The Naval Battle of Guadalcanal, 12–15 November 1942, and Tassafaronga

A Japanese twin-engine torpedo bomber burns over Ironbottom Sound during the attack on Rear Admiral Richmond Kelly Turner’s transports off Lunga Point on the afternoon of 12 November 1942, the opening blow of four days of fighting for the waters off Guadalcanal. U.S. Navy photograph 80-G-32365 (National Archives); public domain, via Wikimedia Commons.

Guadalcanal · Chapter 7 of 9

The Naval Battle of Guadalcanal, 12–15 November 1942, and Tassafaronga

Two night battles at point-blank range, a day of air strikes on a troop convoy, and a battleship duel decided whether Japan could retake Henderson Field by sea.

12–15 November and 30 November–1 December 1942Ironbottom Sound, Savo Island, Tassafaronga~32 min read

By the second week of November 1942 the Guadalcanal campaign had become a contest of shipping. The 1st Marine Division and the Army regiments that had begun to join it held a perimeter around Henderson Field on the north coast of the island. The Japanese 17th Army held the jungle west of the Matanikau River and the coast beyond it toward Cape Esperance. Each side had learned by October that the force which could land men, food and ammunition faster, and stop the other from doing the same, would win the island. The failure of the Japanese ground offensive in late October, told in Chapter 6, did not end that contest. It moved it back to sea.

Admiral Isoroku Yamamoto’s Combined Fleet now planned its largest reinforcement yet: eleven transports carrying about 7,000 troops, most of them from the Army’s 38th Division, with the heavy equipment and food the earlier destroyer runs could not carry. To get them ashore the Japanese would again try to silence Henderson Field with battleship gunfire, as two battleships had done on the night of 13–14 October. The United States Navy, still without an undamaged carrier in the South Pacific after the Battle of the Santa Cruz Islands, was bringing in its own reinforcements and had little to stop a battleship force except cruisers, destroyers and the aircraft at Henderson Field.

What followed between 12 and 15 November is known to the U.S. Navy as the Naval Battle of Guadalcanal and to many sailors who fought in it simply as Friday the 13th. It was really a series of actions: a Japanese air attack on 12 November; a night melee in the small hours of Friday, 13 November, in which two American admirals were killed; the loss of the antiaircraft cruiser Juneau with the five Sullivan brothers; two days of air strikes that sank a Japanese battleship, a heavy cruiser and most of a troop convoy; and a second night action on 14–15 November in which an American battleship sank a Japanese one. Two weeks later, at Tassafaronga, a Japanese destroyer squadron inflicted one of the worst tactical defeats in U.S. Navy history and still failed in its mission. This chapter follows those events in order, using the Office of Naval Intelligence (ONI) Combat Narratives prepared in 1943–44, the Naval History and Heritage Command (NHHC) ship histories and H-Grams, and the later analyses that corrected wartime claims.

The Japanese plan: battleships for the airfield, transports for the island

The Japanese had learned the same lesson the Americans had: aircraft at Henderson Field made daylight movement near Guadalcanal deadly for slow ships. Destroyers of the “Tokyo Express” could run in at night, unload a few hundred men and their rifles, and be back up the Slot before dawn, but they could not carry artillery, vehicles or the tonnage of rice and ammunition that an army of 20,000 men needed. A convoy of real transports could, but only if Henderson Field could be put out of action for a day or two while the ships unloaded. The bombardment of 13–14 October by the battleships Kongō and Haruna had shown that it could be done, at least for a time.

The November plan was larger. NHHC Director Samuel J. Cox’s summary in H-Gram 012 describes a main body under Vice Admiral Nobutake Kondō of four battleships, three heavy cruisers, three light cruisers and 21 destroyers, supported by the carrier Jun’yō, about 125 operational land-based aircraft and about 25 floatplanes. Within it, Rear Admiral Hiroaki Abe’s bombardment group, built around Battleship Division 11 (Hiei and Kirishima), the light cruiser Nagara and eleven destroyers, was to shell Henderson Field on the night of 12–13 November. Vice Admiral Gunichi Mikawa’s Eighth Fleet cruisers, the force that had won at Savo Island in August, would follow up with a second bombardment the next night. Behind them came Rear Admiral Raizō Tanaka’s Destroyer Squadron 2 with twelve destroyers escorting eleven transports. Cox gives about 7,000 troops in the convoy; other accounts give higher figures, and a 2006 article by Mike Dunne for the Naval Historical Society of Australia speaks of 11,000 men, apparently counting the whole reinforcement effort.

