OCT 3 n 1941 THE ADMIRALTIES Operations of the 1st Cavalry Division (29 February — 18 May 1944 ) American Forces in Action Series Historical Division WAR DEPARTMENT Foreword In a nation at war, teamworfi by the whole people is necessary for victory. But the issue is decided on the battle- field, toward which all national effort leads. The country’s fate lies in the hands of its soldier citizens', in the clash of battle is found the fnal test of plans, training, equipment, and — above all — the fighting spirit of units and individuals. AMERICAN FORCES IN ACTION SERIES presents detailed accounts of particular combat operations of United States forces. To the American public, this record of high achievement by men who served their nation well is presented as a preface to the full military history of World War II. To the soldiers who too\ part in the operations concerned, these narratives will give the opportunity to see more clearly the results of orders which they obeyed and of sacrifices which they and their comrades made, in performance of missions that find their meaning in the outcome of a larger plan of battle. DWIGHT D. EISENHOWER, Chief of Staff. WAR DEPARTMENT Historical Division Washington 25, D. C. 1 December 1945 The Admiralties: Operations of the 1st Cavalry Division, 29 February — 18 May 1944, is the eighth of a series called AMERICAN FORCES IN ACTION. This study was prepared by the Historical Division from the best military records available. Although in its published form the book contains no documentation, the original manuscript, fully documented, is on file in the War Department. Photographs are from the following sources: U.S. Army Signal Corps (pp. 22, 24, 25, 27, 32, 35, 42, 49, 70, 72, 73, 83, 94, 110, 136, 139); U. S. Army Air Forces (pp. 16, 72, 86, 106); 8th Engineer Squadron (pp. 68, 88, 91, 98, 101, 130, 138, 147); U.S. Air Transport Command, Overseas Technical Unit (pp. 5, 58, 119, 145); Acme Newspictures, Inc. (pp. 26, 28, 52, 112, 121); U.S. Navy Department (p. 135). The photographs from Acme Newspictures, Inc., may not be reproduced without the approval of the War Department. Readers are urged to send directly to the Historical Division, War Department, Washington 25, D. C. all comments, criticism, and additional data which may help in the preparation of a complete and definitive history of the Admiralties operation. Distribution: AC of S(2); WDSS Div(2); T of OpnsfHist Sec) (20), (I & E Off) (1); Def Comd(Hist Sec) (20); FC(I & E Off) (1); Class III Instls(I & E Off) (1); GH, ea 25 beds(l); CH, ea 25 beds(l); RH, ea 25 beds(l); SH, ea 25 beds(l); Sp Sv Sch 3-6, 9-11, 17(10), 2, 7, (20); USMA(80); A(I & E Off) (1); CHQ (I & E Off)(l); D(I & E Off)(l); R(I & E Off)(l); SBn (I & E Off)(l). T/O & E: 8-550, GH ea 25 beds (1); 8-550T, GH ea 25 beds(l); 8-560, SH ea 25 beds(l); 8-591, CH ea 25 beds(l); Special distribution. For explanation of distribution formula, see FM 21—6. AFA-8 IV Contents Page INVASION PLANNING I The Objective 4 The Attack^ Date Is Advanced 10 Planning the Reconnaissance in Force 12 THE BATTLE FOR LOS NEGROS BEACHHEAD . . . .' 21 The Landing 23 “Hold What You Have Ta\en” 29 Enlarging the Perimeter 38 The Big Night Attach ( 3/4 March ) 43 Reinforcements 50 Enemy Side 53 GAINING CONTROL OF SEEADLER HARBOR 58 Advance North ( 5—7 March ) 60 New Beachheads: Papitalai and Lombrum 64 Supply Problems 68 Landing of the 2d Brigade 71 ATTACK ON MANUS ISLAND 76 Preparations 79 Landing at Lugos Mission (15 March ) . 82 8th Cavalry Moves East (15—16 March) 85 Chec\ at the Airdrome ( 16 March ) 90 To the Lorengau River (17 March ) 93 THROUGH LORENGAU TO ROSSUM 97 Lorengau: An Easy Conquest 97 Locating the Enemy Strength 103 Heavy Going to Rossum 106 The 8th Cavalry Breaths the Enemy Hold 112 v Page ROUND-UP ON LOS NEGROS 117 Attac\ on Hill 260 118 Patrolling by the 12th Cavalry 123 Attack^ in the Papitalai Hills 126 LAST STAGES 135 Capture of Outlying Islands 135 Final Mopping Up on Manus 140 Construction Activities 144 CONCLUSION 148 ANNEX NO. 1: ABBREVIATIONS 151 Illustrations Page Eastern Los Negros 5 Attack Area 16 Men of the Reconnaissance in Force 22 First Wave 24 Second Wave 25 Troops of the Second Wave 26 Signalmen Operated Their Radio 27 Thick Undergrowth 28 Cavalrymen Used Picks 32 Seriously Wounded Soldiers 35 Sixth Army Commander 42 A Multiple .50-Caliber Antiaircraft Gun 49 Supplies Were Dropped 52 Mokerang Peninsula 59 Overgrown Coconut Plantation 68 Salami Plantation 70 Water-Absorbent Coral Sand 72 Unloading at Salami 72 Supplies are Efficiently Handled 73 Engineers Operated the LCM’s 83 Lorengau Air Strip 86 D— 7 Bulldozer 88 Blasted Trees 91 Troop G, 8th Cavalry 94 The Sand Bar 98 vi Page Floating Bridge 101 Number Two Road 107 In Well-Camouflaged Positions 110 105-mm Howitzers 113 The Hills West of Lemondrol Creek 119 L— 3’s Were Used 121 Improvised Ferry 130 Pityilu Island 135 Japanese Naval Gun 136 Cavalrymen Carrying Full Battle Equipment 138 PT (Motor Torpedo) Boats 139 Completed Airdrome on Mokerang 145 Shell Holes 147 Insignia 150 Maps No. Page 1 Control in the Pacific, February 1944 viii 2 Allied Advance Toward the Bismarck Sea, as of 29 February 2 3 Admiralty Islands 8 4 Plans for Initial Landing, 29 February 14 5 The First Day’s Beachhead, 29 February 30 6 Defense of the Airdrome, 2—4 March 40 7 Expanding the Beachhead: Objectives 60 8 Expanding the Beachhead: Operations 62 9 Attack on Manus 77 10 Landing at Lugos Mission, 15 March 78 11 Check at the Airdrome, 16 March 90 12 Advance Through Lorengau, 17—18 March 95 13 Fight for the Rossum Road, 19—25 March 102 14 Hill 260, 14-17 March 118 15 12th Cavalry Probes Inland 124 16 The Last Fight on Los Negros, 20—25 March 128 vii MAP NO. 1 Invasion Planning GENERAL DOUGLAS MACARTHUR, Commander-in-Chief of the Southwest Pacific Area, referred to the conquest of the Admiralty Islands as “putting the cork in the bottle.” Inside the bottle was a major segment of the outer defenses for Japan's Pacific Empire, a seg- ment which barred Allied advance from the Southwest Pacific toward the Philippines and which protected the flank of Japanese strong- holds on the north coast of New Guinea (Map No. 1, opposite). The enemy bases on the islands of the Bismarck and Solomon Seas formed an integrated defensive system. Sea and air power linked the scattered garrisons and furnished the striking forces to meet Allied invasion; many harbors were available for shipping, and a network of airfields permitted quick transfer of planes for support of any threatened point. The capture of the Admiralties was one of the last steps in the campaign which sealed off the Bismarck-Solo- mons area from supply and reinforcement, denied its use to the enemy for effective air and naval operations, and left garrisons totaling over 100,000 troops in isolated impotence. The offensive that produced these results had begun in the summer of 1943, from bases in New Guinea and Guadalcanal which represented the Allies’ first, hard-won conquests (Map No. 2, page 2). The attack was delivered along two widely separated axes of advance. Under direct command of General MacArthur, Austra- lian and United States troops struck northwest along the coast of New Guinea. From Guadalcanal, Admiral William F. Halsey’s forces of the South Pacific Command took a parallel route through islands of the Solomons chain, with objectives and a time schedule set 1 using bayonets attached to 5-foot poles. The garrison’s ammunition supply was low, and all troops had been directed to use it sparingly. It was also discovered after the big night battle that the majority of the dead Japanese had bandages tied around their arms at pressure points, presumably to enable them to continue fighting after the lower portion of their arms had been struck by bullets or severed. The lessened intensity of the night attack^ cind the mounting toll of Japanese dead after 3 March indicated that the beachhead on Los Negros need not fear a coordi- nated Japanese offensive. But the Japanese determination to resist to the last man, and the large number of the enemy thought to be left on the island, promised a continuation of hard fighting for the caval- ry me n. 681583 °— 46— 5 57 Gaining Control of Seeadler Harbor BY 5 MARCH THE BREWER TASK FORCE had accomplished its primary objective of securing a beachhead in the Admiralties. In addition to capturing Momote air strip, the task force had inflicted such serious casualties on the Japanese defenders that the three squad- rons on the Los Negros perimeter, reinforced by the 12th Cavalry Regiment on 6 March, would be comparatively safe from an organ- ized counterblow and could start offensive operations. Therefore, the next phase of the operation was directed west toward the other main objectives, Seeadler Harbor and Lorengau airdrome. 58 MOKERANG PENINSULA forms the northern tip of Los Negros and the eastern limit of Seeadler Harbor. The north shore of Manus (left background ) is paralleled by the chain of small islands that guard the harbor. Plantation growth was removed when our forces built this airdrome. The first step toward these objectives was to gain control of the eastern side of Seeadler Harbor, formed by the Mokerang Peninsula as it curves west to make an acute angle with the north coast of Los Negros (Map No. 7, page 60). Holding this end of the harbor, the Brewer Task Force would have a base for easy shore-to-shore invasion against Lorengau. The most urgent task was to seize the Salami Plan- tation area, with its excellent beach on the west coast of the peninsula, already designated as Red Beach, landing spot for the 2d Brigade Combat Team scheduled to arrive the morning of 9 March. Next, the eastern harbor must be made safe, both for landings and for offensive movement, by neutralizing the areas where enemy garrisons and coastal guns could harass operations in the harbor. 59 MAP NO. 7 Advance North (5~7 March) The units of the 1st Brigade already at Hyane Harbor, as well as those arriving on 6 March, would be used to clear the way for the 2d Brigade's landing at Red Beach. General Swift arrived to take com- mand of the advance forces of the 1st Cavalry Division at 1100, 5 March, and at once ordered the 2d Squadron, 7th Cavalry, to move across the skid way to a point about 500 yards north (Map No. 8, page 62). The attack would be supported by the regimental weap- ons company of the 5th Cavalry, as well as by artillery. The men of the 2d Squadron were in high spirits because of their successful de- fense of the perimeter when first under fire the previous night. 60 The advance did not move out smoothly. The Japanese began harassing attacks at 1120 and followed these up an hour later with a strong attack from both Porlaka and the native skidway, just as the 7th Cavalry was being relieved from their defensive positions by the 2d Squadron, 5th Cavalry. The few Japanese who penetrated the position were killed, and the attack was broken up by mortar and artillery fire. Twenty-five dead Japanese were counted and twelve cavalrymen were wounded. The 2d Squadron was reorganized im- mediately and the advance moved out at 1630. Its progress was slow as the area was heavily mined. The mines, probably placed prior to the enemy attack, were so numerous that his advance and retirement through them must have been difficult. Although they caused some casualties to our troops initially, they were removed without further cost. Japanese dead from the furious counterattacks of 3 March still littered the road and the entire skidway area. At dark the 2d Squadron dug in on the skidway after an advance of 500 yards. During the night there was little enemy activity, although trigger- happy soldiers along the skidway did a large amount of firing. At- tempts at wire-tapping were made by the enemy. Once a voice over the wire pleaded, to the accompaniment of groaning and sighing, “For God’s sake, lift that mortar fire.” No action was taken since the speaker was unable to identify himself. Two Americans were killed during the night; the Japanese casualties were impossible to estimate because the bodies of their dead from previous action covered the area. In the morning bulldozers and burial parties buried the Japanese dead, some of the bulldozer operators wearing gas masks. At 0820, 6 March, the 12th Cavalry with additional artillery, signal, and engineer troops as well as medical detachments began to disembark from four LST’s at Hyane Harbor, bringing with them three light tanks and five amphibian tanks (known as buffaloes or LVT’s, Landing Vehicles, Tracked). The 12th Cavalry, commanded by Col. John H. Stadler, Jr., was to join the attack along the Salami road around noon, following, the 2d Squadron, 7th Cavalry, now attached to the 12th Cavalry, which would lead out first, with Troop E as advance guard. The 12th Cavalry was led to an assembly area in preparation for the attack, and the 271st Field Artillery went into position at the southeast end of the air strip with guns practically hub to hub. At 1030 the 2d Squadron, 7th Cavalry, started north along the 61 MAP NO. 8 62 road, which was hardly more than a trail and ankle-deep with mud. The route soon became congested with troops and mired vehicles. The Japanese had felled trees across the road to impede the advance, and these had to be removed. The engineer platoons, which accom- panied the advance of the 12th Cavalry at noon, had a heavy task clearing the way to Salami. One angledozer built a road around a crater from a 500-pound bomb which it had been necessary to explode. Another dozer had gone forward to aid the advance column. Its crew disposed of crudely constructed booby traps, which the advance guard had side-stepped, and then went to work neutralizing a vehicle trap. This was a ditch about 4 feet deep camouflaged by canvas (stretched flush with the road) on a framework of poles and covered with coral sand. When this ditch was filled, both platoons made their way to where the front elements of the train were mired down. Here a nar- row 150-yard fill had been breached at numerous points by bomb bursts. The engineers worked on this road for the next several days. At noon Troop E was on the Salami road 2/ z miles northwest of the airdrome; it received orders to cover the northwest flank while the 12th Cavalry bypassed to the west, continuing the advance on Salami Plantation to the beaches of Seeadler Harbor. Occasional pill- boxes occupied by one or two Japanese were encountered, but the enemy withdrew on contact. As the 12th Cavalry approached the beach, a few Japanese concealed in buildings and bunkers put up a fight that lasted an hour. Our tanks fired canister into the buildings at ranges up to 50 yards, and 75-mm high explosives into the firing slits of the bunkers, which were so well camouflaged that it was often necessary to get within 30 yards. If the slit was not visible to the tank gunner but was to the troops near the tank, the cavalrymen would indicate its location by firing tracer bullets. At 1630 the three squadrons closed into bivouac at Salami, minus all of their vehicles except three tanks; the rest had mired down 800 yards north of the skidway. All along the route the column had passed Japanese arms and equipment abandoned in hasty retreat. In the har- bor area, large amounts of supplies and equipment were captured: gasoline, ammunition, radios, drafting equipment, charts, food, and propaganda pamphlets. The spoils also included buildings in good condition. The captured equipment belonged to the Japanese Iwaka- mi Battalion which had defended the skidway area. According to documents captured on Salami Plantation and later. 63 this battalion was decimated by 6 March. Six hundred men had been lost in the skidway area and in the attacks on the perimeter. The re- maining 200, with an additional 100 stragglers from other disorgan- ized units, were ordered to retreat through Salami (Red) Beach and across Papitalai Harbor to Papitalai Mission. Natives on Mokerang Peninsula later told the Angau Party that the Japanese retreat de- veloped into a rout. They were panic-stricken; some did not even wait to take paddles for the native canoes that they had appropriated for their escape to Papitalai Mission. Not more than 80 Japanese, fran- tic from fear and exhaustion, arrived at the mission to bolster the force already there. New Beachheads: Papitalai and Lombrnm The foothold on Seeadler Harbor won at Salami Beach on 6 March must be immediately expanded. Control of the western shores of Los Negros was essential to insure the safe landing of the 2d Bri- gade, due to arrive at Salami in three days. Our command thought that the Japanese who had abandoned Salami had probably crossed the harbor to Papitalai Mission and Lombrum Point, already held by enemy troops. Some of the enemy might remain on Mokerang Pen- insula, which, along with Koruniat and Ndrilo Islands, was reported to be occupied by enemy forces with naval guns. These shore areas would be cleared of any possible enemy threat, and the islands inves- tigated. The 5th Cavalry at Momote and the 12th Cavalry with the 2d Squadron, 7th Cavalry at Salami would both send forces over wa- ter to Seeadler Harbor’s southeastern shore and drive out the enemy remaining there. The 12th Cavalry would also clear Mokerang Penin- sula north of Salami and determine the situation on Ndrilo and Koruniat Islands. The 5th Cavalry had begun the work of clearing the southern shore of Seeadler Harbor on 6 March by pushing patrols west from the air strip. The lack of opposition encountered in moving west bore out what the increased count of enemy dead and the abandoned equipment at Salami had indicated: the greater part of the enemy combat strength on Los Negros, if not throughout the Admiralties, had been dissipated. A patrol from Troop F was able to reconnoiter to Porlaka without encountering opposition, whereupon the entire troop established a bridgehead there, protected by the regimental antitank platoon and the mortar platoon of Troop H. Patrols from 64 the 1st Squadron moved south of their forward positions and worked through about a mile of territory south and southwest of the air strip along Porharmenemen Creek, capturing medical supplies, ammu- nition, and documents, as well as disposing of four snipers. Having covered the ground to Porlaka, the 5th Cavalry’s next mission was to cross Lemondrol Creek to Papitalai, while continuing to patrol south and west of the air strip. The 12th Cavalry with the 2d Squadron, 7th Cavalry, would simultaneously seize Papitalai Mission and Lombrum Plantation to anchor the defense of the eastern end of Seeadler Harbor. Although no enemy activity on Papitalai was report- ed from the command post at Porlaka, which had excellent observa- tion on these points across Lemondrol Creek, some Japanese resistance was expected against the 5th Cavalry’s shore-to-shore operation. A reconnaissance patrol of 40 volunteers from Troop B, led by Capt. William C. Cornelius, moved across Lemondrol Creek at 1200 on 7 March. A 15-minute mortar and artillery concentration on the objective preceded the landing, and mortars, heavy machine guns, and antitank guns at Porlaka remained ready to support the patrol. Two waves, 45 minutes apart, were dispatched in canvas pneumatic boats, rubber boats, and engineer assault boats. The opposition at Papitalai was ineffective. An estimated 50 Japanese staged a heated skirmish and then withdrew. Captain Cornelius, leading the first wave, is reported to have killed single-handed 4 of the enemy with rifle fire and grenades while operating 50 yards in advance of the troops. Severely wounded, he died the next day; for his courage and leadership he was awarded the Distinguished Service Cross. The Reconnaissance Platoon held Papitalai while Troop F crossed the creek, using a captured Japanese barge as well as their own boats. Enemy rifle and mortar fire was directed at Papitalai and Porlaka through the night and next morning, but the enemy could not muster enough force to threaten our positions. One small enemy artillery piece caused some trouble for the troops at Porlaka until it was si- lenced by the mortars and antitank guns there. Our losses in the tak- ing of Papitalai were three officers wounded, one officer and one en- listed man killed. The 2d Squadron, 12th Cavalry, was designated as the force to take Papitalai Mission, the next promontory northwest of Papitalai. The move would start from Salami on 7 March. A