Second Operational Phase:
D-plus 1--D-plus 7
(16-22 September)

The events of D-Day revealed to a great extent the pattern of the Japanese defense of Peleliu, which subsequent developments would substantiate. In the main, this conformed to that of previous island defenses, with certain variations and refinements which showed that not all of the lessons of previous disasters had been lost upon the enemy.1

They still displayed that fundamental inability, stemming from lack of coordination among their scattered observation posts, to mass the fires of their heavy weapons. On Peleliu, terror of U.S. air observers provided a further inhibition. Discovering early that spotting of their positions invariably brought down a deluge of heavy gunfire and bombing, they resorted to the practice of keeping their artillery and mortars concealed in caves and other emplacements, running them out to let off a few rounds, then pulling them back to cover in all haste. This weakness was not readily discernable to the amtracks crossing the reef nor to the infantry operating on the open airfield; only after the division discovered how much heavy ordnance the Japanese actually had available did it come apparent how infinitely worse the destruction might have been.2

As always, they displayed their childlike faith in the tactic of infiltration, though their aptness here exceeded anything they had shown previously. Enemy soldiers knew their own sectors thoroughly, and individuals and small groups moved with certainty toward definite objectives over familiar terrain. Their main purpose was to reoccupy emplacements already overrun from which to harass the Marines' rear areas from concealment. That three of them penetrated into the heart of General Smith's command post on the night of D-Day indicates some measure of their success.3 Similar instances occurred throughout the operation, always with disquieting effect; but infiltration it still remained, with negligible influence on the ultimate outcome.

There had been no banzai charges on D-Day, and there were to be none during the entire operation. The counterattacks had been well planned, if poorly timed, and so they continued to the end: usually short, sharp thrusts with limited objectives. Both


RUINED JAPANESE INSTALLATIONS in vicinity of Administration Building. Above ground targets such as these were pulverized by naval gunfire which left many hidden positions untouched.

the Iwo Jima and Okinawa operations would follow this pattern, but because Peleliu set the precedent the seeming phenomenon puzzled many people at the time. Throughout the campaign Marines continued to watch hopefully for some crack in morale to produce that peculiarly Japanese manifestation which had so greatly simplified the mop-up of other islands. But the troops here included some of the finest Japan ever put in the field, under officers who had learned the futility of pitting Bushido against fire power and steadfastly refused to play into their enemy's hands.

The Japanese plan consisted basically of an organized and carefully integrated defense in depth, but it possessed far greater strength and flexibility than anything yet encountered in the Pacific.

This increased strength resulted from skillful utilization of naturally difficult terrain, combined with time and materials for developing defensive positions of maximum effectiveness. As noted in Chapter 1, the fall of the Gilberts and Marshalls had confronted the Japanese with the problem of strengthening hurriedly their second line of defense. Miscalculating our initial objective, they had given first priority to the Palaus at the expense of the Marianas. Thus, Marines attacking Saipan, Tinian and Guam had found the defenses incompletely developed,


with quantities of heavy weapons and fortification materials lying around for lack of time to put them to use.4 The Japanese on Peleliu not only received the wherewithal first, but an added three months grace in which to make the most of it. They did.

The flexibility stemmed from a system of zone defense which combined features of the mobility necessary on a land mass the size of Saipan, and the dug-in defense-to-the-death characteristic of the much smaller islands of Tarawa and the Marshalls. Uncertain exactly where our landing would be attempted, Colonel Nakagawa set up his defenses in four zones (see Map 5), each assigned to one reinforced battalion, with the remaining battalion held in general reserve in the ridges of the northwestern peninsula. Operating on interior lines, troops occupying the unaffected zones were in a position to move swiftly to the reinforcement of whichever zone was attacked once our intentions became known.5

However, the successful drive across the island on the afternoon of D-Day effectively isolated the battalion defending the southern zone, members of which were thenceforth fated to fight to the death in their holes in the conventional Japanese manner, unable to furnish direct support to their compatriots elsewhere. This circumstance, combined with the failure of the major D-Day counterattacks, determined Colonel Nakagawa to shift his strategy to a last ditch stand in the most formidable terrain available. As events soon proved, troops originally posted to defend the eastern zone abandoned some of the most formidable installations on the island (Beach PURPLE) and retired to the ridges, where they redeployed in conjunction with the troops already in that area. Into these same ridges came other Japanese, falling back before Marine pressure, there to make the conquest of the island as costly as possible.

In this, no doubt, they were influenced to some extent by their advance intelligence. This was somewhat better than in many instances: the garrison was alerted as to the probable time and place of the coming attack as early as 3 September, and just prior to D-Day General Julian Smith was identified in command, though placed in the wrong area (Morotai). Within five days the landing force had been correctly estimated as "probably a division." However, the same entry (20 September) reveals a serious misapprehension regarding the invaders' intentions which colored Japanese tactical and strategic conceptions for some time to come:

It seems as if they (U.S. forces) had been planning on taking up their positions in Koror and Palau Island (Babelthuap) following the main forces' occupation of Peleliu. However, the enemy was unable to withdraw its main forces from Peleliu, because they had received heavy losses there. Also, they were constantly harassed by the danger of being attacked by our aggressive artillery from Angaur, and by counterlanding on Peleliu which may be staged by our Angaur forces. Because of these facts the enemy were compelled to change their plan and land their reserve forces on Angaur.6

These developments in enemy thinking could not be known to the invaders on D-Day, nor for some days to come. Hence, throughout the second operational phase the attack continued in accordance with the original plan, even while gains actually achieved fell further and further behind that optimistic schedule.

The commanding general and his staff landed at about 0950 on D-plus 1 and took over direction of the operation, occupying the advance command post which General Smith had set up the previous day. As described in Chapter III, this was located in an overrun Japanese antitank trench a short


SOUTHERN PELELIU, thoroughly fortified and defended by a reinforced battalion, fell to the 7th Marines by the afternoon of D-plus 3.

distance inland of Beach ORANGE 2. By digging into the reverse bank of this sharp declivity, the division's nerve center achieved excellent protection from the ubiquitous enemy shells which continued to come snarling and screeching into the area in slowly diminishing numbers for the next week. These blew up a couple of small ammunition dumps nearby and knocked out a 105mm howitzer, but only a single shell scored on the command post proper, landing in the message center and killing one man.7

Orders called for the assault to jump off at 0800 on D-plus 1 on all fronts, following an intensive preparation by naval gunfire, air strikes, and the massed fires of the artillery emplaced on shore. On this day, and increasingly throughout the phase under discussion, the operation broke down into three distinct patterns in accordance with the missions of the respective combat teams, which will be dealt with in turn as they fitted into the picture as a whole.

Clean-up in the South8

Final dispositions on the night of D-Day had seen two battalions of the 7th Marines in line, facing generally eastward and southeastward. On the right, the flank of the 1st Battalion was anchored firmly on the western


shore well below the extremity of Beach ORANGE 3, and extending inland along the O-A line. But the 3d Battalion was still well short of the eastern shore where Company I remained held up by that blockhouse strong point which it had been unable to reduce before dark, and where Company L had been echeloned to the left rear in order to tie in with 3/5.

Early on the morning of 16 September (D-plus 1), tank support and heavy preparatory fire enabled Company I, attacking eastward, to seize the supporting positions to its front--three concrete dual-purpose gun positions and a barracks area--with no great difficulty, but the blockhouse itself proved to be a very tough nut indeed. Upon its eventual capture, its walls were found to be of reinforced concrete more than five feet thick on all sides. This withstood direct hits by all infantry weapons, and 75mm tank guns merely bounced off it. Flame-throwers, unaided, could make no impression, as the two entrances were below ground level and, like the fire ports, protected by one-inch armor plate.9

This fortification was finally reduced by demolition teams working under the concealment of smoke placed to blanket all of its apertures, and Company I reached the eastern shore just beyond at 0925.10 While this unit consolidated the beach positions against enemy counterlandings or attacks through the mangrove swamp a short distance above, Company K attacked southward in line on the left of the 1st Battalion, with Company L in support.

The terrain over which this drive moved might possibly answer the loose description of "low and flat", which persistent intelligence reports had applied to the entire island, but only in comparison with some vastly worse terrain which the invaders would encounter a little later. Scrub jungle overgrew most of it, hampering even infantry maneuver and greatly complicating the bringing up of support weapons.

And the Japanese had improved on nature. They had reason to believe that U.S. troops might attempt an initial landing in this area, as indeed the planners had seriously considered, and from shore to shore the region was honeycombed with strongly reinforced and cleverly camouflaged defensive installations with cleared interlocking fields of fire: pillboxes, casemates, bunkers, rifle pits, and trenches. These had been blasted out of coral bedrock so adamant as to make digging-in virtually impossible for troops in assault; once off the sand beaches, the men could take cover only behind natural features or by piling up improvised breastworks of such loose rocks and blasted tree trunks as came to hand.11 This characteristic would operate to the attackers' disadvantage throughout all phases of the campaign. To an extent not equalled in any other Pacific operation, natural features pitted the unprotected Marine in the open against the hidden Japanese in a thoroughly fortified position.

The fighting in the south continued grim and deadly, but unspectacular. It boiled down to overpowering, in turn, individual positions more often than not mutually supporting. There was a redeeming feature in this connection, however. Because they had been prepared in anticipation of an amphibious landing, these installations presented their strongest faces seaward, so that the Marines advancing from the north caught them, to a greater or lesser degree, in the flank and rear. But, from whatever direction approached, their occupants had no intention of surrendering. Assigned to the defense of this area was the 3d Battalion,


15th Infantry,12 reinforced, one of the elite units of the Imperial Japanese Army, and its members fought to the death with all the skill which their situation permitted.

After reduction of that massive blockhouse, the 3d Battalion had the easier going during the drive southward. Company K, advancing in line on the left of the 1st Battalion, blasted and burned out successive enemy installations in its path to such good effect that by 1025 the leading elements came within sight of their final objective, the southeastern promontory, its approaches now defended on the landward side by only two pillboxes. By 1200, both of these had been reduced.

However, the combined strain of protracted fighting and debilitating heat (to be described subsequently) were now exacting their toll. The men were desperately in need of water13 before they would be in shape to assault again, so here the advance was halted. When it became clear that the supplies could not be delivered until late afternoon, the company was ordered to "dig-in" behind piled up coral rocks for the night. The remaining hours of the day were turned to good advantage by bringing up tanks to fire directly into enemy defense installations on the promontory and at the same time cover details of engineers clearing out the mines found to be liberally sown across the open sand spit which provided the only approach by land.

