America’s Naval Guns Destroyed Tinian’s Defenses in 2 Hours — Japan’s Key Island Fell in 9 Days

It could not beat two Japanese heavy cruisers. The Salt Lake City was a Pensacola-class heavy cruiser, 14 years old in the March of 1943, fighting with the hull architecture of the decade that built her. Her 10 8-in guns were arranged in four turrets along her center line, and her belt armor was rated to defeat her own shells at combat range.

A designed condition that was about to be tested in the gray waters south of the Komandorski Islands at the western end of the Aleutian chain. Against her, across a closing range of fog-gray bearing sea, were the Nachi and the Maya, two Japanese heavy cruisers together carrying 20 8-in guns between them, with fire control systems that had been updated through 1942, and crews that had been drilling on heavy gun engagements for 15 years.

The arithmetic of what happened when one aging American cruiser met two Japanese cruisers in waters with no air cover and no reinforcement within useful range was not complicated. What Rear Admiral Charles McMorris did with the arithmetic was not complicated, either. He fought for 3 hours and 23 minutes. His cruiser took eight direct hits from Japanese shells, lost power to her screws, and sat dead in the water in the North Pacific while four American destroyers charged toward two heavy cruisers to buy her the time she needed.

At the end of it, Japan’s Aleutian supply convoy had turned north and would not reach Attu. The garrison waiting for it would fight with what it had carried in from the year before. This is the story of a battle America had no reason to survive and of the supply line that vanished because it did. The Japanese garrison at Attu, in the westernmost reach of the Aleutian Island chain, had been isolated since June of 1942.

Its supply line ran from Paramushiro in the Kuril Islands through the open North Pacific to the southeastern coast of Attu. A route exposed to weather, American submarines, and by February of 1943, American surface forces. The garrison numbered approximately 2,600 men. It needed food, ammunition, and fuel on a schedule that the North Pacific’s weather made irregular, and American interdiction was making dangerous.

The convoy that sailed in March of 1943 was intended to resolve that shortage. Two transports carrying troops and material, escorted by two heavy cruisers, two light cruisers, and four destroyers, the largest surface escort Japan had assembled for the Aleutians. It outgunned every American surface unit that had been observed operating in the sector.

The plan was straightforward. The convoy would reach Attu, offload, and return. It did not reach Attu. Not a single ton of what it carried was delivered. The convoy that turned north on March 26th, 1943, was the last realistic supply attempt Japan would make for the Attu garrison. The men waiting for those supplies would spend the next 2 months on what they already had.

On May 29, they made a final charge toward American artillery positions. Approximately 29 of 2,650 survived. The instrument that produced that outcome was not a fleet victory or a strike from the air. It was two cruisers that declined to stop fighting when the numbers said they should. But none of that began south of the Komandorski Islands.

It began 9 months earlier in the North Pacific with a decision nobody in Washington expected and nobody in Tokyo fully explained. On June 3rd and 4th, 1942, while the Battle of Midway was in its opening hours, 1,500 miles to the south, a Japanese carrier task force struck Dutch Harbor on Unalaska Island and damaged its facilities.

The attack was a diversion, the Aleutian element of the Midway operation, intended to draw American attention north while the main assault targeted Midway. It did not draw American attention north. Admiral Nimitz correctly identified the main effort and left the Aleutians thinly covered. What followed was not in the original plan.

Japan’s Aleutian campaign commander, Vice Admiral Hosogaya, landed troops on Kiska on June 6th and Attu on June 7th. The occupation of two American islands, the first seizure of American territory since 1812, was accomplished without American resistance. The strategic rationale was genuinely unclear to Japan’s own planners.

The Army General Staff had opposed the operation. The Navy had pushed it as an extension of the Midway defensive perimeter. What it produced was an obligation, two island garrisons 1,500 miles from the nearest Japanese base at Paramushiro, in waters that were foggy for most of the year and violent for the rest, requiring continuous supply through weather conditions the Japanese navy was not organized to operate in reliably.