shortage of sup- plies at Salami postponed until 8 March the attack upon Lombrum 65 Plantation, assigned to the 2d Squadron, 7th Cavalry. However, sup- plies for the 12th Cavalry’s attack on Papitalai Mission were dropped from artillery liaison planes. Ammunition and water were also loaded on five buffaloes from the supply dumps located at Momote, and three of them arrived at Salami by 1300. Troop G of the 2d Squadron, 12th Cavalry, moved out across Seeadler Harbor shortly afterwards, transported on two buffaloes; the third had burned a clutch, but others arrived for use in shuttling the troops across the harbor. An air strike and a half hour bombardment by the 271st Field Artillery Battalion preceded the attack; fire on Papitalai Mission Point was well adjusted by the observer at Salami Beach. The attacking squadron had to cross nearly a mile of open water, and each wave encountered enemy fire from mortars, machine guns, rifles, and one 75-mm howitzer. The first wave, a platoon of Troop G, led by 2d Lt. Alfred W. Prentice, landed in the face of fire from bunkers and native shacks and held alone for 45 minutes until the next wave landed. The advance elements later broke up a counter- attack by about 30 Japanese. By nightfall, the rest of the troop landed and held a beachhead 50 yards inland and 150 yards in width. As night landings were impossible over the coral reefs, Troop G faced a dangerous period. However, direct support fires adjusted by an ar- tillery liaison officer accompanying Troop G were brought down just 50 yards beyond the defense perimeter, and broke up three determined counterattacks during the night. By morning the Japanese had pulled out from their beach de- fenses at Papitalai Mission. Patrols from Troop G were able to pene- trate 1,000 yards south and southwest, although the Japanese subse- quently reoccupied these areas. The remainder of the 2d Squadron, 12th Cavalry, arrived in the morning and occupied a perimeter 250 yards wide and 90 yards long. They encountered no resistance. The 2d Squadron’s casualties during the attack on Papitalai Mission up to noon of 8 March were 7 dead, 27 wounded or sick. The enemy had put up more fight at Papitalai Mission than at Salami, Porlaka, or Papitalai, and to 8th Area Army Colonel Ezaki sent an account of the resistance at the Mission which was certainly keeping up a good front: Fresh enemy troops initiated a landing at Sabukaleo (Papitalai Mission) by amphibious trucks under cover of terrific gunfire. Enemy strength is ap- proximately 1,000 at present and is increasing. The Garrison Unit and the 66 Sabukaleo Sector Unit are at present engaged in a terrific battle and the count of dead up to 1700 hours has been 100. However, morale is high. We are attacking with the spirit of fighting to the last man. In a later message Ezaki admitted that the Americans had gained a foothold, but he promised an attack the night of 8 March to wipe out the new beachheads at Papitalai and Papitalai Mission: The Sabu\aleo Sector Unit carried out a night attack and did considerable damage, but the enemy successfully increased his strength with support of terrific bombardment and shelling; moreover, they have established a gun position by the water’s edge at Papitalai and are advancing. Tonight the 8th, the Garrison Unit, together with the Sabu\aleo Sector Unit, as well as a platoon of infantry will carry out a night attack against the enemy at Papitalai and Sabukaleo. These night attacks did not develop. The perimeters at Papitalai and Papitalai Mission were undisturbed during the night and, signifi- cantly, the record of Ezaki’s messages to 8th Area Army Headquarters ends at this date. By 1200 on 8 March supplies for the 2d Squadron, 7th Cavalry's attack on Lombrum Plantation began arriving at Red Beach over the difficult road from Momote. More supplies were dropped from B— 24’s. The squadron was issued 2 days’ rations, water, and ammu- nition and ordered to move across Seeadler Harbor and secure a beach- head on Lombrum Point. They were to be transported on LCM’s (Landing Craft, Mechanized), towed to Hyane by the LST’s on 2 March. H Hour was 1420. There was some sporadic fire as the troops hit the beach where two Japanese were killed, but by 1430 Troops E, F, and G had established a perimeter about 100 yards in depth. Patrols sent out immediately encountered no Japanese, and it was the opinion of the commander of the first wave that the Japanese had not occu- pied the area in strength for some time. By 1700 the perimeter was extended to take advantage of adjacent terrain, and all defensive in- stallations were completed. Abandoned enemy equipment here — gas, oil, and bomb dumps — indicated further disintegration of the enemy strength on Los Negros. The southeastern harbor area was now apparently secure, al- though Japanese hidden inland continued to keep the cavalrymen on the alert: on the morning of 10 March, a lieutenant of the 12th Cavalry was found dead near his fox hole. On the east shore of Seead- ler Harbor the 12th Cavalry was fulfilling its mission of securing 61 OVERGROWN COCONUT PLANTATION on Mokerang gave the enemy hiding places, which had to be hunted out by 12th Cavalry troops. Mokerang Peninsula to cover the north flank of the 2d Brigade’s land- ing. On 7 March three light tanks were sent to reconnoiter Mokerang point. The patrol killed a few Japanese and saw a number of pillboxes in the Mokerang area, but encountered no enemy in strength. Supply Problems Equipping the 12th Cavalry and the 2d Squadron, 7th Cavalry, at Salami with enough supplies to carry on their overwater attacks was a difficult and hazardous operation. The single road from Mo- mote to Salami was impassable for most vehicles during the days when the supplies were most urgently needed. Buffaloes got through by going overwater part of the way, but the rest of the essential sup- plies had to be dropped from airplanes or sent in LCM's from Momote around Mokerang Peninsula. The sending of LCM’s into Seeadler Harbor was an operation which was possible only after continued naval efforts from D Day on. Magnetic mines, dropped from our 68 own planes in May 1943, were presumably still in the harbor and had to be removed. To make entry into the harbor safe for our forces, destroyers also had to neutralize the Japanese harbor defense guns, which had already proved effective. The destroyers and minesweepers worked to accomplish these missions, but even by 7 March, when six LCM’s loaded with supplies were to make their way around the point, it was not certain that enemy resistance on the islands guarding the harbor had completely disintegrated. On D + 2 the Mullany had accompanied two minesweepers within the harbor and had been fired upon from Hauwei Island with intense, accurate fire. The Mullany had then secured the assistance of the Warramunga, Bush, and Ammen in an attempt to knock out the batteries, and each had fired 70 rounds from 5-inch guns. Although the effectiveness of much of the fire had been hindered by rain squalls, it was assumed that the defenses had been silenced. On the next day, however, when the Mullany accompanied the minesweeper again, she was driven off by an estimated five 4-inch guns which were so well emplaced that they seemed not affected by the Mullany’ s return fire. At daybreak on the morning of 3 March the Mullany and the Ammen returned for the third time to bombard the shore battery but were met by intense fire which frequently straddled the ships. At this time it seemed to the destroyers that the Japanese had even stronger defenses than had previously been encountered, since guns opened up from positions which had been unobserved before. However, the ene- my was evidently conserving ammunition, as he would not fire when the destroyers were some 10,000 yards away. Accordingly, the two destroyers stationed themselves beyond this range and undertook slow and deliberate bombardment. On 4 March cruisers of Task Force 74 bombarded the positions on Hauwei Island with undetermined results. On 5 March the Nich- olson approached the same positions in an attempt to draw fire, but the Japanese opened up only when the Nicholson got within 850 yards of the island. The enemy guns registered one hit that rendered useless a 5-inch gun, killed three men, and wounded two. Neverthe- less, the Nicholson reported that she put out of commission the gun that fired on her. She was subsequently credited with destroying two enemy guns and determining Japanese shore strength and gun posi- tions, thus making possible the selection of suitable landing points in Seeadler Harbor out of range of hostile gunfire. 69 Air power was also employed against the enemy’s harbor de- fenses. On 7 March seven B-24’s bombed Hauwei and Ndrilo and the next day 17 heavy bombers and 11 mediums got through to bomb and strafe diese islands and other targets. The job of neutralizing the enemy on the islands outposting the harbor had been done so well that expected opposition against the LCM’s did not materialize. On 8 March six LCM's loaded with sup- plies with two destroyers and two minesweepers entered the harbor without drawing any fire. The destroyers bombarded Koruniat and Ndrilo Islands to be sure that the 2d Brigade would land safely on Red Beach the next day. As a last preparation for the 2d Brigade, the 12th Cavalry, less the second squadron, moved 200 yards south of Salami Beach on the morning of 9 March to protect the right flank of the landing, which was to take place early in the afternoon. To assist in covering the north flank, three light tanks were again sent to Mokerang point, while pa- trols were sent north and east. These patrols had only light contact with the enemy. They killed two Japanese and destroyed a Japanese telephone line, but the small number of enemy encountered in the area indicated that the 2d Brigade would have a safe landing. SALAMI PLANTATION on Mokerang Peninsula provided an excel- lent beach for the landing of the large force of the 2d Brigade. Six LST’s were able to move right up to the beach and drop their ramps on the shore. Landing of the 2d Brigade Targets of the naval preparation for the 2d Brigade landing were on Manus Island. The outlying islands had been given a thorough pounding and it seemed that any threat from coastal guns would come from Manus. From the convoy escorting the six LST's and the Liberty ship bringing the 2d Brigade to Salami, destroyers were detached to bombard enemy positions. The Long, covered by the Wil\es and the Swanson, swept the vicinity of Lorengau, and all three ships took under fire the jetty and what appeared to be light gun emplacements. At 1300 the 2d Brigade Combat Team, a force of 4,000 men, be- gan to disembark from LST’s and the Liberty ship in Seeadler Har- bor. The combat team, commanded by Brig. Gen. Verne D. Mudge, included the 8th Cavalry, the 1st Squadron, 7th Cavalry, the 61st Field Artillery Battalion, two antiaircraft batteries, a troop of engineers, a medical squadron, and a clearing company as well as a surgical hospi- tal and signal, ordnance, quartermaster, and labor troops. The first mission of the 2d Brigade was to wipe out the isolated groups of enemy still bothering our troops on Los Negros in the Salami area. One group of Japanese had been located by the natives of Moke- rang Village and reported to the Angau party. The Australians had been hospitably received by the two native chiefs in official and cere- monial regalia, which had been hidden from the Japanese. As the Japanese had forced them to labor for small recompense and had stolen their furniture and canoes, the natives were delighted that the Americans had come. They furnished guides to show a Japanese hideout to a combat patrol of Troop F, 8th Cavalry which was as- signed the mission of mopping up the group. The patrol, led by an Angau man and native police boy guides, stalked the enemy, who was holding out in native shacks in a stretch of jungle between Salami and Mokerang plantations. The patrol surrounded the Japanese and killed with small-arms fire all 15 hiding there. The 8th Cavalry's first casualties in action were one killed and two wounded. This small action stimulated the imaginations of troops holding a jungle area for the first time against an unknown number of enemy. Infiltration had been carefully explained to the cavalrymen and they expected a fight. Under an overcast sky with an occasional flurry of rain, troopers constantly “saw” or “heard” Japanese in the palm trees and in the undergrowth, and all night long the plantation echoed to 71 UNLOADING AT SALAMI. Enemy air power in the Bismarck-New Guinea area had been so well neutralized that beaches were comparative- ly safe from attac\. During the entire action few enemy planes appeared. WATER-ABSORBENT CORAL SAND on Los Negros was good material for surfacing runways. Heavy-duty rollers pac\ sand on the Momote strip. The troops also constructed a control tower (left background). SUPPLIES ARE EFFICIENTLY HANDLED by a conveyor belt and piled up for units making overwater attacks or holding forward lines. 73 the burst of grenades and the spasmodic firing of small arms. Soldiers in shallow fox holes were sprayed with earth and grenade fragments. At dawn a careful search revealed not one dead Japanese or any sign of the enemy, hut a few Americans had been wounded in what the soldiers called the “fire fight.” Although the debarkation of the 2d Brigade occurred under better circumstances than had been anticipated and all vehicles were unload- ed on to the beaches in a few minutes, some miscalculations caused difficulties. The engineer equipment had been loaded well in the rear of the LST’s and consequently was taken off last. The vehicles which were landed first soon made the wet roads a sea of mud, and all traf- fic was at a standstill. According to the report of Troop C, 8th Engi- neers: It was practically impossible to get engineer equipment forward to the “bottlenecks” and many hours were wasted in hacking out new routes. At dark that night many vehicles were still mired in the mud, and work continued the following day clearing the roads. A route for a new road was selected, roped off, and work started cutting corduroy. This proved to be a tedious, back- breaking job, as the only available materials were coconut logs, 8 to 12 inches in diameter, and too heavy to handle by manpower alone. But with the aid of winch trucks, bulldozers, and manpower, logs were placed, sand hauled in with dump trucks, and carryalls, and on 12 March traffic was moving in all areas. The engineers were aided in their first, most urgent job of getting die roads cleared up by a platoon of a quartermaster depot company, which arrived on 10 March. An amphibious truck company and other service elements also came ashore to handle the increasing influx of supplies necessary for further military operations and for the sched- uled construction activities. The obstacles of jungle terrain, on the one hand, and insufficient water transportation on the other made even the routine job of keeping the troops armed, clothed, and fed a tre- mendous task. After finishing its first job on road repairs, the quarter- master depot company began to unload the Liberty ship Etamin, using roller conveyors without which the job would have been impossible with the limited number of troops available. The amphibious truck company unloaded supplies on its 16 dukws. More medical units and supplies came ashore. The medical de- tachments serving the reconnaissance force had been hampered be- cause of scanty protection afforded by the flat terrain. In the first days when casualties were high, they were also severely handicapped in 74 facilities, stripped to the barest necessities because of the nature of the operation. When bulldozers arrived on 2 March it had been pos- sible to build revetments around the hospital tents, and the over- crowding was alleviated a little by transporting the seriously wounded to destroyers bound for the Sixth Army Headquarters. A collecting troop and a platoon from a clearing troop arrived on D + 2 and were on hand for the big attacks on 3/ 4 March. However, the strain on the small medical units continued until a section of the 58th Evacuation Hospital landed in Hyane Harbor on 6 March. In expectation of further operations, it was decided that more medical facilities were needed, so the 27th Portable Surgical Hospital, the 603d Medical Clearing Company, and the 1st Medical Squadron (less detachments) landed at Salami with the 2d Brigade, and by 15 March the surgical hospital was installed and functioning. The Momote air strip, prize of the Los Negros fighting, had been cleared and leveled for use a week after the initial landing at White Beach. A damaged B-25 was the first airplane to land on the strip on 6 March. Eight Kittyhawks of the 76th Fighter Squadron landed on the 4,000-foot strip on 9 March. The repaired strip had an alert area 600 feet long, and 6 dispersal bays. Gaining and securing this objective had cost the 1st Brigade 116 dead, 434 wounded. These sacrifices had not only gained the objective; American forces had exacted about 11 times their own losses in killed from the enemy. The enemy dead on 8 March had reached 1,288, with no prisoners taken. The possibility of any large-scale organized re- sistance on Los Negros had been removed. Our next objective would be Lorengau airdrome on Manus, assigned to the 2d Brigade. Operat- ing from a firm base on eastern Seeadler Harbor, the 2d Brigade would face little of the grave risk that accompanied the landing of the 1st Brigade in Hyane Harbor. 