In its somewhat wider zone of action, the 1st Battalion advanced with Companies C and A in line (right to left), Company B in reserve. Progress was facilitated, as was that of the 3d Battalion, by strong support by heavy weapons: not only land-based artillery, but naval gunfire, called air strikes, and LCI rocket concentrations. But the area was defended in exceptional strength. In addition to the conventional pillboxes, bunkers and casemates, the battalion had to knock out four 5-inch guns and three lighter dual-purpose antiaircraft weapons before the morning was over. Nevertheless, 1/7 gained the southern end of the island proper only a short time after Company K. And there, for the same reason, a halt was called until the men could be in some measure rehabilitated; a halt which was to last through the night, while similar preparations were made for the following day's assault on the southwestern promonotory, or Ngarmoked Island.

Throughout D-Day the 1st Battalion had been harassed by fire from the small unnamed island lying just north of Ngarmoked which naval gunfire and air strikes had been unable to silence. The main gun position there was subsequently spotted from the beach and knocked out by tank fire, with such good effect that a patrol sent across the reef at low tide found no more live Japanese there. However, during the night of D-plus 1 the enemy made several attempts to reoccupy it, evidently with the object of getting once more on the flank and rear of the attackers. Illuminating shells revealed large numbers in the act of wading across from the southwest promontory, making them excellent targets for machine gun and rifle fire. No further trouble was experienced from that direction.14

The 2d Battalion, 7th Marines, unable to come ashore on the evening of D-Day for lack of transportation, landed on ORANGE 3 during the morning of D-plus 1, still in division reserve. The tentative intention of the division command had been to commit it in support of the southward drive of its parent regiment, and Company G and 81mm mortar platoon actually moved forward for that purpose. However, the precarious situation in the 1st Marines zone on the division's left was at last becoming apparent, in contrast to the comparatively satisfactory progress in the south, and shortly after noon the reassembled battalion displaced to Beach


WHITE 2.15 There it was attached operationally to the 1st Marines, in which capacity it served throughout the remainder of this phase of the fighting.16

In the 3d Battalion zone, the plan for D-plus 2 called for Company L, in reserve the previous day and hence comparatively fresh, to pass through Company K and assault the southeastern promontory. The attack was originally scheduled to jump off at 0800, and a preparatory air strike was delivered accordingly. However, the engineers, working farther forward, discovered another enemy minefield laid across the sand spit. The jump-off was postponed until 1000 in order to give them an opportunity to clear this out, which dangerous work the engineers accomplished without casualties under heavy covering fire delivered by the tanks and infantry.

This precaution paid off. Assaulting in the wake of a mortar and artillery barrage, one platoon of Company L, supported by three tanks, gained a foothold on the promontory at 1026 and occupied enemy installations knocked out by tank fire the previous afternoon. With the approach thus cleared, the remainder of the company was rushed across the spit in amtracks against only scattered rifle fire. The southern portion of the promontory was seized first in order to obtain positions from which fire could be placed across the lagoon on the opposite (Ngarmoked) promontory to support the attack of the 1st Battalion. The drive then turned northward across not too difficult terrain, overrunning a series of enemy positions. The entire promontory was declared secure at 1320, although, as is always the case in areas prepared for defense by the

PRIVATE FIRST CLASS ARTHUR J. JACKSON, 3d Battalion, 7th Marines, won the Medal of Honor and subsequent promotion to 2d Lieutenant for his courageous and effective assault on several Japanese pillboxes during the fighting on southern Peleliu.

Japanese, mopping-up proved difficult and arduous. The last twenty-odd enemy survivors were driven into the water, where they were picked off by riflemen while attempting to escape northward across the reef.

No such quick victory awaited the 1st Battalion. The southwestern (Ngarmoked) promontory was much the larger of the two, its terrain higher and more rugged, its defenses more numerous and formidable.17


Here the attack moved out according to the original schedule. Again, the comparatively fresh company which had been in reserve the previous day was committed to the assault, and initial momentum carried one platoon of Company B with one tank across the spit. At 0835 these encountered the enemy main line of resistance and were stopped cold without being able to gain a foothold. The company requested permission to withdraw these elements in order to give heavy supporting weapons an opportunity to work over the ground further. Permission was not granted until 0935, when the withdrawal was effected with some difficulty under galling crossfire from the not yet secured southeastern promontory.18 By this time, Company B had been roughly treated and was passed once more into battalion reserve.

While artillery and mortars poured concentrations into the target, the 1st Battalion pulled itself together for another try. All available tanks and LVT(A)'s were moved into the area, and half-tracks and 37mm guns from the Weapons Company were brought laboriously forward. The early success of the 3d Battalion on the opposite promontory released many of the supporting weapons attached to that unit, as well as substituting friendly supporting fire for the enemy's flanking fire.

Company A attacked at 1430, preceded by a 10-minute air strike and supported by three tanks. This time the Japanese first line of defense was carried. The company gained the high ground and commenced the grim business of slugging its way southward, while artillery and mortar concentrations blasted enemy positions to its front. At 1530, Company I was moved over from the 3d Battalion as a special reserve, and four minutes later Company C, thus released, was committed on the promontory to develop the right flank, here greatly lengthened by the bulge of the western shore. An hour later, Company B was ordered across the spit to positions immediately behind the two assault companies, and thus the entire 1st Battalion was in the fight.

Resistance was stubborn all along the line. Company C progressed to the western shore, in the process reducing the heavy enfilading weapons which had been playing such havoc with the ORANGE Beaches. But when the company attempted to advance southward, it soon came upon a sizeable swamp which threatened to break off its contact with Company A on the left. The defensive line as set up for the night was in the shape of an inverted U, with a shallow penetration in the middle where the swamp intervened, but a substantial hold on the shore on either flank. Four tanks, four armored amphibians, and two half-tracks were on the scene; Company B was in close support, ready to be committed wherever needed, and Company I in reserve just across the spit.

With the end plainly in sight, the 1st Battalion was in no hurry to resume the attack the following morning (D-plus 3, 18 September). While preparatory fires pounded the terrain to the front, dispositions were made with the utmost care. When, at 1000, the advance finally began, all of the armor which had been brought up during the previous afternoon and night was in position to move forward in support.

The plan for this day's fighting called for Companies C and A to circle the impeding swamp, on the right and left respectively, tie in their flanks to the south of it, and drive in line to the end of the promontory. Company B was committed on the extreme left to pinch out the eastward bulge of land which widened the front a short distance ahead of the jump-off line.

Experience of both battalions on the preceding day had indicated that some Japanese were bound to be overrun in the course of such a drive, hiding out either in their dug-in positions or in the numerous caves in this region of jumbled coral high ground. All units in assault were directed to ignore these and press on toward their objectives,


leaving mopping-up to demolition teams supported by reserve elements specially detailed for the purpose.

This soon proved to be far more of a problem than anyone had expected up until now. Company C detached 15 men to cover suspected openings of caves and pillboxes in its rear, only to have these pinned down in short order by Japanese swarming out from the underground in numbers which, at this early stage of the operation, seemed astonishing.19 Eventually it became necessary to commit most of the special battalion reserve (Company I) and the Division Reconnaissance Company, which had been attached to the 1st Battalion since D-Day afternoon, to bring the situation in the rear under control.

The attack developed according to plan, but slowly in the face of last-ditch resistance in every sector. By 1048 Company B, now deployed on a line running north and south, had pinched out the eastern bulge of the promontory. With its flank thus covered, Company A, reached the southern shore in its zone of action at 1224, Company C completing its mission on the right at 1344.

It was now becoming apparent that Company B had drawn the most difficult assignment. The defense system in its area had been strategically designed to cover the entrance to the nearly circular lagoon, or cove, lying between the promontories against an anticipated U.S. landing. The shore approaches were heavily mined, and enemy prepared positions crowded thickly in upon one another. At 1354 the company estimated that it had killed 350 Japanese, and resistance had been constricted to a single hard core only about 50 square yards in area. However, by this time the tanks supporting the infantry had withdrawn to rearm, and the half-tracks had bogged down. Thus, the attack had to wait until a bulldozer could be brought forward to extricate the latter.

As the company waited, significant sounds to its front indicated that at last the Japanese had decided to cooperate--by blowing out their own brains.20 Many others leaped into the sea. When the final position was taken shortly afterward, 15 or 16 of these last die-hards were found to have been officers.

So far as the southern part of the island was concerned, the fighting was over, save for the inevitable mopping-up. A fine Japanese reinforced infantry battalion, in strongly fortified positions, had been sent to join its ancestors. As the two Marine assault battalions straightened themselves out and squared away for a little well earned rest, the regimental command notified Division Headquarters optimistically: "At 1520, O-1 was taken. The Seventh Marines Mission on Peleliu is completed."21

Advance of the 5th Marines22

By far the most substantial gain in the actual ground secured during the Second Operational Phase was to be achieved by the 5th Marines. But there was small indication that this would be the case on the morning of D-plus 1.

The regiment began this day with a near-catastrophe when one Japanese shell struck the field telephone switchboard, wiping out a good portion of the wire section, while a second shell scored a direct hit on the regimental command post. This was located in a section of captured enemy trench23 near the edge of the airfield, and the shell struck


OPEN AIRFIELD area favored tank operations, but dominating ridges to the north demanded fast action.

the parapet, knocking out the R-1, R-3, and a newly arrived replacement major, as well as the naval gunfire liaison officer. The commanding officer, Colonel Harold D. Harris, was almost completely buried by the explosion; miraculously, he escaped being wounded but sustained a severely wrenched knee which kept him partially incapacitated and in great pain for several days.

How seriously this blow impaired the command situation can be understood more clearly when it is remembered that the executive officer, Lieutenant Colonel Lewis W. Walt, had been detached from the staff the previous evening in order to take over the disorganized 3d Battalion. Replacements were brought up hurriedly from the division command post,24 with the result that the attack jumped off on time. Later in the day Colonel Walt returned to his regular post which had become of added importance because of Colonel Harris' injury.

The 5th Marines, it will be remembered, had for its initial mission capture of the airfield. The 1st Battalion, on the left, had spent the night in the woods at the edge of the open ground, along the O-1 phase line, where it had halted of necessity in order to maintain contact with the stalled 1st Marines. On the extreme right 3/5, on the eastern shore and along the southern edge of the airfield, also enjoyed some cover. But the 2d Battalion, which had advanced rapidly in the wake of the repulsed tank attack late the previous afternoon, now lay in line nearly midway across the open airfield. All three units beat off short, sharp counterattacks during the night, and with the coming of dawn became subjected to intensive and effective shell fire.