The supply runs began immediately. Between June of 1942 and February of 1943, Japanese ships made more than a dozen resupply transits to Attu and Kiska. Most were conducted in fog and completed without American interference. The American response in the fall of 1942 was primarily from the air.

Army Air Forces bombers flying out of Adak, Amchitka, and Umnak Islands as those bases were constructed through the late months of the year. Air interdiction in the Aleutian weather was expensive and often ineffective. Fog and icing conditions grounded aircraft for weeks at a time while supply ships moved under the overcast.

The surface interdiction campaign was slower to develop. Rear Admiral McMorris took command of Task Group 16.6 in January of 1943 and began running patrols along the northern approaches to the Aleutian chain, specifically in the waters south of the Commander Islands where the supply route passed through a geographic narrowing. In February, those patrols produced a result.

McMorris caught the Japanese transport Akagane Maru and sank her on February 19th, 1943, the first successful surface interdiction of the Aleutian supply route. Japan’s response was direct. The next supply convoy to sail for Attu would carry a heavy cruiser escort. Two heavy cruisers would ensure that an American patrol force could not repeat February’s result.

It did not repeat February’s result. What it produced instead was different in almost every particular. Vice Admiral Hosogaya’s escort force assembled at Paramushiro in mid-March. The core was his flagship Nachi, a Myoko class heavy cruiser displacing 13,380 tons and carrying 10 20.3 cm guns in five twin turrets with a combat speed of 35 knots.

With her was the Maya, a Takao class heavy cruiser of similar displacement and identical main battery. Together, the two ships put 20 8-in guns in the water, more heavy gun firepower than any American surface unit operating in the North Pacific sector. Supporting the heavy cruisers were two light cruisers, Tama, a Kuma class vessel carrying seven 14-cm guns, and Abukuma, a Nagara class light cruiser with an equivalent battery.

Four destroyers completed the screen, Wakaba, Hatsushimo, Ikazuchi, and Inazuma. The convoy itself consisted of two armed transports, the Sanko Maru and the Asaka Maru, carrying military supplies and reinforcement personnel for the Attu garrison. The total force was two heavy cruisers, two light cruisers, four destroyers, and two transports.

It outgunned any American surface group that had been observed in the sector. Hosogaya’s nine previous supply runs to the Aleutians had all succeeded. This one had an escort that the first eight had lacked. He expected to encounter a patrol force, perhaps McMorris’s group, perhaps another, that would measure the odds and withdraw before forcing an engagement.

The route from Paramushiro to Attu traced south along the Kamchatka Peninsula before turning east through the waters south of the Komandorski Islands, the Komandorskiye Ostrova, a Russian island group just north of the convoy’s track. The passage was not a geographic choke point in the strict sense, but the islands provided a navigation reference in the persistent fog, and the convoy’s route was established by nine prior successful transits.

Hosogaya was confident in it. He had good reason for that confidence. The only new variable was that McMorris was in those waters on a specific morning looking for exactly what appeared on his radar at 07:30. Rear Admiral Charles McMorris was at sea on the morning of March 26 because Task Group 16, Six’s patrol schedule placed it south of the Commander Islands on that date.

His force was sized for the mission as it had been understood through February, intercepting lightly escorted transports, not engaging a fleet that outgunned his flagship by a factor of two. McMorris’s task group consisted of his flagship, USS Richmond, a light cruiser of the Omaha class commissioned in 1923, and carrying 12 6-in guns.

Richmond was fast, could make 34 knots on her best day, and was built for a generation of naval warfare that predated the weight of shell that Japanese 8-in guns threw. Her armor was rated against destroyer guns and little else. Her 12 6-in guns had approximately 1/3 the individual shell weight of a Myoko class heavy cruiser’s main battery.

The task group’s striking power was USS Salt Lake City CA-25, the Pensacola class heavy cruiser that was McMorris’ only answer to Japanese heavy cruiser fire. Salt Lake City had been refitted through 1941 and 1942 with updated fire control equipment, and she could hit with accuracy at ranges up to 20,000 yd. She was one heavy cruiser against two.

