How many caves existed on Peleliu, and how many of these were utilized by the Japanese defenders, will never be known. After the conclusion of the land fighting, the Island Command undertook a comprehensive correlation of all intelligence on the subject, together with a painstaking on-the-scene survey of all caves still recognizable as such, which counted more than 500.¹

An interesting aspect revealed by this study is the fundamentally different approach to cave utilization by Japanese Army and Navy elements. Perhaps characteristically, the Army's principal interest was adaptation of terrain to defensive combat. In contrast, the Navy, at least until shortly before the landings, appeared to regard caves mainly as shelters against attack by sea or air, their defensive possibilities purely incidental.

Thus, the Army made use of all the many natural caves in Peleliu's coralline limestone formations which possessed tactical value, improved on nature as necessary, and constructed their artificial caves as covering or mutually supporting positions. The latter occurred in a variety of forms, depending on local terrain conditions and tactical situations. Most of them were small, owing to limiting factors of time and labor available, so the larger natural caves were mainly adapted, often with great ingenuity, for the emplacement of heavier weapons.

Naval troops, on the other hand, almost entirely ignored natural caves, preferring to construct their own. In this, progress was facilitated by the presence of the 214th Naval Construction Battalion, a tunnel construction unit (Suidotai), made up of personnel who had been miners and tunnel workers in civilian life, commanded by civilian engineers. Their handiwork was characterized by professional competence. Passageways were of ample size and free of the cluttering stalactites and stalagmites inevitable to natural limestone erosion, with cross-tunnels and excavations opening off both sides for the convenience and protection of the occupants.

Most of the Navy caves were located near the northern end of the northwestern peninsula, though a few had been constructed adjacent to Army installations for liaison purposes. Although they were not designed primarily for combat, defensive tactics could be improvised readily, particularly in the larger ones which provided a high degree of interior security owing to the complex of passageways and bays. This was

Diogrom of cave system (lower level only) in northern ridge (TS 159Q)
PELELIU ISLAND

demonstrated by the trouble encountered by the 5th Marines in northern Peleliu. But the failure to integrate these positions into a system of mutually supporting defenses accounts for the comparatively quick securing of the Amiangal ridges in contrast to the long-drawn-out process of attrition which took place in the Umurbrogol.

The accompanying sketch shows the floor plan of the largest and most elaborate Navy tunnel system discovered on Peleliu; so elaborate that many Marines believed it to

be the mine which supplied the raw material for the phosphate refinery, located nearby.² It is a cave (or caves) of what is defined as the "adjoining H's type," with multiple entrances facing in three directions from the nose of the northernmost ridge on the entire island. Dominating the West Road and the reef approaches to Ngesebus, it could be by-passed only by use of the extraordinary concentration of supporting weapons described in Chapter VI. Viewing its complex interior, one can understand more readily the plaint voiced in 5th Marines Regimental Narrative that: "Tank guns, firing point-blank directly into the caves and tunnels, did not even temporarily cause the enemy therein to cease fire."

How the enemy utilized this striking example of the miner's art has been reconstructed in some detail from the description furnished by a prisoner taken in February 1945: one of five living Japanese still lurking there three months after the last fighting had faltered out in the Umurbrogol.³

On 3 September when the prisoner first entered the caves there were over a thousand men in them. . . . On 28 September 1944, (D-plus 13) all the military personnel were organized for an attack on the American forces who were holding the hill directly over the caves. They poured forth from all nine tunnels and attacked. He claims they drove the Americans off the hill but suffered many losses.⁴ When they reassembled in the caves there were only about 50 military men left. . . .

The survivors moved the wounded into tunnels 9, 10, and 11, the workers into tunnels 4, 5, 6, and 8, and the military personnel into tunnels, 1, 2, and 3. On the morning of 29 September our forces assaulted the entrance to tunnel #1 with a tank, machine guns and flame-throwers. The Japanese military personnel attacked through tunnel #1 and all the military were killed except the prisoner who remained behind an improvised barricade at "A" in tunnel #2, and six others who were in lower passages. The flame-thrower used by our forces penetrated to where he was and burned his left arm and leg. He claims the flame-thrower was our most effective weapon in this attack and that it reached point "A" where he was standing and points "B", "C" and "D" killing some of the workmen who were located in these areas. The tank and machine gun fire was only effective in the immediate tunnel it was firing into.

Later, tanks and flame-throwers were used on most of the entrances and when our forces withdrew there were only 30 remaining alive. The wounded in tunnels 9, 10 and 11 were all killed by our flame-throwers which penetrated the entire length of tunnel 7-8 and to points "F" in tunnel 9 and "G" in tunnel #6. Our flame-throwers also penetrated the entire length of tunnel #3 and tunnels #4 and 5 and around to points "N" in tunnel #6 and "H" in tunnel #1. Tunnel #2 was not subjected to the flame-thrower and most of those men who survived had taken refuge in it.

Our forces then closed all entrances and the Japs moved into tunnel #2 which was the most completely closed. They posted men at points "I" in tunnel #5, "J" in tunnel #8 and "K" in tunnel #11. These men were able to pick off a number of "Foolish Yankees" who wandered into the entrances of tunnels #10, 7 and 4 during November and December (TN: During this period American personnel were killed and wounded in and around this cave). According to the prisoner as soon as they fired at anyone they would withdraw into tunnel #2 and as many times as flame-throwers or demolitions were used on the various tunnels they were always safe because the entrance to this tunnel was completely covered and unrecognizable from the outside.

Prisoner states that about 1 January 1945, after they shot an American in tunnel #8, very large explosive charges were set off in all the entrances except tunnel #2 and that the force of these explosions was much more penetrative than the flamethrowers. They reached to every part of the cave and shocked those they did not kill. If charges had been put in tunnel #2 all would have been killed. However, 19 of the remaining 30 were killed and three more were badly injured. On 24 January

1945 American "Seabees" who were excavating in that area stumbled on two of the Japs at "A" in tunnel #2. The prisoner states that they then moved the three wounded into tunnel #6, left two men at "A" in tunnel #2 as guards, and the others went to tunnel #11, passing point "L" in tunnel #9 where our demolitions had caused the tunnel to partially collapse. The flame-thrower used on tunnel #2 at this time killed the two guards. (TN: Probably gasoline that was used by our forces in an attempt to burn out this cave at that time). The prisoner claims that on the night of 1 February 1945 the five Japs who were in good physical condition dug their way out of tunnel #8.

Footnotes

1. Japanese Military Caves on Peleliu, "Know Your Enemy!" CinCPac-CinCPOA Bulletin 173-45. This detailed and elaborately illustrated work provides the basis for most of the material incorporated herein.

2. Actually the abandoned phosphate diggings lay some distance to the south, in the vicinity of Asias village, where mining was by the open-pit rather than the tunneling method. JANIS No. 103 Vol. I, p. VIII, 4.

3. Peleliu Island Command POW Interrogation Report #6.

4. No mention of this counterattack appears in any Marine operation reports at regimental or higher levels, and the participating battalion (2/5) turned in no written report of such counterattack. The CO of this unit recalls heavy fighting and casualties that night, but the Marines atop the hill (elements of Company F) most certainly were not driven off. Oral statement by LtCol Gordon D. Gayle, interviewed 17Feb50.