The terrain here, even more than to the south, might be compared to that of the atoll


FIRESWEPT AIRFIELD; here being crossed by 5th Marines, was naked of cover.

islands of the Gilberts and Marshalls. But with a difference: the airfields on Betio, Roi and Engebi were not dominated by concentrations of heavy weapons posted on high ground and directed from observation posts at elevations as high as 300 feet from which the enemy could "look down the throats" of the attackers. There was little actual ground resistance on the airfield itself, but deadly fire poured from the ridges and from installations in the woods, or scrub jungle, on the north and east which made a terrain compartment of the entire area. While these woods might provide covered approaches to the heights, there were no covered approaches to the woods; they had to be taken by determined tank-infantry attack, supported by naval gunfire, air strikes, and all the artillery and mortars which could be spared from other sectors.

The advance on this day took the form of a wide turning movement toward the northeast, pivoting on the extreme left of the 1st Marines. In the 5th Marines zone of action, the most substantial gains were achieved by the 1st Battalion. Lieutenant Colonel Boyd had Companies A and B in line (left to right), with Company C echeloned 200 yards rearward behind Company A. In this formation, the battalion swept the entire northern portion of the airfield. In a little more than an hour, advance elements reached the main service apron and hangar area. Here they found the Japanese installed in strength in a large antitank trench and several stone revetments and ousted them only by dint of some heavy fighting. This position was consolidated, and during the afternoon the battalion pushed forward to the O-2 line. There was no possibility of digging-in here, however,


so a withdrawal was made to the antitank trench for the night.

Throughout he day the 1st Battalion experienced contact difficulties on both flanks. On the left, the 2d Battalion, 1st Marines, was meeting a somewhat different tactical situation (see sub-chapter following) and had troubles of its own. Colonel Boyd succeeded in tying-in firmly with Company F of that regiment, only to discover a little later that this unit was out of touch with the rest of the 1st Marines. The gap there, estimated as 200-250 yards, was covered by bringing up Company I from 3/5 (see below), but F/1 remained attached to 1/5 throughout most of the day. However, because of the excellent visual contact and the open nature of the ground, physical tie-in was not considered essential.

The 2d Battalion made slower progress. Major Gayle's unit had to cope not only with the shell fire that lashed the open ground, but with the woods that bounded the airfield on the east, giving way along an ill-defined shore line to dense mangrove swamp. These woods were in actuality a nearly impenetrable tangle of scrub jungle, difficult to travel in itself and infested with Japanese who had to be eliminated by the costly hand-to-hand process, supported where possible by tanks operating on the open ground. By nightfall the 2d Battalion had closed up on the flank of the 1st and enjoyed some surcease from the galling fire from the ridges under cover of the captured woods. But with the coming of darkness, the battalion found it expedient to fall back a short distance on the open airfield in order to obtain clear fields of fire to its front.

The 3d Battalion, 5th Marines, played a comparatively minor role in this day's fighting. The simultaneous pushes of its own regiment northward and the 7th Marines first to the east25 and then to the south, each with a flank on the eastern shore, had the effect of pinching out 3/5 for practical purposes. During the morning, the battalion expanded its hold along the shore line behind the advance of the assault units, elements of Company L assisting the flank of the 2d Battalion, Company K cooperating with 3/7. But with the two drives proceeding in opposite directions with all possible speed, both companies were soon overextended. Company I was ordered to fill the gap which was developing between them, but before this move could be effected, about 1200, new orders arrived sending Company I to the north of the airfield to reinforce the battered 1st Battalion.

Because of the urgent need for Lieutenant Colonel Walt in his regular post as regimental executive officer, he was relieved in his temporary command of 3/5 at 1500 that afternoon. Major John H. Gustafson was designated to command the battalion, with Major Hierome L. Opie moving in to take over as Executive Officer. These officers received word of this new assignment only at dusk and had little time to learn the true condition of the unit and its equipment before the battalion was moved up to relieve 1/5 in its forward positions the following morning.26

D-plus 2 (17 September) was a significant date in the story of the operation, as this day brought the invaders squarely up against what was to prove the major terrain feature of the island's defense: the ridge system which began just above the Japanese barracks area and followed the axis of Peleliu's northwestern peninsula, harboring the hard core of enemy resistance. As this lay entirely within the zone of action of the 1st Marines, the fighting there will be discussed in detail in connection with the operations of that regiment (see following subchapter).

The line of advance of the 5th Marines lay across the low ground to the east of the


ridges. Thus, every forward movement brought the elements of that regiment under increasingly heavy flanking fire from the enemy's commanding positions on the left. This, combined with the necessarily slow progress of the 1st Marines, was to be a controlling factor in this area as long as the regiment operated there--and for some time after that.

The situation was especially acute in the zone most adjacent to the high ground where open terrain provided virtually no concealment and little cover. Attacking early on the 17th behind adequate fire preparation, the 1st Battalion reached the O-2 phase line once more by mid-morning against light ground resistance but heavy mortar and artillery fire from elevated emplacements on the left. This time the position was secured, and there 1/5 was relieved by the 3d Battalion which had been alerted for this mission the previous evening.27 But when the latter unit attempted to move forward again in the afternoon, it was pinned down so effectively by that flanking fire that gains on the left were negligible.

On the right, the 2d Battalion had the advantage of excellent concealment in the woods; and its concomitant disadvantage: so excellent was the concealment that the jungle was all but impenetrable. Thus, the men were relatively free from that galling fire from the left, though Japanese observers in the hills promptly called down concentrations whenever any personnel showed themselves in the open. This was particularly the case during the morning when a tank attempted to cooperate along the jungle's edge; it drew such volume of fire as to cause more harm than it could do good and had to be withdrawn.

Resistance on the ground was limited to the sporadic fire of a few scattered snipers. However, after the experience of the previous day in similar territory, this could not be foreseen, and the troops combed through the area in a skirmish line, hacking their individual ways as best they could. The brutal heat which now gripped the island was beginning to exact its toll. Heat prostration cases mounted, and it was necessary to call frequent halts to rest the men. Supply difficulties, especially getting water forward, did nothing to help.

Nevertheless, the 2d Battalion's advance passed well beyond the O-2 line for a gain of 600-700 yards. Company E, on the left, dug in in contact with the 3d Battalion where Company K, having found some woods in its own area, had pushed forward to the edge of a coconut grove near the point designated "Omaok" on the map: a native village at one time but now an area of demolished Japanese barracks and utility buildings. Company F anchored the battalion's right flank on the edge of the mangrove swamp with Company G, in reserve, extending southward along the shore from there, facing that dismal morass from which nobody knew what to expect.

Progress continued slow on the left of the 5th Marines throughout 18 September (D-plus 3). The regimental zone of action in this area was bounded by a tactically important thoroughfare which became known as the East Road, which here began skirting the base of that hill system identified on the maps as Umurbrogol Mountain.28 The 1st Marines had gained a foothold on the most adjacent ridge late the previous afternoon (see sub-chapter following), but fire from innumerable other elevations farther to the west and north continued to hamper the advance of troops traversing the open ground. This, combined with the necessity for keeping tied-in with the adjoining regiment, now heavily engaged with both the


enemy and the terrain, halted the forward movement of the 3d Battalion around midday.

Conditions on the right, however, were very different, and soon changed the entire nature of the 5th Marines' mission.

Jumping off at 0700, the 2d Battalion continued to hack its way through the concealing jungle, relatively immune to enemy shellfire and encountering only scattered ground resistance. This advance was aided greatly by a half hour's concentration of artillery time shell (air burst) fire on the woods and mangrove swamp to the front and flank by 2/11. Within two hours advance elements of Company F reached an improved road leading eastward to the Ngardololok area,29 beyond which lay PURPLE Beach and the ill-defined shore contours and semi-islands which constituted the northeastern peninsula, or shorter prong of the Peleliu "lobster claw", seizure of which constituted an important part of the regiment's mission.

A short distance to the east of where the Marines first reached it, the Ngardololok Road was so closely bordered by swamps as to become in effect an open causeway: a perilous bottleneck which threatened to check any advance in force beyond this point. Aerial observation had revealed major Japanese installations both in the RDF area and the vicinity of PURPLE Beach, but the nature of the resistance encountered during the past two days indicated that the enemy might have withdrawn. To test this possibility, at 1040 the commanding officer of Company F was ordered to take a patrol across the causeway and see what happened. The patrol accomplished its mission by 1120 without drawing any fire whatever, whereupon an air strike on Ngardololok was called for to prepare the way for a crossing in force to set up a bridgehead.

Now began a series of misadventures which, alone, hampered the progress of the 2d Battalion on this day. The air strike, staged at 1245, missed the target altogether. To make up for this, an artillery barrage was placed on the area, and at 1335 Company F, reinforced by elements of Company G, commenced the perilous crossing of the causeway. Before this could be accomplished, a second and wholly unexpected air strike30 came diving down to strafe the advancing troops. The bridgehead was set up, nevertheless, but Company E and the battalion command post, moving forward in support, were caught in the open approaches to the causeway by a misplaced artillery concentration, and a little later on the causeway itself by friendly mortar fire.

Even before the crossing of the causeway, a new tactical factor had arisen to change the nature of the 5th Marines' mission. The long, shallow arm of the sea which separated the island's northwestern and northeastern peninsulas had been reached which, in effect, split the regimental zone of action well over toward the left. All that remained on the west was the strip of ground between the East Road and the deep swamp that fringed the inlet, a front which could be maintained adequately by a single company rather than requiring an entire battalion. Company L, therefore, was left here to tie-in with, and protect the flank of, the 1st Marines, while the rest of 3/5 was hurried eastward across the causeway to aid in exploiting the gains of the 2d Battalion. At 1700 that evening Companies I and K dug in on the right (south) of 2/5, which had expanded its initial bridgehead to the north and east: holding all of the shore line in its area and facing the main Ngardololok installations across open ground.


It will be remembered that this was the day (D-plus 3) when the 7th Marines secured all of the remainder of southern Peleliu. This accomplishment released additional personnel of that regiment, with the result that 3/7 took over most of the swamp line still held by the reserve battalion of the 5th Marines, enabling 1/5 to move up in closer support of its parent regiment in preparation for the drive to come. In the positions they occupied that night, the 2d and 3d Battalions were poised for the completion of the regimental mission, which proceeded systematically and uneventfully from this point on. An air strike paved the way for the 2d Battalion to secure the remainder of the RDF area during the morning of 19 September. Opposition was limited to a scattering of Japanese, evidently stragglers, who lurked among the demolished installations. They inflicted a few casualties but for the most part showed a greater disposition to hide than to fight, lending weight to the earlier supposition that the first class enemy troops originally assigned to this area had been withdrawn.