75 Attack on Manus Island The 1st Brigade had broken large-scale organized resistance on Los Negros and could confine its operations to mopping-up actions against remaining enemy groups holding out in the hills southwest of Papitalai and Lombrum. An estimated 150 to 200 Japanese were left in the 5th Cavalry sector, and it was thought that many survivors had fled to Manus. Wiping out the bulk of the enemy garrison on Manus would be the job of the newly landed 2d Brigade. The number and disposition of the enemy on Manus were matters of some uncertainty. A rough estimate was that 2,700 Japanese were concentrated there and would probably make a stand at Lorengau. Although it was confidently expected that the Japanese could do no more than put up a last losing fight, the lack of knowledge of their numbers and disposition on the largest island of the Admiralties re- quired that plans be made carefully for an invasion in strength. Fight- ing into the interior where the mountain range would provide defen- sive positions and through the swampy forests covering the rest of the island would not be easy. Only four roads winding about the island and converging at Lorengau would be suitable for moving vehicles, and long stretches of these red-clay roads would be impassable in rainy weather. Surrounding jungle and swamps would confine any large- scale movement to the roads. Therefore, the plan was to pin the enemy to the coast where it was hoped he had concentrated his main strength. Plans were drawn to invade Manus Island on beaches near Lugos Mission, 2J4 miles west of Lorengau, where the airdrome was the main objective (Map No. 9, opposite page). Lorengau was known to be fortified. The village and the airdrome had been worked over by in- termittent bombing and strafing attacks, naval bombardment, and 76 artillery since 29 February, but the results were not certain and the sea approaches were thought to be mined. Therefore, the landings were planned for the area of Lugos Mission, where no local defenses had been observed, although troops landing there would be within range of any light guns at Lorengau. Two suitable beaches on each side of the Lihei River were designated as the assault landing points: Yellow Beach 1 east of the river and Yellow Beach 2 on the west (Map No. 10, page 78). Troops A and C of the 8th Cavalry would make the assaults on these beaches. After both beachheads were secured, the remainder of the 8th Cavalry would come ashore, followed by the 7th Cavalry which would relieve Troop C on Yellow Beach 1. The 7th Cavalry would remain in reserve inshore from the beach, with one troop de- fending the right flank of the beachhead. Troop C of the 8th Engi- neer Squadron would land on Yellow Beach 1, make necessary im- provements on either beach, and construct crossings of the Lihei River to connect the two beachheads. MAP NO. 9 77 78 MAP NO. 10 When the 8th Cavalry had secured a beachhead to include Lugos Mission, each squadron would move east, the 1st Squadron along the coast (Number Three Road), and the 2d along an inland road (Num- ber One Road) to execute a wide sweep toward Lorengau. If a size- able garrison were concentrated at fortified Lorengau, the inland arm of this attack would prevent the enemy from escaping along the only road leading west to the mountain areas. As a preparation for the attack it was decided to take several of the small islands lying a few hundred yards off the north coast. These islands would be reconnoitered with a view to emplacing artillery to support the attack, a procedure which had produced outstanding re- sults at Kwajalein. Artillery fire added to the customary preinvasion concentration of naval gunfire would undoubtedly help reduce the cost of landing against possible resistance in strength. The invasion of Manus was set for 1 3 March ; the expeditions to the islands would be undertaken on 11 March. Preparations From Salami Beach three patrols from the 302d Cavalry Recon- naissance Troop, accompanied by artillery officers, set out on recon- naissance of Bear Point on Manus and the islands of Butjo Luo and Hauwei. Bear Point was found to be free of the enemy, but unsuit- able for artillery. The northern island of Butjo Luo provided excel- lent sites for artillery emplacements, and both islands were apparent- ly unoccupied. The patrol on Hauwei Island ran into trouble. A platoon in strength, the patrol moved out from Salami on an LCV and a PT (Motor Torpedo) boat and landed without opposi- tion on the western end of the island. Maj. Carter S. Vaden, S-3 of the 99th Field Artillery Battalion, threw two grenades into a well- camouflaged bunker about 10 yards from the beach and the party progressed inland. Immediately after the grenade explosions, a mortar shell landed on the beach in the rear of the party. From three sides, machine-gun, rifle, and mortar fire opened up. Fortunately, the gre- nades had evidently sprung the ambush prematurely and the party was not cut off from the rear. Using their submachine guns, the men of the patrol were able to hold back the enemy. They were aided by some fire from the PT and the .30-caliber machine gun of the LCV. After 2/ z hours they 79 managed to withdraw to the water's edge. The PT boat had returned to its tender after the commanding officer was wounded, and the cox- swain of the LCV shouted and motioned for the cavalrymen to get on board, but only five of them reached the boat, as the others were engaged with the enemy on shore. The LCV had troubles of its own; enemy mortar and machine-gun fire found its range and wounded all on board except two. Then the LCV grounded on submerged coral and sank 200 yards offshore after receiving a direct hit from a mortar. The wounded men, put into lifejackets, floated about in the water. When the remaining members of the patrol on shore were wad- ing out toward the LCV, six Japanese attempted to set up a machine gun on the beach. The men cut them down with their submachine guns. Still under fire, the cavalrymen swam out to the survivors of the LCV, making a group of 18 men out in the water. Failing to attract the notice of the destroyers nearby, the tired men had to stay in the water for 3 hours until a PT boat picked them up. At the same time a destroyer closed in to the island to draw hostile fire and bom- bard enemy positions on the western end. The losses in this unexpectedly difficult operation were high. Six men of the reconnaissance troop and two artillerymen, one of whom was Major Vaden, were killed. Three men were missing and every member of the patrol received wounds of varying seriousness, as well as second degree burns from long exposure to the sun and water. It was now apparent that Hauwei, which was ideal for artillery posi- tions, would have to be taken by a larger force, even though it would mean postponing the attack on Manus. The Navy also recommended a delay to permit clearing all sea approaches. Butjo Luo would be se- cured by a cavalry troop, although no opposition was anticipated there. The 2d Squadron, 7th Cavalry, relieved at Lombrum Point on 11 March by elements of the 12th Cavalry, was designated as the force to secure Butjo Luo and wipe out resistance on Hauwei. In the afternoon Troop F, with artillery personnel, moved out on an LCM and secured both islands of Butjo Luo without opposition. On 13 March the entire 99th Field Artillery Battalion, except for a rear echelon left at Salami for maintenance of vehicles, was moved to Butjo Luo. Twelve howitz- ers and six 37-mm antitank guns were carefully placed in position. To guard against detection, no trees or underbrush on the beach facing Manus were cut until the evening before the attack. 80 Hauwei was to be taken by the remainder of the 2d Squadron, 7th Cavalry on 12 March. Destroyer gunfire and rockets as well as 105-mm concentrations from the 61st Field Artillery Battalion at Mokerang point preceded the operation. Kittyhawks of the 77th Pursuit Squadron (RAAF) at Momote airfield also bombed and strafed the objective. The cavalrymen reported, however, that the covering fire was not very accurate. Small-arms fire was encountered by Troop E, landing on the west shore; some machine-gun fire from bunkers opened up on Troop G trying to move in from the south beach. The cavalrymen began to make their way inland and successfully bypassed without casualties a few mines activated by trip wires. Upon reaching a point 300 yards inland on a line running north and south, both troops were pinned down by heavy rifle, machine-gun, and mortar fire. Casualties suffered up to this time were 3 killed and 10 wounded. Under the protecting fire of the Troop H mortars, which landed shortly after the assault waves, Troop E was able to advance east, although Troop G could not move forward along the south coast in the face of continued heavy enemy fire. The squadron commanding officer then decided to withdraw Troop E and dig in for the night, as contact was being lost and the island was too wide for complete coverage by two troops. Troop C, which had been ordered to proceed from Salami, arrived at 1800 and took up a support position on the small perimeter. A medium tank would also arrive the next morning to give impetus to the 2d Squadron’s attack. Plans were made for overcoming the difficulty encountered on the first day. Japanese bunkers along the coast had caught the cavalry- men advancing along the axis of the narrow island (1,300 by 400 yards) in a cross-fire; the next day the bunkers along the edges of the island would be attacked at the same time that the line was advanced in the center. During the night the 61st Field Artillery Battalion from the north end of Mokerang Plantation fired -1,000 rounds to harass the enemy. This fire was directed so accurately and carefully that it was brought down on the enemy within 100 yards of the cavalrymen’s positions without any casualties to our troops, and helped minimize enemy activity during the night. At 1000 the next morning Troops C, E, and G advanced abreast in a line from north to south, with the tank operating on the south flank. The enemy defended the island with skill and tenacity. Well- 81 constructed ami carefully placed bunkers covered all avenues of ap- proach, with fire lanes for all automatic weapons. Enemy snipers were accurate; of two troopers killed and eight wounded by sniper fire, all were shot in the head or chest. However, the defenders had no weap- ons to counteract tank action, and by 1200 the three troops of cavalry with the effective support of the tank had succeeded in mopping up all centers of resistance. An enemy bunker on the south shore, manned by eight Japanese using rifles and two heavy machine guns, was knocked out by tank action and troop envelopment from the north. Before being silenced, this bunker withstood four direct mortar bursts, and four point-blank hits from the tank’s 75-mm. Another Japanese trench position, containing light machine guns and knee mortars, was surrounded and its occupants annihilated. Forty-three enemy dead, all naval personnel, were counted. They had been in good physical condition and were well equipped. Our casualties for this small oper- ation were 8 killed and 46 wounded. The advantages won by this operation were soon evident. At 1500 on 13 March the 61st Field Artillery Battalion began debarking on Hauwei from LCM’s. The next day the artillery battalion put its how- itzers into position on the southwest side of the island. The 271st Field Artillery arrived and placed its guns on the west. The firing positions of the 105-mm batteries, on the flank of our own troops who would at- tack to the east after landing on Manus, satisfied an artilleryman’s dream, since very close support of troops could be maintained by minor shifts of deflection, thus eliminating the danger of short rounds. The 99th Field Artillery Battalion, on Butjo Luo, was also in position to support any southward movement of our troops in pursuit of the ene- my. By dawn of 15 March the three artillery battalions on Hauwei and Butjo Luo had their guns registered for high burst adjustment on the north shore of Manus, ready for the 2d Brigade’s attack. H Hour was set at 0930. Landing at Lugos Mission (15 March ) At 0700 the three artillery battalions opened fire from Hauwei and Butjo Luo. Their targets were in the Lorengau Township area, chosen to divert the enemy’s attention from the landing beaches around Lugos Mission. Four destroyers, assigned to support the inva- sion with a bombardment beginning at dawn, raked the shore from 82 the Tingau River to east of Lorengau with their 5-inch guns. Arriv- ing in 3 hours from Nadzab in New Guinea, 18 B-25’s appeared at H — 30 minutes and bombed and strafed Lugos Mission and the beach areas until H — 5 minutes. Between 0907 and 0925, they dropped eighty-one 500-pound bombs, strafed with more than 44,000 rounds of machine-gun fire, and cleared the target just as the troops were coming ashore. Offshore, the assault troops of the 2d Brigade on LCM’s and LVT’s moved in toward the beach with the slow-moving buffaloes in the first wave. They were protected by an umbrella of fire from three rocket boats (two LCV’s and one LVT) operated by amphibious engineers. ENGINEERS OPERATED THE LCM’s (above) and LVT’s which brought the first tvaves carrying the assault squadrons of the 8th Cavalry to Manus. The jourth wave was carried up to the beach on an LST. 83 The wave of buffaloes drew up to the beaches, and it appeared that they were going to make it without opposition until an enemy machine gun opened up from east of the landing beaches. Immediate- ly the buffaloes responded with machine-gun fire, and two PT boats moved close in. With 100 rounds from a flak boat added to this fire the machine gun was silenced. Two waves of landing craft followed the buffaloes to the shore in rapid succession and the narrow beaches, backed either by jungle or a high bluff, were soon congested with as- sault troops. On the east beach, men of Troop A began a vigorous drive toward Lugos Mission and Number Three Road, which follows the coast eastward to the 1st Squadron’s objective, Lorengau airdrome. Led by Capt. Raymond }. Jennings, the troop stormed the bluff where Lugos Mission was located and overran the mission before the few Japanese who had survived the bombardment could put up effective opposi- tion. Predesignated groups worked their way into the mission com- pound and cleared out the area with grenades. Subsequent groups mopped up bunkers which were sited to cover the approaches from the beach. Troop A was soon on its way along Number Three Road leaving behind it some 20 dead Imperial Marines, without the loss of a single cavalryman. From the western beach. Troop C drove through the jungle with- out opposition. On a ridge 800 yards inland, they stopped and dug in, having accomplished their mission of establishing a defensive per- imeter to protect the landing of successive elements. Additional waves of the 1st and 2d Squadrons quickly followed the assault troops ashore and moved toward their objectives. The 1st Squadron followed Troop A down Number Three Road east toward the airdrome while the 2d Squadron pushed inland directly south toward Number One Road, its intermediate objective. Col. William J. Bradley, the regimental commander, came ashore with the third wave to set up an advance CP and establish communication with the fast separating squadrons. The decision to split the regiment into two forces bound for ini- tial objectives 2 miles apart was a bold one, but was considered nec- essary. Attack along the coast road might be stopped by heavy oppo- sition from Lorengau, or the coast might prove impassable for vehi- cles. Furthermore, by sending one force toward Lorengau from Num- ber One Road, the 2d Brigade could prevent the escape of the Japanese garrison west to the mountain areas of the interior. 84 8th Cavalry Moves East (15~16 March ) Leading elements of the 1st Squadron, advancing down the coast road, found movement slow and difficult on the red-clay trail which had a glue-like consistency from the recent rains. Troop A reduced enemy bunkers along the road and progressed about a mile from the beach. At 1120 the troop was stopped cold. Three mutually supporting bunkers covered a section of the road which was hemmed in by the beach on one side and a heavy mangrove swamp on the other. With no room to maneuver, the attackers could approach these bunkers only from the front. A squad led by S/Sgt. D. H. Yancey, former sta- ble sergeant and one of the few old Regular Army noncoms in the troop, rushed the position without waiting for orders from Captain Jennings. Before the squad was able to get within grenade range, Ser- geant Yancey was killed and the attack faltered. The enemy bunkers were clearly a job for the artillery. One of the forward observers of the 271st Field Artillery Battalion was now called on to bring in artillery fire from the battalion’s how- itzers on Hauwei. First, the area would be swept with time fire to clear out snipers protecting the position, and then the bunkers them- selves would be softened up with a battalion concentration of delayed fuze. Troop B had been alerted to go through Troop A and attack the bunkers. However, shell fragments from time fire bursting less than 100 yards away kept the men in their fox holes. Moreover, tanks which had been called for did not show up. Strong leadership got the men going shortly after the artillery lifted, but the bunkers had withstood the artillery pounding; again enemy fire broke the attack. Again artillery fire was called for, and 81-mm mortars went into action from positions on the very edge of the ocean. An air strike was also requested of P-40’s which had maintained an alert since H Hour. Each plane carried a 500-pound bomb in addition to its regular arma- ment. When the position had been turned into a waste of craters, T roop B advanced again. This time there was no opposition. Exam- ination of the position showed that the Japanese had remained in their bunkers until the massed fire power pulverized the position. The row of bunkers had evidently been the final defense of the western edge of the Lorengau air strip. Without any further opposition, the 1st Squadron came suddenly out of the jungle into the clearing and 85 LORENGAU AIR STRIP was only 3,000 feet long and was used as an auxiliary field. Number Three Road runs along the south side of the strip and is joined by Number One Road, which comes out of the jungle (lower right). Clearing (right) contains some buildings of the village, the rest of which lies across the Lorengau River (far right). at 1700 occupied a commanding ridge in the coconut palms overlook- ing the southern edge of the airdrome. There the tired soldiers dug in for the night. Our casualties were 2 killed and 11 wounded; about 40 Japanese had been killed in the advance down Number Three Road. At 1800 Troop C joined the squadron, having been relieved from its beach- 86 head position by the 7th Cavalry. Throughout the night the cavalry- men near the airdrome were harassed by a number of snipers occupy- ing a fringe of coconut palms north of the dispersal area. On the tortuous track leading south to Number One Road, the 2d Squadron with Troop F in the lead made its way over a continuous succession of ridges. Sniper fire as well as the difficult terrain made 87 D— 7 BULLDOZER operated by Private Webster j. Ough, Troop C, 8th Engineer Squadron, enabled tan\s to move through the jungle to Number One Road. Bulldozers were also used extensively to demolish bunkers. the cavalry’s advance, followed by borrowed artillery tractors towing ammunition and supplies, extremely slow. But by 1500 the leading dements had reached Number One Road. There three bunkers, situ- ated to protect both the track and the road, obstructed the column. Scattered mortar fire from undetermined positions in the jungle also began to be troublesome. It seemed advisable to establish a perimeter for the night and attack the bunkers on the following day. Artillery fire placed on the general area of the bunkers silenced them for a time. Under cover of the protecting fire, the 2d Squadron dug in. The 7th Cavalry forming the brigade reserve had come ashore on the second trip of an LST and completed its landing on the west 88 beach at 1635. Engineers accompanying the first waves ashore had discovered that the Lihei River could be forded by men and vehicles using the sand bar at its mouth. However, by the time the 7th Cavalry arrived the sand bar was washed out, and this necessitated the ferry- ing of all light equipment, and the towing of heavy equipment by D— 7 bulldozers. Troops of the 7th Cavalry forded the Lihei River, about waist-deep, and moved to Lugos Mission to establish a perimeter de- fense covering the west half of the entire beachhead while Troop C, west of the river, protected the right flank. The balance of the 8th Cavalry defended the left flank, with a perimeter extending from Yellow Beach 2 to the eastern edge of Lugos Mission. The 1st and 2d Squadrons radiating east and south would be well protected from the rear. In the morning the 2d Squadron got off to an early start, under the direct command of the brigade and regimental commanders. The position which had delayed the advance the previous night was over- run with the aid of one light and two medium tanks. A D— 7 bulldozer was absolutely necessary to keep them going over the difficult terrain. The dozer cut down grades, cleared heavy jungle growth, and towed the tanks. Bulldozer operator Pvt. Webster J. Ough did his work un- der sniper fire with mortar shells falling close by. Another engineer, Sgt. Sammie C. Mandel, with the squad protecting the dozer, was wounded in the foot by a sniper as he climbed on one of die tanks to show the tankmen the route. Once on Number One Road, the squadron moved slowly north- cast. Camouflaged bunkers, sited to give fields of fire up and down the road, made the advance difficult. The only maneuver space was in heavy jungle, yet by using it for cover the cavalrymen isolated eight bunkers. Then, with the aid of one tank that remained with the squadron up Number One Road, the troopers destroyed the bunkers with grenades. No artillery was called for, and at the end of the day the squadron halted at a position on Number One Road about 100 yards from Lorengau and the eastern end of the air strip. The squadron had sustained only seven casualties in its move up the inland road. On the other wing of the 2d Brigade’s push east, the day had told a different story. Although the 1st Squadron attacked the air strip all day, the enemy still held the eastern half of that objective on the evening of 16 March. 