The four destroyers USS Bailey, USS Coghlan, USS Monaghan, and USS Dale carried 5-in guns and torpedo tubes. Against Japanese destroyers in a fair engagement, they were approximately even. Against Japanese heavy cruisers, they were not a rated threat at gun range. The task group had no air cover. Adak Island’s air facilities were more than 500 mi to the east.

No strike aircraft were positioned for the patrol sector McMorris was working. If his ships were damaged in what was about to begin, nothing was coming for them in any timeframe that mattered to the outcome. American signals intelligence had given McMorris general notice that a supply convoy was in transit on the northern route.

It had not accurately characterized the escort strength. McMorris knew a convoy was coming. He did not know the Nachi and the Maya were its escort until his radar operator counted the blips. At 07:30 on the morning of March 26th, 1943, the radar operator aboard USS Richmond reported contacts to the northwest.

Large returns, multiple. McMorris asked for a count. There were eight contacts. The geometry of what McMorris was about to do required a decision that is not obvious in the accounts that emphasized the heroism of what followed. He could not outrun the Nachi’s 35 knots on any heading that put open water between them.

He could not call for air support that was not in range. He could not withdraw without abandoning the mission that had placed him south of the Commander Islands in the first place. And a withdrawal in contact with a superior force would invite exactly the pursuit he could not survive. What McMorris chose in the few minutes between the contact count and the moment his ships came to battle speed was to fight.

The decision that changed the engagement’s outcome was made approximately 15 minutes after McMorris chose to fight. It was not a decision to engage. It was a decision about range. The conventional response to an unfavorable gunnery situation is to open range. To increase the distance between your force and the enemy’s, accepting lower accuracy from both sides while preserving maneuver room.

McMorris did the opposite. He closed range. The logic was this. At 25,000 yd, the Nachi’s fire control system could engage with its accuracy advantage maximized. Firing at a target that offered a predictable course in clear conditions. At 18,000 yd, both sides were within the reliable envelope of their own systems and the advantage compressed.

At 12,000 yd and below, McMorris’s four destroyers were within torpedo range of the Japanese formation. Information that Hosogaya had to account for with every gunnery solution. A Japanese admiral simultaneously in a gun duel and watching four American destroyers hold torpedo range could not give full attention to either problem.

The second element of the decision was invisible to the Japanese at the time and is not in most accounts of the battle. From the moment of first radar contact at 07:30 until the engagement was decided, McMorris’s ships were transmitting continuously contact reports to Adak, to Pacific Fleet Command, to any station that could relay the task group’s position and the Japanese fleet’s composition.

The content of those reports was accurate. What Hosogaya’s radio intelligence team intercepted was the volume and frequency of American transmissions, not their content. To the Japanese signals officer on the Nachi’s bridge, the American radio traffic looked like the preliminary coordination sequence of a land-based air strike, the kind of communication pattern that preceded a bomber attack, not a surface engagement.

There were no bombers. None were in range, but the interpretation of McMorris’s standard contact reporting procedure as air attack coordination was plausible and it sat in the background of Hosogaya’s tactical decision making for the full 3 hours and 23 minutes of what came next. At 08:39 in the morning, USS Salt Lake City opened fire at a range of approximately 20,000 yd, a little over 10 mi. The range was long.

The purpose was to establish the engagement and force Hosogaya to respond. Hosogaya responded by sending Nachi and the destroyers on an intercepting course while the two light cruisers kept the convoy on its original heading under their escort. The intention was to drive off the American patrol force while the transports continued toward Attu.

The first Japanese salvos fell close. The Nachi’s fire control system was operating at the peak of its accurate range. By 0900, McMorris’ ships were maneuvering hard to spoil the fire control solutions while Salt Lake City’s guns kept up a rate of return fire that forced the Japanese heavy cruisers to maneuver in turn.

At approximately 0910, Salt Lake City absorbed her first direct hit, a shell below the waterline that caused flooding in her forward engine room. The damage was controllable. What it signaled was the engagement’s direction. In the hour between 0930 and 1030, Salt Lake City took four additional hits. Each time, the damage control parties responded.