The timing rested on Yamamoto’s reading that the Americans had no operational carrier in the area. That was nearly true. Hornet had been sunk at Santa Cruz on 26 October, and Enterprise, damaged in the same battle, was being repaired at Nouméa. The American intelligence picture was nonetheless good. By 12 November, the ONI narrative records, the Americans estimated that two carriers, four battleships, five heavy cruisers, about 30 destroyers and a convoy of transports were gathering in the Rabaul–Buin area, and they expected air attacks on Guadalcanal on 11 November followed by a bombardment and possibly landings.

Turner’s convoys and the air attack of 12 November

Vice Admiral William F. Halsey Jr., who had replaced Vice Admiral Robert Ghormley as commander of the South Pacific Area on 18 October, was running reinforcements of his own. Two convoys under the overall command of Rear Admiral Richmond Kelly Turner carried about 5,500 Army and Marine troops, according to H-Gram 012 (the ONI narrative says about 6,000), along with ammunition, aviation fuel and food. Rear Admiral Norman Scott, the victor of Cape Esperance in October, escorted three transports with the antiaircraft cruiser Atlanta and three destroyers; they arrived off Lunga Point on 11 November and were attacked by dive bombers that morning. Turner’s own group, four transports escorted by the heavy cruisers San Francisco and Portland, the light cruisers Helena and Juneau and ten destroyers, anchored off Kukum at 0530 on 12 November.

The Japanese air attack came early in the afternoon. The ONI narrative records torpedo planes attacking at 1405, with the action over by 1419; it counted 20 to 25 Mitsubishi Type 1 bombers (the G4M, Allied code name “Betty”) and eight Zero fighters. H-Gram 012 gives sixteen G4Ms escorted by thirty Zeros and states that only two of the bombers returned to Rabaul. Marine and Navy fighters from Henderson Field met the bombers on the way in, and the transports and warships, which had got under way to maneuver, put up heavy antiaircraft fire as the bombers came in low over the water. No torpedo struck a transport. The American loss in aircraft, according to ONI, was three F4F-4 Wildcats and one Army P-39.

One of the stricken bombers crashed into San Francisco. The cruiser’s Navy ship history (DANFS) records that at 1416 a damaged torpedo plane released its torpedo off the starboard quarter and then struck the after control station, “Battle II,” before plunging into the sea, destroying the station and three 20 mm mounts. DANFS counts fifteen killed, 29 wounded and one missing; H-Gram 012 gives 24 killed and 45 wounded, among the wounded the ship’s executive officer, Commander Mark Crouter. The destroyer Buchanan was hit by American antiaircraft fire in the same action. The transports went back to their anchorages at 1525 and continued unloading until they sailed that evening.

By then American search planes and coastwatchers had reported the Japanese surface forces coming south. Turner knew a bombardment group with battleships would be in Ironbottom Sound that night. He decided to withdraw the transports through Lengo Channel toward Espiritu Santo and to send every warship he could spare back into the Sound to meet it. The ONI narrative records that Task Force TARE left Savo Sound at 1815.

Callaghan’s column

Turner formed the combat group, designated Task Group 67.4, from the cruisers and destroyers of both convoys and put it under Rear Admiral Daniel J. Callaghan. H-Gram 012 notes that Callaghan was fifteen days senior to Scott, and that Turner’s choice of Callaghan, who had never commanded in a surface action, over Scott, who had won one a month earlier, remains debated. Scott went along as second in command in Atlanta. Callaghan flew his flag in San Francisco, the largest ship present and his own former command. Her search radar was the older SC set, as was Atlanta’s; the newer SG surface-search radar, which gave a far clearer picture of surface contacts, was fitted in Portland, Helena, Juneau, Fletcher and O’Bannon.

Callaghan arranged his thirteen ships in a single column: the destroyers Cushing, Laffey, Sterett and O’Bannon in the van; then Atlanta, San Francisco, Portland, Helena and Juneau; and in the rear the destroyers Aaron Ward, Barton, Monssen and Fletcher. The formation resembled the one Scott had used at Cape Esperance, and it had the same weaknesses: a long line was hard to control by voice radio at night, the destroyers at the ends could not easily use their torpedoes without fouling the cruisers, and the ships with the best radar were not the flagship. Cox writes that no battle plan by Callaghan survived, if one was written. Some historians have argued that the column was chosen for simplicity, because the ships had never worked together and there was no time to train; others regard it as a misuse of the destroyers and of the SG radar. The record does not show Callaghan’s reasoning.

