight resistance, by 4 June the division had reached the soggy banks of the Minatoga River at the southern base of the peninsula. On the right of the advance, III Amphibious Corps attacked the Oroku Peninsula, which was across the bay from Naha (captured 27 May) and contained Naha airfield. Here it had much harder going than the left flank, for the Okinawa Base Force composed of naval units and home guards put up a stubborn fight with mines, machine guns, and 40-mm. antiaircraft fire, aided by the mud that prevented the marines from using their tanks. But clearing weather and an amphibious landing enabled the marines to clear the peninsula by 12 June. In the center of the general advance that began 1 June, XXIV Corps by 5 June had come up against the last Japanese defense line, dug into the deep caves of the formidable Yaeju-Dake—Yuza-Dake escarpment. There XXIV Corps faced an enemy that was determined to fight to the last man.

Supply by Water

When the roads were washed out in late May and early June, Tenth Army got its supplies to the front by the same means


employed by Eighth Army during the rainy season on Mindanao—air and water. Most of the airdrops were to the 1st Marine Division (561 out of a total of 830 plane loads), which between 30 May and 9 June was supplied almost entirely by air. On 6 June, when XXIV Corps was getting ready to jump off in its large-scale attack on the southern escarpment, 76 plane loads of supplies were dropped to XXIV Corps and III Amphibious Corps, chiefly water, rations, small arms and mortar ammunition, and grenades. But the bulk of the ammunition for both corps came by water, along the east coast for XXIV Corps and the west coast for III Amphibious Corps.38

The west coast had the advantage of the Hagushi beaches. The III Amphibious Corps had early established a supply route using DUKW's from Hagushi to the mouth of the Asa Kawa River, where the main advance dump was located. On 1 June after the roads went out, efforts were made to survey and buoy channels at that point so that LCT's and LSM's could be used and also to clear away the hulks in the harbor of Naha to permit LCM landings. These were not entirely successful. The main dependence at Naha continued to be an offshore LST and an AK (the Wesleyan Victory) loaded with ammunition, plus a fleet of about seventy DUKW's operating out of Hagushi. South of Naha the troops were supplied by thirty-four LVT's making a daily run from Hagushi.39

On the east coast, by the end of May two discharge beaches were in operation on Nakagusuku Bay—Awashi and Kuba. From these beaches water, fuel, and ammunition to support XXIV Corps were carried by LCM's and LCT's to Yonabaru, where two platoons of the capable and experienced 644th Ordnance Ammunition Company were on hand to set up an advance dump. When the roads from the rear became impassable, Island Command brought up a Victory ship, the Berea Victory, loaded with 7,200 tons of ammunition, and anchored it off Yonabaru to serve as a floating ASP. Thereupon, the XXIV Corps Ordnance officer then instituted an effective "amphibian" adaptation of the normal ammunition control system. A crew from the 644th was sent aboard the Berea to select and place aboard the lighters the required types of ammunition, on shore-to-ship radio orders from the XXIV Corps ammunition officer. When the ammunition arrived on the beach, troops were waiting there to receive their loads directly from the lighters, thus cutting down the handling to such an extent that in several instances the ammunition reached the gun positions within six hours after the request had been made. Later an LST was also brought to Yonabaru as a floating ASP. After the Chinen Peninsula was cleared, two LST's served the same purpose at the little port of Minatoga, which was close behind the XXIV Corps front. By these means, XXIV Corps was assured enough ammunition to mount its final attack.40


Photo:  Entrance to a cave held by the Japanese

THE PROBLEM AND THE SOLUTION. Above, entrance to a cave held by the
Japanese; below, a flame-throwing tank fires into an enemy-held cave.

Photo:  A flame-throwing tank fires into an enemy-held cave.


Ordnance Class II supplies were also sent on the east coast water route from Kuba to Yonabaru, mainly in LCT's. The first attempt to use DUKW's very nearly ended in disaster, for it was made on the morning of 4 June when a typhoon east of Okinawa kicked up heavy seas. A convoy of six DUKW's loaded with Ordnance, Medical, Signal, and Quartermaster supplies and commanded by Capt. W. A. Gore, the assistant Ordnance officer of 7th Infantry Division, ran into rough water and after several hours of fighting wind and waves lost one of the DUKW's and had to return to shore. Luckily no men were lost.41