Subsequent operations of this day substantiated the supposition beyond further doubt. The 3d Battalion, pushing southward through tangled jungle following artillery and mortar concentrations, killed a handful of Japanese with virtually no resistance. Company I reached the ill-defined shore line across the mangrove swamp from the airfield area where the 2d Battalion had encountered so much trouble two days earlier, and Company K pushed a strong patrol to the east coast just below PURPLE Beach. Later that afternoon, Company G occupied the southern end of the beach, and commenced vigorous patrolling to the northeast.

The abandoned defenses of PURPLE Beach proved to be very formidable indeed as they faced seaward, confirming further the wisdom of the planners in deciding early against attempting a landing there. The beach itself was heavily mined and obstructed with tank traps and barricades which slowed progress much more than mopping up the few scattered enemy remnants. Nevertheless, by late afternoon of the following day (20 September) Company G had secured the northern tip of the northeastern peninsula proper and sent a patrol to the off-lying island (designated "Island A" in current reports) which was found to be unoccupied.31

Simultaneously, a patrol from Company I moved southwestward along the narrow projection below the beach. Like the 2d Battalion elements, this patrol employed war dogs, one of which was credited with alerting the troops to an enemy ambush32 during the late afternoon. A sustained fire fight failed to prove decisive by nightfall, and the men of Company I dug in short of their goal. Night infiltration attempts weakened the Japanese defenders, however, and the objective was secured the next morning with no great difficulty.

The remainder of the 5th Marines story for this phase is quickly told. The 2d Battalion continued its northward push, hampered only by natural obstacles. On the morning of 21 September, Company B seized Island A which the previous day's patrol had found unoccupied. Company F then passed forward and secured the larger adjacent island of Ngabad, again without opposition. The battalion command post kept pace with the advance, and Company E, in reserve, moved up in the rear to help in organizing the newly occupied territory. On 23 September, Company G's seizure of the small unnamed--and undefended--island due north of Ngabad and midway across the opening of the Peleliu "lobster claw" completed the regiment's mission in this zone of action, isolating


all enemy resistance on the northwestern peninsula.

Behind the rapid advance, the 3d Battalion (less Company L) completed mopping-up and set about organizing PURPLE Beach for defense against anticipated counterlandings by Japanese troops from the central Palaus. On 21 September the 1st Battalion, which had been in regimental reserve farther south, moved up to Ngardololok and undertook similar work thereabouts. From then until the end of the Peleliu operation, the entire region was designated a defense area and was occupied in turn by various assault units relieved from the lines for a brief rest.

During this far-reaching expansion, Company L had remained tied-in with the 1st Marines to the west of the inlet which divided the 5th Marines' zone of action, and its movements during this period were governed by those of that regiment. The position held by Company L on 18 September, the day that the rest of the 3d Battalion crossed the causeway, described a rough arc from the East Road across to the southern end of the swamp which fringed the inlet, facing northward. On 19 September, a forward surge by the right of the 1st Marines carried to the summit of the dominating elevation in that area,33 and Company L tied-in with the advance by pivoting on its right flank to form a north-south line along the edge of the swamp.

The 1st Marines, however, failed to hold the ground thus gained, and with the retirement of the survivors the following morning, Company L fell back to the position previously occupied. There it remained without participating in any further movements of importance until, with commitment of the 5th Marines in the Third Operational Phase of the campaign, it rejoined its parent regiment and played its part in the operations to the northward, as narrated in Chapter VI.

Difficulties of 1st Marines34

The over-all scheme of maneuver had anticipated that the 1st Marines would encounter the most difficult opposition. This forecast had been more than substantiated on D-Day, and the appalling extent to which it was true became increasingly evident.

As previously recounted, the regiment's situation on the evening of D-Day left a good deal to be desired: with most of the left flank company isolated and a gap in what should have been the beachhead line where the Japanese still held the northern half of that formidable coral ridge which barred progress inland. Jumping off to the attack on the morning of D-plus 1, Lieutenant Colonel Honsowetz's 2d Battalion, on the right, made excellent progress, crossing the west turning circle of the airfield within half an hour. But on the left of the 3d Battalion zone the line was unable to advance at all.35

Tactically, this immobility was not serious in itself, for the over-all plan called for the northward drive to pivot on the left flank. But the fact, that the immobility was enforced, with, in effect, an enemy salient driven into the line, called for cogent measures. The regimental reserve (1st Battalion) attacked strongly in the early afternoon with Company C, the last fresh element, in assault and one platoon of Company B in reserve. Two tanks had been brought up, and with their support the infantry succeeded in capturing a 500-yard segment of the ridge.36 With this high ground as a lever to pry the left of the line, the remainder of Company B was able to advance and gain contact with Captain


CLOSE INFANTRY ACTION through the rubble of building area north of airfield.

Hunt's command on the point which by then had been cut off for thirty hours.

A brief digression here to complete the saga of Company K. With the sea at their backs and Japanese in every other direction, the survivors of the two platoons had withstood sharp counterattacks throughout the night of D-Day. An amtrack operating across the reef carried out their wounded and brought in supplies and ammunition, and eventually the platoon which had been pinned down in the tank trap. The enemy were determined to recapture that bitterly contested point, but they were unable to muster forces for a major attack until about 2200 on the night of D-plus 1.

This took the form of a frontal assault against the high ground by an estimated 350 Japanese, together with a secondary, or diversionary, attack along the narrow flat at the shore line. At one stage during the previous night, Captain Hunt had been down to 18 men and must surely have been overwhelmed. Now most of Company B37 was in position, as well as what remained of his own command; he was tied-in with the battalion line, and had telephone communication with the mortar platoons of both companies, which were already registered on the ground to his front.

Thus, the Japanese faced all the disadvantages attendant upon assaulting adequately manned prepared positions, covered by well placed defensive fire. What developed was a shambles, and by 0200 on 17 September a great tranquillity lay upon the area.

When the advance jumped off on the morning of D-plus 2, Company I passed through


"A JOB FOR THE INFANTRY." Typical terrain in Peleliu's coral limestone ridges.

the position and Company K was ordered into reserve. Of Captain Hunt's complement of 235 men who had landed two days previously, 78 were left.38

Elsewhere in the 1st Marines zone, progress during the second day was the most satisfactory the regiment was destined to experience during its operations on Peleliu. The arbitrary line demarking this zone from that of the 5th Marines was so drawn that the two regiments, closely tied-in as they were, encountered essentially different terrain problems. Whereas the 5th Marines advanced almost entirely across open ground, the 1st, after cutting across one corner of the airfield, entered the building area to the north. Many of the structures here had been made of reinforced concrete, now blasted to rubble by the heavy naval gunfire. These ruins provided at least some cover from the brutal shelling pouring from the high ground. Of course, they also provided cover for the defending Japanese infantry, with the result that the area became the scene for some savage close-in fighting.

The center of the regimental front, comprising the right company of the 3d Battalion and the left elements of the 2d, broke through and pushed on to gain control of a valuable stretch of road running east-west. On the right, however, the Japanese stubbornly defended the junction of the main road which circled the base of the ridge


beyond the airfield headquarters, then angled off to run the whole length of the northwestern peninsula.

This road, marking the O-2 phase line, was still 200-300 yards beyond reach when the time came to set up positions for the night. The extreme left, not having moved beyond the point where Hunt's men had been fighting for two days, was still more than 100 yards short of the O-1 line.

All the trials which the 1st Marines had suffered up to this point were compounded by the protracted hell which commenced for the regiment on D-plus 2. For this was the day its personnel made the acquaintance of the Peleliu ridges at first hand. Preliminary aerial photography had done scant justice to this outlandish formation. At that time scrub jungle cloaked the high ground, blurring the outlines and softening contours, so that it might have passed for pleasantly undulating country of hills and dales. Now naval gunfire, if it had accomplished little else of great importance, had sheared away vegetation like a giant scythe, and the true features of the ridges stood revealed in all their stark horror--or so they appeared to the eyes of military men ordered to attack them.

First Marines Regimental Narrative (pp. 7, 8) describes the terrain thus:

The ground of Peleliu's western peninsula was the worst ever encountered by the regiment in three Pacific campaigns. Along its center, the rocky spine was heaved up in a contorted mass of decayed coral, strewn with rubble, crags, ridges and gulches thrown together in a confusing maze. There were no roads, scarcely any trails. The pock-marked surface offered no secure footing even in the few level places. It was impossible to dig in: the best the men could do was pile a little coral or wood debris around their positions. The jagged rock slashed their shoes and clothes, and tore their bodies every time they hit the deck for safety. Casualties were higher for the simple reason it was impossible to get under the ground away from the Japanese mortar barrages. Each blast hurled chunks of coral in all directions, multiplying many times the fragmentation effect of every shell. Into this the enemy dug and tunnelled like moles; and there they stayed to fight to the death.

General O. P. Smith (Personal Narrative) adds:

Ravines, which on the map and photographs appeared to be steep-sided, actually had sheer cliffs for sides, some of them 50 to 100 feet high. With nothing else on your mind but to cover the distance between two points, walking was difficult. . . . There were dozens of caves and pillboxes worked into the noses of the ridges and up the ravines. It was very difficult to find blind spots as the caves and pillboxes were mutually supporting. . . . These caves and pillboxes housed riflemen, machine gunners, mortars, rockets, and field-pieces. The Japanese technique was to run the piece out of the cave, fire, then run the piece back in the cave before we could react.39

Even before encountering the ridges, the 1st Marines received reinforcements. The inability of the division reserve (2d Battalion, 7th Marines) to land late on D-Day for commitment in the southern area made it available for the more critical northern zone, and this unit was moved to the WHITE Beaches on the morning of D-plus 1 and attached operationally to the 1st Marines, as described in previous sub-chapter. Initially the battalion was held in special reserve,40 but it was destined to play its active part all too soon.

Before the attack could jump off on the morning of D-plus 2, some reorganization was necessary. The regiment's casualties now exceeded 1,000, and the units were so depleted that it was necessary to use all three battalions in line. Owing to the difficulties on the left, the 1st Battalion had been committed in the middle of what had been originally designated the 3d Battalion zone. Now 1/1 was ordered to move into the center of the regimental line at dawn, while 3/1 tied-in in its own sector. The 2d


Battalion retained its position on the right, in contact with the 5th Marines, and one company of the newly arrived reserve battalion (Company E, 7th Marines) was moved up in close support.