681583 °— 46— 6 89 Check at the Airdrome (16 March) In defense of Lorengau airdrome, the enemy put up the heaviest resistance yet encountered on the Manus beachhead. Japanese in well-defended bunkers farther along the southern edge of the air- drome were to prove a more difficult obstacle for the 1st Squadron than the snipers in the coconut grove north of the strip who had harassed them during the night. In the morning, squadron com- mander Maj. Moyers S. Shore sent a platoon from Troop A north across the strip after the snipers. No attack would be made along the open strip until this danger was removed, but Troop C was ordered at the same time to advance along the south edge and wipe out any enemy encountered there (Map No. 11, below). Before Troop C had moved 200 yards over a series of small ridges studded with coconut trees, the leading platoon ran into 90 BLASTED TREES at the edge of the Lorengau air strip show effects of our heavy artillery fire. Japanese in bunkers held up our advance here. heavy machine-gun fire from commanding ground 150 yards to its front. Capt. Winthrop B. Avery, commanding Troop C, emplaced his attached platoon of water-cooled machine guns and the regimen- tal 81-mm mortars and attempted to launch a coordinated attack by maneuvering one platoon around the south flank of the enemy position. When Major Shore ascertained that Troop C was held up by the enemy on the ridge south of the air strip, he committed the balance of the squadron to move around Troop C’s left flank. Since the snipers north of the strip had been cleared out by noon, an attack could be made in the open. Troop B, supported by light tanks on its left, was to go down the strip, followed by Troop D and squadron headquarters. Troop A was directed to cross the strip and drive east along the north edge. This attack moved out at 1300. Meanwhile Troop C was making progress against the stubborn enemy position. Although a frontal attack at the enemy-held ridge failed, the platoon sent out to flank the position from the south had succeeded in getting literally on top of the enemy. Led by 91 S/Sgt. Ervin M. Gauthreaux after the platoon leader had been wounded, the platoon destroyed two bunkers with grenades and drove several Japanese from the connecting tunnels of the bunkers into the open. Suddenly the platoon was pinned down by friendly fire coming from the direction of the airfield. Troop B had been stopped in its advance down the open strip by fire from enemy in the area undet attack by Troop C. Thereupon, the cavalrymen on the strip fired south at the position to cover the removal of their wounded to a protective grove some 200 yards away. This fire was hitting Sergeant Gauthreaux’s platoon as it assaulted the same position from the other direction. The tanks, advancing along the north side of the strip, fired their machine guns diagonally across the front. An enemy machine gun in a coral cave on the south side of the strip opened up, but the tank crews held their fire because they knew that Sergeant Gauth- reaux’s platoon was immediately beyond the cave. They pulled back to a position from which they could fire at the cave without en- dangering the Americans on the other side, and succeeded in knocking out the machine gun. Unfortunately, the tanks were not fitted with two-way radios, so they were unable to stop Troop B’s fire against Troop C's platoon. Troop C's commander realized what was happening and made a difficult decision. He ordered Sergeant Gauthreaux to withdraw to avoid further casualties from Troop B’s misdirected fire, although the platoon was on the enemy’s main position and was mopping it up. Captain Avery then called for artillery on the position whose fire had successfully held up the advance of the entire squadron. Four hundred shells were put down on the well-defended ridge. As the fire lifted, Troop C advanced, this time in a frontal move. Heavy fire from some still untouched bunkers again broke up the attack and Troop C reformed on its line of departure. Under cover of Troop C’s attack. Troop B was able to disengage and reorganize in the palm grove north of the strip. The unsuccessful attack on the bunkers cost the 1st Squadron 9 killed and 19 wounded. Their line extended from the beach across the air strip at about midpoint, and over to the high ground on the south. General Mudge ordered the 1st Squadron, 7th Cavalry, in reserve at Lugos Mission, to relieve the exhausted men of Troop C holding the high ground south of the air strip. During the relief, 92 the 7th Cavalry suffered 5 killed and 15 wounded from the still active enemy bunkers to the front. The 2d Squadron, 7th Cavalry established a perimeter defense back of the front lines, which were held by the 1st Squadron, 7th Cavalry on the right and the remaining 8th Cavalry units on the beach side of the air strip. Both regiments dug in for the night. To the Lorengau River (17 March) After the unfortunate day at the air strip, careful plans were laid for a coordinated attack the next morning to take the strip and push through to Lorengau, lying beyond the Lorengau River in a cup-shaped valley surrounded by jungle hills rising to 400 feet. To link up the two columns and coordinate the advance, the 2d Squadron, 7th Cavalry would move south of the strip and make contact with the inland force (Map No. 12, page 95). Both these squadrons would then advance to the river along Number One Road. All troops on the left wing at the air strip came under the 7th Cavalry commander, Col. Glenn S. Finley, who ordered the 1st Squadron, 7th Cavalry to attack the bunkers which had held up the 8th Cavalry advance. Assault squads from the 8th Engineer Squadron were attached to the 7th Cavalry for use against the enemy bunkers. Preparation the night before the attack was extensive. The Navy put down harassing fire; the artillery battalions, from their excellent island locations, brought fire down on Japanese bunkers as close as 50 yards from our own positions. In the early morning a barrage from twenty-four 81 -mm mortars preceded the attack, and two towed 37-mm's and two light tanks fired at point-blank range into the enemy bunkers. Troop D of the 8th Cavalry gave excellent support to the 7th Cavalry attack. When the troop commander, 1st Lt. Donald D. Taylor, observed the enemy across the strip working on a bunker which had been neutralized the day before, he obtained a direct hit with an 81-mm mortar which completely demolished the position, later discovered to have been defended by 15 Japanese with a .50-caliber and several .30-caliber machine guns. Troop D also destroyed an enemy machine gun which disclosed itself by firing at our troops. As the mortar barrage ceased, the automatic weapons opened up to cover the assaulting troops of the 1st Squadron, 7th Cavalry. 93 TROOP G, 8TH CAVALRY comes out of the dense jungle bordering Number One Road and cautiously advances through a clearing near thici ( growth on the west side of the Lorengau River. Alert for any Japanese facing them, the men ta\e advantage of every opportunity for cover. They came out of their fox holes at 1033, with shouts of “Garry Owen," the regiment’s traditional battle cry. The 1st Squadron met little resistance, as the artillery fire had practically wiped out the bunkers. A secondary enemy position similar to the first was en- countered on a ridge further east. An artillery and mortar concentra- tion was placed on this position and the cavalrymen occupied it at 1130. All remaining pillboxes were demolished by the engineer squads. While the 1st Squadron pushed across the pulverized air strip, the 2d Squadron, 7th Cavalry moved around the south flank to the right and by noon contacted the 2d Squadron of the 8th Cavalry along Number One Road near Lorengau. The 8th Cavalry had en- countered little resistance on the road, destroying whatever bunkers were encountered with the aid of fire from medium tanks. The light tank accompanying the squadron had a track blown off by the first Japanese antitank mine encountered in the operation. 94 MAP NO. 12 95 At 1300 contact had been established between all frontdine troops, from the air strip to Number One Road, and another forward movement was started. Because the backbone of the enemy resistance had been broken at the previous positions, little enemy opposition was encountered on this drive. Covering the 1,500 yards to the river took 2 hours because the Japanese had hurriedly put in large quan- tities of antitank and antipersonnel mines. They had been emplaced, however, with no prepared plan, and many were poorly camouflaged or left lying completely on the surface. The Japanese had also dug holes and refilled them without emplacing any mines. As the caval- rymen slowly made their way to the river, they suffered a few casual- ties from the mines. As soon as the 7th Cavalry reached the river at 1500 the Recon- naissance Platoon was sent across. The patrol immediately drew fire from bunkers on the hills overlooking Lorengau, and withdrew. Mortar fire was placed on the bunkers that could be spotted. Since it was evident that the Japanese were well entrenched around Lorengau, the troops dug in for the night on the west bank of the river and established perimeter defense against expected counter- attacks. These did not develop. Landing craft bringing in supplies to the jetty northwest of the river entrance received fire from the bunkers on the hills above the village, but they were able to land the much-needed supplies. Attack by the 2d Brigade on the next morning would determine how much Japanese resistance was left at Lorengau, long considered to be the best defended area in the Admiralties. 96 Through Lorengau to Rossum In view of the estimated number of Japanese on Manus, enemy resistance to the 2d Brigade’s attack had seemed insignificant. Lugos Mission and its beaches were left practically undefended, and the positions along Number One and Number Three Roads, which might easily have been reinforced, were lightly held. Although Japanese in bunkers at the airdrome had carried on successful delay- ing actions, some of the positions were not even manned. However, there was reason to believe that Lorengau might be the spot chosen for a determined stand. On the body of a dead Jap- anese officer killed in the fight for the airdrome, maps were discovered showing the location of all defensive installations on Manus. These maps indicated that the enemy defenses were more extensive than previously estimated. Not only Lorengau but the road leading inland to Rossum was heavily fortified. If a substantial number of Japanese chose to fight it out in this area, they would have good positions for defense. Lorengau: An Easy Conquest Well-defended Lorengau is sheltered in a valley surrounded by jungle-covered hills rising to 400 feet. The Japanese had evidently considered that attack would come from the sea, for many of their defenses were sited to cover the shoreline. The offshore approaches and the three roads entering the settlement from the east, south, and north were also covered by defensive positions. From the position held by the 2d Brigade, the only approach to the town was over a sand bar at the mouth of the Lorengau River, a slow-moving stream 97 20 yards wide and 10 to 20 feet deep, except at the sand bar. The beach area, which our troops would then have to cross, was heavily protected. Naval demolition mines, containing 20 pounds of black powder and detonated by a master switch located in a control bunker up the hillside, covered the beach. About 100 yards inland from the beach a series of rifle pits and gun emplacements were constructed on the side of a dike which paralleled the shore. On the hills surround- ing the harbor, beach, and town about a dozen heavy bunkers were used to reinforce the fires of beach defenses and to add depth to the position. As usual these bunkers were carefully camouflaged, although the concentration of naval and artillery fire as well as repeated bombings had disclosed the location of several. Favoring the attacker, however, was the observation over the whole area afforded by the position of the 2d Squadron, 8th Cavalry on the ridge northwest of the town. A certain amount of cover was furnished by the rolling ground, and there were many excellent mortar sites from which both the mortar platoons of Troop H and the Regimental Weapons Troop could furnish close support. The range to the furthermost enemy position did not exceed 1,800 yards. THE SAND BMW/ the Lorengau River mouth allowed an easy crossing. 98 On the morning of 18 March the 2d Squadron, 8th Cavalry would move down from its ridge positions to the mouth of the broad river, cross over the sand bar, and attack Lorengau under cover of mortar and artillery fire. The Regimental Reconnaissance Platoon was to lead out, followed by the 2d Squadron and the Headquarters Troop marching in column of troops. At noon Navy minesweepers would come in to remove obstacles in Lorengau Harbor, so that supplies could be quickly brought over water to back up the attack. The 7th Cavalry, which would not take part in the attack, was given the responsibility of covering the naval operations by neutralizing any fire received from the high ground south of Blue Beach. Preceding the attack on 18 March and continuing through the initial phases, artillery and 81-mm mortar fires were directed at the whole enemy position. Since no personnel were visible, the automatic weapons did not join in the preparation except for harassing fire on the slits of those bunkers which could be seen. At 1000 the attack began. The Reconnaissance Platoon led out in single file through the heavy grass and crossed the sand bar in the river mouth to the opposite shore, followed by Troops E, F, and G. This maneuver 99 appeared to present a perfect target for the enemy, yet due either to dead spaces or the enemy’s disinclination to disclose the positions of his guns until the attackers were within the shortest range, the advance progressed almost without interruption. Scattering fire from several light machine guns was the only enemy resistance. The Recon- naissance Platoon’s mission was to clear the rifle pits, which had been badly battered by the initial fire preparation. After neutralizing the pits the platoon was to take a position astride the track leading into the settlement from the east. Advancing in a vigorous assault along the beach, the platoon drove the defenders off the beach posi- tions and killed those who vainly attempted to stay the advance by remaining at their guns. In clearing the lower beaches, the platoon discovered and cut off the master cable connecting the beach mines with their detonator. Later on, a single Japanese, dead from con- cussion, was discovered in a small bunker overlooking the town, his hand still clutching the exploder. While the Reconnaissance Platoon was clearing the shore area, the 2d Squadron troops made their way across the river. Occasionally a man would be hit and sink under the waist-deep water to be hauled ashore by his comrades, but the crossing continued in good order. Upon reaching the eastern shore the troops deployed and reorganized to continue the advance. Troop E was ordered to drive a wedge into the enemy center, while Troop F came up on its flank echeloned to the left rear. As Troop G gained the enemy bank, it directed its attack against the high ground on the right. Once more the supporting weapons facilitated the advance, and as the troops drew near the safety limits of the artillery and heavy mortars, their automatic weapons and 60-mm mortars took up the Are. With complete confidence in their comrades manning the guns and mortars, little groups led by officers and noncoms assaulted bunker after bunker. Scarcely waiting for the mortar smoke to clear, the troopers were on the position with grenades and tommy guns before the bewildered occupants could man their guns. The entire defenses of the town were swiftly overrun and 87 of the en- trenched enemy were killed. Comparison of the casualties marks the effectiveness of the 2d Squadron’s attack, in which not one man was killed and only seven soldiers were wounded. The weakness of the enemy resistance at Lorengau indicated that the earlier estimates of the Japanese strength on Manus were 100 too high. Instead of a garrison of over 2,000, it now appeared that not more than 500 troops remained to offer resistance to the 2d Brigade. In the afternoon the entire 8th Cavalry Regiment occupied the Lorengau area, while Troop E was sent out to secure the road to Rossum. Patrols were also sent east to Salesia Plantation. While one squad advanced along the Salesia road, which followed the shore line, it was paralleled by another squad in a buffalo. The fire support furnished by the vehicle probably caused the Japanese to vacate a series of seven bunkers and leave behind an antitank rifle in perfect operating condition as well as a 20-mm machine gun. These bunkers were destroyed by captured naval mines used for demolition charges. The 7th Cavalry remained on the ridge overlooking Lorengau, where they had established a perimeter the previous night. Supplies soon started arriving from Lugos Mission, as the engineers had been at work clearing the Number Three Road for several days. Given a rush order to complete the road on the afternoon of the 17th, three D-7’s worked all night to clear the road and demolish pillboxes. Although the road was ready for traffic by 1100 on the 18th, it was FLOATING BRIDGE across the Lorengau River is laid by Troop C, 8th Engineers. Pneumatic floats (6-ton) placed next to the embankments and half -boats in the center formed the base. Treadways laid across these made a bridge with an 8-ton capacity. Completed on 21 March. MAP NO. 13 102 necessary to station bulldozers at points which frequently became impassable after heavy traffic. The engineers removed great quantities of mines encountered in the eastern drome area and in the road south of the drome to the river. The bangalore type did not prove highly effective; a tank which hit one suffered only a damaged tread. Anti-invasion mines on the Lorengau beach were all removed by the afternoon of 18 March. Locating the Enemy Strength Since the easy conquest of Lorengau, the estimate of enemy strength had been considerably scaled down. The next task was to determine how the remaining troops were disposed. According to the captured maps, strong positions guarded the road to Rossum, so it was thought that most of the Japanese force would be found there. To test the situation, patrols were sent out on 19 March along Number Two Road as well as on the other roads remaining in enemy hands. Troop B of the 8th Cavalry was directed to reconnoiter along the coastal track as far as McElroy Plantation; a patrol from Troop A was sent along Number Two Road which led through Old Rossun to Rossum (Map No. 13, opposite). Patrols from both regiments were sent inland along Number One Road. The patrols along Number One Road made no contact with the enemy; one went as far inland as the trail junction just north of Tingo. Troop B also had good luck on its mission. Except for scat- tered sniper fire, the troop met no opposition as it moved into the McElroy Plantation accompanied by light and medium tanks. Facing inland, the troop dug in along the track at its entry to McElroy Plantation and spent an uneventful night. The Rossum patrol was not so fortunate. When 1st Lt. James M. Concannon was shown the enemy map of the area his patrol was to reconnoiter, he exclaimed, “Well, here goes Concannon to get himself a Purple Heart!” Lt. Concannon’s platoon was ordered to go along Number Two Road which went to Rossum and from there to Kelaua Harbor, through the heart of the Japanese defenses indicated on the captured maps. Hardly had the platoon crossed the outpost line still held by Troop E along Number Two Road when it came under fire from a pair of bunkers sited to cover the track. Lieutenant Concannon and 103 one of his men were wounded. The platoon sergeant assumed command and attempted to flank the position. When the attempt was not successful, the sergeant decided to withdraw and report, as his mission was one of reconnaissance. Upon receiving the infor- mation that the enemy was in position not 200 yards from the outpost line, Colonel Bradley directed Major Shore to send Troop A back up the track to neutralize the enemy position. The troop was rein- forced with machine guns and two 37-mm antitank guns. The leading elements of Troop A encountered sniper fire as soon as they passed the outpost line. Machine-gun and small-arms fire was sprayed on the tree-top area and artillery fire was adjusted on it. Five Japanese were killed. The troop resumed its approach march until it reached the bend in the track where the enemy bunkers had stopped Lieutenant Concannon’s platoon. Three or four machine guns in bunkers, and riflemen in trees and fox holes, proved an effective barrier to the troop’s advance. Captain Jennings called for his 60-mm mortars to go into action, under cover of machine-gun fire. Antitank