When the last major overland supply route for XXIV Corps disintegrated in the mud, the Engineers concentrated on keeping the roads running south from Yonabaru to Minatoga in operation. From dumps at these two east coast ports the corps was mainly supplied for the rest of the campaign. The 7th Infantry Division's 707th Ordnance Light Maintenance Company was moved to Yonabaru the second week in June. The 96th Division's Ordnance company, the 796th, was not sent there because of a lack of shipping space to make the voyage around from the west coast. Later, clearing weather enabled the company to move overland to Kamizato, south of Yonabaru, where it was close behind the XXIV Corps lines.42

Bloody Finale

From the last great escarpment to the cliffs that mark the southern shore of Okinawa, the Japanese disputed every rod of territory with machine gun fire, extensive mine fields, and effective use of their dwindling supply of 47-mm. and other artillery pieces. They no longer had any hope of winning the battle, only of prolonging it, and prolong it they did until 22 June, killing more than 1,500 Tenth Army men and wounding more than 6,600. Among those killed was the Tenth Army commander, General Buckner—the only army commander to die in action in World War II.

It was the terrain that enabled the enemy to hold out as long as he did—the deep caves and rocky cliffs that had to be burnt out and stormed—but he could not indefinitely withstand the full weight of armament that Tenth Army brought to bear in this small area. After the rains stopped on 5 June the southern tip of the island lay open under the spring sunshine to the bombers, to the Navy guns offshore, and to the Long Toms emplaced on the heights of Shuri, so that targets could be hit at will. By 10 June the roads had dried from mud to dust, therefore the tanks and direct fire artillery could be brought up.43

Relatively free from the antitank artillery fire that had done so much damage before Shuri, the American tanks could operate more freely and aggressively. The "blowtorch and corkscrew" method of attacking cave openings was materially improved when the men of the 713th Armored Flame Thrower Battalion attached to the flame thrower a 200-foot hose to spray


napalm on places inaccessible to the tank. Also (in a reversal of their former roles) flame throwers and other tanks were used to protect infantrymen from the bands of machine gun fire that were causing heavy casualties. The tanks carried combat reinforcements to the front lines in their hulls, as well as supplies of blood plasma, water, and ammunition, and brought back the wounded.44

The Japanese resisted to the water's edge. Many chose to commit suicide rather than surrender. The last act of the bloody drama took place on the seaward side of a rocky cliff on the southern shore. There before dawn on the morning of 22 June the commander of the Japanese forces on Okinawa and his chief of staff emerged from the entrance of their last cave headquarters and committed hara kiri.45 This marked the end of organized resistance on Okinawa. The mop-up phase took ten days longer. Reaching the southern end of the island, XXIV Corps and III Amphibious Corps turned north to the Naha-Yonabaru valley, cleaning up pockets of resistance and destroying enemy soldiers who might escape to engage in guerrilla warfare in the north, as the Japanese had been ordered to do by their general in his last written order, dated 18 June. In the final days of the campaign and in the mop-up phase, infiltration by Japanese—many disguised in Okinawan kimonos—who were being driven north by the combat troops was a real danger to men in rear installations, some of whom, reported the 96th Infantry Division historian, "got a chance to put notches in their seldom-used firearms." A few Japanese were killed in the sector of the division's light maintenance company, but by setting up outposts on surrounding hills the Ordnance men kept the infiltrators out of their company area.46

During the last two weeks in June, 80,000 Okinawan civilians crawled out of the caves on the southern tip of the island. A third or more were wounded. There were few able-bodied men among them; they were mostly women, old men, and children. As they plodded north in long columns they saw that the Okinawa they had known was gone. Villages had been leveled, the city of Naha laid waste, the ancient capital of Shuri reduced to coral-stone and red-tile rubble, a "crater-of-the-moon landscape"47 of utter desolation. Around the ruins of Shuri Castle remained only remnants of its massive ramparts and only blackened skeletons of its great trees. Nothing was left of the great gates, the temples, the gardens.48

In the "typhoon of steel" (as the Okinawans called it)49 that had swept over the island, about 110,000 Japanese had lost


their lives and 7,400 had been taken prisoner. The American casualties—Army, Marine, and Navy—were 11,689 killed and missing and 34,559 wounded. The ratio of American to enemy losses was not so high as it had been at Tarawa or Saipan or Iwo Jima, but those were smaller, shorter campaigns. On Okinawa more American lives had been lost than in any other campaign against the Japanese.50