The 2d Battalion was the first to meet the ridges. Pushing forward in the wake of the artillery and naval gunfire preparation, the assault companies quickly overran the important road that marked the O-2 phase line and captured the enemy position at the vital road junction which had held up the advance on the previous afternoon. This placed the battalion astride the island's second main traffic artery: the East Road, which ran from the administrative area through the village of Asias and along the narrow flat between the swamps and the eastern base of the high ground, then angled across the peninsula to join the West Road near the northern radio station (Road Junction 15--see Map 3).

The foremost units had progressed along the road for about 150 yards when they were stopped abruptly by a combination of brief loss of contact with the 5th Marines on their right, and the intensity of fire from the rugged coral ridge which now loomed menacingly on their left.

This ridge attained an elevation of 200 feet (henceforth known as Hill 200) and paralleled the East Road, thus thrusting a formidable salient ever deeper into the zone of action of the 2d Battalion as that unit advanced. Its flanks were steep, precipitous in places (much worse would be encountered) and honeycombed with caves which held observation posts, from which fire on the airfield area was directed, and many of the weapons which delivered this fire.

The battalion wheeled left and took the ridge under attack, scaling the steep slopes in the face of withering small arms fire and point-blank salvos from mountain guns and dual-purpose pieces which were run in and out of cover. At such close range this fire was extraordinarily accurate. Casualties among the assault troops mounted, and tanks and LVT(A)'s moving up in support were knocked out right and left. By nightfall virtually all of the crest had been taken. The men dug in41 as best they could against the fire that now poured on them from elevations farther forward and from a second, slightly higher ridge (Hill 210) which paralleled Hill 200 on the west. The narrow, steep-sided ravine separating the two teemed with Japanese who effectively prevented further assault in that direction.

According to the Japanese version, as set forth in Tada Record, 18 September: "At 1300, under protection of heavy naval gunfire, an enemy unit composed of two tanks and approximately two companies of infantry successfully advanced up to a high spot on the east side of Nakayama" (Japanese designation of Hill 200). Such apparent discrepancies in date appear frequently and might be accounted for either by faulty communications or the commander's natural reluctance to report a reverse until convinced that it was irremediable. At this time Nakagawa's CP was farther north in the same ridge, and the success of the Marines in gaining a foothold on the crest may have been the determining factor in causing him to displace later that day to the cave in the heart of the enemy final pocket of resistance which he occupied henceforth.

The capture of this strategic height had two immediately important effects. First, it greatly reduced the flanking fire which had been so hampering the 5th Marines, enabling that regiment to proceed with its main mission. Second, it exactly reversed the previous situation: instead of confronting a deep salient thrust into its own lines, the battalion now held an equally formidable salient driven into the Japanese defensive stronghold.

In the meanwhile, the other elements of the 1st Marines had commenced the day with an equally swift advance, a pace they


SCENE OF SOME OF 1st MARINES HEAVIEST ACTION on D-plus 2, 3. Troops on Hill 200 (left) exchanged fires across gulch with Japanese still holding Hill 210 (right).

were able to sustain considerably longer. In the center, 1/1 had to cope with an obstacle which, according to its officers' not unreasonable reckoning, should not have been there. This was a formidable blockhouse, with reinforced concrete walls four feet thick and protected by twelve pillboxes, all connected by a maze of tunnels. Unlike the hidden installations in the hills and jungles, this building showed clearly on the prelanding aerial photographs and was pinpointed on the map issued to all ground troops; yet although Admiral Oldendorf had reported before D-Day that the fire support ships had run out of profitable targets, the blockhouse had not been so much as nicked, much less reduced.42

This omission was quickly rectified, however. The old battleship USS Mississippi had been assigned in direct support of the


SOUTHERN RIDGES, air view from southeast. Long razorback at left is Hill 200, with Hill 210 just behind it. Five Sisters in center, Walt Ridge right foreground.

battalion, and very shortly after the shore fire control party was able to adjust her 14-inch guns, the blockhouse began disintegrating. The big armor piercing and high capacity shells crumbled the walls, and their terrific concussion killed those Japanese missed by fragmentation. Smaller naval guns, tanks and infantry support weapons concentrated on the protecting pillboxes, and presently the advance overran the position and rolled on into dense jungle to the O-1 line.43

During this action the front had become over-expanded, necessitating the forward movement of two platoons of the reserve in order to protect the flanks, and here the battalion paused to regroup.

By now the 1st Battalion, too, was face to face with the ridges. The arbitrary phase line O-2 had been laid along the main road, here running northwest-southeast parallel to the battalion's front. Just across the road began that irregular, unpredictable rising of the land which was such a familiar feature of the landscape of Peleliu's northwestern peninsula. No sooner had 1/1 halted to reorganize than the line was assailed by a fury of fire from the high ground, and casualties began to mount alarmingly.

As quickly as it could be managed, reorganization was completed and the men surged forward again, slugging their way upward much in the manner the 2d Battalion was storming Hill 200. There was no other way: clawing up and over razorback crests, shinnying coral pinnacles, plunging down into sheer-sided gullies and ravines, dodging behind boulders. By evening they had gained the forward slope of the first series of hills and were firmly, if uncomfortably, established.44 One significant feature of this advance was that in the course of it the infantrymen were obliged to knock out 35 separate Japanese-manned caves, existence of which had not been noted in their advance intelligence.45


Writing of this mission long after the event, Lieutenant Colonel Davis, in his letter previously cited, described it as "the most difficult assignment I have ever seen. . . . Company A depleted itself on the bare ridge on the right as Company C became seriously over-extended on the left and was faltering. Everything was thrown in to exploit Company C's success. Remnants of Companies A and B, Engineer and Pioneer units were committed to fill the gaps as darkness came. Headquarters personnel were formed into a meager reserve. . . . Company C had moved 800 yards during the day, and we had been able to hold on although the cost was extremely heavy."

On the 1st Marines' left the story was somewhat happier. Lieutenant Colonel Stephen V. Sabol's 3d Battalion, which had had the most difficult going up to now, was destined to have the easiest henceforth. For its course lay along the coastal flat (or comparative flat), with the high ground to its right. Slow to get started, the battalion rolled on against minor resistance for an average advance for the day of 700 yards, and all that stopped it then was the danger of over-extending its front through the necessity of keeping tied-in on the right. As with the fighting in the ridges, this condition was to set the pattern of the regiment's advance for several days to come.

The varying fortunes of the different units resulted in an odd-appearing and far from reassuring contour to the front line on the evening of 17 September. Because of the inability of the 2d Battalion to secure Hill 210, the right of the 1st Marines' zone somewhat resembled a constricted letter "W". And there were two ominous features. While the Marines held a deep salient in enemy territory along the crest of Hill 200, this was precarious in the extreme. For in Japanese hands, the parallel ridge to the west formed an equally deep salient in U.S. territory.

Furthermore, in adjusting the lines in this area after dusk, a gap was created between the flanks of the 2d and 1st Battalions. This the night-prowling enemy soon discovered and commenced to infiltrate in force. To cope with this situation, it was necessary to commit Company F, 7th Marines, and that unit was obliged to fight its way into position before the gap could be closed. Pressure on Hill 200 was so sustained and heavy throughout the night that at 0200 on the morning of 18 September, the 2d Battalion was obliged to bring up Company G, 7th Marines,46 to bolster the position there. However, any overwhelming counterattack which might have made the situation something worse than precarious was effectively interdicted by concentrations of well placed artillery and naval gunfire on the approaches.

This brings the narrative of the 1st Marines to the morning of D-plus 3 (18 September). That day, as has been noted, was when the 7th Marines completed the securing of southern Peleliu, and the 5th Marines, leaving only Company L to maintain contact, set off on its quick overrunning of the eastern and northeastern portions of the island. But no such decisive results were in store for the hard-fighting 1st Marines, that day or ever during this campaign.

Casualties had been heavy, especially among the front line units, as was to be expected. For example: of 473 total effectives remaining in the 3d Battalion, 200 were headquarters personnel. Total losses for the regiment ran 1,236. Yet there appeared to be grounds for satisfaction with the progress made. A good foothold had been gained in the high ground, and the stiffest resistance the Japanese had been able to offer (so far) had been overcome. In any case, Colonel Puller reported to the


PERILOUS APPROACH. Japanese emplaced in precipitous Hill 200 (background) poured flanking fire into troops attempting to assault the Five Sisters (right, partially hidden by tree).

division command post with an optimism which was contagious. The general feeling seemed to be that a break-through was imminent, whereat enemy resistance would collapse, or at worst disintegrate, as had happened on Saipan, Tinian and Guam after a certain point had been passed.

The trouble with this reasoning was that on the other islands the collapse had occurred upon U.S. troops reaching favorable terrain and had been heralded by at least one suicidal banzai charge. But there had been no banzais on Peleliu, and the terrain was becoming worse instead of better.

Optimistically, necessary regroupings were ordered, preparatory to resuming the attack on the morning of 18 September. Unit commanders put every available man in the lines, including headquarters personnel not absolutely indispensible elsewhere. Company D, 1st Pioneer Battalion, sent up 115 replacements who were divided among the assault units. At 0600 the reserve battalion (2/7) took over the center from 1/1 which was relieved for a brief spell in regimental reserve (very brief indeed in the case of Company B), and the assault jumped off at 0700 behind a 30-minute preparation by planes, artillery and naval gunfire.

D-plus 3 combined the worst features of D-plus 2 with a lot of new ones, and the best that can be said for the day's progress


was that in general the alignment of the front was rectified. The enemy salient on Hill 210 was finally pinched out by dint of savage and costly fighting on the part of 2/1 and 2/7. The Japanese, in turn, concentrated a murderous fire on 2d Battalion elements on the forward nose of Hill 200. Following up with powerful counterattacks, they made the position untenable and forced a short withdrawal from the ground so painfully won the previous afternoon.

At 1400 Lieutenant Colonel Honsowetz reported the 2d Battalion situation as desperate and requested immediately reinforcements and smoke barrage to screen the position. Company B was ordered up from the 1st Battalion's reserve area, attached to the 2d Battalion, and thrown in assault against Hill 205, slightly forward and to the right of the ridge to which it was hoped to gain direct access by this means. Somewhat to the surprise of all hands, they secured the hill with comparatively light casualties by 1630. However, it proved to be isolated from the main ridge system, hence of value only as an observation point and for relieving some of the pressure. Pushing on from there, Company B was decisively thrown back before the most formidable terrain obstacle yet encountered: an incredible complex of up-ended peaks and palisades which was to gain evil fame under the name of the Five Sisters.47

Progress was somewhat better on both flanks. In the 3d Battalion's zone on the left, the advance, as on the previous day, met only minor opposition, but the assault elements were obliged to halt after moving only a few hundred yards in order to maintain contact with 2/7, which was operating with great difficulty in the high ground. Those elements of 2/1 on the extreme right, moved across the comparatively level flat below Hill 200 to what had been the village of Asias, tying in there with the left company of the 5th Marines,48 the rest of whose zone would now be separated from that of the 1st Marines by swamps and a long indenting arm of the sea. Thus at dusk the regimental front as a whole formed a shallow "U", a situation sounder tactically than the constricted "W" of the previous evening. But in the center the day's gains looked something less than imposing on the map, and on Hill 200 there was actual retrogression.

