Planning and Preparation

Higher Level Plans

Admiral Nimitz issued CinCPOA "Joint Staff Study, Revised," for Operation STALEMATE (seldom was a more appropriate code name applied to any operation) under date 10 May. This designated Commander Third Fleet as over-all commander; Commander III Amphibious Force as Joint Expeditionary Forces Commander; and Commanding General III Amphibious Corps as Commanding General Joint Expeditionary Troops. Target date was set tentatively as 15 September.

These echelons commenced preliminary study at once, handicapped by the previously noted lack of reliable information. The regular staff of III Amphibious Corps was already committed to the Guam phase of the Marianas operation, plans for which had assumed virtually final form. Thus, it was necessary to set up a separate provisional staff when, on 18 May, a dispatch arrived from ComSoPac directing Corps to send a staff planning representative to Noumea, New Caledonia, to confer with Commander Third Fleet.1

As a result of this conference, Commander Third Fleet issued a Warning Order for the operation on 26 May. The following day, however, he was obliged to cancel this upon notification by CinCPOA Dispatch that the order was to be issued by that headquarters. Thus, the staff representatives of III Amphibious Corps returned to Guadalcanal with nothing definite accomplished, though much miscellaneous information had been compiled and a tentative scheme of maneuver worked out. In accordance with a CinCPOA dispatch dated 23 May this planning group was ordered to proceed to Pearl Harbor upon departure of CG, IIIAC for the Marianas.2

The CinCPOA Warning Order, in favor of which Halsey's had been cancelled, was issued on 29 May. This envisioned seizure of the entire Palau group in an operation on a larger scale than either Saipan or Guam, and nearly as large as the two combined, with 8 September designated as the target date. Major General Geiger, CG, IIIAC, was designated also Commander Expeditionary Troops and Landing Force. Four divisions, organized into two corps, were to be employed. The IIIAC, consisting of the 1st Marine Division and 81st Infantry Division, was to assault the southern islands of Peleliu and Angaur. Simultaneously, the XXIV Corps, U.S. Army, 7th and 77th Infantry Divisions, would assault the


main island of Babelthuap, second largest in all Micronesia. Still another Army division, the 27th, was designated to stand by in reserve on New Caledonia.3

Once Third Fleet and IIIAC planning staff arrived at Pearl Harbor,4 the project began to shape up in earnest. XXIV Corps was already in the area, so the several staffs could work in close conjunction, and 1st Marine Division sent a representative (Lieutenant Colonel L. J. Fields, D-3) to join the conference on 1 July. Joint Army-Navy Intelligence Service (JANIS) prepared hydrographic-geographic surveys of the target area. Aerial and submarine photographs were made, and the providential capture on Saipan of the Japanese order of battle for the entire 31st Army area greatly simplified the intelligence problem.

Yet the plan soon began to run into trouble. The 27th Infantry Division, designated as general reserve, was removed from the picture by being committed on Saipan. Then came the long delay in the Guam operation,5 to which both IIIAC and the 77th Infantry Division had been committed. Nor was there any visible solution to the transportation problem, that perennial limiting factor in all Pacific operations, especially acute here because of the large commitments of shipping in the Marianas.

Other factors were developing which tended to throw a different light upon the whole Western Carolines operation as conceived in the 30 May order.6 Additional intelligence cast serious doubt on the advisability of attacking Babelthuap at all. The big island was known to be strongly garrisoned and probably very powerfully fortified, whereas the terrain was so rugged as to offer limited possibilities for the development of air facilities. In contrast, Peleliu boasted a fine airfield long in operation, with an auxiliary strip building on the off-lying small island of Ngesebus which, once in our possession, would make neutralization of all forces and facilities on the big island a simple matter.

As a result of these several factors, on 7 July CinCPOA issued a second warning order, cancelling the previous one and setting forth a new concept under the designation STALEMATE II. This substituted the 96th Infantry Division for the 77th in the XXIV Corps, and set the target date for the southern Palaus back to 15 September.

Major General Julian C. Smith was placed at the head of the planning group which had been detached from IIIAC and which was now designated X-Ray Provisional Amphibious Corps, to exercise command over the 1st Marine Division and 81st Infantry Division. At this time General Smith was Deputy Commander, FMFPac, and he continued to function in both capacities until his appointment as Commander Expeditionary Troops and Landing Force for STALEMATE II, since General Geiger, whose continued absence on the Guam operation had left this important post unfilled.7

STALEMATE II split the over-all operation into two phases, to be carried out by two separate attack forces. Plans for the southern Palaus were retained intact, but


the objectives of the XXIV Corps were shifted from Babelthuap to Yap and Ulithi, with target date tentatively set for 5 October.

Here is the command set-up as then designated:

THIRD FLEET--Admiral Halsey
CTF 31: Third Amphibious Force--VAdm Wilkinson.
CTF 36: Expeditionary Troops and Landing Force--MajGen Julian C. Smith.
CTF 32: Western Attack Force--TG 36.1: Western Landing Force (III AC). RAdm Fort.
TG 32.1: Peleliu Attack Force--TU 36.1.1: Peleliu Landing Group (1st MarDiv).
TG 32.2: Angaur Attack Force--TU 36.1.2: Angaur Landing Force (81st InfDiv).
CTF 33: Eastern Attack Force--TG 36.2: Eastern Landing Force (XXIV Army Corps). VAdm Wilkinson.
  TU 36.2.1: 7th InfDiv.
  TU 36.2.2: 96th InfDiv.

The 77th Infantry Division was designated Floating Reserve, to be embarked at Guam, while the newly arrived 5th Marine Division was to be held as General Reserve in Hawaii.8

As far as concerned Phase I, this organization remained till D-Day, with only one major change. Upon completion of the Guam operation, General Geiger was ordered back to Guadalcanal where, on 13 August, he and his staff from IIIAC relieved General Julian Smith in command of Western Landing Forces (TG 36.1). X-Ray Provisional Amphibious Corps, as such was deactivated, but most of the staff was retained by General Smith who continued in command of the higher echelon (CTF 36), now designated Expeditionary Troops Third Fleet.9

As had become standard practice long since in amphibious assaults, over-all plans paid due attention to all off-lying Japanese bases and facilities which might conceivably interfere with the attack on the target islands. One part of Third Fleet's mission was defined thus:

Seek out and destroy hostile air and naval forces which threaten interference with STALMATE II operations, in order to inflict maximum damage on the enemy and to protect our own forces.10

Accordingly, the plan called for powerful strikes against the Bonins (Iwo Jima and Chichi Jima) on 31 August-2 September, Yap on 7-8 September, and Mindanao 9-14 September. Because of the astonishing success of the initial attack against the latter island, and the weakness of resistance encountered, the follow-up strikes originally planned were hurriedly shifted from there to the central Philippines, where the Visayas were hit on 12 September with even more devastating results.11

Because these unexpected developments were destined to have a profound effect on the operation immediately under discussion, it may be well to step out of strict chronology here in order to discuss briefly what actually happened in that connection.

The Western Attack Force completed its preparations on schedule and was en route to the target when, on D-minus 1 (14 September) the command ship intercepted a message from CinCSoWesPac to CinCPOA carrying the rather startling implication that abandonment of the attack on Yap was under serious consideration (see Appendix B).

On the morning of 16 September, while Marines of the 1st Division were fighting desperately for Peleliu, Admiral Wilkinson was alerted by commander Third Fleet to be prepared to use the one regimental combat team which comprised the corps reserve for the seizure of Ulithi. This was substantiated that evening by orders from the same source to carry out this mission ". . . as early as practicable . . . with resources at hand." Then, as though to make all this official--and explicable--Admiral Halsey arrived in


Map 3

person the following day aboard his flagship, USS New Jersey.12

In brief, the decision had been made to cancel entirely the second phase of STALEMATE II as originally planned in favor of a new and daring conception of the Philippines campaign.13 The Eastern Attack Force, designated, for the Western Carolines, reached its advance staging area, Eniwetok, only to be ordered immediately to report to General MacArthur's command at Manus in the Southwest Pacific. Thus, the XXIV Corps sailed right out of the plan and out of the theater; and IIIAC, at grips with the enemy on Peleliu and Angaur, found the only reserve on which it had been led to think it could count committed to the seizure of Ulithi, hundreds of miles away--with results discussed subsequently as they apply chronologically.

Planning on Division Level14

This planning by the higher echelons was largely generalized in nature and concerned mainly with top level strategic considerations. Because of the improvised nature of the X-Ray staff and the absence of IIIAC on the Guam operation, the actual detailed, down-to-earth planning developed upon the fighting echelons to a degree perhaps unprecedented since Guadalcanal: the 1st Marine Division and the 81st Infantry Division as concerned the Western Phase (because of its subsequent cancellation, discussion of the Eastern Phase has no place in this narrative). The 81st, based in the Hawaiian area, was able to work in close conjunction with the other echelons, but the isolated 1st Marine Division was largely on its own.