Flooding was managed. One hit destroyed an anti-aircraft gun mount and killed three sailors. One hit opened hull plating below the waterline and added to the forward engine room flooding. The ship was hurt. She was not stopping. McMorris had arrived at a tactical understanding that he did not communicate directly to his ships because to communicate it in those terms would have changed the character of the engagement.

The battle was no longer about defeating the Japanese escort. It was about time. The convoy, the Sanko Maru and the Asaka Maru, had turned north when the shooting started. Every minute that McMorris remained in the fight was a minute that the transports were sailing away from Attu. Every Japanese shell fired at Salt Lake City was a shell fired by a ship that was not escorting the transports southward.

At 09:30, the convoy was 40 mi northeast of the fighting and moving north at its transit speed. At 11:00 o’clock, it would be 80 mi northeast. The calendar that the Attu garrison was living on was being consumed by the same engagement that was consuming Salt Lake City’s hull integrity. And the question of which one ran out first was not a tactical question.

It was a question of what McMorris’s ship could absorb and keep moving. Between 10:00 o’clock and 11:00 o’clock, McMorris used Richmond’s smoke-making capacity to screen Salt Lake City from the worst Japanese firing solutions. Not withdrawing, but buying the cruiser cover between hits. He positioned his destroyers on Salt Lake City’s beam, visible to Japanese fire control officers, constituting a standing torpedo threat that kept Hosogaya from closing to the range where his gunnery advantage became decisive.

At close range, under 8,000 yd, a destroyer torpedo salvo was a risk Hosogaya could not dismiss. At the same close range, Salt Lake City’s 8-in shells were firing into Nachi’s belt armor at angles where the armor’s protection degraded. Hosogaya wanted decisive range. McMorris was making decisive range expensive.

It took 8 hours to get every name and number in this segment right. The after-action reports, the loss tables, the squadron rosters. If that work is worth something to you, hit the like button. It’s the only signal that tells us to keep doing it this way instead of the fast way. At 11:03 on the morning of March 26th, 1943, the flooding in Salt Lake City’s forward engine room reached the point where the propulsion system lost prime.

The screws stopped turning. USS Salt Lake City, eight hits taken and power gone, lay dead in the water in the Bering Sea with eight Japanese warships closing from the northwest and four American destroyers the only material between her and the Nachi’s guns. McMorris gave the order. The destroyers were to charge.

USS Bailey, USS Coghlan, USS Monaghan, and USS Dale turned toward the Japanese formation at full speed. They were not capable of defeating the Nachi or the Maya at gun range. A 5-in shell is not a threat to a heavy cruiser’s armored belt at any engagement range. What they carried that was a threat were torpedo tubes.

And a salvo from four destroyers with multiple torpedo tracks converging at angles the Japanese destroyers could not fully cover was a threat that required immediate response. A destroyer torpedo at close range is not a problem that resolves itself with superior gun power. The torpedoes are faster than any evasion available once they are in the water.

USS Bailey was hit twice by Japanese gunfire before her torpedoes were in the water. Five men were killed on her bridge and at her fire control station. Bailey fired five torpedoes from her tubes and turned away. Coghlan fired four. Monaghan and Dale pressed their approach before turning. None of the torpedoes reached the Japanese formation.

The charge was not designed to hit. It was designed to force a decision. While the destroyers charged, Salt Lake City’s engineering crew worked in conditions that no training exercise had replicated. The flooding in the forward engine room was pumped cleared and the propulsion system reprimed. At 11:54 in the morning, 51 minutes after she had lost propulsion, USS Salt Lake City’s screws turned again.

She was moving at 17 knots, making smoke, and turning southeast. At 12:03 in the afternoon, Vice Admiral Hosogaya ordered his fleet to break off the engagement and turn north. The reasons behind that order are not simple and are not fully recoverable from the documents available in English. The most immediately significant was material.