It should be borne in mind that position lines shown on the operations maps during this phase of the fighting are intended only as approximations. Owing to the inadequacy of the maps in this particular region and the incredibly broken nature of the terrain, exact orientation was all but impossible, and many misunderstandings resulted. As a participating battalion commander puts it:

There was no such thing as a continuous attacking line. Elements of the same company, even platoon, were attacking in every direction of the compass, with large gaps in between. When companies were asked for front lines they were apt to give points where the Company Commander knew or thought he knew he had some men. . . . There were countless little salients and countersalients existing.49

Irresistable Force Versus Immovable Object

After surviving a night of constant harassing artillery and mortar fire, what was


left of the 1st Marines (plus 2/7) resumed the attack on 19 September (D-plus 4). Artillery and mortars opened preparatory fires at 0620, naval gunfire at 0645, and the assault jumped off at 0700. Again, the 3d Battalion found comparatively easy going across the coastal flat on the left and advanced about 400 yards before being obliged to halt in order to maintain contact.

But in the 2/7 zone of action in the center: "Progress was extremely slow and costly. The terrain was rugged, visibility extremely poor, there were only poor fields of fire and it was extremely difficult for the companies to use their mortars and automatic weapons to full advantage."50 Here the advance moved in a direction varying from northeast to almost due east as the Japanese core of resistance hardened in the higher ridges.

On the right 2/1 slugged its way forward under generally similar conditions until about noon, when it was brought up short by the same obstacle which had repulsed Company B's abortive effort the previous afternoon: the Five Sisters. Although there was no way of knowing it then, these elements had reached the southern face of what was to prove the final pocket of Japanese resistance on Peleliu.

This curious formation, in contrast to most of the ridge topography, was oriented generally transverse of the island. Thus, in effect, it formed a wall barring passage northward, its southern face almost sheer. Even if the summit could be reached, working along it from peak to peak would be next to impossible owing to the precipitous drops between them. The Sisters were not destined to be permanently secured until November, after most of their supporting positions had been knocked out. However, at this period the Tables of Organization did not provide for any crystal-gazers at regimental level; so, once the advance positions had been consolidated, the battalion optimistically set about mounting another assault.

All ground weapons were brought into play where possible including tanks, bazookas and flame-throwers, both portable and self-propelled, and mortars and machine guns blanketed the targets. Losses were terrible, and the attack against the main objective got nowhere. However, on the open low ground on the extreme right, 2/1, with Company C and the Division Reconnaissance Company attached, succeeded in pushing some distance forward past the eastern flank of the ridges, though at such cost that Companies F and G had to be combined with a squad from the 4th War Dog Platoon to make up a single skeleton company. But the day's operations resulted in the deepest penetaration to be achieved by the 1st Marines in this zone, seizing ground which, following a subsequent withdrawal, was not destined to be regained for nearly two weeks.

Company C began this action with 90 men. About noon, Captain Everett P. Pope, its commanding officer, was ordered to seize Hill 100,51 a steep, seemingly isolated elevation which dominated the East Road and the swampy low ground to the battalion's right front. Company C approached through the swamp and reached the road at the base of the hill, where two strong pillboxes were discovered. The men were in the process of assaulting these when they were suddenly pinned down by machine-gun fire at a range of about 50 yards from across a small pool on their right. Unable to get at these assailants, the captain obtained permission to withdraw the way he had come, pass to the left of the main lagoon and attack up the road with the support of tanks.


HORSESHOE CAUSEWAY, where two tanks came to grief supporting Company C, 1st Marines, on D-plus 4.

The road at this point angled sharply to the east, traversing the main lagoon, or sink hole, on a causeway across the mouth of a wide draw which was to gain an evil reputation under the name of Horseshoe Valley, or just "the Horseshoe." Beyond that, it skirted the base of the objective hill past the positions where the company had previously been pinned down, then angled to the northeastward.

The withdrawal was effected successfully, but it was late afternoon before the new advance began. The first tank to attempt the causeway slipped over one side and became immobilized. A second tank, in an effort to extricate it, slipped off the other side,52 thus blocking the narrow approach to further supporting arms. Nevertheless, Company C crossed it in squad rushes, paused momentarily at the foot of the hill, then assaulted with only mortar and machine-gun support. The men's rush carried to the summit, where they were disconcerted to discover that, instead of having seized the crest of an isolated hill, they had merely gained the nose of a long ridge where their own position was dominated by a higher knoll only some 50 yards to the front.53

The Japanese reacted with promptitude and vigor. On that open ridge crest, the men of Company C were exposed not only to fire from the commanding ground to their front, but crossfire from the parallel ridge that rimmed the Horseshoe on the west: the Five Brothers (see Chapter VII). Counterattacks were incessant throughout the night


CAPTAIN EVERETT P. POPE, who was given the Medal of Honor in recognition of his valorous conduct in the abortive seizure of Walt Ridge on D-plus 4.

and produced some savage close-in fighting that degenerated into a matter of fists and rocks and throwing the enemy bodily over the cliffs. But at dawn, with only 15 men and one officer left, Captain Pope was ordered to withdraw to the main line, a movement again effected successfully through the swamp.54

It had not been Colonel Honsowetz's intention to send Company C on this mission alone, and the Reconnaissance Company had been ordered to support it on the right. However, Captain Pope had already carried the summit before the 2d Platoon of the latter unit, supposed to tie in here, could get into position to assault, and "There is some doubt as to whether Captain Pope realized this platoon was supporting him during the entire night to follow."55 With no flank contact with the troops atop the ridge, the platoon also received some rough treatment throughout the night, and at dawn fell back to the cover of the swamp's edge before an attack in force. For some reason, the Reconnaissance Company did not receive the withdrawal order until half an hour after the retirement of Company C and came under additional fire from the reoccupied positions on the ridge crest. Falling back at last via the swamp route, the survivors reached the main line safely, where soon afterward the depleted unit was relieved.

Concurrently with the action in the eastern zone, 2/7, tying in on the left of 2/1, slugged its way through the jumbled high ground in an effort to draw abreast of the western flank of the Five Sisters. Some progress was made, but as a flanking movement the attempt was fruitless in that Japanese resistance kept the assailants from reaching any commanding ground from which to bring supporting fires to the assistance of 2/1's attack. Company A, 1st Marines, passed through elements of 2/7 in a turning movement from the west with 56 men in assault. These gained a little ground along a ridgetop which was swept by machine-gun fire, only to find themselves suddenly stopped by a sheer drop of 150 feet. Only six of the Marines who participated in this fight were still unwounded when they finally got back to 2/7 lines.

The attack on the morning of D-plus 5 (20 September) had for its main objective the recapture of Walt Ridge. Once more 2/1 assaulted with the 1st Battalion (less Company A) and the Division Reconnaissance


FORWARD POSITION of 1st Marines in lee of Horseshoe causeway. Southern nose of Walt Ridge in background.

Company attached. Every resource under the regiment's control was employed. Every officer and man who could be spared from regimental and battalion headquarters was attached to the companies.56 Additional tanks, LVT(A)'s, half-tracks, 37mm guns, and mortars were brought forward as far as possible. Deadly Japanese crossfire knocked out many, and those remaining proved inadequate to support the infantry successfully against the combination of cruel terrain and devilishly prepared positions.

This was an all-out effort; there was nothing in reserve. Its result is described as well by the Japanese as in any Marine reports, and a great deal more succinctly:

Since dawn the enemy has been concentrating their forces, vainly trying to approach Higashiyama (Walt Ridge) and Kansokuyama (Hill 300, at this time considered by Marines as part of the Five Sisters formation) with 14 tanks and one infantry battalion under powerful aid of air and artillery fire. However, they were again put to rout, receiving heavy losses.57

Although some advance was made, the low ground proved untenable, and the exhausted survivors fell back to a position where they could cover, at a respectful distance,


the mouth of the Horseshoe. After six terrible days, it had to be admitted that the 1st Marines had been definitely stopped. More days were to elapse before graves registration men could get to the dead at the points of fartherest penetration.58

In the center 2/7 advanced slowly almost due east, battling sheer cliffs as well as heavy fire, and by mid-afternoon Company F, on the right, succeeded in gaining the crest of a ridge (then designated Hill 260) facing the Five Sisters from the west across the mouth of a narrow, steep-sided declivity.59 The battalion front had been narrowed and shifted farther to the right by having 3/1, on the western coastal flat, extend its front across the road to the foot of the rising ground, which movement once more held up the advance of the latter battalion.

First Marines casualties reported on 21 September totalled 1749. That the preponderance of these were among the front line units further served to lower combat efficiency far beyond the danger point.60 Although the 3d Battalion continued in its zone under regimental control for two days longer, pushing patrols as far forward as the coastal village of Garekoru, the 1st Marines, as an assault unit on the regimental level, had ceased temporarily to exist. In accomplishing its mission to this point, it had killed an estimated 3,942 Japanese and had reduced the following major enemy positions and installations: 10 defended coral ridges, 3 large blockhouses, 22 pillboxes, 13 antitank guns, and 144 defended caves.

Defensive positions on southern Peleliu were taken over by artillery elements, and the 7th Marines (less 2d Battalion), after due reconnaissance of its new positions, completed the relief of 2/1, 1/1 and 2/761 at 1820 on the afternoon of 20 September. The battered remnants of the regiment came down out of the ridges at last, to recuperate as well as conditions permitted in the more or less static eastern defense zone (RDF area, PURPLE Beach, and the northeastern islands) which had been set up by the 5th Marines, whom they relieved.

But the arrival of comparatively fresh troops at the main scene of action had no immediate effect on either the tactical plan62 or the results achieved.

The 1st Battalion, 7th Marines, took over the zone on the right from 1/1 and 2/1, and the 3d Battalion relieved 2/7 in the center, with 3/1 and the 1st Marines command post remaining in position on the left for the time being. The 1st Battalion resumed the attack promptly in the morning of 21 September, in the same direction as the final assault the previous day, and with the same objective: recapture of Walt Ridge. The jump-off was postponed until 0800 owing to delay in the arrival of the supporting tanks, giving artillery concentrations and an air strike plenty of time to work over the target.