guns were also laboriously dragged by hand up toward the position. In overcoming the position, the action of the machine gunners was outstanding. At the first fire from the bunkers, Cpls. Peter J. Armstrong and Armando V. Valencia set their heavy machine gun in position at a range of 30 yards from one of the bunkers. Having disclosed his position by laying a continuous burst of fire on the bunker, Corporal Armstrong was wounded by sniper fire. Armstrong continued firing until a grenade knocked him unconscious. Although the gun had been hit in several places and was leaking a steady stream from its water jacket, Corporal Valencia took over and continued firing at the bunker slits until the machine gun froze; then he fired with his submachine gun until it was knocked out of his hands by an enemy bullet. In the meantime the mortars had obtained four direct hits on the bunkers, and the antitank guns had made two. Corporals Armstrong and Valencia were later awarded the Distinguished Service Cross, and for its action during the day Troop A was given a unit citation. After neutralizing the stubborn position, Troop A moved along the road slowly. The day was clear and hot — 86.6' F. with humidity 82 — and there were many cases of heat prostration. Enemy snipers continued to harass the troopers, and at a point about 500 yards 104 beyond the outpost line Troop A halted. Troop C came up to relieve the weary men, who returned to Lorengau. Shortly after Troop C had begun organizing the position, a counterattack was launched by eight Japanese using rifle grenades and small arms. Considering the size of the attacking force, the casualties of the troop were heavy: 2 killed and 10 wounded. All the Japanese were killed, largely due to a quick adjustment by the forward observer of the 61st Field Artillery Battalion followed by four volleys fired for effect. Despite the fact that dusk was approaching when the counter- attack was broken, it was decided to attempt an evacuation of the wounded. Violating the iron-bound rule of not moving after dark taught an old, bitter lesson. The litter squads were fired on in the dark and four of the bearers were killed, six wounded. During the night Troop C was continually harassed by sniper and machine-gun fire, although no attacks were made on its positions. The artillery had adjusted a tight box barrage around the perimeter, but it was not found necessary to call for artillery during the night. The day’s action had confirmed the captured map’s description of the Japanese disposition and defenses. It was clear that whatever enemy strength was left on Manus was concentrated down the Rossum Road. However, extensive areas to the east had to be cleared in order to prepare for the eventual linking with the 1st Brigade, as intelligence summaries had reported that the Japanese formerly opposing the 1st Brigade on Los Negros were withdrawing across Loniu Passage to Manus. Accordingly, it was decided at a conference on the evening of the 19th that the 8th Cavalry would have the mission of clearing the eastern end of Manus while the 7th Cavalry would get the tough Rossum Road assignment, as well as the western area. This would give the 8th Cavalry a comparative rest after its strenuous action since the Lugos Mission landing. Patrols to western Manus were to set out on 20 March. A re- inforced platoon from Troop B, on an LCM, was to reconnoiter the northern coastline as far as Bundralis Mission (Map No. 3, pages 8-9). Number One Road as far as Drano was the area assigned to Troop A for a 4-day reconnaissance. Both patrols heading west, inland and along the coast, were successful. Troop F, however, the first 7th Cavalry unit to have a try at the Rossum Road, ran into difficulties which were not altogether unexpected. 681583 °— 46— 8 105 Heavy Going to Rossutn Number Two Road, situated along a spur of the island’s main east-west ridge, is an easy route to defend. The track is a narrow corridor, flanked on both sides by precipitous banks which descend into deep, jungle-covered ravines. The first unsuccessful attempts to progress along the track indicated that the enemy fully recognized the defensive advantages of the terrain. Snipers found cover every- where and well-constructed pillboxes, camouflaged and placed on commanding ground, contained automatic weapons carefully sited to deny the road to the attackers. Therefore, the problem of going through this territory would always present two difficult alternatives. Flanking the road through the jungle would be long, arduous, and strength-sapping, while a frontal attack would be costly because of the small number of troops that could be brought along the road to bear on the enemy positions. Troop F’s mission for 20 March was to move along the road south to the Rossum — Tauwo — Kelaua Ffarbor area, reconnoiter for supply dumps and enemy positions, and return in three or four days. In the morning Troop F, with one light tank and one bulldozer, passed through the positions of Troop C, 8th Cavalry, which was ordered to secure the rear of Troop F’s advance. Small patrols of the advancing party, operating to the front and flanks, uncovered numerous shacks and supply dumps. At 1152 the patrol met enemy pillboxes at a position about 800 yards down the road, where two troopers were wounded. From this point on, the advance was difficult. At 1250 Troop F leached the north edge of Old Rossun where it encountered strong resistance from mortars and machine guns in bunkers, situated both to the left of the road and in front. The position looked so strong that Capt. William C. Frey, the troop commander, asked the regi- mental headquarters whether he should attempt to push through. He was told to use the light tank and knock out the opposition. The troop destroyed two bunkers, but the enemy resistance showed no signs of weakening. Patrols sent to the flanks ascertained that ap- proximately 150 Japanese, in the strong position extending across the road, blocked Troop F’s advance. At about 1600 Captain Frey told the regimental commander, “It’s too hot for me here,” and asked for permission to return 106 to Lorengau. Captain Frey and 10 other men were wounded and 5 had been killed. Under the protection of Troop E, sent forward along the road, Troop F withdrew to Lorengau. Withdrawal of the vehicles along the narrow track surrounded by jungle was extremely difficult; they included, in addition to the NUMBER TWO ROAD runs inland from Lorengau to Rossum through heavy jungle which gave excellent concealment for bunkers and snipers. tank and D-7 dozer, a TD-9 (International tractor) towing a 1-ton trailer used for hauling supplies, and a radio jeep for communication. Before the tank could turn around, it had to back up several hundred yards. In retracing their route, both the tank and TD-9 hit mines, which blew tracks on each. Either the Japanese had come in behind the troops and mined the road, or the vehicles had failed to pass over the mines the first time and had followed a different track on their return. Under enemy fire, the dozer operators hooked on to the tank and towed it back to Lorengau. The other vehicles were abandoned after the jeep’s radio was removed. When Troop F had drawn inside the defense perimeter around Lorengau for the night, Troop G of the 8th Cavalry was placed astride the Rossum Road about 800 yards north of Old Rossun to hold the territory already gained in this direction. During the night artillery placed harassing fire on the Japanese positions in front of Old Rossun which had forced Troop F to withdraw. It was now apparent that more extensive preparations and plans were required to carry on successful operations against the enemy blocking the road at Old Rossun. First, several patrols would move out to look for trails that could be used to flank the enemy holding the road. Other patrols would fan out to locate further enemy positions and prepare for future moves inland. After these prelimi- naries, which would take place on 21 March, an entire squadron, guided by natives from Angau, would attack the stubborn enemy positions in the area of Old Rossun. In the morning of 21 March, Troop E took over the perimeter held by Troop G, 8th Cavalry and immediately pushed patrols out 600 yards toward Old Rossun. Patrolling throughout the day gained little new information. Some supply dumps were located, but there were no Japanese anywhere except along the Rossum Road. No trails for flanking the road positions could be discovered. The area north of Old Rossun was bombed by a scheduled strike of P— 40’s at 1800, after the patrols were all out of the area. Smoke shells fired from mortars and adjusted by an artillery observation plane guided the planes to the target, obscured in the jungle. At 0820 on the clear morning of 22 March the 1st Squadron (less Troop A) moved down the Rossum Road. An artillery and mortar barrage pieceded the attack, and a rocket concentration was fired on the enemy’s rear area. A medium tank accompanied the 108 squadron to clear the way for the troops, but after the D— 7 dozer accompanying it had a track damaged by a mine, the progress of the tank was slow and cautious, since it had to be preceded by a mine- detector squad. To give this squad protection, the tank had to stop often and spray the trees and area in front with machine-gun fire. At 0905, the squadron called for mortar fire to aid the tank in clearing out snipers and grenade-throwers encountered about 400 yards north of Old Rossun. Troop C was withdrawn slightly to per- mit a concentration of 75-mm shells from the howitzers at But jo Luo to be placed on three bunkers holding up the advance. Bazookas were brought forward to fire against bunker ports. After overcoming this position, Troop C again met determined resistance only 50 yards further on from snipers, grenade-throwers, and machine guns. It was sent out to try to work around this position from the east, while Troop B was brought forward on the road. The second center of enemy resistance was impossible for Troop B to overcome without support, so all troops withdrew to allow both artillery battalions to concentrate on that point. With adjustments made by air observation, the 99th and 61st Field Artillery Battalions fired time fuze to silence snipers, and delayed fuze to penetrate bunkers. Troops C and B advanced abreast, this time receiving no return fire. They destroyed five pillboxes and by 1850 reached the northern edge of Old Rossun, where they formed a perimeter for the night. In an advance of only a few hundred yards, the squadron had suffered a loss of 11 killed and 29 wounded, and had not yet wiped out enemy resistance along Number Two Road. Having spent a comparatively quiet night, harassed only by a few snipers, the 1st Squadron set out at 0730 to again try to destroy the enemy blocking the way to Old Rossun. This time two troops ad- vanced along the ridges on each side of the road to avoid another costly frontal attack along the route covered by the enemy’s auto- matic weapons. Troop B moved forward on the right and Troop G on the left. Troop C remained in reserve. A light tank with its accom- panying bulldozer moved along the road. During the night the enemy had evidently put more mines in the road. Before moving along very far, the bulldozer hit a mine which blew a track. The enemy had also effectively blocked the road by setting fire to the TD-9 and trailer abandoned by our forces on 19 March. The armored attack bogged down and the cavalrymen made 109 little progress. At 1150 artillery and mortar concentrations were placed on the enemy positions at Old Rossun, and Troops G and B pushed forward again, slowly cutting their way through the heavy growth. An advance patrol of Troop G moved 700 yards southeast and found 5 unoccupied bunkers. By the afternoon the troops had reached Old Rossun and were still moving slowly forward. At 1342 both units were pinned down by heavy sniper and mortar fire. When mortar fire was placed on enemy positions, Troop G was able to push ahead and by 1800 the entire squadron had pushed 200 yards south of Old Rossun on a 200-yard front. In the hard, slow advance the 1st Squadron had destroyed 21 bunkers and suffered a loss of 1 killed, 27 wounded. IN WELL-CAMOUFLAGED POSITIONS like this, the Japanese would hold their fire until the foot troops were almost on top of the bunker. This procedure made the placing of 81-mm mortar and artillery fire on the bunkers extremely difficult. The Japanese evidently planned alternate positions, as many more bunkers were constructed than were occupied. 110 Even though the squadron was now past the defenses at Old Rossun, it had taken 2 days to move less than 1,000 yards, and the gain had cost 68 casualties. It appeared that the same kind of fighting would have to continue, for ahead of the exhausted troops lay more bunkers, manned by Japanese ready to use their excellent positions to make our advance as costly as possible. The sixth consecutive day of fighting along the Rossum Road began with a heavy artillery concentration, as well as rocket and mortar fire, although a requested air strike was cancelled. Another attempt was also made to find a possible route for flanking the enemy road positions south of Old Rossun, by a 14-man officer patrol sent out for this purpose. They were to follow the Lorengau River for about a mile and then to turn left. While the flanking possibility was being explored, the troops along the road pushed forward at 1142. Troop E, brought up from reserve, was to envelop enemy positions on his left flank, while Troops B, C, D, and G advanced on a 300-yard front with G and B astride the road as in the previous day’s action. In spite of the heavy artillery preparation, the attack on 24 March fared no better than on previous days. One difficulty was that the hundreds of 75-mm shells poured onto the positions did not have a sufficient penetrating effect to reach the entrenched bunkers through the heavy jungle growth. Even though delayed-action fuzes were used, reducing tree bursts to 10 percent, many of the bunkers remained untouched. As soon as the cavalrymen began their advance, Japanese came out of their holes, quickly set up their automatic weapons with supporting snipers, and poured fire upon our troops. With no visibility through the heavy forest, the troops advancing blindly under fire needed strong leadership to keep them moving forward. Troop B received the brunt of enemy fire and its commander, Capt. Roman D. Hubbell was killed. A medium tank was employed to try to force the enemy position. However, the tank commander refused to move forward until pre- ceding troops had probed for mines. This procedure drew enemy fire, which was extremely costly to the men near the tank. In addition, there were no means of communicating with the tank commander except through the pistol port. In the attempt to improve communica- tions, the platoon leader accompanied the tank on foot until he was wounded by a mortar shell and had to be evacuated. The 7th Cavalry thought that the tank was more of a hindrance than a help. Ill At 1250, Troop E, moving around the west flank, was pinned down by heavy enemy mortar fire and the entire attack bogged down. The squadron had progressed about 250 yards south of the night’s perimeter, and Lt. Col. John B. Maxwell, II, regimental executive officer, who had been placed in charge of the squadron on 23 March, reported to brigade headquarters that he would be unable to com- plete his mission without air support. Lt. Col. Kenneth L. Johnson, commanding officer of the 99th Field Artillery Battalion, also re- ported the disappointing results of his 75-mm howitzers. The 1st Squadron had suffered 4 killed and 20 wounded and Colonel Maxwell believed he could no longer hold in his present location. He informed the brigade commander that he would have to withdraw to the previous night’s perimeter, but the squadron held until after 1600, marking its forward lines with smoke pots to direct the requested air strike. Then the troops pulled back to their previous positions just south of Old Rossun. At 1730 strafing attacks were made and twelve 500-pound bombs dropped. In the late after- noon, the 105-mm guns of the 61st Field Artillery Battalion began registering on the Japanese bunkers and interdicted the road where the 1st Squadron had been stopped. The 8th Cavalry Breaks the Enemy Hold Attempting to overcome the enemy’s delaying action along Number Two Road was assuming the proportions of a major action. The 1st Squadron, 7th Cavalry had suffered severe casualties in its repeated attacks, and the front line had been moved less than 2,000 yards from Lorengau in 6 days of fighting. Troop B had lost two troop commanders, and 87 percent of its officers were casualties. Fresh troops were needed for a renewed assault. The main force of the remaining enemy defenders was well dug in on high ground. South of Old Rossun, this ridge was occupied by approximately a company, with tree snipers covering the track as well as all possible lines of approach to the bunkers. A wide flanking movement from the west seemed to be an impossibility, as the officer patrol that had been sent along the Lorengau River to find a route had been unable to make its way through the heavy jungle east of the river. The same line of advance, close to the road, would have to be followed as in the earlier attacks, but heavy artillery and air 112 105-MM HOWITZERS support the 2d Brigade on Manus. During the initial stages, howitzers of the 61st Field Artillery Battalion fired from positions on Hauwei. To give better support to the attac\ along Number Two Road, the battalion moved to the Lorengau air strip on 26 March. preparation as well as tank and mortar support were expected to facilitate the cavalry advance. The 1st Squadron, 8th Cavalry was ordered from Salesia Planta- tion to make the attack. In the early morning these troops would move by boat and buffalo to Lorengau. At 0800 the P^lO’s were to arrive and strafe all the road positions south of Old Rossun. The artillery would follow with a two-battalion concentration for a half hour. Under cover of this fire the 1st Squadron would leave the line of departure in box formation on either side of the road, with Troop A 113 leading the left column and Troop C the right. Troop R was to follow Troop C along the right side of the road and envelop the hill objective from that flank. Promptly at 0800 the first flight of RAAF P-40’s appeared over the target. Twelve planes, each flying two missions, dropped twenty- four 500-pound bombs and strafed the area with thousands of rounds of .50-caliber ammunition. As the last planes flew over, the artillery concentrations by the two supporting battalions came down and the 1st Squadron began its approach march, passing through 7th Cavalry troops some 500 yards behind their bivouac positions, from which they had withdrawn in order to clear for the air strike. The squadron’s progress was slowed by the jungle growth and was further impeded by continuous sniper fire, even though the troops were well dispersed to protect against snipers. Dispersion made con- trol over the troops difficult and the cavalry had to keep their pace regulated to the speed of the tanks, which inched along behind mine detector squads. As a result, the cavalry failed to cover the distance to the objective under cover of artillery preparation. At 1335 the squadron was reaching the objective area, having knocked out two bunkers and four machine guns. Troop B dis- covered that a precipitous cliff and a swamp prevented an envelop- ment on the right of the hill, so it took up a position in support of Troop C, which was on the crest of the objective ridge by 1440, await- ing a counterattack. Troop A on the left waited on the forward slope of the objective ridge for tank and mortar support, as it was re- ceiving mortar fire from the reverse slope. Troop A had suffered 24 casualties from knee mortars and from bunkers still intact. One tank, damaged by a mine, was pulled out of the way by a bulldozer to allow the two others to go after the bunkers, plainly visible on the ridge, which were holding up Troop A’s advance. A 105-mm howitzer concentration was placed on these positions and at about 1600 Troop A moved up the ridge, its right flank protected by Troop C upon the ridge. Fire from the twelve 81-mm mortars, com- manded by Capt. Walter M. Hart, who established his OP well ahead of the combat elements, gave excellent support. Captain Hart adjusted his mortars so that 160 rounds of HE, heavy, with delay fuze was delivered on the exact crest of the hill into an area not larger than 50 by 100 yards. Under this cover, and using their own machine guns against the firing slits, the troops advanced within grenade- 114 throwing range. Both the medium and light tanks made their way around the disabled tank and directed their machine guns and larger weapons against Troop A’s objective. When the foot troops were on top of the positions, the tanks, assisted by flame throwers, went forward to finish off the bunkers. Bulldozers followed them up by filling in all bunkers, trenches, and fox holes. Japanese trying to escape were cut down by rifle fire, and no attempt was made to explore the interior of bunkers as it was felt that investigating troops would be endangered. By 1700 all enemy resistance had ceased. The 1st Squadron dug in on top of the ridge, which gave excellent observation over the road to the south. There was no sign of further enemy positions to the front. The day’s successful action had cost the 8th Cavalry 7 killed and 29 wounded, but almost 100 of the enemy had been destroyed. The last concentrated resistance on the Rossum Road was broken. The difficulties of destroying an enemy force well entrenched in jungle terrain had been fully demonstrated on the Rossum Road. The 1st Cavalry had gained valuable experience, particularly in the methods for using supporting weapons and air coordination. The last successful fight had involved maximum use of supporting fires to aid the infantry, and the results had brought out both limitations and advantages. The air strikes, artillery, bazookas, flame throwers, and tanks did not succeed in wiping out the enemy positions to the extent desired ; nevertheless, the combined employment of these weap- ons helped the riflemen, covered by the BAR man and closely sup- ported by mortar fire, to get close to the remaining bunkers and finish off the stunned occupants. P-40’s, used as bombers, had not proved as accurate as B-25's. An improvised bombing method forced them to commit themselves to their bombing run while some distance from the target. Although cavalry lines had been withdrawn 500 yards preceding the air strikes on the last enemy positions south of Old Rossun, the bombs were so dispersed that some fell in close proximity to the bomb line. The artillery was little better in putting the bunkers out of commission. After 1,455 rounds of 105-mm artillery fire had been poured on the same positions, an observer could obtain evidence of only 2 bunkers destroyed by it. Nevertheless, the artillery seemed to have accom- plished two of its missions. It prevented the enemy from reentering the bombed area and pinned him to his fortifications. 