Abe’s group came down from the north in a formation that, according to postwar Japanese accounts used by Richard Frank and James Hornfischer, had been thrown into disorder by rain squalls on the approach. His battleships carried high-explosive bombardment shells ready for Henderson Field rather than armor-piercing rounds for ship targets. Neither admiral had a clear picture of the other. The Americans had radar but no plan for using it; the Japanese had no radar but excellent night optics, trained lookouts and the Type 93 torpedo.

Friday the 13th: the first night action

At 0124 on 13 November Helena’s SG radar picked up the Japanese at 27,100 yards, about 13.5 miles, northwest of the American column off Lunga Point. Over the next quarter hour the contacts were reported by voice radio (the Talk Between Ships, or TBS, circuit), which was crowded with reports and requests; Callaghan, without an SG plot of his own, had to build his picture from what he was told. He turned the column toward the contacts. At about 0141–0142 Cushing, leading, sighted Japanese destroyers crossing ahead at close range and turned away to avoid collision, and the ships behind her had to turn too. The column began to bunch and break up.

At 0145 Callaghan ordered “stand by to open fire.” The Japanese struck first. At about 0148 Hiei and the destroyer Akatsuki switched on searchlights and fixed Atlanta at a range the Pacific Wrecks record gives as 1,600 yards. Callaghan then issued the order that became famous because it showed how little either side could see: odd-numbered ships of the column were to fire to starboard and even-numbered ships to port. For the next forty minutes the fight was a melee among two dozen ships at ranges that sometimes fell below a thousand yards, in which ships fired at whatever they could see, often their own side.

Atlanta, in the searchlights, fired back and was hit almost at once by gunfire and a torpedo, probably from the destroyer Inazuma or Ikazuchi, which knocked out her power. In the confusion, San Francisco, firing to port at a ship she could only partly see, put several 8-inch salvos into Atlanta’s superstructure. Rear Admiral Scott and three of his four staff officers were killed on Atlanta’s bridge. Both H-Gram 012 and the Pacific Wrecks record attribute the fatal hits to San Francisco’s fire.

The American destroyers in the van fought at extraordinary ranges. Laffey, according to Cox, passed under Hiei’s bow at about twenty yards, too close for the battleship’s main guns to depress, and raked her upperworks with 5-inch and machine-gun fire; DANFS records that she fought “at point blank range” with her remaining main battery guns, fired torpedoes, and machine-gunned the battleship’s bridge. According to later accounts, her torpedoes were fired too close to arm. Japanese shells and a torpedo then wrecked Laffey, and ONI records her blowing up at 0202–0204 as her crew abandoned ship. Cushing was hit repeatedly and left burning; Barton, ninth in the American order, was struck by two torpedoes at 0155 and sank within moments, taking most of her crew. Monssen, near the rear, was illuminated and wrecked by Japanese gunfire.

The cruisers suffered as heavily. San Francisco exchanged fire with Hiei and was then hit by heavy shells on her bridge. The ONI narrative states that Callaghan was killed when Hiei’s salvo smashed the bridge; some later accounts credit Kirishima or a cruiser, and the record cannot settle the question. Captain Cassin Young, San Francisco’s commanding officer and a Medal of Honor recipient for his conduct at Pearl Harbor in command of the repair ship Vestal, was killed with him. DANFS states that the hit on the bridge killed or badly wounded every officer there except the communications officer. Portland was hit by a torpedo, which H-Gram 012 attributes to the destroyer Yūdachi, at about 0158 and began to steam in circles with her stern damaged, though she kept firing. Juneau took a torpedo on her port side near the forward fireroom; the cruiser’s DANFS history credits the destroyer Murasame. The shock buckled her deck and knocked out her fire-control equipment and power, and about seventeen men were killed instantly according to the report of her senior surviving officer, Lieutenant Roger O’Neill of the Medical Corps.

The Japanese were hurt too. Hiei, the largest and most visible target, was hit again and again by cruiser and destroyer gunfire at ranges where even 5-inch shells could wreck a battleship’s upperworks; Cox gives more than 85 hits. Her steering was damaged and her bridge struck, and Abe was wounded. Akatsuki, the destroyer whose searchlight had illuminated Atlanta, was sunk by gunfire with few survivors. Yūdachi was disabled, and Amatsukaze was heavily damaged but escaped. At about 0200, with Hiei damaged and his formation scattered, Abe called off the bombardment and ordered a withdrawal, even though Kirishima was almost undamaged. Cox notes that Abe did not know that only Helena and Fletcher among the Americans were still fully fit to fight.