This unit was in something less than the best condition, physical and mental, of its distinguished career. Following the long and exhausting, if not especially bloody, Cape Gloucester campaign, it had been removed from New Britain early in May to Pavuvu, in the Russell Islands, about sixty miles northwest of Guadalcanal. The purpose of this move was officially given out as "rest and rehabilitation," terms whose implications became increasingly ironic as all hands grew acquainted with the more salient attractions of that rat-infested mudhole.15 Also, more than two-thirds of the personnel had been in the Pacific in excess of twenty-four months, had served through two exhausting jungle campaigns, and had been led to expect early relief under the new rotation of service plan. To assure that ample replacements were received in plenty of time, the division commander, Major General William H. Rupertus, and Colonel John T. Selden, chief of staff, obtained orders to proceed Stateside to place their needs before Headquarters Marine Corps, and hence were absent during the initial planning phases.

CinCPOA warning order of 29 May reached the Division on 2 June. Brigadier General Oliver P. Smith, Assistant Division Commander, put the division staff to work immediately, to such good effect that a complete study with two alternate schemes of maneuver was ready for submission to the commanding general upon his return on 21 June. The alternative determined upon then became the division plan. Since the new conception of STALEMATE II (CinCPOA warning order dated 7 July) did not affect the operations on either Peleliu or Angaur, this same plan was approved in turn by Commander Western Attack Force (CTF


AKARAKORO POINT, northern tip of Peleliu, was separated from Ngesebus by reef in foreground and from the northeast peninsula (background) by a reef-floored arm of the sea which tapered off into dense mangrove swamps. On shore at right is phosphate plant before its destruction. Photo taken during carrier strike, 31 March 1944.

32--Rear Admiral George H. Fort), and by CG, IIIAC upon General Geiger's return from Guam to resume command on the corps level16 (TG 36.1--Western Landing Force) on 13 August.

Description of this plan in essential detail follows.

Intelligence Phase

Earliest reports on Peleliu showed the island to be about six miles long, shaped roughly like the claw of a lobster, and surrounded by a fringing reef up to 1000 yards in width. The southern portion where the airfield lay, below what might be called the hinge of the claw, achieved a maximum width of about two miles. The whole was described as "low and flat", a definition which persisted throughout these surveys although even the early charts showed estimated elevations up to 260 feet17 (in contrast to "low, flat" Tarawa where maximum elevation reaches perhaps 20 feet above high tide).

First systematic aerial photographic coverage occurred during the devastating strike of the Fast Carrier Task Force (TF 58) against the Palaus on 30-31 March, 1944, as one of the covering actions for the Hollandia operation in New Guinea. This attack proved to be even more successful than its perpetrators could appreciate at the time the Palaus were permanently crippled as a naval base of real importance, and an estimated


160 Japanese planes were destroyed in the air or on the ground.18

Photographic coverage, however, was incomplete and generally unsatisfactory for planning purposes. To supplement this material and keep it up to date, subsequent sorties were flown at behest of X-Ray staff, starting 2 July and continuing until late August, by both Navy carrier planes and New Guinea-based planes of the Army's Fifth Air Force, taking both vertical and oblique shots. Between 23 and 28 June the submarine USS Seawolf lay off Peleliu and photographed the profiles of all possible landing beaches. From this material a workable and generally accurate19 map was compiled by 64th Engineer Topographical Battalion, USAFICP, on scale 1:20,000. This, with a target area grid superimposed, was the standard map used throughout the operation, special portions being blown up to 1:10,000 and 1:5000 for the benefit of unit commanders operating ashore.20

Complete as this photographic coverage was, it fell short of entire adequacy. The division aerial photographic interpreters were able to spot many ground installations--but not nearly so many as the size of the garrison and length of enemy occupancy would give reason to believe existed. And a number of installations spotted could not be identified.21

The map as finally issued showed the southern part of Peleliu to live up generally to the JANIS description: "low and flat" (though even this proved to be a considerable over-simplification). The same could be said of the northeastern peninsula, or shorter prong of the lobster's claw. Here in a small village with the jaw-breaking name of Ngardololok the Japanese had a radio direction finder, power plant and lesser utilitarian installations, all in the open and tastefully landscaped. Beyond this the terrain of the peninsula deteriorated into a series of smallish islands, separated from each other and from the longer northwestern peninsula by a complex of swamps and shoal coral, some of it dry at low tide.

The airfield was an unusually good one, with bomber and fighter strips of hardpacked coral, served by ample taxiways, dispersal areas and turning circles. Immediately to the north of it lay a sizeable area of utilitarian rather than defensive installations: barracks, hangars, large water cisterns, machine shops, another power plant, a radio station, a large two-story administration building of the type encountered previously on well developed Japanese air bases. Just beyond this area jungle masked the details at what appeared to be the base of rising ground. A dense mangrove swamp bordered the airfield to the east, while to the west and south lay scrub jungle liberally interspersed with wild coconuts and occasional grass-grown clearings.

"There was never any question in the minds of the 1st Division planners but that the high ground north of the airfield was the key terrain feature of the island," declared the officer who headed those planners during the early stages.22 By this, he referred to the northwestern peninsula, or longer pincer of the lobster claw. That this was high ground had been noted in the early JANIS studies. Profile photographs taken by the


SOUTHERN NGESEBUS, with its air strip and causeway connecting to Peleliu, as it appeared on 25 July, 1944.

submarine substantiated this, but without providing reliable means of estimating how high or how rugged it might be. Vegetation cloaked the slopes; sparse, scraggly vegetation, as it turned out, but sufficient to mask the contours from aerial observation. Not a single defensive installation of any importance was spotted in this region, although previous experience with the Japanese in similar situations indicated that there were doubtless many such.

As finally drawn up, the map indicated the high ground to be a single more or less continuous ridge system running about two-thirds of the way up the peninsula, flanked by a good road on either side. Here the ground flattened out briefly, and the East Road angled through a wide draw to converge with the main, or West Road, which continued past another Japanese radio station to the phosphate refinery which had constituted the island's principal industry. Inland from these rose a narrower ridge and several abrupt, more or less isolated hills, one of them mounting a radar installation. About 500 to 700 yards beyond the northwestern tip of the peninsula, and connected with it by a wooden causeway across the shallow reef, lay the semi-joined smaller islands of Ngesebus and Kongauru where the auxiliary airstrip was building.

With this somewhat speculative intelligence, the planners had to be content, since it would be manifestly impossible to send pre-invasion patrols ashore on a land mass so small and so strongly held. For detailed data concerning reef and beach conditions they must wait on the operations of the underwater demolition teams on D-3, D-2 and D-1.23

From the foregoing it may be seen that, while the terrain of southern Peleliu might possibly be compared with an atoll island of the Tarawa type, there was a second school of thought which uneasily sought other comparisons for that northwestern peninsula. The most immediate possibility was Saipan,24 and officers from the division were dispatched to interview anybody they could find who had fought on that island. But the two operations followed one another too closely to permit compilation of a thorough intelligence study of Saipan, and the little information obtained came mostly from men wounded there and evacuated to hospitals within reach.25

And there was another difference between Tarawa and the Palaus. Betio had been garrisoned by 4,836 Japanese, with no other substantial enemy forces within reach. Even the notably low early estimates placed 9,000 Japanese on Peleliu, and in estimating the


enemy's potential the Division Intelligence Section was acutely aware that upward of 25,000 additional troops were posted on islands within practicable reinforcing distance.

Our preliminary knowledge of the enemy's strength resulted from one of those providential accidents of war which might be considered as bordering on the miraculous in the case of any enemy other than the security-unconscious Japanese, with their predilection for keeping voluminous records and then letting them lie around. Saipan had been headquarters of the 31st Army, and with its fall U.S. forces captured a considerable part of that echelon's files,26 together with a Japanese intelligence officer to help interpret it.

This remarkable combination of related documents designated the units stationed in the Palaus and listed the approximate strength of each, but did not indicate definitely their disposition. There were sufficient clues, however, to provide competent intelligence officers with the means for shrewd analysis. X-Ray Intelligence handled most of this work, passing its findings along to the division, modified from time to time as additional information became available. These estimates proved remarkably accurate from the first in the case of Peleliu, where a final check at the end of the operation disclosed discrepancies of only three minor units not identified on the preliminary estimates.27 The fact that more Japanese were encountered on the island than the estimates indicated resulted from their partially successful efforts to reinforce the garrison by barge from the large reservoir of troops on Babelthuap.

The backbone of the force defending the Palaus was the 14th Division, Imperial Japanese Army. This was one of the oldest and best units in all of the enemy's armed forces, with a history dating back to the Sino-Japanese War (1894-95) and an outstanding record in the Russo-Japanese War (1904-05). Formerly a part of the famed Kwantung Army, the division had been rushed to the Pacific in the spring of 1944 when the fall of the Marshalls had awakened Japan tardily to the peril facing her second line of defense. Its component infantry regiments were the 2d, 15th and 59th. Also on hand were the 53d Independent Mixed Brigade, and the usual miscellany of smaller units so dear to the Japanese military mind, together with sizeable contingents of naval personel, both combatant and construction.