115 Bazookas were not very effective against bunkers. The training allowance of bazooka rockets had been limited, so the effectiveness of the weapon's use depended on the operator’s individual aptitude. Since most of the bunkers attacked were old and covered with vege- tation, the slits were difficult to locate. When the slits could be seen from a distance of 25 yards, bazooka operators could place a rocket through the slit or right beside it. If the rocket went through the slit it would kill all occupants; if right beside it, the enemy would be stunned and a grenade could be put through the slit. The move- ment of tanks was limited and, according to the observer, Col. Marion Carson, they were not used in the assault until the troops had advanced slightly beyond the enemy position. The terrain offered no suitable positions for artillery observation posts, so the forward observers had control both of the initial con- centration and the fires on enemy positions. Artillery communica- tion was by radio, but sound-powered telephones were generally used between the cavalry commands. Although radio was available, the sound-powered telephone was more reliable and the enemy could not listen in. Conversations could be carried on in a lower voice, and there was no antenna to get tangled in vegetation or betray locations by moving shrubbery. The fight on 25 March represented the last large, coordinated action on Manus. Only remnants of the enemy remained, to be dealt with by units of troop size or less. The Rossum Road action had proved that jungle fighting, even against a numerically inferior and disorganized enemy, could be more difficult and dangerous than fight- ing larger numbers in open country. Both regiments together had lost a total of 36 killed and 128 wounded in the week of fighting to break the block on Rossum Road. Because bunkers were destroyed without searching and the jungle made an exact count impossible, the number of Japanese killed could not be determined, but prob- ably no more than 200 Japanese had held up our units of squadron size. If the small groups of Japanese left on the island decided to give as determined resistance in well-chosen jungle positions, they could perhaps exact the same expensive ratio of casualties from our large and victorious force. Mopping up Japanese in the jungle was not an operation to be taken lightly, but was a necessary final meas- ure, if the projected naval and air installations were to be safe from snipers and small group raids. 116 Round-up On Los Negros After 7 March, while the 2d Brigade was undertaking the conquest of Manus Island, the fighting on Los Negros continued on a reduced scale, in attempts to wipe out remaining enemy pockets of any strength and in patrol clashes with small isolated groups. The perimeters at Momote and Salami Beach began to be extended as patrols found that the surrounding jungle was free of Japanese. By 11 March the daily patrols had determined that the entire area from Mokerang Peninsula to Southeast Point and as far west as Porlaka was generally free of the enemy. The 1st Squadron, 5th Cavalry pushed its perimeter at Momote west and south to the water barriers of Lemondrol Creek and Ihon Lagoon. From the 12th Cavalry beachhead at Salami, continuous patrols to the interior encountered so few Japanese that it was clear the enemy lacked sufficient num- bers to harass our troops on the peninsula. West of Lemondrol Creek it was a different story. From the beachheads at Papitalai, Papitalai Mission, and Lombrum (Map No. 8, page 62), the patrols had run into strong enemy resistance. Hills inland from these perimeters contained enemy pockets which would require the efforts of groups larger than patrols to wipe out. As in the Rossum Road action, the enemy here seemed determined to hold out and inflict as many casualties as possible on our superior force. The hilly jungle terrain lent itself to ambush and sniper action, as well as dug-in defense, and the Japanese utilized all these possibili- ties. A few efforts at infiltration through our perimeters were at- tempted, but the final stages of the battle for Los Negros would be chiefly a job of forcing the enemy from his excellent holding-out po- sitions in the hills and jungles of the interior. 117 Attack on Hill 260 While the troops holding the beachheads from Lombrum to Papitalai waited for supply channels to be established, they probed to locate the center of enemy resistance. The 12th Cavalry was re- sponsible for the beachheads at Papitalai Mission and Lombrum (Map No. 14, below). The 2d Squadron had captured Papitalai Mission and on 11 March the 1st Squadron took over Lombrum Point, relieving units of the 7th Cavalry for use in the attack on Manus Island. The 5th Cavalry held the Papitalai perimeter with Troop F. When that regiment extended its control over the area east of Lemondrol and Porharmenemen Creeks on 11 March, the 2d Squad- ron moved into the Porlaka area while the 1st Squadron on the left flank was concentrated near Southeast Point. It was planned that the 12th Cavalry would do most of its fighting in the northern sector, MAP NO. 14 118 THE HILLS WEST OF LEMONDROL CREEK held strong enemy pockets. Looking westward across Momote airdrome, Lemondrol Creek, can be seen as a barrier betiveen the Momote area and the rugged hill country west of the cree\ • The 1st Brigade held perimeters beyond the cree\ at Papitalai and Papitalai Mission, above right end of Momote strip. moving inland from the coastal positions to cut off the Japanese hammered at by the 5th Cavalry as it pushed west on both sides of Lemondrol Creek. Since 8 March every Troop F patrol from Papitalai had run into enemy resistance in the hilly country to the west. Although small patrols of squad size or slightly larger were held up day after day by resistance about 800 yards west of Papitalai, these first encounters did not reveal the full strength of the enemy pocket. When an at- tack by a platoon of Troop E, reinforced by two tanks, two bazookas, and flame throwers, was repulsed on 11 March, the strength of the enemy strongpoint became clearer. The enemy was evidently well dug in along the ridge, marked by a series of knobs, running west from Papitalai. The highest point, Hill 260, about 2,500 yards from Papitalai, was probably the center of defense. However, on two knobs 119 cast of 260, well-defended outposts prevented any movement farther into the interior. The position 800 yards from Papitalai which had repulsed Troop E was defended by a platoon with knee mortars and at least 3 machine guns. The total enemy strength facing the 5th Cavalry was estimated at 150 to 200 Japanese. Artillery and mortar concentrations preceded the 11 March attack, but the advancing cavalry did not have effective tank sup- port. The approaches to the first hill west of Papitalai were impos- sible for the tanks. A path to the crest runs along a narrow hog-back that only two men can travel abreast. The tanks had to try the bor- dering jungle, where they were soon stopped by the sloping ground, the heavy growth, and a sticky clay which clogged the tracks so that they would not grip. Leaving a few riflemen with the tanks, the pla- toon continued to advance and reached the top of the hill, only to be pinned down by rifle and machine-gun fire from the reverse slope. Neither the flame throwers nor bazookas had kept pace, but these weapons would have been of little value in the thick rain for- est which reduced visibility to less than 10 feet. Fire from hidden Japanese covered the trail from all directions and the platoon was forced to withdraw to Papitalai with a loss of three wounded. To clear the way for the 5th Cavalry’s push west, a squadron would have to neutralize the enemy defending along the ridge toward Hill 260. The plans called for a main effort by the 1st Squadron, 5th Cavalry, which was to move up from its position at Southeast Point, cross Lemondrol Creek and advance west from Papitalai on the morning of 14 March. Diversionary efforts would be achieved by sending combat patrols out on the right and left flanks of the 1st Squadron’s attack. The 12th Cavalry would patrol in the north and the 2d Squadron, 5th Cavalry, back at Momote airdrome, would send patrols to investigate the situation south of Lemondrol Creek. Four objective lines were drawn up for the 1st Squadron. The second included the outpost hill where Troop E had been stopped, the third, the next knob west, and the fourth, Hill 260. The squadron was ordered to continue attacking until 1500, when they would stop, hold all ground gained, and prepare night positions. At 1030, after a concentration of 75-mm howitzer fire by the 82d Field Artillery Battalion and a regimental mortar concentration, the 1st Squadron jumped off. Troops A and B reached the first objec- tive without opposition at 1120, where they were ordered to remain 120 until 1200 while fighter-bombers, scheduled to strike at 1245, bombed and strafed the further objectives. However, after four bombs had been dropped on the outpost hill, the strike was called off and at 1300 Troop A, supported by artillery and mortar concentrations and using Troop B as a base of maneuver, took up the attack. This time fire from knee mortars and machine guns as well as from snipers was encountered. The front-line cavalry fired red rock- ets, and the artillery, adjusted by forward and air observers, placed fire 50 to 150 yards in front of the advance elements. Two platoons of Troop A were able to move slowly forward in a double envelop- ment of the hill. A platoon of Troop B also moved forward, as the balance of Troop A could not go through them. Several pillboxes were encountered and teams were organized to knock them out. On the left flank, open brush country allowed a bazooka to knock out a machine-gun nest at a range of 75 yards, killing 8 Japanese. Where a L— 3'S WERE USED by our air observers to direct artillery fire. Here they are protected behind an embanfment built for that purpose while Seabees wor\ on the Momote air strip with their heavy equipment. good line of sight could be obtained, bazookas were used on pill- boxes; otherwise riflemen and machine gunners worked on them. The attack on the pillboxes was hampered by enemy mortar fire, soon silenced by our countermortar fire. At 1440 the 3 advance platoons had captured the hill on the second objective, having lost 2 killed and 20 wounded. Although the outpost hill was in our possession, the remaining Japanese evidently intended to make a firm stand farther along the ridge. At 1555 the enemy counterattacked Troop A and was repulsed by artillery and mortar fire. From 1930 until midnight the artillery and mortars fired alternately every half hour on the next hill and the trails leading from it. Nevertheless, the enemy attacked again at 2100. This time the cavalrymen broke up the attack with their rifles and machine guns, with support by mortars and artillery. At 0730 on the 15th, Troop A moved forward again. After ar- tillery and mortar concentrations, the troop advanced to the third objective without any resistance. Troop A dug in there and Troop B sent out patrols 200 yards to the front. Still no opposition was en- countered. Difficulties of supplying the troops over an extended supply line which consisted of l l / 2 miles of narrow, rutted, and slippery trail prevented further advance. Troop C, aided by a section furnished by the 82d Field Artillery Battalion, took 5 hours for a round trip. The 1st Squadron’s last objective was the largest knob, Hill 260, on which it was now estimated were 100 well-entrenched Japanese. By 17 March sufficient supplies had been brought up to enable Troop C, which had relieved Troop A, to push on toward this knob. After the usual artillery and mortar preparation, Troop C, protected in the rear by Troop B which was dug in on the third objective, advanced to within 50 yards of the hill crest before being stopped by machine-gun and rifle fire. Squadron commander Lt. Col. Charles E. Brady then dispatched Troop B north to envelop the enemy from his left flank. Although Troop B had to cut its way laboriously and noisily through the jungle, the envelopment was highly successful. The Japanese put up little resistance and both troops moved onto the hill and secured it by 1310. About 40 or 50 dead Japanese were counted, although the total, which was impossible to determine in the jungle, was undoubtedly much higher. The 1st Squadron’s losses in the day’s attack were four killed and seven wounded. 122 Patrolling by the 12th Cavalry In its attack on Hill 260, the 5th Cavalry had hoped for strong diversionary attacks to the north by the 12th Cavalry. On 13 March, however, the 12th Cavalry was placed in division reserve to be ready to support the 2d Brigade’s attack on Lugos Mission, and was per- mitted to do only a limited amount of patrolling. Until the regiment was released from its alert, activities in the Papitalai-Lombrum area would be confined to small patrolling missions. These would endeavor to determine enemy strength and to locate trails along which Japanese might try escape to the west. It was also hoped that these patrols might give the enemy the impression that our forces were moving westward on Los Negros and thus divert Japanese attention away from Manus. Patrol reports indicated by 14 March that not more than 200 Japanese would be encountered in the hilly region southwest of Papitalai Mission and Lombrum (Map No. 15, page 124). While the 5th Cavalry began its attack on the central ridge where a portion of these Japanese defended, 12th Cavalry patrols clashed with enemy groups in the hills to the north. On 14 March, D Day for the attack on Hill 260 and the day before the invasion of Manus, the 2d Squad- ron at Papitalai Mission sent Troop G on a patrol along the coast toward Lombrum, with the intention of indicating to the enemy that our forces were moving west. At Chaporowan Point Troop G was to contact a 1st Squadron patrol, but swampy terrain delayed this group; instead, Troop G ran into Japanese with knee mortars, machine guns, and grenades about 500 yards east of the point. An enemy flanking maneuver put Troop G, right on the harbor and with restricted possibilities of movement, in a difficult position. All available machine guns, mortars, and 37-mm’s in the Lombrum perimeter were used to help extricate them and cover their with- drawal. They were finally able to break clear and return to Papitalai Mission with 15 wounded. The troop had lost 2 killed, and it was sub- sequently determined that 31 Japanese had been killed by our con- centrated fire. After the failure to connect the perimeters by moving along the coast, Troop A was sent out on LCM’s to Chaporowan Point on 16 March. Followed by Troop B and a portion of 1st Squadron Head- quarters, Troop A’s landing was unopposed. As these troops dug in for the night, the mortar platoons of Weapons and D Troops in 123 position on Lombrum Point concentrated their fire on positions in- land from Chaporowan Point where patrols had encountered resist- ance. This fire succeeded in protecting the new perimeter from attack. On 17 March, 1st Squadron patrols inland from the Lombrum and Chaporowan perimeters continued to meet resistance. The coco- nut groves of Lombrum and the Mission were comparatively free of the enemy but the jungle hill country southwest of the perimeter, and the coastal areas between the three perimeters, continued to be contested. West of Lombrum there was little opposition. Although their supply situation was poor and morale deteriorating, the Jap- anese had certain advantages against our patrols. The entire country being probed was well known to the Japanese and entirely strange to the cavalrymen. The Japanese were more security conscious than in the early days of our advance and no more maps or diagrams could be found. In addition the enemy themselves were engaged in active 124 patrolling, so that, although this brought them into contact with our troops, their base positions could not be accurately determined. Regions reported clear one day might be occupied a few hours later. The experience of two 1st Squadron patrols is typical of many during this period. S/Sgt. Lindal L. Barrett was sent in the morning of the 17th with a patrol to reconnoiter trails to the south and west of Chaporowan Point. He returned at 1140 without encountering any enemy. S/Sgt. Rex C. Clark had been sent out the same morning with the mission of reconnoitering a creek to the south before turning- west. To accomplish this mission he had to cross, 100 yards south of the perimeter, the same trail that Sergeant Barrett’s patrol had taken. At that point he ran into an ambush, 15 minutes after Sergeant Barrett had passed through without drawing fire. Sergeant Clark’s platoon was forced to withdraw and at 1330 Sergeant Barrett's platoon was sent back into the area by the same track that it had originally taken. This time Barrett’s platoon also ran into the ambush and was com- pletely surrounded by the Japanese. Sergeant Barrett ordered his men to withdraw and, despite the fact that he had been wounded in the arm, stood and fired on visible targets, killing two Japanese and making the withdrawal possible. While the 1st Squadron patrols hunted out the elusive enemy, the 2d Squadron at Papitalai Mission was released from division reserve and was ordered to support the 5th Cavalry attack on Hill 260. On 18 March the 2d Squadron would make a reconnaissance in force against the enemy left flank at a point 200 yards northwest of Hill 260. Although the 5th Cavalry had secured Hill 260 on 17 March, evidently earlier than expected, the 2d Squadron’s attack toward that area was not called off. The enemy was located close to the 5th Cavalry supply trail and his snipers were giving trouble. A platoon of Troop F commanded by 1st Lt. Arthur L. Allen, Jr., preceded the squadron as point. Just as it entered the jungle 600 yards from the perimeter, the platoon received heavy machine-gun and rifle fire from various directions. When two enemy machine guns were spotted, Lieutenant Allen personally placed his light machine guns and riflemen and directed their fire. Lieutenant Allen was killed, but his platoon held its ground, and enabled the remainder of Troop F following down the road to deploy. After a sharp fight it became obvious that a very strong position had been encountered, and the squadron withdrew to the perimeter with 37 casualties. The 125 continual patrol clashes and the failure of the 2d Squadron attack in this area indicated that it would take more than patrols to mop up the enemy in the hills southwest of Papitalai Mission. Attack in the Papitalai Hills After its successful attack on 17 March, the 5th Cavalry held a salient west from Papitalai, as far as Hill 260. West and northwest of their position all advance was strongly contested. 