American intelligence predicted the fight for Peleliu would last four days. This was a catastrophic miscalculation. The battle would grind on for over two months, a direct consequence of a failure to understand a fundamental shift in Japanese defensive doctrine. Colonel Kunio Nakagawa of the IJA 2nd Infantry Regiment, a veteran of combat in China, commanded the island’s defense. He discarded the strategy of annihilating landing forces at the water’s edge. Nakagawa ordered his 11,000 men, a mix of the 14th Infantry Division and naval troops, to concede the beaches. The real battle would be fought inland. His defensive scheme centered on the Umurbrogol Mountain, a jagged spine of coral and limestone ridges. For months, Nakagawa’s troops, with Korean and Okinawan laborers, transformed the mountain into a fortress. They engineered over 500 interconnected caves and tunnels, creating a labyrinthine defensive network. Archival analysis (NARA Record Group 127) shows these were not simple shelters. They were sophisticated, mutually supporting positions with concealed entrances and firing embrasures designed to survive naval bombardment and permit defenders to reoccupy cleared positions.
The battle for Peleliu became one of the most vicious campaigns for U.S. forces in the Pacific. The 1st Marine Division, under Major General William Rupertus, landed on September 15, 1944. They immediately faced intersecting fire from artillery pre-sited on high ground. The division suffered 1,111 casualties on the first day. The fight for the Umurbrogol, which Marines nicknamed “Bloody Nose Ridge,” was savage, close-quarters combat. Temperatures reached 115 degrees Fahrenheit. Water shortages led to widespread heat exhaustion. The coral ridges tore apart boots and uniforms. Advancing Marines had to clear cave systems one by one with flamethrowers and explosives, often finding that Japanese soldiers would emerge from a hidden tunnel to attack them from the rear. The 1st Marine Regiment, under Colonel Lewis “Chesty” Puller, was effectively destroyed as a fighting force in six days, suffering 1,749 casualties. The U.S. Army’s 81st Infantry Division was brought in to relieve the battered Marines. The battle devolved into a slow, grinding siege that officially ended on November 27, 1944. U.S. casualties exceeded 9,600. Nearly the entire Japanese garrison of 10,900 was killed.
The fight did not end there. Post-battle examinations revealed the grim nature of the final days. Colonel Nakagawa, having commanded the defense for 74 days, performed ritual suicide on November 24th after burning his regimental colors. His attrition-based defense would be studied and replicated on Iwo Jima and Okinawa. The aftermath also brought to light evidence of atrocities. Operational logs confirm the discovery of mutilated American dead. Several Marines were found tied to trees, apparently used for bayonet practice. Japanese forces on Peleliu took almost no prisoners, with the vast majority of the 301 captured being non-combatant laborers. The violence continued for years. It was not until April 1947 that the last organized group, a party of 33 soldiers led by Lieutenant Ei Yamaguchi, was convinced by a former Japanese admiral that the war was over and formally surrendered.
This entire ordeal was made possible by a logistical success for the Japanese in the summer of 1944. A series of convoys successfully delivered the men and material that translated directly into the tactical nightmare for the Americans. This was not a minor reinforcement. It was a substantial, deliberate influx that proved decisive. Archival analysis shows these shipments were part of a hurried reinforcement of the Palau Islands, using veteran units transferred from the Kwantung Army in Manchuria. The core of the new Peleliu garrison became the 2nd Infantry Regiment of the 14th Division, commanded by Colonel Nakagawa. His force of roughly 6,500 combat troops, supplemented by nearly 4,000 naval personnel and laborers, arrived with the tools needed to enact the new defensive doctrine.
The failure of American interdiction to stop these deliveries remains a key point of analysis. The convoys brought not just rifles and men, but the heavy materials for sophisticated fortifications. Post-battle surveys reveal the direct impact of these supplies: immense quantities of cement, steel rebar, heavy machine guns, 81mm and 150mm mortars, and anti-tank mines were offloaded. These were the building blocks for turning the island’s interior into a fortress. The delivered materials enabled Colonel Nakagawa to execute his attrition-focused plan with devastating effectiveness. He ordered his men to concede the beaches and draw the Marines into the Umurbrogol. The jagged spine of coral and limestone was systematically transformed. Over 500 natural caves were expanded, reinforced, and connected. Many of these positions were former mine shafts adapted for military use. The delivered cement and steel created hardened embrasures for machine guns and artillery, with entrances angled to deflect flamethrowers. Operational logs show these positions were designed as a “honeycomb” system with interlocking fields of fire, allowing defenders to evacuate one position and reoccupy another through the tunnel network. This engineering effort, fueled by the unhindered supply convoys, turned what American planners saw as low hills into a fortress that would bleed the 1st Marine Division for months.