Commander of the Palaus Sector Group, which included all the Palaus and the island of Yap to the northeast, was Lieutenant General Sadae Inoue, who was also commanding general of the 14th Division. His headquarters were at the town of Koror, on the small centrally located island of that name, pre-war administrative capital of western Micronesia. The greater portion of the troops, estimated upward of 25,000, were on the big island of Babelthuap, immediately to the north. This force included the Sea Transport Units of the 1st Amphibious Brigade, 1,349 men. Since many of the assault elements were known to have received special training in counterlanding operations, the potential of the enemy for reinforcing the Peleliu garrison appeared very real indeed.

Early reports had Angaur, the southernmost island and objective of the 81st Infantry Division, garrisoned by some 4,000 Japanese. Subsequent intelligence scaled this figure down to include only one reinforced battalion of the 59th Infantry, an independent AA artillery company and a few smaller units: an estimated total of 1,400-2,100 combat troops.28


Since this monograph is concerned mainly with the action on Peleliu, presented herewith is a detailed breakdown of the final preinvasion intelligence estimate of the Japanese forces on that island and semi-joined Ngesebus, which proved so accurate.29

14th Div.  
  2d Inf Reg including 1 Arty Bn 3,283
  One Bn 15th Inf30 including one Arty Bty (4x75mm) and one Mortar Co. (10x81mm) 1,030
  Tank Unit, less 1 Plt (12 tanks) 100
  Sig Unit, 1 Radio Squad 10
  Intendance Duty Unit 30
  Fd Hosp 250
53d Ind Mixed Brig  
  One Ind Inf Bn31 685

Navy, Combatant
45th Guard Force Det 200-400
114th and 126th AA Units 600

Navy, Labor
204th and 214th Const Bns } 2,000-2,200
Elements of 43d and 235th Const Bns }

Navy, Airbase Personnel
PELELIU 1,270  

Army, Combatant 5,300  
Navy, Combatant 800-1,000  
  6,100-6,300 6,100-6,300
Navy, Labor 6,100-6,300
Navy, Airbase Personnel 2,220
Navy, Airbase Personnel

The Japanese over-all commander on Peleliu was reported to be Major General Kenjiro (or Gongiro) Murai. Precisely what part this officer played in the campaign remains a mystery to this day. Not until after his death was tangible evidence found that he was even on the island (see Appendix F). The one undisputable fact is that all important captured orders concerned with one of the most stubborn defensive actions in history were signed by Colonel Kunio Nakagawa, commanding officer of the 2d Infantry, Imperial Japanese Army. And a dangerous and resourceful opponent he proved to be.

Operational Planning

On Peleliu there was no dearth of wide sand beaches suitable for landing operations. The width of the fringing reef made it apparent that the landing technique developed in the Marshalls and Marianas would have to be adapted here, and this premise was accepted in all studies concerned with working out the scheme of maneuver.

Four sets of beaches were taken into consideration.

  1. Beach PURPLE presented the greatest natural advantages. The reef here was nowhere more than two hundred yards wide, while at one short stretch it appeared practicable (with some improvements, this later proved to be the case) to bring vessels as large as LST's directly to the beach. However, these advantages were equally obvious to the Japanese, and their strongest defense system was set up here. The real clincher in rejecting such a landing was the presence a short distance inland of a dense mangrove swamp which in two places narrowed ingress to the main part of the island to a corridor barely wide enough to accommodate a road a double bottleneck ideal for defense which the Japanese could be depended upon to exploit to the limit.

  2. A projected landing on the SCARLET Beaches, including and overlapping the southern promontories, in conjunction with a coordinated landing on the ORANGE Beaches on the opposite (southwestern) shore was discarded because of the tactical disadvantages


    of assault elements converging on each other, plus the fact that late photographs showed the southern faces of the promontories to consist of coral ledges exposed to the surf and the inlet between them to be choked with concrete tetrahedrons.32

  1. Beach AMBER lay along the northwestern peninsula where the reef is widest, the northern flank enfiladed by the off-lying island of Ngesebus, known to be strongly held. Inland of the beach, a comparatively level shelf was dominated by high ground at ranges of 100-300 yards. A successful assault here would have the advantage of placing what was recognized as the key terrain in the division's hands at the outset. But failure of the initial momentum to carry the ridge line would leave the troops on the constricted low ground without room to maneuver or to emplace artillery, with the enemy "looking down their throats."

  2. A landing on the WHITE and ORANGE Beaches followed by a drive straight across the island to seize the airfield and split the defenders was the scheme of maneuver ultimately determined.33

The soundness of this decision is best attested by hindsight. Officers who participated in the operation, reviewing it in the light of much more thorough knowledge of the island's terrain and defenders, are unanimous in the opinion that the course chosen was the best one.34

The plan as finally formulated called for the landing of the division's three infantry regimental combat teams abreast. The 1st Marines (coded "Spitfire") would put ashore one battalion in assault on each of the WHITE Beaches, with the remaining battalion in regimental reserve. Their mission was to drive inland through the barracks area to a predetermined point, then wheel left and attack the nose of the ridge which extended up the northwestern peninsula.

The 5th Marines ("Lonewolf"), in the center, would land in similar formation on Beaches ORANGE 1 and 2, the battalion on the left tying-in with the 1st Marines, the other driving straight across to Peleliu's eastern shore. The support battalion, landing at H plus one hour, would move in between the other two, attack across the lower end of the airfield, and then participate in the turning movement to the northward. Following securing of the airfield, the regiment's mission became seizure of the comparatively flat northeastern peninsula and its outlying islands.

The 7th Marines ("Mustang"), less one reinforced battalion to be kept afloat as the division reserve, would land in column of battalions on Beach ORANGE 3, drive across to the eastern shore on the flank of the 5th, then wheel right and mop up the now isolated enemy contingents in a drive to the southern tip of the island.

The advantages of this scheme of maneuver were obvious; the disadvantages during the planning phase, somewhat less so.

On the credit side, it would be possible to land on a wide front, with a straight and clear approach to the beaches across a not-too-bad reef which attained a maximum width of about 700 yards. The airfield and comparatively open ground immediately to the south of it was suitable for tank operations, and it was anticipated that the drive across the island could be completed in short order. This would not only provide elbow


room for emplacing artillery, but should enable the 5th Marines to take Beach PURPLE in the rear, potential typhoon conditions making it imperative that beaches on both shores be secured as soon as possible to facilitate the continuous flow of supplies.

The disadvantage of driving across open, "flat" (more or less) terrain commanded by the high ground to the north was clearly--if perhaps not fully--realized. Although the contours of this high ground were still hidden by the vegetation, past experience made it a foregone conclusion that the Japanese would have plenty of heavy weapons emplaced there. It was understood, too, that in having this high ground as its objective the 1st Marines had drawn by far the toughest assignment. But it was anticipated that the 7th would complete its initial mission the first day and become available to regroup in support of the 1st, so that the striking power of two combat teams would be concentrated against the key terrain within a short time. In the meanwhile, it was planned to support the 1st Marines with a maximum concentration of air, naval gunfire, artillery and tanks.

And, according to the original over-all plan, there was what appeared to be an ample margin of safety: the 81st Infantry Division was not to be committed on Angaur until the situation on Peleliu was well in hand.35


The problem of transporting the troops to the target, putting them ashore across coral reef, and supplying them while there followed the general pattern resulting from the lessons learned at Tarawa and further developed during operations in the Marshalls and Marianas. This well known technique need be reviewed here only as it applied particularly to the Palaus operation.

Every type of vessel which had proved its value in such an operation was to be utilized, and in the manner it had previously proved most useful. As had become standing operating procedure by now, the foremost assault waves and the amphibian tractors (LVT's) which were to carry them in across the reef were loaded on LSD's, as were the armored amphibians (LVT(A)'s) which had been designated to serve as an advance wave and lead ashore the waves carrying assault troops. Development of the new LVT(4) with integral ramp astern (first used at Saipan) made it possible to land 75mm pack howitzers fully assembled and ready for action, and an adequate number of these vehicles were assigned to the artillery for their initial mission. So were the Army amphibious trucks (DUKW's) equipped with an A-frame unloading device capable of handling a 105mm howitzer, a method developed by the 7th Infantry Division at Kwajalein.

The tanks and their crews traveled aboard LSD's, the big, seagoing, drydock vessels originally designed to effect repairs on smaller craft at sea but now adapted to assault purposes. The tanks would be preloaded in LCT's which would move under their own power into the flooded drydock compartment. This would then be pumped dry for the run to the target area, there to be flooded once more and the capacious maw opened to emit its occupants in the transport area. The LCT's would then run in to the reef's edge and there discharge their tanks, waterproofed for the purpose, to get the rest of the way to the beach under their


Map 4
Scheme of Maneuver

PELELIU'S REEF, up to 700 yards wide in the landing areas, required that supplies be manhandled or carried in amphibian vehicles until pontoon causeway was built.

own power, guided by specially assigned LVT's.36

The remainder of the troops and their supplies would make the journey in the more conventional types of assault transports: AKA's (personnel) and AKA's (cargo), combat loaded as it should go without saying. Admiral Fort and Generals Geiger and Julian Smith, with the staffs of the higher echelons, would be embarked in USS Mount McKinley, one of the specially designed command ships (AGC's) which had proved so useful; General Rupertus and the bulk of the division staff in USS DuPage, an APA equipped (somewhat sketchily) to serve as an auxiliary command ship; and General


SEABEES AND ENGINEERS, working under fire, had this causeway fully operative by D-plus 6 to bridge the reef. Angaur Island in background.