12th Cavalry patrols made little headway in wiping out the determined Japanese resistance groups north of the spearhead, and 5th Cavalry patrols sent west also met resistance. Therefore General Chase ordered a brigade attack for 21 March to complete the occupation of Los Negros. The 12th Cavalry, less two troops, was released from division reserve on 20 March and was ready to take full part in the last big attack on Los Negros. The 1st Squadron, 5th Cavalry would attack north from its sector and meet the 2d Squadron, 12th Cavalry attacking southwest (Map No. 16, pages 128-129). Then, with the area north of Hill 260 cleaned up, the two squadrons would move westward abreast, to attack the high ground still in enemy hands west of Hill 260. Mean- while the enemy would be encircled by the 12th Cavalry forces at Lombrum and Chaporowan, who would move inland to trap the Japanese escaping west under pressure of the other attack. Troop C, 12th Cavalry, plus one platoon of heavy machine guns, would move southwest from Chaporowan Point and then southeast to intersect the north— south trail and take a position on the high ground there. Farther west at Lombrum, the 1st Squadron, 12th Cavalry would send Troop B, reinforced, on a circling maneuver to the high ground slightly west of Troop C’s objective. This move would catch any enemy escaping Troop C’s wedge west of the Papitalai hills pocket. The Reconnaissance Platoon would wipe out a strongpoint just south of Lombrum Bay and then protect Troop B's left flank. Southern Los Negros would be cleared by the 2d Squad- ron, 5th Cavalry, concentrated at Southeast Point, which would move 2(4 miles westward into previously probed territory and make an advance north of Palapi Hill. On 21 March the attacking elements of the 12th and 5th Cavalry regiments pushed out. Troop F, 12th Cavalry met determined 126 resistance and by nightfall reached a point 1,700 yards southwest of Papitalai Mission, where they dug in. Although the 12th Cavalry troops were only about 250 yards from the 5th Cavalry, they were not able to contact each other because of the dense jungle. Troop C of the 5th Cavalry had patrolled 700 yards west of Hill 260 where they ran into fire from machine guns hidden in thick patches of cane. Troop C withdrew to allow artillery and mortar fire to be placed upon the position, then tried again with no better results. From the western positions, the 12th Cavalry attacks fared better. Troop C progressed 1,000 yards south of Chaporowan Point, where it dug in on a hill overlooking a Japanese village reported by earlier patrols and called “Juarez” by the cavalrymen. From Lom- brum Troop B fought its way inland about 2,500 yards through swamps and jungle to a trail on the high ground south of Lombrum and dug in across the trail. The 12th Cavalry troops had accounted for 31 known dead Japanese, demolished defensive installations, and captured quantities of stores in their push inland. The night was not a quiet one for Troop B. A group of Japanese from the village of “Juarez,” evidently pushed out by the pressure of Troop C’s attack, moved west toward Troop B’s perimeter block- ing the east-west trail. Troop B held its fire until the trail was crowded with the enemy, and then opened up. The Japanese dispersed but succeeded in surrounding the troop and attacked the perimeter all through the night. They did not break through the tight defense. In the morning Troop B, aided by a perfectly placed concentra- tion from the 271st Field Artillery Battalion, succeeded in driving through the Japanese surrounding the perimeter and worked its way northeast to join with Troop C. Then, after another artillery preparation on the Japanese village “Juarez,” Troop B occupied the village. Troop A, at Chaporowan Point, then joined the rest of the 1st Squadron at the village. It had been relieved by Troops G and H which were moved from the Mission to Chaporowan Point by boat. The squadrons advancing from the east did not succeed in closing up the gap between their lines and the 12th Cavalry sweep south of Chaporowan Point. The 1st Squadron, 5th Cavalry moved forward at 0900 with Troops C and B abreast. They attacked west to the hill where Troop C had been held up the previous day. Heavy fire again stopped the advance. An encircling movement around the 127 Papilalai Mission THE LAST FIGHT ON LOS NEGROS 20-25 MARCH Planned movements Enemy lost stond 50£^^^I000 YARDS MAP NO. 16 128 129 IMPROVISED FERRY operated by Troop A, 8th Engineer Squadron carried troops and supplies from Momote through dangerous shallows and reefs to the 2d Squadron, 5th Cavalry on the southern coast of Los Negros. enemy position was attempted, but the thick cane and dense under- growth held troops to as little as 200 yards advance in an hour. The resistance was too strong and the 1st Squadron was forced to with- draw to Hill 260. The position which had held up the 5th Cavalry’s advance west for two days was estimated to be held by one rein- forced platoon, well concealed and dug in, defending with three machine guns, knee mortars, and grenades. Continuing its advance southwest from Papitalai Mission the 2d Squadron, 12th Cavalry pressed 800 yards farther inland on 22 March. However, they were still unable to make contact with the 5th Cavalry at nightfall. The situation was made worse because Japanese patrols had cut the 2d Squadron’s supply line. Although they were com- pletely isolated, during the night Troops E and F beat off a deter- mined counterattack on their position, killing 15 Japanese, 2 of whom were officers. Some of the Japanese were wearing American fatigue clothes and helmets. Four Americans were wounded. 130 South of Lemondrol Creek the forces pushing west had met lighter resistance than the squadron attacking in the Papitalai hills. The enemy offered no real fight to the 2d Squadron’s patrols until 16 March when the Japanese entrenched at Puwas, about 3 miles west of Southeast Point, forced back a reinforced troop with machine- gun fire. Artillery and mortar concentrations pounded these positions, and on 23 March the 2d Squadron occupied an area from Puwas to the west side of Palapi Hill with a loss of three killed and three wounded. Since the initiation of the brigade attack on 21 March, the efforts of two regiments had not been notably successful in wiping out the stubborn enemy pockets in the Papitalai hills. Our losses from 21 March to noon on the 23d, were 14 killed and 78 wounded. Although the 2d Squadron, 12th Cavalry had cleared most of the area of the Papitalai Mission peninsula and, together with the 1st Squadron troops blocking the enemy escape west, had accounted for about 100 enemy dead, the general line of advance to the west had been pushed only a few yards from the 5th Cavalry position on Hill 260, gained on 17 March. The primary reason for the slow advance was the terrain. Jungle far denser than that encountered in the coastal areas gave the Jap- anese excellent concealment and hindered the coordination and for- ward movement of our troops. Difficulties of supplying even a slow advance were immeasurable, and often the attack had to stop to wait for supplies. However, the fierce resistance of the enemy holding out in the jungle led our command to revise its estimate of the Japanese remaining in the area. It was now believed that 600 to 800 Japanese defended the area from the Papitalai hills to Loniu Passage. All day of 23 March was devoted primarily to improving the supply route to the forward troops. The isolated 2d Squadron, 12th Cavalry finally made contact with the 5th Cavalry at 1130 and received food and ammunition from them. Two platoons of Troop G, 12th Cavalry were sent from Papitalai Mission to carry supplies to the squadrons over the 5th Cavalry’s supply trail from Papitalai. The wounded were evacuated back over the trail, and supplies of ammu- nition, rations, and water were immediately built up in the forward supply dumps behind the two squadrons. Plans were made for both squadrons to attack abreast on the high ground west of Hill 260 early in the morning of 24 March. 131 With the 12th and 5th Cavalry in contact and holding a continu- ous front, coordination could be achieved. Troops A and C of the 5th Cavalry and E and F of the 12th Cavalry would advance abreast, with the two troops on the outside echeloned to the right and rear. The advance was to be a leapfrog movement. Any platoon held up would cover the advance of supporting platoons. Preceded by a 15- minute artillery preparation, the attack jumped off from Hill 260 at 0830. Progress continued with light resistance until the troops encoun- tered machine guns emplaced in a group of huts several hundred yards to the front. All troops halted and were undecided about what to do. At this juncture Capt. Henry B. Greer, the ranking officer in the front line, quickly met the situation by calling all troop commanders together and quietly giving them directions. He ordered that all heavy machine guns be brought to the center and emplaced to fire from hill positions where they would dominate the village and at the same time be under cover. The machine guns opening up on the village would be the signal for every man to move forward either through the village or around it. The plan worked well. When the order to open fire with the heavy guns was given, every man moved forward, firing from the hip with carbines, M-l’s, BAR’s, and light machine guns. The volume of fire was so great that the movement carried through and beyond the village, scattering and blanketing all enemy fire and killing all the Japanese in the village. At 1400 the men dug in on the high ground (west of Hill 260) where the enemy had delayed the advance for so long. Fifty Japanese had been killed in the attack. Our casualties were 3 killed, 12 wounded. Although neither squadron had more than a day’s supply of rations and ammunition, it was decided to continue the push west on 25 March. At 0830 both squadrons moved west toward their objective, a trail leading south from Chaporowan Point to Folach Passage. Artillery preparation had been extensive and during the advance a total of 2,844 rounds of 105-mm were poured on the objective area, effectively wiping out enemy resistance. The attacking force had killed 33 Japanese, and 20 dead from artillery fire were counted. Inadequate supplies forced the troops to stop at a position 100 yards short of the trail. The terrain over which the troops had advanced was hilly and covered with dense undergrowth. It had no 132 trails and was impassable for vehicles carrying supplies. A new route had to be found and supply bases moved forward. The squadrons maintained their positions throughout the night and were reached in the morning by a troop bringing supplies from the 12th Cavalry at Chaporowan. After the successful attack of 24 March, enemy activity against forward troops was negligible. It began to appear that the high ground won by the 1st Brigade was the last position on Los Negros to be effectively defended. Between 26—31 March the fate of the Japanese defenders became clearer. The survivors still had defensive positions west of the Papitalai hills, but without ammunition or supplies they could not put up any fight against our troops closing in from both east and west. Many elaborate defenses, some of them recently con- structed, were discovered in the area west of Hill 260. One emplace- ment was found that had a shaft 18 feet deep, which tunneled into the side of a hill; similar constructions had shafts 8 to 10 feet deep. However, as early as the night of 23/24 March the enemy’s disastrous ammunition shortage was displayed to our troops when Japanese had thrown sticks and mud clods into the fox holes of the 1st Squad- ron, 12th Cavalry. A diary found on a dead Japanese suggests the fate of all the survivors: 28 March. Last night’s duty was rather quiet except for the occasional mortar and rifle fire that could be heard. According to the conference of the various unit leaders, it has been decided to abandon the present position and withdraw. The preparation for this has been made. However, it seems as though this has been cancelled and we will firmly hold this position. Ah! This is honorable defeat and I suppose we must be proud of the way we have handled ourselves. Only our names will remain, and this is something I don't altogether like. Yes, the lives of those remaining, 300 of us, are now limited to a few days. 30 March. This is the eighth day since we began the withdrawal. We have been wandering around and around the mountain roads because of the enemy. We have not yet arrived at our destination but we have completely exhausted our rations. Our bodies are becoming weaker and weaker, and this hunger is get- ting unbearable. 31 March. Although we are completely out of rations, the march continues. When will we reach Lorengau? Or will this unit be annihilated in the moun- tains? As we go along, we throw away our equipment and weapons one by one. 1 April. Arrived at native shack. According to a communication, friendly troops in Lorengau cannot help but withdraw. Hereafter there is no choice but to live as the natives do. 133 By 30 March extensive patrolling activities by both regiments revealed conclusively that there were not enough Japanese remaining on Los Negros to cause our forces any trouble. Most of the combat troops could settle down to a much-deserved and needed rest. Al- though many dead Japanese were discovered along the jungle trails, no count ever reached the high estimate of 23 March, which allowed for a force of over 500 on Los Negros. Undoubtedly many Japanese dead escaped detection in the jungle. The remainder was believed to have fled to Manus, where remnants estimated at 400 were thought to be scattered. On Los Negros 1,917 enemy dead had been counted since 29 February. The 1st Brigade had lost 143 killed and 408 wounded. For most of the troops on Los Negros the combat phase of the Admiralties operations was over. The 1st Squadron, 5th Cavalry took up positions on the southern coast, and the 2d Squadron occupied Loniu Village (Map No. 3, pages 8-9). After the 12th Cavalry had cleared out the northern area west to Loniu Passage, the entire regiment reassembled at Lombrum Point on 29-30 March. Here the troops cleaned clothing, equipment, and arms and relaxed on the excellent beach. The 5th Cavalry had been in action for 31 days, the 12th for 24 days. Final mopping-up operations on outlying islands were assigned to the 12th Cavalry. PITYILU ISLAND was covered with coconut plantation and dense ram forest when it was invaded by the 1st Squadron, 7th Cavalry. The air strip was completed in January 1945, making a total of five in the Admiralties. Last Stages Capture of Outlying Islands Overwater assaults on the lesser islands of the Admiralties, along with mopping-up actions on Manus, comprised the last phase of combat operations. In light of the losses on Hauwei and with the expectation that landings would be opposed, attacks on the outlying islands would be made in considerable force. Pityilu Island, 3 miles north of Lugos Mission, was the first scheduled for attack on 30 March by the 1st Squadron, 7th Cavalry, reinforced (Map No. 3, pages 8-9). The other islands were objectives of the 12th Cavalry; the 1st Squadron, reinforced, would seize and mop up Koruniat and Ndrilo Islands on 1 April and two days later the 2d Squadron, teinforced, would attack Rambutyo. Pityilu Island, thought to be defended by 60 Japanese, is nearly 3 miles long and varies in width from 250 to 650 yards. The beach chosen for the assault is the only one suitable for landing; it is of white, hard sand and is located about the center of the southern shore. Six waves transported in LVT’s, LCM’s, and LCV’s would make up the assaulting force, which would move inland through the coconut plantation covering the western two-thirds of the island. JAPANESE NAVAL GUN (120-mm, type 3, 1914), captured on Hau- wei, was repaired by mechanics of the 61st Field Artillery Battalion and used to fire 40 rounds on Pityilu. Although the gun’s telescopic sights had been removed by the enemy, the open sight for direct laying could be used in indirect laying by calculating deflection from an aiming point. 136 Naval gunfire, artillery, and air strikes had been used against this island at various times before the attack date. Destroyers had first bombarded it to keep down hostile fire when the approaches to Lorengau and Lugos Mission were being cleared of mines. A Japanese naval gun captured on Hauwei Island had been put back into commission and used to augment the fires on Pityilu. Prepara- tion for the attack began at 0630 on 30 March by 2 destroyers which fired 30 rounds each until 0730. A spectacular air strike followed the naval fire. For the first 10 minutes P-40’s dive-bombed the landing beach; the next 10 minutes P-AO’s and Spitfires strafed the entire island. Immediately afterwards the 61st Field Artillery Battalion, which had registered the previous day from positions on the south side of the Lorengau air strip, pounded the island with a heavy concentration of 105’s. When the artillery barrage was lifted, two LCS's (Landing Craft, Support) on either flank of the beach opened up with their rockets. By this time the assault waves were approaching the shore, and the rockets searched the island in front of the first wave. The successive waves landed unopposed and the troops estab- lished a beachhead with Troop C as the left wing, Troop A in the center, and Troop B on the right. At 1000 some patrols sent out immediately after landing reported no contact with the enemy. The Reconnaissance Platoon moved by buffalo to the west, while Troop C in reserve on the beach sent patrols west into the interior. Troops A and B, with one medium tank leading the way, began an advance east toward the rain forest which covered that end of the island. As the troops advanced they ran into light machine-gun and sniper fire which was easily silenced. Enemy guns in a hut, encoun- tered by Troop B after moving 1,000 yards along the south coast, were neutralized by the tank, which blew up the entire position. At 1212, after progressing 1,500 yards, Troop A ran into heavy resistance from dug-in positions midway between the north and south shores. Troop A started to withdraw to permit an artillery concentration to be placed on the position, but the Japanese followed the withdrawal so closely that it was impossible to evacuate our wounded until a light tank was brought up to cover this operation. Then a 45-minute artillery concentration was placed on the enemy bunker, after which Troop A, aided by the light tank, attacked the position and killed 14 Japanese. 