The Nachi had fired more than 700 8-in shells in the course of the battle and was approaching the end of her ready magazine. That figure was not available to McMorris. The Americans did not know it until Japanese records were translated after the war. A fleet that broke off because its flagship was nearly out of primary ammunition would have described the withdrawal differently to its commanding officer.

What Hosogaya reported was his concern about American air attack. The radio traffic his signals team had been monitoring for 3 hours looked to that team like air coordination. The same volume and frequency pattern that preceded a land-based strike. The sky was clear. No aircraft came, but the interpretation was plausible enough to a commander whose ships were not fresh, whose magazine state was not comfortable, and whose supply convoy had been running north for 4 hours on a heading that was taking it away from its destination at the same speed

it had been approaching at. There was also the calculation of the destroyers, four of them, two undamaged, were still in the water between Hosogaya and the Salt Lake City. Closing to the range where the Nachi’s depleted magazine was no longer the constraint, under 8,000 yd, where gunnery became decisive, meant threading his flagship through a torpedo salvo from boats that had already demonstrated they would charge directly toward a heavy cruiser.

The exchange rate on that geometry, with a near empty main battery magazine, was not one Hosogaya could calculate in his favor. He turned north. The convoy, 80 mi northeast and moving, continued north. The Sanko Maru and the Asaka Maru would not turn south again. McMorris did not pursue. Salt Lake City was making 17 kn on a repaired propulsion system, had taken eight hits, and was not in condition for a chase against a fleet that still outgunned every ship he had.

He turned east for a dog. The battle had lasted from 08:39 to 12:03, 3 hours and 23 minutes. It was at the time the longest daylight surface engagement in American naval history since the era of sail. What happens in the next 12 minutes is the part most documentaries skip entirely.

If you want the rest of this war told the same way, from the records, not the legends, subscribe before it starts. What Japan lost at the Komandorski Islands on March 26th, 1943, was not a ship. It was the arithmetic of the supply line. The Sanko Maru and the Asaka Maru reached Paramushiro without delivering their cargo. The supplies they carried, food, ammunition, and reinforcement personnel for the Attu garrison, were offloaded back at the base they had sailed from.

They were not reshipped. The window for resupplying Attu before the American amphibious landing that Japan’s intelligence had been anticipating was measured in weeks. And that window was consumed by the duration of the engagement. If Hosogaya had driven off McMorris in the first hour and continued the convoy south, which the Japanese fleet’s material advantage made entirely plausible, the supplies would have reached Attu in time to change what happened there.

Instead, the Attu garrison received its last effective supply delivery in December of 1942. From December through May of 1943, approximately 2,600 men lived on what they had and what their few surviving fishing vessels could provide from the surrounding waters. Their ammunition was rationed. Their rations were reduced.

Their medical supplies ran out months before the American landing. Vice Admiral Hosogaya was relieved of command on March 31st, 1943, 5 days after the battle. He was not given another sea command for the remainder of the war. The navy’s internal judgment was that a force of his strength should not have broken off against two American cruisers, regardless of magazine state or air threat assessment.

He was the only Japanese flag officer relieved of command after a battle in which his ships were undamaged and his force was materially superior to the one that had faced him. The judgment was never formally reversed. Japan attempted to maintain the Aleutian supply line after Komandorski by other means. Submarines were assigned supply cargo runs.

An I-class submarine could carry approximately 20 tons per transit, a fraction of what a transport convoy delivered, and the cost of taking fleet submarines off combat patrol. The submarine supply pipeline delivered perhaps 1/10 of what the convoy system had been providing. It was not sufficient to change the garrison’s situation.

On May 11th, 1943, American amphibious forces landed on Attu with approximately 11,000 troops. The Japanese garrison, numbering approximately 2,650, had prepared extensive defensive positions in the mountains and fought effectively despite being outnumbered more than four to one. The battle for Attu lasted 19 days. On May 29, Colonel Yamazaki Yasuyo led the surviving garrison, approximately 750 men still capable of movement, in a final charge that broke through the American perimeter and reached within a few hundred yards of the American artillery ammunition

supply before being stopped and surrounded. Approximately 29 Japanese prisoners were taken from the entire garrison. The remainder were dead. Kiska, the other Aleutian island Japan had occupied in June 1942, was secretly evacuated on July 28th, 1943. Japanese cruisers extracted 5,183 men from Kiska in dense fog in less than an hour, an operation of considerable skill that removed the garrison without American detection.