O. P. Smith and remainder in a similar vessel: USS Elmore. Two APH's (hospital transports) armed and claiming no immunity in combat, would be on hand to care for the wounded, four regular hospital ships to arrive as early as practicable.37

The formation of the assault was also conventional, following the general pattern set in the Marshalls and Marianas. The LST's would discharge leading echelons in their launching area, from whence the vehicles would proceed to the line of departure where they would be formed into waves and advance successively to the beach. Troops from the transports would move to the line of departure in landing craft (LCVP's and LCM's) where the waves would be formed. From here they would continue in formation to the transfer control line, just beyond the reef's edge, to rendezvous with the returning LVT's and DUKW's which would shuttle the men and cargo in across the coral. Patrol craft and submarine chasers were stationed along both the line of departure and transfer control line to coordinate boats and vehicles in the ship-to-shore movement and to facilitate communications from ship-to-shore and among the various landing elements involved.

The methods of getting equipment and supplies ashore in the wake of the troops presented a few more or less novel aspects, which were to prove satisfactory in varying degrees. The amphibious trailer was used in quantity at Peleliu for the first time. Equipped with wheels and a metal cover which could be bolted down to make it water-tight all around, this vehicle was of a size which could be handled conveniently by a ship's crane. Hoisted from the hold of a transport and lowered over the side, it floated low in the water and could be towed as far as the reef's edge by any powered craft available, and lugged across the coral and on shore by an amtrack or DUKW, or even a tank or bulldozer.

Some 60 of these trailers were used. Apportioned among the infantry battalions, regimental headquarters, the artillery and division headquarters units, they were loaded with high expenditure rate items: mortar and machine gun ammunition, flame-thrower fuel, medical supplies, signal equipment and other technical gear easily damaged. The light artillery brought in their first unit of fire in them. As a means of bridging the gap in time between arrival of assault elements at the beach and establishment of the shore party ashore, the amphibious trailer proved its worth, for there were supplies at hand, drawn into cover on the beach and available if needed, with no shore party assistance required. True, the hitch provided was valueless while waterborne, but this was cured by a towing bridle of manila line that could be passed readily from boat to trailer


and could be cut should the trailer be hit, and such an arrangement did not affect the amphib's maneuverability. While a one-shot item of equipment, for re-loading at ship's side proved impractical due to low freeboard, the amphibian trailer did permit immediate response to a unit's need for supplies.

The use of pallets was another Army innovation, tried first at Kwajalein. The pallet was simply a heavy sled, measuring four by six feet, with wooden runners. Pallets were to be loaded in the staging area not to exceed limits of 3,000 pounds weight and three feet in height, the load being lashed down firmly with 1¼ inch steel strapping. The division used some 2200 at Peleliu to palletize a representative portion of all elements of the bulk cargo load. While results were generally satisfactory, the conclusion was reached that only barbed wire, screw pickets, and heavy shell lifts, or items not readily transportable in a cargo net, would thereafter be palletized. The objection lay in the requirement for a fork lift to handle pallets in each hold where they were to be stowed aboard ship, unless square-of-the-hatch space was used; and this space, generally, had to be held for high-priority vehicles and equipment.

The plan for handling supplies was based on all bulk cargo moving from ship to dump either in cargo nets or on pallets with cranes to be used at all points where the net or pallet had to be loaded, transferred, or unloaded. Commodore Buchanan, the Transron Commander, in one high-powered raid on the Advance Base Construction Depot on Banika, procured 36 Trackson cranes and six P&H five-ton crawler cranes for the division. At Guam crawler cranes had been placed on the reef. There they presented relatively stationary targets and some were knocked out. The division's veteran pioneers had used a crane mounted on a barge as far back as Guadalcanal and found it good. So why not mount cranes on self-propelled barges to establish a cargo transfer line 1,000 yards or so off the reef? The barges could move in or out depending on fire received from the beach; and so another shore party technique was born.

Pontoon barges were seven cells long by three wide, bolted together to form a single unit powered by an outboard motor. They were to be carried to target secured to the sides of LST's, from which they could be launched simply by cutting the lashings. Under their own power, each was to proceed to a designated ship to receive its crane. Special hold-downs were prepared in advance, and the crane's boom was rigged as the barge chugged to its appointed station. A landing craft loaded with netted or palletized cargo came along one side, a DUKW or LVT along the other, and the crane swung the load from the boat to the vehicle which carried it across the reef and to the dump where the load was picked out by crane. Nine barges were used at Peleliu with excellent results.

Six other LST's would arrive on the scene with pontoon causeways attached, one to each side. These were made up of 2x30 pontoon cells to a length of 175 feet. Once the beach had become sufficiently secure, they would be assembled to build a dry roadway across the reef from the shore to deep water where trucks could load direct from LST's.

Supplying water to the troops ashore constituted a particularly acute problem in the case of this operation. Peleliu had no surface water, and the heat was known to be peculiarly enervating. The Japanese used large cisterns to collect the frequent rain, but until the Marines should have facilities for collecting or otherwise producing drinking water, it would be necessary to supply all requirements from the ships. Five-gallon combat cans were gathered in quantities to be rushed ashore, and a reserve supply was put into 55-gallon oil drums which had been scoured out for that purpose. The latter expedient proved unsatisfactory; too often the scouring was insufficient to make the water palatable, and it was found that unless the drums were filled entirely flush the inside would rust and pollute the water.

Fortunately, it was discovered soon after landing that even shallow holes scooped in


PONTOON BARGES transshipping cargo from boats to amphibian vehicles near reef's edge.

the sand near the shore would yield drinkable water, and with the drilling of real wells and setting up of distillation apparatus, the problem passed the critical stage within a few days.

Preliminary Bombardment

Fast Carrier Task Forces had attacked the Palaus as early as 30 March, and they struck again in July and August. They were also to participate in the pre-D-Day bombardment and to furnish close air support for the landing. The Escort Carrier Groups, whose initial mission was to protect the transport convoys en route to the target, were to join in the close support for the landing. Once the assault troops were ashore, all naval aircraft in the area were to be available for special support missions on call.

The pre-landing naval gunfire objectives were the conventional ones: to soften up the target by knocking out enemy aircraft and artillery installations, and destroying as many strong points as possible. Because of the nearness of large enemy ground forces, a third naval gunfire mission assumed unusual importance: to destroy all ships, barges and sampans to be found among the islands to the northward, and to be prepared to intercept any attempts to reinforce the garrison from that direction.


The original plan called for only two days of pre-landing bombardment. Subsequent intelligence caused this to be revised upward to three days. Not all of this was fired, however. Although events were to prove the naval gunfire preparation at Peleliu the least adequate for its purpose of any Pacific operation with the possible exception of Tarawa, before the morning of D-Day Admiral Oldendorf, Commander Western Gunfire Support Group (TG 32.5), was complaining that he had run out of profitable targets.38

This view was justified to the extent that gunfire had demolished most of the more conspicuous aboveground installations: barracks, hangars, utility and administrative buildings, larger guns in open emplacements most of which had already been damaged to varying degrees by previous air strikes. Also, it had sheared a large part of the jungle from the northwestern ridges. But the Japanese, with ample shell-proof cover and operating according to a new tactical concept designed to cope with naval gunfire, suffered negligible casualties. And the sad truth is that the effect of this vast expenditure of ammunition was disappointing, insofar as disrupting Peleliu's basic defense system was concerned, (see Appendix C).

A more immediate factor contributed to this inadequacy. Unlike units which had been operating in the Central Pacific, the 1st Marine Division had never had to cope with this peculiar problem before and had no Naval Gunfire Officer organic to the organization.39 At the last moment Lieutenant Marvin P. Morton, USN, was flown in from Pearl Harbor to take over this function and performed excellently once the troops were ashore and liaison was all-important in controlling call fires. But the lack of an experienced officer in the post proved a severe handicap during the planning and preparatory stages.

Essentially the schedule and control of naval gunfire adhered to the pattern which had become standard long since. Pre-landing fire on D-Day morning would be concentrated on the beaches, shifting inland, then to the flanks, as the leading waves approached the shore. Special missions would be fired on call once the troops were safely landed. Eighteen LCI gunboats mounting rocket launchers were to blanket the beaches immediately ahead of the assault elements, while four other LCI's mounting 4.2" mortars were to concentrate on the rugged ground on the left flank and inland of the northern beaches from which trouble was anticipated, beaching on the reef if necessary to steady their aim.