The plan called for an advance up the East Road. Because of the narrowness of this approach, only one company was placed in assault: Company C, with Company A following close to provide infantry support for the tanks and be in ready position to pass through should the situation demand.63

Smoke shell concentrations from the battalion's 81mm mortars helped to screen the approach, and initial opposition was described as light. The tanks were unable to


Map 7
Second Operational Phase

negotiate the still blocked causeway across the mouth of the Horseshoe, but they succeeded in circling the swamp, or sink hole, to the northward, thus entering that ill-famed valley itself, and rejoined Company C on the road beyond, where it skirted the base of Walt Ridge.

The flat here was narrow, hemmed in on the east by that fringing swamp which comprised a major terrain feature along most of this flank of the northwestern peninsula. No enemy resistance was encountered from there, but fire from the ridge intensified as the troops moved abreast of it. And as the advance elements wheeled to the left and commenced the steep ascent, they were caught in a vicious mortar barrage that blanketed the entire eastern face, supplemented by grenades and small arms fire from caves higher up the slope itself.

WHERE THE NORTHWARD DRIVE WAS HALTED. East Road skirts base of Walt Ridge, which faces Five Brothers across floor of the Horseshoe. Five Sisters and Hill 300 on left, with weird contours of China Wall barely discernable at upper left. Swamp in foreground channeled assault approaches.


Mounting losses made further advance impossible, and the ground gained quickly became untenable. To commit Company A at this juncture would have been fruitless, and by mid-morning both companies had rejoined Company B which had remained on the defensive line from which the attack had originated. There the 1st Battalion was relieved in position by the Regimental Weapons Company at 1800.64

On this same morning (D-plus 6) the 3d Battalion attacked from its new positions deeper in the ridges on the left of 1/7 behind naval gunfire, air, artillery and mortar preparations. Three medium tanks and an LVT flame-thrower were in support but could be of little use because of the broken nature of the ground. Companies I and K in assault (left to right) reported a gain of about 100 yards at 0918, but "the advance for the rest of the day was slow and tedious and measured in yards."65 The battalion's left tied in with the flank of 3/1, the last element of the 1st Marines remaining committed, on the western slope of the ridge system, but a gap began to develop between 3/7 and 1/7 on the right. It was now apparent that the only substantial gains were likely to be achieved on the western coastal flat, and the 3d Battalion's front was accordingly narrowed in order to make some of its elements available for support in that zone.66 This was accomplished by ordering the 2d Battalion, which had reverted to parent control at the time of its relief the previous day, back into the center of the line to maintain contact.

The 7th Marines attacked again on 22 September. On its narrowed front, the 3d Battalion moved out in column of companies, Company L passing through Company I to take the lead (Company K had been pinched out the previous evening and gone into battalion reserve). The terrain became rapidly worse; mortar and machine-gun fire pinned the troops down repeatedly. The maximum gain for the day was recorded as only 80 yards, with even less on both flanks.67

This retarded advance limited, in turn, the progress of the 3d Battalion, 1st Marines, on the extreme left. The right flank of this unit was now also operating in the high ground, where it was having troubles of its own. Not until late afternoon did Company L, by dint of hard fighting all day, succeed in seizing a crucial elevation as an anchor for a tenable defensive line for the night. By that time 3/1 was so depleted and exhausted that it was necessary to bolster the line with elements of the 1st Battalion, 1st Marines, which, reorganized into two abbreviated companies, had been moved up in reserve.68

In the afternoon still another head-on assault was launched against the southern face of what it was becoming apparent was the enemy main line of resistance. At 1530 the 1st Battalion, 7th Marines, which had spent the night in bivouac north of the airfield, moved forward to the left of the defensive line sector held by the Weapons Company in an effort to seize the Five Sisters. Company B, in reserve throughout the battalion's fruitless attempt on Walt Ridge the previous day, was now in assault, to be supported as necessary by Companies A and C. Company E, tying in the right of the 2d Battalion's defensive line with the flank of the Weapons Company was also made available in case of need.

Artillery concentrations by 75's and 105's prepared the way, and a platoon of tanks


DEATH VALLEY was first entered by 7th Marines on D-plus 7, in an attempt to seize Five Sisters (right) from southwest.

from Company B, 1st Tank Battalion, advanced with the infantry. The Weapons Company contributed a fire feint against the objective of the day before: Walt Ridge and the lower end of the Horseshoe. Mortars placed so effective a smoke screen that the assault company advanced 250 yards over ground so rocky that it slowed the tanks to a crawl without encountering anything worse than "moderate" sniper fire.

In short, what should have been adequate conventional fire support for such an attack had been provided; yet at this point Company B began to receive machine-gun fire from its front and the ridge on the left,69 which became progressively worse as the advance continued. A few minutes later, 1/7 Historical Report records the time as 1630, the troops reached the mouth of a narrow draw, on the left lip of which elements of the 2d Battalion were in position.


RIDGES WHERE 1st AND 7th MARINES FOUGHT D-plus 2--D-plus 7. Approximate front held at end of phase sketched in.

The tanks fired pointblank into the cave openings which pocked the steep walls. A platoon of Company A moved up to reinforce Company B, now suffering severe casualties. But the assault had not moved far before discovering that the draw was walled in on the north by a high, steep coral formation. The depleted units found themselves in a topographical funnel, with sheer sides from which mutually supporting dug-in enemy positions covered the low ground with fire from every angle,70 and which were impervious themselves to infantry assault.

Once more, an untenable position had nullified a day's substantial gains. At 1749 Company B commenced its tortuous withdrawal. The canyon was again blanketed with smoke, and the tanks and Company A closed up to help in covering the extrication. By 1830 what was left of the 1st Battalion, which had begun this action with only 60% effective strength, was safely back in its bivouac area behind the lines. Only there was not much left.

Again, the Japanese version of the action is terse and to the point: "This afternoon


the enemy attempted to demolish our positions around Kansokuyama and Higashiyama. Attempt unsuccessful, and only a part of the enemy advanced to a hill west of Kansokuyama."71 The last allusion is not wholly clear, as the Marines claimed no advance. It may possibly refer to the seizure the previous day of the northern extension of Hill 200 where it bounds the western mouth of Death Valley, news of which might have been delayed in reaching the Japanese command post.

The Japanese estimates of their own accomplishments up to this point contain a few interesting items, but perhaps are not much more out of line, at that, than Marine estimates of 6,880 Japanese killed.72

Enemy's total losses (except for those during landing operations):
Casualties to personnel at least over 5,000
Tanks including alligators over 120
Artillery guns approx. 1573

Marine reports of the same date listed personnel casualties of 3,946 (including landing operations) and one 105mm howitzer lost to enemy action (three others lost when DUKW's carrying them sank en route back to ship); 19 out of 30 tanks were operative, only six being total losses during the entire period of 1st Tank Battalion's participation in the operation.74

Repeated Japanese use of the term "alligator" evidently stems from one of their early intelligence bulletins which identified and described the LVT(1) (now obsolete) which was designated by that title. It is not clear whether Nakagawa was lumping all tracked vehicles with the tanks, or including only LVT(A)'s where the comparison is reasonably apt, but his figure is rather startling in any case. No accurate reports are available regarding total amtrack losses up to this date, but only 22 LVT(A)'s were permanently knocked out during the entire operation.75

With this day's action, a definite phase of the Peleliu operation came to an end. The immovable object had stopped the irresistable force--for the time being. But there was another day coming up; many other days, in fact.


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


1. For further detail on the military thinking underlying the Japanese defensive concept on Peleliu, see Appendix C.

2. IIIAC OpnRpt, 2-4. This report gives a particularly cogent and lucid analysis on Japanese tactics on Peleliu.

3. O. P. Smith PerNar.

4. Hoffman op.cit., incompleteness of Japanese defense installations in the Marianas is noted in action reports of most U.S. units participating.

5. It is believed that many of the troops used in the Japanese counterattacks against the 5th Marines on D-plus 1 were drawn from the Beach PURPLE Area. Harris.

6. Tada Record. This persistent belief that the action on Peleliu served to thwart a planned attack on islands to the northward may have been wishful thinking, but it appears to have had a real effect on attempts to reinforce and supply the beleaguered garrison. See Appendix F.

7. There were repeated hits and casualties in adjoining areas, however, one of these within 50 yards of the CP proper. Ramsey. The Japanese appeared to have certain weapons registered on primary beaten zones in the vicinity of the beach, for "The Division CP and 2-11 CP operated with relative impunity for several days immediately adjacent to an area approximately 50 x 150 yards in size that received, by actual count, over 200 major caliber shells in a two hour period." Ltr LtCol N. P. Wood to CMC, 14Mar50, hereinafter cited as Wood.

8. This account synthesized from basic reports of 7th Marines previously cited, except as otherwise noted.

9. 3/7 WD, 6. This source contains a detailed description of this formidable structure.

10. Ibid. This maneuver had the effect of pinching out most of the elements of 3/5. For their subsequent disposition, see sub-chapter following.

11. A minor classic of descriptive understatement was contributed by a young lieutenant of the 5th Marines. Surveying the terrain with increasing disapproval, he turned to his battalion commander and observed, in a voice out of deepest Texas: "Majah, this sho' is po' land."

12. The defending unit was identified definitely as early as 1222 on D-plus 1. 7th Mar R-2 Journal, Sheet #21.

13. Some indication of the men's conditions is provided by a terse dispatch, CO 3/7 to CO-7, logged at 1324: "Out of water. Troops having dry heaves." Ibid., Sheet #22.

14. Gormley.

15. 2/7 War Diary, 3, gives 1010 as the hour Co G returned to battalion control, 1245 when Bn CO reported to 1st Marines. The 81mm mortar platoon was not released by 7th Marines at this time and did not rejoin the battalion until 1730 the following day, 2/7 Unit Journal.

16. 7th Mar R-2 Journal, Sheet #20. 1st MarDiv SAR, II, 4. Subsequent activities of 2/7 are treated herein in sub-chapter dealing with operations of 1st Marines.

17. As in the case of the southeastern promontory, a narrow sand spit provided the only dry approach. Under cover of darkness on D-plus 1 a demolition crew disarmed the numerous mines sown across it, making possible the use of tanks and half-tracks in the subsequent attack. Gormley.

18. "The heroic action of stretcher teams, evacuating wounded across the narrow sand spit under intense enemy fire, merited particular praise."--Ibid.