When American and Canadian troops landed on Kiska on August 15th, 1943. They found the island empty. The Aleutian campaign that Japan had begun with the seizure of American territory in June of 1942 ended with empty islands in August of 1943. The engagement that determined its trajectory was a 3-hour gun battle in the Bering Sea in which the losing side refused to stop.

American losses in the Battle of the Komandorski Islands were lighter than the engagement’s duration suggested they would be. USS Richmond was not hit during the battle. USS Salt Lake City sustained eight direct hits from Japanese 8-in shells, suffered progressive flooding throughout the action, and lost propulsion for 51 minutes.

Seven men were killed aboard her and approximately 20 were wounded. She remained afloat, returned to Adak under her own power, and was repaired and returned to service within weeks. USS Bailey took two direct hits during the destroyer charge, which killed five men and destroyed her forward fire control station. She fired her torpedoes and withdrew.

USS Coghlan, Monaghan, and Dale were not hit. Total American losses, 12 men killed, approximately 20 wounded, no ships sunk. Japanese casualties were approximately comparable in number. Nachi sustained approximately 14 killed and multiple wounded during the gun action. The other Japanese vessels in the formation suffered light casualties or none.

The Japanese fleet withdrew with its ships structurally undamaged, its supplies still aboard, and its convoy task unaccomplished. The symmetry of the losses in men, 12 Americans, approximately 14 Japanese, was nearly exact. The asymmetry of the strategic cost was not. The convoy that turned north carried supplies that the Attu garrison would need in May.

Salt Lake City’s dead were on a ship that kept fighting and produced an outcome they would not see. The garrison’s dead, 2,621 of approximately 2,650, died in the mountains of Attu in conditions that were, on most of the days of that battle, indistinguishable in their cold and fog from the conditions in the Bering Sea on March 26.

The distance between those two outcomes was measured in a convoy that ran north for 4 hours and did not turn south. The chronology of the Komandorski engagement follows the calendar of the Aleutian campaign. June 3rd-7th, 1942. Japanese forces occupy Kiska and Attu. The first seizure of American territory since 1812 proceeds without effective American resistance.

June through December 1942. Japan conducts more than nine supply runs to the Aleutian garrisons, completing most of them under fog cover without American interference. American air bases are built at Adak, Amchitka, and Umnak through the fall. Air interdiction disrupts, but does not halt the supply schedule.

February 19th, 1943. McMorris’s task group intercepts and sinks the transport Akagane Maru. First successful surface interdiction of the Aleutian supply route. Japan responds by adding heavy cruiser escort to the next convoy. March 26th, 1943. Battle of the Komandorski Islands. Task Group 16.6 intercepts Hosogaya’s Fifth Fleet south of the Commander Islands.

3 hours and 23 minutes of running gun battle end with Hosogaya’s withdrawal. The convoy turns north. No supplies reach Attu. March 31st, 1943. Vice Admiral Hosogaya relieved of command. No replacement supply convoy is organized. May 11th, 1943. American forces land on Attu with approximately 11,000 troops. May 29th-30th, 1943.

Colonel Yamasaki’s final charge. Approximately 29 Japanese survivors from 2,650. July 28th, 1943. Japan secretly evacuates Kiska’s garrison. 5,183 men in under an hour under fog cover. August 15th, 1943. American and Canadian forces land on Kiska. The island is empty. The campaign ended where Japan’s Aleutian plan had always been logistically fragile in the supply run.

The garrison at Attu died with the supplies that would have sustained it still aboard ships that had turned north in March. Four factors separated the Komandorski Islands engagement from the surface battles that surrounded it in 1943. The first was the convoy’s northward course. When the engagement began at 08:39, the Sanko Maru and the Asaka Maru turned north away from the fighting.