Troop Training Program

Training of the 1st Division for the new operation proceeded along conventional lines, handicapped by certain complications which were something less than conventional.

One of these had to do with personnel. The division had withdrawn from Cape Gloucester to Pavuvu in May following an exhausting campaign which had protracted itself throughout the entire inhospitable northwest monsoon season on New Britain. Now all hands found themselves dumped on a rain-soaked, rat-infested hunk of real estate where virtually nothing had been done to prepare for their arrival.40 Thus, instead


of the "rest and rehabilitation" upon which a vigorous training program should be predicated, all hands were obliged to turn to on the construction of roads and camp facilities; and, as many men will never forget, the disposal of mutiple tons of fallen coconuts and palm fronds which had accumulated in the long-neglected plantations41--and which would continue to fall, as many a dented skull would testify.

The incidence of sickness shot upward, while morale plummeted to the lowest point it ever reached during the Pacific service of this elite outfit. The food situation was only one of several causes, but by no means the least cogent. Rations had been bad throughout the protracted Cape Gloucester operation, as is to be expected in an active combat area. At Pavuvu, the amply stocked 4th Base Depot on neighboring Banika was only ten minutes away by cub plane, the big base at Guadalcanal only about 60 miles, yet any change for the better was purely coincidental. Fresh meat was available for only one meal a week on the average, fresh eggs twice during this four-month stay.42

Casting about for indigenous sources of fresh food led to the discovery of large herds of cattle on this and neighboring islands which had been used by the departed planters to graze in the groves in order to keep the ground at the base of the coconut palms free of undergrowth. A call was sent out for all men experienced as cowboys and butchers. A roundup was held and a slaughterhouse constructed at cost of considerable labor and expense which was capable of supplying all hands with fresh beef two days a week. But just as this got into full production, as General Smith puts it in his Personal Narrative: "The Australians decided they did not want to engage in that type of reverse lend-lease operations and we were directed to cease."

Efforts of enterprising individuals to supplement their diet with alligator steaks had to be curtailed when the hunters lost direction in the dark (this was purely a nocturnal sport) and shots began falling in the bivouac areas. Fishing the lagoons with hand grenades and TNT was likewise forbidden after several men had been severely injured. The giant clam, captured in its prime, would feed a dozen men, but its meat proved no great improvement over the uninterrupted diet of spam and dehydrated potatoes twice a day, varied by powdered egg and an occasional can of C-Ration.

While the service troops on Banika and Guadalcanal had a reasonably adequate supply of beer with which to sustain their sagging spirits, there was none at all on Pavuvu


during the first weeks.43 Even after a regular ration was instituted, men considered themselves lucky to average three cans a week. Motion pictures were limited mostly to the poorest "B" films which troops in the rear areas, who had some choice in the matter, could not be bothered showing for themselves. The only USO show to reach this miserable hole was not scheduled to come at all; it arrived only by dint of the personal efforts of Bob Hope and at considerable inconvenience to his troupe, who managed to sandwich in a morning performance between rear echelon engagements shortly before the division shoved off for Peleliu.

Conditions were especially disheartening to the 24-months men, who still comprised a sizeable majority of the division, and who had been led to hope for early relief under the rotation plan, or at worst a return to Australia. By the end of July, a total of 4,860 replacements44 had arrived, making it possible to relieve slightly more than half the two-year veterans, which was very cheering to those actually returning to the U.S., but did little to raise the spirits of those left behind, many of them key specialists for whom competent replacements might never be available. And the new men, fresh from the amenities of Stateside life and requiring intensive special training, were poorly conditioned for the wretched life into which they had been so rudely pitchforked.

Yet, despite all the discouragements and handicaps, such was the spirit and resiliency of this veteran combat outfit, passed on by the old-timers to the infusion of new blood, that morale rebounded progressively as the prospect of going into action again became more imminent. Or perhaps, as many firmly believed, all hands were simply so disgusted with Pavuvu that they welcomed the opportunity to vent their resentment on the Japanese.45 Whatever the case, the number of sick in hospital had been reduced to 150, and the spirit of the division left little to be desired in the matter of fighting edge when the time came to embark for the operation.

A second factor hampering the training program was terrain: both its nature and extent.46

Except for fringing reef, Pavuvu shared practically no features in common with Peleliu. From the time of arrival it had been apparent that the island simply lacked the area of suitable ground necessary for the training of a unit the size of a reinforced division, and the weeks spent in making the place habitable (more or less) emphasized this feature as well as cutting seriously into actual training time. Units as small as battalions, maneuvering over terrain supposed


PACK HOWITZERS, first artillery emplaced on D-Day, were landed and supplied by amphibious vehicles.

to simulate the target, found themselves dodging among heads and mess halls and tripping over the guy ropes of their own tents. Conditioning hikes, so useful in hardening the men physically for the ordeal to come, were possible only by marching the troops monotonously around in constricted circles, usually colliding with other units doing the same thing. Weapons and combat ranges were constructed which were adequate in themselves, but neither large enough nor numerous enough to serve adequately the number of men who had to use them.

If Pavuvu was inadequate for infantry training, it was something worse than that for the supporting arms with bulky equipment. Yet despite handicaps of space, the 1st Tank Battalion spent a full day with each infantry battalion, during which time each squad practiced coaching movement and fire by tank-infantry telephone and visual signals, an expedient which went far to foster mutual understanding, cooperation and esteem.47 But the artillery "was reduced to the pitiful expedient of firing into the water with the observers out in a boat or DUKW."48

Altogether, the situation was one calling for all the resourcefulness and ability at improvisation for which the Marine Corps is


noted. When large unit training proved unfeasible, the emphasis was placed on small unit exercises until transportation could be obtained for rehearsals on more suitable terrain elsewhere (see below). Intensive schooling was instituted for junior officers and NCO's. Congestion of the ranges was largely solved by adhering to an exceedingly tight schedule calculated to wring out the last ounce of usefulness from the limited facilities.

That a division containing 40% replacements should emerge even half trained under the circumstances would have been a minor wonder. That actually the ground training of all hands proved something more than merely adequate constitutes a high tribute to the efficiency and cooperation of officers and men of all units on all levels.

Shortage of Equipment

The amphibious phases of training, in particular, were further handicapped by a sad lack of practically everything necessary to carrying them out; shortages which, constituting a mere retarding factor at the beginning, grew to be a matter of major concern as the time for action approached.

Between the jungle and the monsoon, the four-months Cape Gloucester operation49 had been especially debilitating to the 1st Marine Division's rolling stock, tracked vehicles and communications equipment;50 and many expendable items had been left on New Britain under an exchange arrangement with the Army. This plan proved to be something less than a boon to the division since many of these items proved later to be in short supply in Army depots at Guadalcanal. The demands of the long-drawn-out Marianas campaign further delayed the acquisition of replacement equipment interminably, or so it seemed to the division's worried officers.

Yet once more energy and resourcefulness overcame all obstacles--or a good many of them, anyhow. Few deficiencies in equipment or technique were apparent on D-Day and the critical days that followed.

It should be remembered that up until this time the 1st Marine Division had never made an assault landing across a wide fringing reef. Although most of the veteran personnel were familiar with the various landing craft and vehicles employed in such an operation, considerable training was required in the specialized adaptation of these to this particular problem: launching of amphibian tractors and DUKW's from LST's at sea, transfer of troops from landing craft to vehicles in deep water, advance in waves across the coral, and quick formation of the land assault once disembarked. Thus, the shortage of all this equipment presented a serious problem. Its solution involved a striking example of that improvising on which the Marine Corps has had to depend so often.

The 1st Amphibian Tractor Battalion was still organic to the division51 upon arrival in the Russells, but it had few LVT's still operative. To supplement these, IIIAC sent over Army DUKW's. In mid-June 40 LCVP's arrived from Tulagi. These were surveyed craft but adequate for their present jobs, however inadequate in numbers. No LST's became available until August, but with this skeleton set-up the division undertook a program of piecemeal training of


all personnel in operations across reef. It was while observing such a landing drill that the commanding general fell from an amtrack and sustained the severely fractured ankle which nearly caused him to be left behind on the operation.53

Then, to complicate this makeshift set-up even further, in early July the 1st Division was directed to form two provisional amphibian tractor battalions within the organization: the 3d Armored Amphibian Battalion (Provisional) and 6th Amphibian Tractor Battalion (Provisional), "utilizing personnel of the 1st Amphibian Tractor Bn., augmented by personnel from units of the division."54

Here was a task of daunting magnitude. The only trained LVT men available were those of the single unit already in existence; thus the 1st Amphibian Tractor Battalion was obliged to divide itself into thirds, so to speak, in order to furnish each of the new units with a cadre of experienced officers and men, and at the same time retain a similar cadre for itself. Utilizing "personnel from units of the division" to build the three battalions up to strength was easier said than done: units preparing for combat are not prone to part with any of their men without a fight. And even when sufficient personnel had been obtained for all three battalions, it would be necessary to train them in a number of vehicles grossly insufficient for a single battalion and already working full time on the landing exercises of the assault troops.