Archival patrol logs for the Gato-class submarine USS Crevalle (SS-291) during its fourth war patrol, from June to August 1944, place it directly in the path of Japanese convoys steaming toward the Palaus. Under the command of Lieutenant Commander Everett H. Steinmetz, the Crevalle operated as part of a coordinated attack group, a wolf pack, tasked with interdicting this flow of material. On the night of July 25, 1944, this wolf pack located a large, escorted Japanese convoy. A running battle ensued for the next 30 hours. The choice facing Steinmetz was not a simple one of attack or not attack. It was a complex tactical problem weighed against the survival of his boat and 80-man crew. Every periscope observation and sonar reading carried the potential for catastrophic failure.
The operational environment in the South China Sea had become increasingly lethal. By 1944, the Imperial Japanese Navy had made significant improvements in anti-submarine warfare (ASW). They had established the Grand Escort Fleet in late 1943, creating dedicated surface and air units to protect convoys. Japanese escort vessels, the Kaibōkan, were being built in greater numbers and equipped with improved hydrophones. Their air assets, sometimes flying from escort carriers, now employed airborne radar for night searches. For a commander like Steinmetz, this meant the window for a successful submerged attack was shrinking. Lingering at periscope depth to refine a torpedo solution exposed the boat to detection. A premature dive to escape a hunter-killer escort could mean losing the convoy entirely. During the 30-hour engagement, Crevalle and its pack mates were repeatedly forced deep by the convoy’s escorts, enduring barrages of depth charges. A submarine commander was under orders to prioritize the safety of his vessel, a strategic asset, over a single tactical victory. Losing a fleet submarine and its highly trained crew was a far greater blow to the war effort than a few enemy transports slipping through.
The physical environment of Peleliu was an adversary as formidable as the Japanese garrison. A review of operational logs shows the invasion on September 15, 1944, occurred as the northwest monsoon season was tightening its grip. Torrential downpours frequently reduced visibility to mere yards, nullifying American air and sea power. Carrier-based F4U Corsairs and spotter planes were often grounded. When they could fly, cloud cover and rain obscured targets. The sea itself became hostile. High winds and unpredictable squalls generated heavy swells that made the transfer of men and materiel from transports to smaller landing craft a perilous exercise. The constant exposure to salt spray and extreme humidity accelerated equipment failure, corroding weapons and short-circuiting radios.
A jagged, fringing coral reef surrounded the island. This reef extended up to 700 yards from the designated landing beaches, creating a wide, shallow moat impassable for conventional landing craft and the Landing Ship, Tanks (LSTs) that carried the bulk of the division’s armor and artillery. Every piece of equipment had to be laboriously transferred from the LSTs to the churning tracks of amphibious tractors (LVTs) for the long, slow, and dangerously exposed journey to the shore. This created a logistical bottleneck under the direct fire of pre-registered Japanese artillery.
The reef itself was an instrument of destruction. The coral was not a flat shelf but a maze of sharp, irregular heads that ripped open the thin hulls of LVTs and DUKWs. On the first day of the invasion, Japanese fire destroyed 60 of these amphibious vehicles. A disabled vehicle could block one of the few viable channels through the coral, creating a traffic jam of stationary targets. Compounding this hazard was the poor quality of hydrographic intelligence. The charts available were often based on old, unreliable data. Underwater Demolition Teams (UDTs) performed reconnaissance, but their efforts were hampered. The murky, churned-up water concealed coral heads and shallows invisible to both aerial photography and swimmers. As a result, LVT drivers went in partially blind, discovering new obstacles at the cost of their vehicles and the lives of the men inside.
American submarine patrol logs from 1944 reveal a persistent vulnerability. The technology for a Gato-class submarine’s navigation was dangerously primitive for the demands of the war. Submarine commanders relied almost entirely on dead reckoning and celestial navigation. This meant tracking their position by plotting course, speed, and time from their last known good fix. That fix was weather dependent. It required a clear view of the sun or stars for a sextant sighting. During the relentless cloud cover of the northwest monsoon season, a submarine crew could go for days without a reliable positional fix. The ship’s magnetic and gyrocompasses accumulated errors. Unseen ocean currents could push a submarine dozens of miles off its calculated course.
This navigational uncertainty bled directly into combat effectiveness. The American Torpedo Data Computer (TDC), an advanced electro-mechanical analog computer, could calculate a brilliant firing solution, but it was wholly dependent on the quality of the data fed into it. A commander had to input his own vessel’s precise course and speed, information corrupted by errors in the dead reckoning plot. If the submarine’s actual position was miles from its assumed one, the firing solution would be flawed, sending a spread of Mark 14 torpedoes into an empty patch of ocean. Heavy seas made periscope observation a nightmare. A commander attempting to gather data on a target would have his view constantly blinded by waves. The violent motion of the submarine made keeping the crosshairs steady an impossibility. The incessant drumming of rain on the hull and the churning of stormy seas created a wall of ambient noise, degrading the ability of hydrophone operators to detect the faint propeller sounds of a distant convoy or the quiet approach of a Japanese escort.