19. Although they did not comprehend its full significance at the time, the Marines were encountering the Japanese tactic of "passive infiltration," which was to assume the proportions of a major problem during later phases of the Peleliu operation. See Chapters VII and VIII.

20. 7th Mar R-2 Journal, Sheet #43.

21. Ibid., Sheet #44.

22. Derived from basic reports of 5th Marines previously cited. Supplementary sources specially noted.

23. A zig-zag communication trench, without revetments. A map captured later indicated that it served as a registration area for a Japanese battery. This was the third time since its establishment that it had come under heavy fire. On this particular occasion, 15 to 20 rounds were received, the two noted doing the most damage. Ltr Capt L. T. Burcham to CMC, 7Mar50.

24. The R-1 replacement was Captain Paul H. Douglas, division adjutant, who thereupon entered a colorful and distinguished phase of his military career which has gone down in the traditions of the 1st Marine Division. He was wounded in turn during a later stage of the operation and again, more seriously, on Okinawa. A former Professor of Economics at the University of Chicago, LtCol Douglas at this writing (April 1950) is serving as U.S. Senator from Illinois.

25. 3/7 did not start sweeping southward until completely across the island in its zone of action; hence, its initial attack crossed the front of the position organized by 3/5 on D-Day night. Walt.

26. Ltr Maj J. H. Gustafson to LtCol G. D. Gayle, 8Nov49, hereinafter cited as Gustafson. Up to this time, Major Gustafson had been Executive Officer of the 2d Battalion, Major Opie Bn-3 of the 1st Battalion.

27. This relief was commenced in mid-morning and effected with great difficulty as 1/5 was receiving heavy fire from the left flank and suffering many cases of heat prostration. Relief was completed about 1230. Gustafson. 5th Mar WD. 1/5 Bn-3 Journal gives hour of completion as 1410. The 1st Bn. passed into regimental reserve where it remained throughout this phase of the operation.

28. For further description of the tactical parts played by both the East Road and the Umurbrogol, see Chapters VI and VII.

29. A Japanese radio direction finder was located here, leading the region to be labeled on the maps, and frequently referred to in reports, as the "RDF Area".

30. The second air strike was ordered by regimental command, but word failed to reach the Battalion CP which was then in the process of displacing forward. Harris. Nearly all of the 34 casualties sustained by the 2d Battalion during this day resulted from friendly fire.

31. 5th Mar R-2 Reports states that this patrol "encountered no resistance," but adds that it "killed three enemy." The patrol returned to the main island for the night at 1730. 5th Mar WD. (2/5 Journal gives hour of return as 1615).

32. 5th Mar WD, confirmed in detail by Walt. The extensive jungle and swamp patrolling during this phase of the operation provided the dogs with almost their only chance to prove their usefulness on Peleliu. Mistaken attempts to employ them elsewhere led to their virtual annihilation. The Action Reports of the 4th and 5th War Dog Platoons make pitiful reading.

33. Subsequently designated Walt Ridge (see Chapter VII). The attack which reached its summit on 19 September was that by Captain Everett P. Pope's Co C, 1st Marines, described in sub-chapter following.

34. Synthesized from basic 1st Marines sources previously cited except as otherwise noted.

35. 1st Mar Nar.

36. Davis.

37. "At this point the 1st Battalion was tied-in in the middle of the 3d Battalion with Company K on our left and Company I on our right. . . . I had just finished a final check of our tie-in when the counterattack came. Hunt bore the brunt of the attack, but flanking fire from Companies B and C plus all our mortars no doubt aided him in beating it off." Ibid.

38. Hunt, op.cit., 139.

39. For a more particularized description, see Chapter VII of this monograph; also Appendix D.

40. At this stage 2/7 was earmarked for the shore-to-shore operation to seize the off-lying island of Ngesebus, to the North. According to the Commanding General's optimistic time schedule, the ground should have been prepared for this move within two or three days; hence, it was desired to keep the battalion intact. The fight in the ridges, however, necessitated a hurried revision of the entire Ngesebus plan.

41. The impossibility of "digging in" in any accepted sense of that term cannot be over-emphasized if a true picture of the fighting on Peleliu is to be given. Repeatedly it was the controlling factor in nullifying costly gains.

42. This blockhouse was not identified as such on the map, showing merely as a large building. The difficulties under which the admiral labored at this time go far toward explaining what appeared a serious oversight: "At the time, in spite of all my previous recommendations. . . . I had only a cruiser division operating staff, consisting of four officers. . . . During the preliminary bombardment and until several days after the landing, my entire staff were on the sick list, only my flag lieutenant remaining on his feet." Oldendorf.

43. Activity in this blockhouse had been reported late the previous day, and the naval gunfire mission was delivered prior to the jump-off on D-plus 2. Major Davis had his CP in what was left of the blockhouse by 0930, "mingled with about 20 freshly dead Japanese." Davis.

44. Hist 1st Mar, 10. The Japanese, however, appeared to disagree: "The enemy began to attack our lines along the hills of Tenzan (southwestern ridges) . . . but were repulsed by our timely firing." Tada Record, 18Sept44.

45. Experience in the Marianas campaign had convinced intelligence officers that there were doubtless many caves in the Peleliu ridges, so the division was not wholly unprepared. But nothing in this or any other war gave any hint of the magnitude of the problem which developed here. These caves and the methods developed for reducing them are dealt with in detail in later chapters and in Appendix D. For the time being suffice it to quote Time, 16Oct44, which characterized them, without being accused of overstatement, as "the incarnate evil of this war."

46. 2/7 Unit Journal. 1st Mar Nar erroneously identifies this as Co E. One platoon was virtually wiped out while fighting with 2/1. Berger.

47. This was the original. "Bloody Nose" ridge, so christened by the troops who first came up against it. The designation immediately became popular, however, and from reports which have come in it would appear that virtually every battalion in the division had a different conception of exactly what terrain feature "Bloody Nose" applied to. Communiques and news dispatches of this period usually referred to "Umurbrogol Mountain", a name appearing on the early maps but applying generally to the peninsula's entire southern ridge system. See subsequent notes on Peleliu place names.

48. This was Company L. As described in the previous sub-chapter, D-plus 3 was the day when the main strength of the 5th Marines began the long stride which resulted in securing all of eastern and northeastern Peleliu.

49. Berger. In this connection it may be significant that a study of 2/7 War Diary indicates that some of the target areas on which preparatory fires were ordered on the morning of 19Sept lay behind or squarely upon positions shown on some unit overlays as having been occupied the previous day.

50. 2/7 WD.

51. To differentiate this tactically important elevation from several others also known as Hill 100, it will be referred to hereinafter as Walt Ridge, the name by which it came to be designated on a subsequent map. Lack of indigenous place names on Peleliu necessitated some hurried improvisations in order to orient terrain features, and some titles selected gave rise to much disgruntlement. In the interests of simplicity, this monograph will use the names, however unjustly applied, by which they came to be known most generally. See Chapter VII.

52. The Japanese apparently believed that they had knocked out these tanks, for they wasted no more fire on them. Both were subsequently retrieved undamaged.

53. Unit reports deal with this action only in general terms. Detail herein derives from Ltr Maj E. P. Pope to CMC, 8Mar50, hereinafter cited as Pope.

54. "Only seconds after we had abandoned the summit, the enemy had moved in, putting in action a LMG." Pope. Capt Pope was the only company commander in the 1st Battalion to retain his post through the operation. He was awarded the Medal of Honor for this desperate fight, and four of his officers and men received the Navy Cross.

55. Ltr 1stLt R. J. Powell, Jr., to CMC (undated) received 21Mar50. As no reports of the Recon Co are extant, this document provides the only source for details of the action described.

56. Indicative of conditions in the assault units is the following entry for this day in 1/1 Unit History (p. 28): "The Battalion 2 Section, cooks, jeep drivers and men from general duty of Hq. Company were organized into machine gun platoons under Lt. Stanfield, the Battalion Adjutant."

57. Tada Record, 20Sept44.

58. The bodies of Captain Pope's dead atop Walt Ridge were not recovered until the final securing of that elevation on 3 October.

59. "Death Valley" (see below).

60. Over-all regimental casualties after the last elements had been relieved came to 56%: 71% in 1st Bn., 56% in 2d Bn., 55% in 3d Bn., 32% in Regimental Headquarters and Weapons Company. Of the nine rifle platoons in the three companies of the 1st Battalion, 74 men--and no original platoon leaders--remained. 1st Mar Nar, 12.

61. In three days and nights of steady fighting with the 1st Marines, 2/7 had sustained more than 300 casualties while achieving a maximum gain of 800 yards. Berger. The battalion reverted to control of its parent regiment upon relief.

62. For discussion of the tactical decision confronting the Commanding General at this time, see Chapter VI.

63. 1/7 Historical Report, supplemented by 7th Mar R-2 Journal. These two documents provide the only official sources currently available on this particular action.

64. With a few offensive sorties to lend variety, the Weapons Company was destined to occupy this containing position for the next two weeks, until the jutting jaw of Major Joseph E. Buckley, its CO, became as familiar a terrain feature as Walt Ridge.

65. 3/7 WD, 10.

66. At this time it was contemplated that the 5th Marines would be employed to exploit the soft spot on the left, and FO #2-44 dated 21Sep44 was issued accordingly. This plan was revised the following day in FO #3-44. 1st MarDiv SAR, II, 5.

67. 3/7 WD, 10.

68. 1st Mar Nar, 12. 3/1 patrols operating forward of the battalion position worked more than 1000 yards up the coast to the village of Garekoru on this day or the one following. Ibid., and 7th Mar R-2 Journal disagree on the date, and 3/1 Record of Events makes no mention of the movement.

69. This was the northerly extension of Hill 200, called "Nakayama" by the Japanese. Although friendly troops occupied the crest, they were unable to get at the enemy still emplaced in strength in caves along the precipitous slope. The relative positions as of 24 Sept are clearly indicated on a Japanese sketch map captured subsequently. This route of attack remained in peril from the flank until the final wiping out of the holed-in enemy on 3 Oct.

70. This declivity was christened "Death Valley", at that time (D-plus 7) and continued to be so designated throughout the operation. They could not know it then, of course, but the Marines had penetrated within 100 yards of Colonel Nakagawa's CP, the nerve center of Japanese resistance and the last strong point on the entire island to be secured. See Chapters VII and VIII.

71. Tada Record, 22Sept44.

72. 1st MarDiv SAR, II, 7.

73. Tada Record, 22Sept44.

74. All figures from 1st MarDiv SAR, II, 7; Anx H; Anx J.

75. Boyer Monograph.

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