Sufficient LVT's were obtained in time for the operation, but competent personnel remained a problem throughout. Because of the comparatively small size of the target, it was not anticipated that transportation on shore would constitute a major problem; hence only one company (Company A) of the 1st Motor Transport Battalion was to bring in its organic equipment, including trucks and repair facilities. This made it possible to detach Company C in toto and assign its men to the 1st Amphibian Tractor Battalion, in which connection they served throughout the operation.55 But even with the addition of this manpower, the latter unit was still so under strength that it proved necessary to reduce the basic crew of each vehicle from three men to two.56

Thus, the new amtrack crews were being broken in at the same time that the infantrymen were being trained for beach assault, and in the same vehicles. And such were the demands on this overworked and insufficient equipment that the artillery had a minimum of time to practice their own new assault technique: loading and unloading of their 75mm and 105mm howitzers in LVT-4's and DUKW's, respectively; also, the loading and unloading of radio jeeps in DUKW's, using the A-frames.

The problem of organizing the armored amphibian battalion multiplied these complexities. The LVT(A), or "amphibian tank," was a newly developed vehicle which had been used in combat heretofore only in the Central Pacific. Few officers of the 1st Marine Division had so much as laid eyes on one.57 And at the outset none were available


here even for demonstration purposes. Thus, not only must sufficient personnel be obtained, but these must be trained initially from blueprints in the operation of a vehicle they would be called upon to take into battle.

The man appointed to this seemingly insuperable task was Lieutenant Colonel Kimber H. Boyer, until then commanding officer of the 1st Motor Transport Battalion. For the first ten days his entire officer complement consisted of one first lieutenant and two warrant officers, and to initiate training he had to borrow a weapons expert from the 7th Marines.58 His personnel arrived in driblets from various sources: rear echelons of other armored amphibian battalions and tank battalions, and men shipped out from the Tracked Vehicle School in the United States. Few had had any combat experience at all; of the 36 officers and 800 enlisted men who finally comprised the battalion, only approximately 50 had ever fought an LVT(A).

Intensive training began with delivery of the first six vehicles: LVT(A)(1)'s. By the first week in August the battalion had acquired 50 per cent of its machines and 60 per cent of its personnel, when disturbing word arrived that its remaining quota of vehicles would be of a different model: LVT(A)(4)'s, mounting 75mm howitzers in place of the 37's with which the men had been trained up to this point, with embarkation only about three weeks away. That the battalion should turn in an outstanding performance after such unpropitious beginnings might well rank as one of the minor miracles of the campaign.59

Shortages of critical equipment were not limited to heavy assault vehicles. They embraced everything from tank spare parts and communications equipment to bazookas, BAR's, and spare machine gun barrels, to such a degree as not only to hamper training but to cause serious concern over the forthcoming operation. This was notably true in the case of the pack type flamethrowers which arrived too late to be distributed according to plan. Final allotments in many categories arrived barely in time to be combat loaded, and a few had to be crammed aboard the transports after the troops had embarked.60

In general, it may be said that the only items in which shortages did not exist were the basic arms of the individual Marine. And even here there was some minor difficulty: the decision to replace carbines in infantry units with rifles and submachine guns could not be implemented immediately for lack of the latter weapons.61

The equipment finally delivered was the most up-to-date available, much of it improved over earlier models with which the 1st Division was familiar.62 Only two entirely


new items worthy of mention were issued at this time.

Most notable was the Navy Mark I flamethrower, an experimental weapon developed by the Seabees at Pearl Harbor, which was capable of throwing a stream of blazing Napalm up to 150 yards. The first three of these were furnished the division in July, together with four LVT-4's on which to mount them. Their primary function was thought to be immediate action against beach pillboxes in support of the assault waves, though actually they proved their greatest usefulness in subsequent developments farther inland. The flame units were installed on three of the LVT-4's as prescribed, the fourth amtrack being adapted as a supply carrier to service them in the field.

The second new weapon was the 60mm shoulder mortar, a standard mortar adapted to a light machine gun mount for direct, flat-trajectory fire against caves and pillbox openings. Some of its parts proved not rugged enough for sustained use and had to be replaced nearly as often as the poor devils who were obliged to fire the contraption from their shoulders. Essentially its functions duplicated those of the bazooka, and its adoption resulted from the frequent failure of bazooka rockets to detonate in the soft mud of Cape Gloucester. However, there proved to be very little soft ground amid

LONG RANGE FLAME-THROWERS were first used at Peleliu. LVT mount was thereafter replaced by tank.


FINAL CRITIQUE, following last rehearsal at Guadalcanal. Assembled Marine and Navy Officers are here addressed by First Division chief of staff. Front row: Generals O. P. Smith and W. H. Rupertus, and Captain T. B. Brittan, USN, commander of the transport group.

the coral of Peleliu and, although the shoulder mortar was used effectively, it came to be considered less practical all around than the weapon it had been designed to replace.63

The training program culminated in two full-scale rehearsals, held on 27 and 29 August respectively, in cooperation with Task Force 32.1 (Peleliu Attack Force). By this time the assault troops were all embarked on the vessels which were to carry them to the target (see below) and standing by off Guadalcanal, and the terrain selected for these maneuvers was the Tassafaronga region of that island.

No coral reef exists here, but this was readily simulated by setting up an arbitrary transfer control line at the proper distance from the beach where the landing craft carrying the later waves halted to rendezvous with the returning LVT's which shuttled the men in from there. Once ashore, the assault units moved out at once on a reasonably accurate facsimile of their missions, and an


advance division command post was set up to coordinate their movements.

The first of these exercises had been described as a "pre-dress rehearsal."64 Naval gunfire and air support units went through their respective motions, but there was no firing. Everything progressed with pleasing smoothness up to the point where the division command post attempted to establish communication with the regiments. This was eventually achieved by field telephone but the radio proved useless, the new equipment having arrived too late to be calibrated accurately and set up in accordance with the plan.65 This was quickly rectified, however, to such good effect that the second rehearsal, using token quantities of live ammunition, aerial bombs and rockets, came off so well that a critique held the following day at IIIAC headquarters produced little that was constructive because "everyone was pleased with everyone else."66

A shore party exercise at Tetere Beach on 3 September, mainly to perfect the setting up of communications, completed the training phase. For better or for worse, the 1st Marine Division and its reinforcing troops were as ready for combat as they would ever be.

Loading and Mounting Out

Availability of shipping had always been one of the critical controlling factors in the U.S. Pacific offensive, involving as it did such vast distances between base and target. Not since the early stages of the war, when ships in sufficient numbers simply did not exist, was the shortage felt so acutely as just prior to the Peleliu operation. Again, responsibility could be attributed to the long delay in the Guam invasion. The sortie of the Japanese Fleet into the Philippines Sea might be rated as a major enemy disaster on the strategic level, but as a means for providing headaches for the 1st Marine Division it proved an unqualified success.

Nobody knew when the necessary shipping would arrive, except that it became increasingly apparent that this was not going to be until the last moment. The elaborate loading plans on which such an operation is predicated were meticulously worked out on the general level, but detailed planning was impossible in advance owing to lack of knowledge as to exactly which ships would be available and the individual capacities and peculiarities of each. And the inevitable misunderstandings among the diverse elements in this highly complicated business necessitated changes, compromises and improvisations up to the last minute.

The problem was not simplified by the fact that loading operations had to be conducted at five widely separated points: Pavuvu, Banika, Guadalcanal, Tulagi, and Espiritu Santo (New Hebrides), the last named being the base of the Marine air squadrons designated to operate from the field on Peleliu once it had been secured. Several of the ships, indeed, were scheduled to pick up portions of their cargo at two or more of these separated points. Careful coordination was necessary to avoid wasted effort and backtracking. Especially at Pavuvu and Banika, beaches, lighters and dockage facilities were insufficient to load more than a few vessels at a time; thus, when most of them appeared simultaneously with little time to spare, they were obliged to take their turns in accordance with an exceedingly tight schedule and keep out of each other's way while doing so.

It was a step by step process, entailing close coordination and prodigious labor. The LST flotilla had to be assembled from several areas of the Central Pacific, and the first units did not begin arriving until 15 August. The last ones appeared only on the 26th, by which time the troops had been embarked in preparation for their landing exercises. Despite this, all 30 LSD's, in addition to 17 transports and two LSD's, had been combat loaded by the 31st.


Following the rehearsals, the convoy stood by off Tetere Beach while the 81st Infantry Division, which had assembled in the area by 26 August, took over Tassafaronga for its own final rehearsals for the assault on Angaur. While awaiting completion of these, the Marines were landed daily for recreation and small unit exercises, and the transports moved by turn over to Tulagi for fueling.