The ultimate consequence of navigating blind was the risk of running aground. A close examination of hydrographic records from the era shows that charts for areas like the Palau Islands were dangerously unreliable. These charts were riddled with inaccuracies and vast, uncharted sections. The waters around Peleliu were a maze of submerged coral reefs and shoals. A navigator’s accumulated error of even a few miles could mean the difference between being in a deep channel and unknowingly steering the boat at 17 knots onto a reef. The loss of the USS Darter (SS-227) in October 1944 illustrates this danger. While maneuvering for an attack in the Palawan Passage, the Darter was forced to rely on a dead reckoning plot over 24 hours old. Believing they were in safe water, the crew drove the submarine hard onto Bombay Shoal. Stuck fast, the crew was forced to abandon and scuttle their vessel. Six US submarines were lost during the war to grounding or accidents, a direct reflection of the intersection of poor charts and imprecise navigation.
A US Navy fleet submarine in 1944 was a strategic weapon. Official US Navy doctrine and the unwritten code of the submarine service both placed an exceptionally high premium on the preservation of the boat itself. A Gato-class submarine was a complex system of diesel engines, electrical batteries, high-pressure air flasks, and delicate hydraulic systems. Combat damage or a simple mechanical failure at the wrong moment could lead to the vessel’s loss. A hydraulic leak in the stern plane controls could make the submarine sluggish and unable to execute a rapid depth change. A failure in the main induction valve, a known design flaw, could lead to catastrophic flooding. The stress placed on the hull by repeated deep dives to evade depth charges could compromise its integrity. For a commander, pressing an attack against a well-defended convoy meant exposing his boat to this gantlet of threats, knowing that even non-fatal damage could leave him stranded hundreds of miles from a friendly base.
The preservation of the crew was an equally powerful imperative. A submarine’s crew of roughly 80 officers and enlisted men was a highly specialized and irreplaceable asset. The training pipeline to produce a qualified submariner was long and rigorous. Losing a full crew represented a devastating blow to the operational capacity of the submarine force, a loss of expertise that could take more than a year to replace. A sustained depth charge attack was a uniquely terrifying experience. The percussive blasts would violently shake the hull, shatter light bulbs and gauges, and cause rivets to pop, while the crew could only wait in the dark. A commander’s responsibility to bring his men home was a core tenet of the service.
Every decision was viewed through the lens of long-term operational readiness. The primary strategic mission of the US submarine force was not to win individual battles but to execute a sustained campaign of unrestricted warfare to sever Japan’s maritime supply lines. This was a war of attrition. A submarine that survived an engagement, even if it failed to sink its target, lived to fight another day. It could return to base, refit, rearm, and go back on patrol. A submarine at the bottom of the ocean could not. This strategy proved exceptionally effective. By war’s end, US submarines, which constituted less than 2% of the Navy’s personnel, had sunk over 55% of all Japanese merchant shipping and a large portion of the Imperial Japanese Navy. The decision by a commander to break off a risky attack was not an act of cowardice but a calculated act of strategic discipline.
Submarine Force Pacific Command directives from late 1944 show a direct response to the operational lessons from engagements like the one involving USS Crevalle. The failure to annihilate the convoy that reinforced Peleliu was not seen as an isolated command failure but as a systemic data point. It reinforced a shift in naval interdiction strategy. The understanding grew that while Japanese anti-submarine warfare was doctrinally inferior to Allied methods, it was becoming more dangerous through sheer numbers and improved hardware. This increasing risk forced a strategic re-evaluation. The lesson was not to become more timid, but to become more lethal through superior coordination and technology.
The US Navy’s response was an institutionalization of learning from failure. Every patrol report was dissected to identify points of failure, whether human, environmental, or mechanical. The loss of the USS Darter to grounding on Bombay Shoal in October 1944 was an exclamation point on the dangers of imprecise navigation. This incident, along with countless others, accelerated the fleet-wide installation of the new LORAN (Long Range Navigation) system. Analysis of LORAN deployment records shows that chains of transmitting stations were rapidly constructed across the Pacific, giving submarines the ability to get an accurate positional fix in any weather, reducing navigational errors from miles to a fraction of a mile. Similarly, the frustration of inconclusive convoy battles led to the wider deployment of improved SV-type airborne radar on aircraft and more advanced surface-search sets for submarines, allowing wolf packs to better track and coordinate attacks. The lessons from Peleliu's reinforcement were integrated into a feedback loop that armed American submarines with the technological superiority to finish the war on shipping.