From that moment, the battle’s strategic outcome was no longer determined by which fleet prevailed at sea. It was determined by the duration of the engagement. Every additional minute McMorris remained in the fight was a minute that the transports were further from Attu and closer to a position from which turning south again would mean arriving at the unloading point in darkness.

In waters where American surface forces were known to be present, by noon, the convoy was 80 mi north of Attu’s approach corridor on a heading that could not be reversed before dark. McMorris did not need to win the gun battle. He needed to fight it for long enough. The second factor was the torpedo threat.

Four American destroyers positioned on Salt Lake City’s beam and visible to Japanese fire control officers constituted a standing torpedo threat that Hosogaya could not dismiss at any range below 8,000 yd. Closing to decisive gunnery range meant threading his flagship through a torpedo salvo that his destroyer screen might not fully intercept.

The destroyers were not a decisive striking force. They were a range constraint that kept the engagement at a distance where Japanese gunnery was imprecise enough for Salt Lake City to keep firing. When McMorris ordered the destroyer charge at 11:03, the charge did not need to hit the Japanese fleet. It needed to force Hosogaya’s attention and maneuver away from the ship regaining her propulsion 50 1 minute away.

The third factor was the Nachi’s ammunition state. Hosogaya had loaded his flagship for an escort mission against an American patrol force that he expected would disengage in the first hour. A patrol force that did not disengage, that absorbed Japanese shells and returned fire for 3 hours and 23 minutes, consumed the Nachi’s ready magazine at a rate that the pre-battle planning had not anticipated.

More than 700 8-in shells fired in a single engagement left the flagship in a position where continuing to close range, which was the only way to complete the mission, meant doing so with a near empty main battery. The ship was undamaged. The weapon the ship carried was nearly spent. Hosogaya calculated that the trade was not acceptable.

The decision to withdraw was made by an admiral whose fleet was intact and whose magazine was not. The fourth factor was the radio. McMorris’s continuous contact reporting from 0730 onward produced a radio signature that the Japanese signals team on the Nachi interpreted as air coordination traffic. It was not.

It was standard fleet procedure for a task group that had found something it needed Pacific Fleet Command to know about. The side effect of that standard procedure was 3 hours of a flag officer in the back of his mind watching for aircraft that never appeared. That background concern did not determine Hosogaya’s withdrawal decision alone.

It shaped the threshold at which the ammunition state and the torpedo threat and the convoy’s northward position became, collectively, enough. Strip away the radio traffic and the ammunition ledger and the 4 hours of the convoy running north and what remains is a single reversal. Vice Admiral Hosogaya assembled a force sufficient to destroy everything McMorris had in the sector and two cruisers that refused to stop fighting held him in an engagement long enough to make the mission those cruisers were supposed to be protecting

impossible to complete. Folded inside that reversal was a second one, quieter, that the battle reports did not record. The Nachi’s main battery was not empty when Hosogaya turned north. It was nearly empty. An approximation that a more aggressive officer might have judged differently. The men who relieved Hosogaya of command five days later judged it differently on his behalf.

Their judgment was that the supply mission was worth the risk of a near empty magazine at close range against destroyers that were willing to charge. Hosogaya disagreed. He was relieved for disagreeing. The garrison at Attu would have agreed with his relievers if there had been 2,621 more of them to ask. USS Salt Lake City was decommissioned in 1947 and assigned as a target vessel for Operation Crossroads, the nuclear test series at Bikini Atoll in July 1946.

She survived both Test Able and Test Baker, the only ship at Komandorski that survived not only a Japanese cruiser fleet but two atomic detonations in the central Pacific. She was finally sunk by naval gunfire off San Diego in May 1948, three years after the war ended. The water south of the Komandorski Islands is 3,000 m deep.

The Japanese records of the Nachi’s ammunition state on the morning of March 26, 1943 were not translated into English until 1953, a decade after the battle and seven years too late to matter to anyone still waiting. If you want to understand how America actually won its war, not the version in the highlight reels, subscribe.

There’s a new one every few days.