With everything in readiness at last, the slower moving tractor groups of the two divisions, aboard LST's (average speed, 7.7 knots), departed Guadalcanal with their respective screening forces early on the morning of 4 September. The faster transports and LSD's (averaging 12.1 knots) left on the 8th to rendezvous with them during the early hours of D-Day off the target which the fire support group would have been softening up for three days by then. Their course took them northwestward through the Solomons, then across the Equator on an approach generally parallel to the northern coast of New Guinea: a distance of approximately 2,100 miles. The sea was smooth throughout, the run uneventful--and miserably uncomfortable for the men, as such trips inevitably are in the tropics.

Previous to departure from Guadalcanal, a sealed envelope had been issued to each of the civilian news correspondents assigned to cover the expedition, and to the troop commanders aboard several of the ships, marked with explicit instructions that it was not to be opened until D-minus 1, at which time the contents was to be broadcast to all hands. On the morning of 14 September, therefore, the seals were broken and the recipients learned, to the considerable astonishment of many of them,67 that the division's commanding general expected the operation to be very tough but very short, comparable to Tarawa, and that Peleliu should be secured within four days: perhaps the most striking manifestation of that preoccupation with speedy conquest at the highest division level which was to color tactical thinking ashore for a month to follow.68


Table of Contents ** Previous Chapter (I) * Next Chapter (III)


1. At this time Admiral Halsey held both of these titles as a dual command. VAdm J. H. Newton relieved him as ComSoPac on 15 June, and Halsey removed headquarters, Commander Third Fleet, to Pearl Harbor on the 17th.

2. III Amphibious Corps Report on Palaus Operation, hereinafter cited as IIIAC OpnRpt; Enc. A, 1.

3. Third Amphibious Force (CTF 31) Report on STALEMATE 2 Operation, 1, 2, hereinafter cited as CTF 31 Rpt.

4. This group was then headed by Col Dudley S. Brown. It was detached from IIIAC on 30 May in accordance with Corps Special Order #25-44 and was operative at Pearl Harbor on 12 June. It was subsequently designated X-Ray Provisional Amphibious Corps (see below) and still later Expeditionary Troops Third Fleet, under which title its Report Palau Operation was issued. Hereinafter cited as ExpTrps OpnRpt. Source of above material is Enc B, 2, of this report.

5. Result of unexpected difficulties on Saipan and sortie of the Japanese fleet, culminating in First Battle of the Philippine Sea. Attack originally scheduled for 18 June did not take place until 21 July, and Guam was not declared secure until 10 August.

6. As early as 13 June a dispatch arrived from the Joint Chiefs of Staff requesting the views and recommendations of top Pacific echelons on the feasibility of by-passing the Western Carolines and other intermediate targets in order to advance the date for striking major objectives closer to Japan; notably Formosa. See Appendix B.

7. CTF 31 Rpt, 2, 3. IIIAC OpnRpt, Enc A, 2. ExpTrps OpnRpt, 2, 3. Interview LtGen J. C. Smith, 23Nov49, hereinafter cited as J. C. Smith Interview.

8. CTF 31 Rpt, 4.

9. ExpTrps OpnRpt, Enc B, 4.

10. Hq Commander Western Pacific Task Forces, Operation Plan 14-44, 1Aug44.

11. 3d Flt AR, 3, 4.

12. CTF 31 Rpt, 8.

13. 3d Flt AR, 4. For further detail regarding this dramatic change in plans, and the thinking that led up to it, see Appendix B, this monograph.

14. Discussions of one phase or another of planning and preparation occur in reports of all the higher levels of command previously cited. In accordance with the policy of this monograph of drawing on the echelon most directly concerned, this account, except as otherwise noted, derives mainly from 1st Marine Division Special Action Report, Palau Operation, hereinafter cited as 1st MarDiv SAR. This work is divided into two parts: Phase I--Planning and Training (Annexes A to E inclusive); Phase II--Operational Phase (Annexes A to L inclusive).

15. A graphic and accurate description of life on Pavuvu appears in 1st MarDiv unit history: The Old Breed, by George McMillan (Infantry Journal Press, 1949), Chap. 16. Hereinafter cited as The Old Breed.

16. As previously noted, General Julian Smith retained command of the next higher echelon: Expeditionary Troops Third Fleet, which included both IIIAC and XXIV Corps until removal of the latter from the area with cancellation of Phase II of STALEMATE II.

17. JANIS #103.

18. Interrogations of Japanese Officials, U.S. Strategic Bombing Survey (Pacific), Vol II, 432.

19. Shortcomings of this map in one particularly important small area led to criticisms which have tended to obscure the fact that for the greater part of the island it was excellent.

20. ExpTrps OpnRpt, Enc B, 5, 6.

21. To facilitate such identification in future operations, the chief API officer, Captain Jerome J. Foley, set up two special teams whose mission after landing was to photograph, sketch and describe fully all enemy installations immediately after they were captured. These teams were made up of trained intelligence personnel, with photographers specially attached from the Public Relations Section. The result was a remarkably complete coverage of enemy installations and weapons showing them both as they appeared on the ground and as they looked in the aerial photographs from which identification must be made, issued under title "Peleliu: Its Terrain and Defenses".

22. BrigGen O. P. Smith in special Memorandum, Marine Corps Schools, Quantico, dated 4 February 1947.

23. Work of the UDT's at Peleliu was especially fine: reporting on reef conditions, destroying obstacles and mines, and scouting approaches to the beaches themselves. IIIAC OpnRpt, Enc C, 2. An earlier attempt had been made in mid-July by volunteer UDT men operating from a submarine, but they were able to reconnoiter one beach only. In a similar attempt at Yap three UDT men were lost and subsequently discovered to have been captured by the Japanese. CTF 31 Rpt, Enc A, 1.

24. Actually a closer parallel was Biak (17 May-26Jul44), in the Southwest Pacific, where the 41st Infantry Division had to cope with terrain problems strikingly similar to those encountered on Peleliu. That could not be foreseen in advance, however, and there was insufficient time for thorough evaluation of the lessons learned in that operation.

25. Oral statement to the author by MajGen O. P. Smith, 13Jan49. General Smith added, in effect, that while he did not think that many responsible officers of the division gave much weight to the Tarawa comparison, certainly even fewer, himself included, ever imagined that Marines would still be fighting on Peleliu a month after the landing.

26. ExpTrps OpnRpt, Enc B, 4.

27. These were a temporary mortar company, a machine cannon company and some sort of a heavy rocket unit about which nothing definite was ever learned. All three are believed to have been improvised on the spot from personnel detached from the infantry units.

28. The lower figure squares neatly with the final estimate of the 81st Division, which reported 1,338 Japanese killed on Angaur, 59 taken prisoner. 81st Infantry Division Operation Report, Phase I, 69; hereinafter cited as 81st OpnRpt.

29. Above tabulation transcribed literally from F-2 report dated 28Aug44. ExpTrps OpnRpt, Enc B, Appendix 10, 1.

30. Further identified subsequently as 3d Bn, 15th Infantry. 2d Bn same regiment also participated in defense of Peleliu though not on the island on D-Day. See Chap V.

31. Further identified subsequently as 346th Independent Infantry Battalion.

32. This figured in the planning for some time, nonetheless, and was to be the last alternative discarded. From enemy dispositions on D-Day, there is reason to suspect that they may have got wind of it: Whereas elsewhere on the island, Japanese sector defense units displayed a high degree of mobility, on southern Peleliu they allowed a strongly reinforced battalion to be cut off and annihiliated. MajGen O. P. Smith, interviewed 4Apr50.

33. A detailed and scholarly analysis of the many factors considered in the several alternatives was prepared by BrigGen O. P. Smith in a special memorandum dated 4Feb47.

34. As evidenced by the following studied opinion: "None of the remaining beaches which might permit a landing in force would allow the rapid development of an adequate beachhead which is so essential in a landing operation. . . . The correctness of the decision to land on the WHITE and ORANGE beaches is hardly open to question." Ltr BrigGen W. A. Wachtler to CMC, 1Mar50, hereinafter cited as Wachtler.

35. The original plan drawn up by X-Ray staff called for operational attachment of one regimental combat team of the 81st Division to serve as division reserve, all three Marine regiments to land in assault. From the first General Rupertus displayed that marked reluctance to use Army troops which was to color the first week of the operation. Accordingly, the division plan called for landing three regiments abreast, less one BLT which would be the sole division reserve. This was predicated on the presumed availability of at least some part of the 81st Division, one RCT of which had been designated corps reserve. J. C. Smith Interview.

36. This highly successful innovation was tried first at Peleliu, as a result of tank casualties and landing delays in the Marianas. An officer of the 1st Tank Bn explains it as follows: "An LVT was placed on each LCT to lead the tanks ashore. These LVT's were used to test the depth of the water, and as long as they propelled themselves along the bottom the tanks would follow, but if the LVT's became waterborne the tanks would stop until the LVT's could reconnoiter a safe passage. . . . Fuel, ammunition and maintenance supplies were loaded on these LVT's which enabled the tank units to have a mobile supply dump available to them upon reaching the beach." Ltr Capt G. E. Jerue to CMC, 2Mar50, hereinafter cited as Jerue.

37. Each APA on the scene was also prepared to care for a limited number of wounded (average 25) who would be carried out direct from the beach until adequate hospital facilities could be set up on shore. See Chap. IV.