681583 °— 46— 10 137 l&ftTLr ■ ^*11 II frills ftWr | ^ CAVALRYMEN CARRYING FULL BATTLE EQUIPMENT cross between Mokerang Peninsula and Ndrilo and Koruniat in native canoes. Troop B came upon a hastily constructed trench containing 21 Japanese, who gave their position away by loud chatter. When 2d Lt. John R. Boehme and two privates went out to investigate the posi- tion, they were wounded by fire from the group. In spite of his wound, Pvt. Paul A. Lahman advanced on the position, firing clip after clip from his BAR. He was credited by Lieutenant Boehme with the destruction of practically the entire force. At 1720 the squadron withdrew on regimental orders to a position on the western edge of the rain forest and established a perimeter for the night. The Recon- naissance Platoon patrolled the western end of the island and returned to report no contact. Although the attacking force then did not know it, all the Japanese garrison had been killed or wounded. After a bombardment the next morning, the squadron advanced and dis- covered more dead Japanese, which made a total of 59 killed against 8 cavalrymen killed and 6 wounded in the mopping up of Pityilu. Seizing Pityilu was an expensive operation compared with the other small islands, which turned out to be either unoccupied or 138 harboring only a handful of Japanese who hid out in the interior. Until 4 April our command did not have knowledge of the enemy document of 2 March that had ordered the calling in of the garrison units from Rambutyo, Peli, Pak, and Pityilu. In preparation for the 1st Squadron, 12th Cavalry’s attack on Ndrilo and Koruniat, which had earlier been suspected of harboring naval guns, PT boats and a rocket boat as well as the customary artillery took the islands under fire. Bombing and strafing by the RAAF and heavy mortar fire from Mokerang Point also preceded the assault, which moved out on 1 April, transported in native canoes. No Japanese were found on Koruniat or on Ndrilo, and the next day the 1st Squadron re- turned to Lombrum Point. PT (MOTOR TORPEDO) BOATS helped support invasion of the small islands with their fire. They also landed small groups on other islands and patrolled the coast line of Manus, sinking enemy barges. 139 Rambutyo, an island 11 miles long and 5 miles wide, was also invaded by a strong force. On 3 April the 2d Squadron, 12th Cavalry, with two platoons of engineers and two platoons of the shore bat- talion, were loaded on destroyers at 0600. After a bombardment the troops landed at noon and met no resistance. Some Japanese were thought to be hiding in the rugged interior, so the 2d Squadron’s task was a mopping-up operation through unusually steep country and heavy jungle, where water was hard to find. By 23 April, with the aid of an Angau warrant officer and native guides and patrols, 30 Japanese had been killed and 5 more captured. Further mopping up was left to the native police recruited by Angau. Meanwhile on 9 April the 1st Squadron of the 12th had occupied Pak Island. Three prisoners, captured on 10 April, had escaped from Rambutyo by native canoe. On 12 April more Japanese were found on the east end of Pak and destroyed, eight of them in a large coral cave from which they fought until the cave was blown in on them by a demolition section from the 8th Engineers. Mopping-up actions on the small islands completed the combat operations of the 1st Brigade Combat Team, which had initiated the invasion of the Admiralties and whose units had fought almost continually for over a month. On 4 April the 5th Cavalry began moving to a semipermanent bivouac area on Koruniat Island, and after its operations were completed the 12th Cavalry moved to Salami Plantation. The troops of the 1st Brigade would settle down to housekeeping, rehabilitation, and training, with existence in the tropics made more bearable by the arrival of refrigeration, bakery, and laundry companies. The 2d Brigade continued to rout out the remnants of Japanese defenders on Manus. Final Mopping Up on Manus After the hard struggle to break the last enemy holding positions on Manus, the 2d Brigade Combat Team pushed inland after the scattering enemy, and penetrated much of the thick jungle, swamps, and high mountains covering the interior. The Angau and native guides helped the 7th and 8th Cavalry patrols make their way through this rugged terrain, while the 302d Cavalry Reconnaissance Troop, trained for overwater patrolling, probed coastal areas around the entire island. 140 Eastern Manus was the area assigned the 8th Cavalry to clear of Japanese, many of whom were thought to have fled to this area from Los Negros. Less Troop A which had suffered high casualties, the 1st Squadron established a perimeter on the high ground at Rossum. From Number Two Road and from the 2d Squadron positions along Loniu Passage, aggressive patrols then covered the entire network of native trails. The Japanese encountered were always in small groups of stragglers, for the most part wandering aimlessly across country trying to exist on the raw flesh of dogs and on what they could steal from the native gardens. Constant pressure was maintained on the remaining enemy, although the pursuing troops were frequently rotated for rest and rehabilitation. Up to 2 May when the regiment was withdrawn to bivouac at Hauwei Island, the 8th Cavalry had accounted for 285 enemy dead after suffering a total loss of 4 troopers killed and 7 wounded. Although the attack on Pityilu limited the 7tn Cavalry’s patrol- ling efforts, by 30 March its patrols operating from Lorengau had covered all the ground west of Number Two Road to the Tingo- Hiwal trail, making a wide area around Lorengau secure for the building of the naval base. After the 1st Squadron completed the occupation of Pityilu, it carried on patrolling activities from Lugos Mission. The 2d Squadron was assigned the difficult task of moving inland along Number One Road as far as Kawaliap. A base recon- naissance camp would be established at Yiringo by a reinforced troop while the other troops established themselves at designated points along the road to facilitate shuttling of supplies, secure the route, send out local patrols, and set ambushes. Supplying an entire squadron, ordered to move as far as 15 miles inland, presented a serious problem. Continuous rains had made the red clay Number One Road impassable for all vehicular trans- portation but a D— 7 dozer with trailer, and beyond Tingo nothing could get through. On 2 April Troop E moved 500 yards west of the Lihei River and established a supply base for the planned patrols. A dozer moved supplies to the east bank of the river, but troops had to take over from there, ford the river waist-deep, and then climb up a series of steep hills that were almost impossible for men carrying loads. Because this supply base was so difficult to reach, a new one had to be set up and by 8 April another suitable route was devised. Buffaloes carried supplies up the Tingau River to a point 141 north of Sabon, and from there native carriers were ready to relay them to the inland patrols when troops of the 2d Squadron reached their assigned positions. Since natives could carry a 41 -pound load from Sabon to Yiringo in about 354 hours, while the same task took the average soldier 2 or 3 days, the new supply arrangement eased the burden on the troopers considerably. By 10 April the troops of the 2d Squadron had stationed them- selves at Yiringo, Drano, Lundret, and Sabon. On the previous day a large group of enemy stragglers, attempting to gather in the Drabito— Metawarri area south of the 2d Squadron’s farthest point of penetration, had been bombed by P-40’s from the 73d Wing RAAF. The southern coastline as far as Patusi Bay had been combed by the 8th Cavalry, so the 2d Squadron’s efforts were directed to a thorough clearing of the area north and west of the Number One Road. Living under terrible conditions of rain and heat, the squadron patrolled the interior until 17 April when it began to move to Hauwei Island, its mission completed. The squadron had accounted for 82 enemy dead and had lost 1 man killed and 4 wounded. However, the heavy rains contributed to a high disease rate and 74 men had to be evacuated. The 1st Squadron kept its perimeter at Lugos Mission until 5 May, and the Japanese accounted for by its patrols, aided by the natives, raised the 7th Cavalry’s total of Japanese killed during the mopping-up period to 102. Seven prisoners were taken. After 5 May the entire 7th Cavalry established a semipermanent camp at Hauwei Island. The 1st Squadron was relieved at Lugos Mission by the 1st Squadron, 8th Cavalry, which continued mopping- up operations until the close of the official campaign on 18 May. The 2d Squadron was also assigned to further patrolling on 17 May. While the 8th Cavalry concentrated on the northeastern area, the 302d Cavalry Reconnaissance Troop continued its extensive patrolling along the coast. Natives furnished by Angau served the 8th Cavalry not only as bearers, scouts, and canoe paddlers, but also as police boys and “hired assassins” who tracked down and killed Japanese. The troopers also became adept at hunting Japanese, often acting as if it were a mere sporting proposition. On one occasion a native reported to an 8th Cavalry patrol that three unarmed Japanese were in a hut to their front. Three members of the patrol dropped their arms, moved into the hut and, after a scuffle, killed the three with knives. The total number of Japanese killed by the 8th Cavalry 142 in this last period of mopping up was 133, with 15 captured and 18 additional bodies discovered. Many of the dead discovered during the period had perished from starvation. In the final destruction of the Japanese, the 302d Cavalry Recon- naissance Troop demonstrated its skill in hunting and ambushing an enemy in unknown and difficult country. The unit, consisting of 8 officers and 177 men, had been intensively trained before the opera- tion, and the experience gained saved many lives. Aside from losses suffered in the ambush at Hauwei, the troops sustained only two casualties from 10 March to 12 May, during which time at least a part of the troop was in daily contact with the enemy. A consider- able part of the success of the troop lay in its mastery of the art of concealment. It was essentially a “sneak and peek" organization which made stealthy landings in small boats by daylight and dark- ness; then concealed the craft and obliterated all traces of the landing. All the troopers were adept at mapping and map correction and the establishment of trail-watching posts. When the enemy ceased using the trails and moved through virgin jungle in the later phases of the campaign, the troopers learned how to find them. By tactful handling of the natives and by mastering their language quickly, the troopers won indispensable assistance. When Japanese came to the villages hunting for food the troopers were immediately informed. Messages from patrols were sent chiefly by native runners, by pigeons, and by code on radio sets. Although the native runners and the pigeons were reliable, the pigeons would not fly when soaked in a tropical downpour and waterproof covers had to be provided for their cages. Sometimes the weight of the radio hampered small moving patrols. Simple codes were improvised and six Sioux Indians who could speak their tribal language fluently improved on these codes, sending messages in the clear with impunity. Although the missions of the troop were chiefly of reconnaissance, they killed 48 Japanese and captured 15 prisoners in their extensive patrolling of the island during May. Enemy dead accounted for by all the units involved in the last phase of operations on Manus reached a total of 586. Prisoners, cap- tured mostly in the last days when the starving Japanese were more disposed to surrendering, totaled 47. During the mopping up, our forces suffered the relatively small losses of 6 killed, and 12 wounded. At the close of the official campaign it was estimated that 150 143 additional Japanese were still hiding out in the jungle. Patrols and armed natives continued to hnd these, but as most of them were dying of starvation, they were no longer a threat. Construction Activities Service troops and construction equipment had come ashore along with combat troops immediately after the reconnaissance in force had seized its beachhead, when every man was needed to defend the initial hold against counteroffensives. The Brewer Operation was conceived with the prime objective of developing an advance naval and air base to neutralize Japanese power in the Bismarck area. There- fore the planned immediate establishment of two airdromes and light naval facilities called for a high priority on service troops who were to be employed at these tasks as soon as practicable. In addition to their use for strategic objectives, service troops also had aided tactical operations in the Admiralties immeasurably. After account- ing well for themselves in their emergency combat role, helping to defend the early perimeter, the first service units ashore took up the varied tasks of building air and naval facilities as well as performing the indispensable services for the combat troops. The quick repair of the Momote air strip brought fighter planes almost to the front lines. Engineers directly helped in the destruction of the enemy; without the bulldozer, combat operations inland on muddy roads, especially with armor, would have been impossible. As the fighting reached the mopping-up stage and more service units came ashore, the activities of building a more habitable base involved more men than did services for troops engaged in combat. After the initial stages of operations, signal communications between Army, Air Corps, and Naval forces were greatly improved by the arrival of the 99th Signal Battalion at Salami on 16 March. This unit took over the switchboards and was responsible for all wire communications on Los Negros until 20 April. On 9 April a radio station and message center was set up at Salami Plantation by a section from the 832d Signal Service Company. COMPLETED AIRDROME ON MOKERANO has two 8,000-foot coral strips ( the right one is darker), and a parallel taxiway (right). 144 ~ 145 In transforming the devastated island into a naval and air base, health precautions were as essential as construction. Early in the operations the 52d Malaria Control Unit began searching for mos- quito breeding places and carried on an extensive oiling program. This work was also kept up by the 28th Malaria Survey Unit which on 22 March made the first blood survey of natives in Mokerang village. Natives were of great use in both tactical operations and con- struction work. The Angau Detachment recruited them for pack trains, for scouting, for police work, and for actually hunting and killing the enemy. By 11 May the entire population of 17,000 was under the control of Angau and 5,000 were being rationed and cared for. In 5 camps, 1,231 native laborers (a few of whom were women and children) were employed on various projects. Many were as- signed to cavalry units and to the air forces. The 58th Evacuation Hospital used a large number in malaria control, and the greatest part were engaged in various overhead duties, including improve- ment of the villages. The 40th Naval Construction Battalion, which had defended the air strip with the cavalrymen on 3-4 March, was joined by other Seabees of the 17th, 46th, 78th, and 104th Naval Construction Bat- talions. The Seabees were to work on the Momote air strip as well as on dock and road construction. Other sections working on roads and maintenance were taken from the 8th Engineer Squadron and the 592d Engineer Shore Battalion. On 16 March the Momote air strip was in use and the airdrome well on its way to completion; however, the captured Lorengau airdrome was discovered not to meet the requirements for the second airdrome planned for the Admiralties. Therefore, at Mokerang Plantation, a coral-surfaced modern airdrome 100 by 8,000 feet was put under construction. By 18 May the Momote airdrome was extended to 7,000 feet and surfaced with coral, completely equipped with taxiways, hardstandings, and storage areas. Many naval facilities, begun soon after the landings, were fin- ished by the end of the official campaign. A floating Liberty dock, a fixed Liberty dock, and a pile dock at Mokerang were ready for use. An LST pile dock was almost complete and a pipeline jetty had been built at Porlaka. Facilities for a full-fledged base were begun: storage for 7,000 barrels of bulk petroleum was ready at Momote 146 and space was being built for 30,000 barrels at Mokerang. Channels into the harbor were also improved and buoyed and a crib dock con- structed. Areas around the air strips were becoming more habitable. Camps had been built for the RAAF and AAF, and more living quarters could be constructed from the lumber turned out by a saw- mill operated by the engineers. Drainage was the biggest obstacle to providing healthful living conditions. Tropical rains kept troops busy and often made them repeat their agonizing efforts. Just as the en- gineers had nearly filled a swamp about 100 yards wide on Hauwei Island, the rains came and the swamp spread to three times its original size. Nevertheless, by 18 May, eastern Los Negros and the Lorengau area of Manus as well as some of the outlying islands looked very little like the lush and dangerous territory invaded by the 1st Cavalry Division a month and a half before. SHELL HOLES from our intense artillery preparation on Hauwei Island had to be filled before the engineers could transform the island into a recreation area for the 2d Brigade. The west end was completely cleared, and a baseball diamond and drill field set up on the east. 147 Conclusion With an initial force of 1,000 men to the enemy’s 4,300, the 1st Cavalry Division had overwhelmed the enemy and, with the arrival of its full complement, had exacted a vastly disproportionate number of casualties from the defenders. Although the official count of their dead did not include those removed for burial by the enemy, the total was 3,280 Japanese killed and 75 captured, which almost equaled the original G— 2 estimate of the garrison’s size. American losses were 326 killed, 1,189 wounded, and 4 missing. The 10-to-l ratio between Japanese and American dead does credit to the cavalrymen fighting their first engagement against seasoned troops in fortified jungle positions. General Krueger partially explained the heavy enemy losses in the Admiralties operation by his claim that, “Our troops were gain- ing superiority on the ground against an enemy whose tactical knowl- edge envisioned only the offensive.” Allied tactics of guaranteeing naval, air, and artillery superiority to our troops in each operation were making the heavy proportion of Japanese casualties an expected result in the Pacific. In the Admiralties invasion, fire from destroy- ers kept the enemy under cover during the landing and our artillery gave the troopers an enormous advantage against an enemy who pos- sessed only two 75-mm mountain guns and one 70-mm howitzer. Bad weather had greatly restricted air operations during the first week after the invasion, and the weather probably accounted in part for the weakness of enemy air defense throughout the campaign; but the constant pounding of Japanese air bases within range of the Admiralties was a more important factor. 148 General MacArthur's decision to send a limited number of men and ships to take an enemy stronghold far in advance of Allied-held territory, and within striking distance of enemy planes, had proved worth the risks involved. At small cost our forces had placed them- selves squarely across the enemy supply lines to the Bismarcks from the west. The difficult ground assault on Kavieng was now unneces- sary, and both Rabaul and Kavieng, almost ringed by our air bases, could be neutralized and then kept in a condition of helpless check- mate. From the new base in the Admiralties, Allied air and naval forces could now launch surprise attacks on the Dutch New Guinea coast and could threaten essential enemy sea lanes within a 1,500-mile radius including the Marianas, the east coast of Mindanao, and the southern limits of the Celebes Sea. Finally, with the neutralization of enemy positions flanking the Admiralties, these islands could be- come the largest staging area for the Philippines invasion. 149 1st Cavalry Division U. S. GOVERNMENT PRINTING OFFICE : 1946 681583 150 Annex No. 1: Abbreviations AAF Army Air Forces APD High Speed Transport (Destroyer) BAR Browning Automatic Rifle CP Command Post HE High Explosive LCM Landing Craft, Mechanized LCP (R) . Landing Craft, Personnel (Ramped) LCS Landing Craft, Support LCV Landing Craft, Vehicle LST Landing Ship, Tank LVT Landing Vehicle, Tracked OP Observation Post PT Motor Torpedo Boat RAAF . Royal Australian Air Force TD-9 International Tractor 151 ) X ♦ /