Preparing for Japan

Even before Okinawa was officially declared secure on 22 June, the Ordnance men were preparing to support the invasion of the Japanese homeland. The first task was to re-equip the infantry divisions, giving priority to the 77th, which was already loading out for Luzon to go to Japan with Sixth Army. Colonel Welch of Island Command had set up at the Ordnance service centers a policy of straight exchange —a serviceable item in exchange for an unserviceable one, whether watch or gun. Most of the vehicles turned in had been damaged by the deep mud, which was particularly hard on brake linings. Otherwise—in striking contrast to almost every other campaign of the war—the mortality on trucks and other vehicles at Okinawa had not been great. There had not been long distances to travel in this siege warfare operation; maintenance discipline had been effectively enforced in Tenth Army by spot inspections; and, for once, spare parts had been relatively plentiful. On weapons, the only serious cause for concern was a shortage of 105-mm. howitzer tubes. These had been requisitioned by Tenth Army, but were short because shipment had been slow.51

While the work of re-equipping the divisions was going on, Ordnance base troops had been landing to do their part in the development of Okinawa as the last great base of World War II. A visitor from the United States, landing on the hot and humid afternoon of 19 July, found the island "a beehive of activity—air and ground," and was "immediately impressed with the extent of roads well paved and the mass of busy traffic," noting in his diary that "tented groups are all over the place and the usual air of American business and ingenuity. What people we are to come out here and be so well established so quickly."52

This observer was the ubiquitous Brig. Gen. William A. Borden, director of the New Developments Division, War Department General Staff, making his fifth visit to overseas theaters—his first was to Tunisia in the spring of 1943. His mission on Okinawa was to let the commanders know


what was being done in the United States to provide them with new equipment and techniques for the assault on Japan. With regard to the Ordnance Department, he could tell the commanders that a cave warfare program had been started on 25 May, sparked by a memorandum from General Stilwell, then commanding general of Army Ground Forces, to General Marshall, stating that the terrain in Japan lent itself to dug-in positions in depth.53

In initiating the program, Maj. Gen. Gladeon M. Barnes, chief of the Ordnance Research and Development Division, made the point that Ordnance already had powerful weapons applicable to cave warfare. Foremost among them were self-propelled artillery—the 90-mm. gun on the M36 carriage and the 155-mm. on the M40 carriage—and the M26 Pershing tank. To supplement these weapons for particular situations were the 57-mm. and 75-mm. recoilless rifles, which could be manhandled into position, and a series of mortars for indirect fire. The catch was that except for twelve Pershings that were on their way to the Pacific, most of these weapons were still in Europe or the United States. The first step in the cave warfare program was to get them out to the Pacific; and Barnes warned the Ordnance division chiefs that "we do not have a couple of years in which to do it." The second step was to get out an ASF booklet on cave warfare. In its final form, dated 24 June 1945, to which not only Ordnance but Engineers and Chemical Warfare Service contributed, other Ordnance items listed were the M24 light tank (highly maneuverable in rough terrain) and several experimental weapons such as the self-propelled 8-inch howitzer, the 4.5-inch multiple rocket launcher, and a ground-launched, rocket-propelled, 250-pound bomb.54

A booklet on how to fight cave warfare could only seem ironical to veterans of the Okinawa campaign; however, the commanders were generally receptive to the Ordnance recommendations on desirable weapons. At Tenth Army headquarters, General Borden encountered General Stilwell himself, who had arrived on 23 June to replace General Buckner. Stilwell, wearing his old campaign hat, said that he was in favor of heavy self-propelled artillery and the Pershing tanks. General Hodge, the XXIV Corps commander, whom Borden interviewed at his headquarters near the ruins of Shuri Castle, had more specific ideas, based on his experience throughout the campaign. He wanted self-propelled artillery for precision fire against point targets because he had discovered that mass artillery fire against cave positions only drove the Japanese into their holes and wasted ammunition. He wanted the Pershing tank, the main armament flame thrower tank, and the recoilless rifle. He also stressed the importance of flares and star shells, the latter in calibers up to 155-mm., for night operations. Hodge's recommendations on weapons were given point by a tour Borden afterward made of the island—from the north, where combat teams of the 27th Division were still killing Japanese, to the south, where the 7th Divi-


sion was also still flushing them out of caves. The tour convinced the visitor that the terrain of Okinawa was "far more rugged than anyone who had not seen it can imagine."55

On 24 July General Borden departed for Manila and spent the following three weeks in the Philippines, visiting Leyte, the southern islands, and the mountains of northern Luzon, where fighting was still going on, and interviewing commanders of Sixth and Eighth Armies on the weapons they wanted for the assault on Japan. On 15 August as he was leaving his hotel for the airport to return to the United States, he was told that a radio had been received from President Truman stating that the Japanese had accepted the U.S. surrender terms. The war was over.56







Previous Chapter        Next Chapter

Return to the Table of Contents

Search CMH Online
Last updated 11 January 2007