38. "I recall vividly the reaction of General Rupertus and LtCol Fields upon hearing CTG 32.5's plain-language voice transmission to the effect that he had run out of targets. Their reaction was that of incredulity." Ltr LtCol F. A. Ramsey, Jr., to CMC, 28Feb50, hereinafter cited as Ramsey. "The dispatch sent by Adm Oldendorf was not only a surprise but was not understood by any of us on the Division Staff in view of the study and requirements we had submitted, and the plans which had been so carefully prepared and agreed to as essential." Ltr LtCol L. J. Fields to CMC, 17Mar50, hereinafter cited as Fields. See also Chaps. III, IV, VIII, this monograph.

39. 1st MarDiv SAR, I, Annex K.

40. It was General Geiger's idea to base the division in the Russells in order to avoid its having to furnish large working details, as was the rule on Guadalcanal (average of 1,000 men a day had been the case with the 3d Division), which would interfere seriously with the training program. Reconnoitering for a suitable spot in the immediate area, the general selected Pavuvu, where there was a battalion of Seabees to prepare facilities. Unfortunately, his reconnaissance was made by air and failed to disclose the more serious terrain difficulties. And the Seabees, who were due for early transfer Stateside, apparently misunderstood their orders. They had commenced construction of an excellent steel pontoon pier and laid out a rudimentary road, but that was about all.--Oral statement, BrigGen J. T. Selden, 4Mar49. This is corroborated by Gen. O. P. Smith who adds that when the true situation became known, General Geiger offered to have the division transferred to the 3d Division's camp on Guadalcanal upon that unit's departure for Guam. However, there was a dearth of shipping available to effect such a transfer, and by that time many of the difficulties on Pavuvu had been overcome. Thus, it was felt that another move would serve only further to retard the training program. BrigGen O. P. Smith Personal Narrative, hereinafter cited as O. P. Smith PerNar.

41. The plantations had been abandoned for nearly three years and many of the nuts were in advanced stages of decay, pervading the island with a faint odor of rot which is difficult to describe and equally difficult to forget. Many a veteran of Pavuvu remains allergic to coconuts to this day.

42. "The men were not in very good physical condition. The purpose of the move to the Russell Islands was to rehabilitate the Division for future operations. Authority was granted to increase the ration twenty-five per cent. It was assumed that fresh provisions in reasonable quantities would be furnished. We built reefer boxes on the beach to receive these stores but the reefer boxes were generally empty." O. P. Smith Op. Cit.

43. There were a few exceptions where individual officers were able to get over to the base depot and persuade the powers there to let them buy out of their own means enough to furnish all hands within their own commands a token drink. This was the first beer most of the men had tasted since their departure from Melbourne nearly nine months before.

44. As the division was finally shaken-down for the Peleliu operation, approximately 30% of the men had been in the Pacific in excess of 24 months and served through two debilitating jungle campaigns; another 30% had 12 months and one campaign under their belts, while the remainder were replacements in various stages of seasoning. As embarked, the division was about 5% above T/O strength. 1st MarDiv SAR, I, Anx. A.

45. The theory is held in some military circles that good treatment only softens first rate assault troops; that the rougher they get it when not in combat, the tougher they will be when they are in combat. Thus, the belief prevailed among many of the officers and men of the 1st Division that the ubiquitous discomfort and privation which fell their lot was the result of deliberate policy, rather than neglect and/or callous indifference. It was certainly true that other Marine combat units had enjoyed comfortable base camps in such pleasant places as New Zealand and the Hawaiian Islands, and even the comparative civilization of New Caledonia and Guadalcanal; whereas the 1st Division, except for the interlude of its happy sojourn in Australia, never got out of the boondocks from the beginning to the end of the war.

46. "It took no great discernment to see, even from the air, that the place was entirely unsuitable. . . . With the exception of an overgrown and abandoned coconut grove, the island was a veritable jungle in which a troop unit even as small as a platoon would have great difficulty maneuvering." Wachtler.

47. Ltr LtCol A. J. Stuart to CMC, 25Apr50, hereinafter cited as Stuart.

48. Ltr LtCol L. F. Chapman, Jr. to CMC, 5Mar50.

49. This operation had lasted throughout the entire season of the northwest monsoon, when rainfall up to nine inches was recorded for a single 24-hour period. In addition to vital combat equipment, nearly everything the division owned had been ruined, including canvas tentage, field kitchens and clerical gear.

50. As a component of the VI Army, the division had operated at Gloucester under Army communication plans and instructions. Not only were its officers wholly unfamiliar with the Central Pacific Communications Plan prescribed for the Peleliu operation, but because of the division's isolation from its higher echelons there was a long delay in obtaining copies of this plan for instructional purposes. Ramsey.

51. The 1st Division reorganized under the new T/0 while at Pavuvu. This decreed, among other provisions, that amphibian tractor battalions should be detached from the divisions and pass under corps control.

52. A further handicap here was that the DUKW is by no means an adequate substitute for the amtrack in reef crossing. It is easily stopped by rugged coral and must pick its way ashore with great care, thus breaking the wave formation and falling more or less automatically into column.

53. "General Geiger visited me at the CP during the first rehearsal (27 August). He had not realized that General Rupertus was unable to make the trip ashore and was somewhat concerned about him going on the operation. I assured General Geiger that in the two weeks remaining before the landing I felt General Rupertus' ankle would mend sufficiently to permit him to carry on." Smith PerNar.

54. 1st MarDiv SAR, I, 5.

55. Company B personnel served as stretcher bearers and relief drivers. Ltr Maj R. B. McBroom to CMC, 13Mar50. Ltr Maj G. J. DeBell to CMC, 20Mar50, hereinafter cited as DeBell.

56. Ltr Maj A. F. Reutlinger to CMC, 10Mar50, hereinafter cited as Reutlinger. This letter contains a wealth of graphic detail regarding the problems encountered in this reorganization.

57. "His (the CO's) only knowledge of the vehicle he was to fight was a picture seen in ONI 226." LtCol K. H. Boyer, monograph prepared at Marine Corps Schools, March 1947, hereinafter cited as Boyer Mono.

58. This was Maj J. E. Buckley, CO Wpns Co, 7th Marines. In a letter to CMC, 10Mar50, LtCol Buckley (Ret.) wrote: "Kimber Boyer did one of the greatest training jobs I ever saw or heard of."

59. Col Boyer himself viewed the "miracle" in a somewhat different light: "It would be an insult to the intelligence of a military minded reader to say that the Battalion was sufficiently trained and indoctrinated to carry out its mission. It can be considered some sort of a miracle that the initial phase of the Operation was not seriously disrupted." Boyer Mono, 17.

60. A full allowance of new dry cell batteries for the vital TBX radios, issued at the last moment, were discovered to be essentially all dead. Fortunately, 4th Base Depot was able to duplicate the issue the same day. Ramsey. "Belts of machine gun ammunition had rotted . . . powder rings on mortar ammunition were disintegrating and bourrelets rusted, shotgun shells swollen or if brass corroded. All ammunition had to be unstowed, inspected and in large part replaced and restowed at the last minute." Ltr LtCol S. S. Berger to CMC, 19Mar50, hereinafter cited as Berger.

61. "Carbines were retained generally by personnel assigned to crew-served weapons and others who formerly carried a pistol. The attitude of the battalions was that the carbine without bayonet was an adequate substitute for a pistol, but not for an M-1 rifle with bayonet." 1st MarDiv SAR, I, 5. "M-1's and TSMG's were substituted in 2/7 for carbines. Machine gun squad leaders carried TSMG's." Berger.

62. The 1st Tank Bn, for instance, which had begun the war with gas-engine light tanks and subsequently acquired one company of gas-engine mediums for the Cape Gloucester operation, now found itself equipped entirely with diesel-engine mediums. Stuart.

63. "As presently constructed, they (shoulder mortars) are too heavy and certain parts are too weak. . . . After firing two to four rounds, it is necessary to replace the gunner. Units feel that the shoulder mortar as now constructed is not of sufficient value to include it in the authorized weapons." 1st MarDiv SAR, II, Annex A, 5.

64. O. P. Smith PerNar.

65. 1st MarDiv SAR, I, Annex E. For amplification see Ramsey.

66. O. P. Smith, op.cit.

67. "I doubt that anyone with the possible exception of the Chief of Staff was consulted before the letter was written and distributed. . . . I was given a complete rough draft of the letter and merely supervised its reproduction." Ltr LtCol W. E. Benedict (D-1) to CMC, 27Feb50, hereinafter cited as Benedict.

68. Most officers believed this unusual document intended in the nature of a "pep talk." This was not its effect on the news correspondents, however: many of the 36 accredited to the division did not come ashore at all, and only six (one of whom was killed) chose to stay through the critical early phases. Hence, news coverage of the operation was sketchy, often misleading, and, when quick conquest failed to materialize, tinged with biting criticism.

Transcribed and formatted for HTML by Jerry Holden for the HyperWar Foundation