By about 0230 the firing had stopped. Captain Gilbert C. Hoover of Helena, the senior officer still in action, assumed command and gathered the survivors able to steam. In San Francisco, Lieutenant Commander Bruce McCandless, the communications officer, took the conn and, with Lieutenant Commander Herbert E. Schonland directing damage control below as the senior surviving officer, kept the ship in formation and fighting. All three, with Boatswain’s Mate First Class Reinhardt J. Keppler, who fought fires and removed wounded until he died of his own wounds, would receive the Medal of Honor.

The order of events in the scrubber follows the ONI narrative and H-Gram 012; the precise minute of several events, especially the hit on San Francisco’s bridge, is uncertain, and accounts written from different ships do not agree. The ONI narrative, written in 1944 from American reports, also claimed that a Japanese light cruiser “of the Natori class” blew up at 0149 and that several Japanese cruisers and destroyers were sunk. Postwar Japanese records show that no Japanese cruiser was sunk that night; the explosions seen were probably American and Japanese destroyers.

The wreck sonar above plots the ships lost in the first night action on the floor of Ironbottom Sound, where most of them still lie. On the collection hub the same display can be switched to the second night action (“nbg2”) and to Tassafaronga (“tassa”), which together account for many of the wrecks between Savo and Cape Esperance. The diving and survey history of those wrecks is covered in Chapter 9.

The cost of the first night

Daylight on 13 November showed Ironbottom Sound littered with burning and drifting ships. Atlanta lay disabled off Lunga Point; her crew fought all day to save her, but with no power and flooding spreading she was scuttled on the orders of her captain, Captain Samuel P. Jenkins, and sank at 2015, about three miles west of Lunga Point. Her wreck lies on its port side at about 430 feet. Cushing and Monssen burned out and sank during the day. Portland, still circling, sighted the derelict Yūdachi, whose crew had been taken off, and sank her with 8-inch gunfire. Aaron Ward, damaged and dead in the water, was towed to Tulagi. Boats from Guadalcanal and Tulagi picked up hundreds of American sailors from the water.

Cox’s table in H-012-1 counts, ship by ship, 1,439 American dead from the first night action and the loss of Juneau that followed it, besides the 24 killed in San Francisco by the crashing bomber on the 12th. DANFS gives slightly different figures for San Francisco (77 killed in the night action, 105 wounded and seven missing). The cruiser had taken 45 hits, none below the waterline. The chart below uses Cox’s figures.

Bruce McCandless’s citation records that, wounded and knocked unconscious when enemy fire killed or disabled the admiral, the captain and most of the bridge personnel, he “promptly assumed command of the ship and ordered her course and gunfire against an overwhelmingly powerful force.” Schonland’s citation honors the damage-control effort that kept San Francisco afloat with fires raging and compartments flooded; Keppler’s honors a boatswain’s mate who, mortally wounded, kept fighting fires and carrying wounded men. San Francisco reached Espiritu Santo on 14 November and San Francisco itself on 11 December. Her shell-riddled bridge wings were later set up as a memorial on the headland at Lands End above the Golden Gate.

Juneau and the Sullivan brothers

USS Juneau (CL-52) was a new ship. Commissioned at the New York Navy Yard on 14 February 1942 under Captain Lyman K. Swenson, she was one of the Atlanta-class antiaircraft cruisers, built around sixteen 5-inch guns. At Santa Cruz on 26 October she had screened Hornet, firing 665 rounds of 5-inch ammunition that day according to DANFS. Among her crew were five brothers from Waterloo, Iowa.

After the night action Juneau, down by the bow and able to make about 18 knots, joined Hoover’s retiring group of Helena, San Francisco, Sterett and Fletcher, steaming southeast toward Espiritu Santo. Lieutenant O’Neill and three corpsmen were sent across to help with San Francisco’s many wounded. At about 1101 the Japanese submarine I-26 fired a spread of torpedoes at San Francisco. They missed her, but one struck Juneau on the port side, close to the earlier hit. Her magazine exploded. Witnesses in San Francisco, quoted in O’Neill’s report, saw her break in two and both halves disappear in about twenty seconds; the DANFS history of The Sullivans says about forty-two seconds. Most of her crew died in the explosion.

Hoover had to decide whether to stop. He had two crippled ships, San Francisco and Sterett, one undamaged cruiser and one destroyer fit to screen against submarines; a search would take at least two hours. To observers on the ships, nobody could have lived through the explosion. Rear Admiral J. C. Wylie, who in 1942 was executive officer of Fletcher, wrote in Naval History in 1992 that he and his captain agreed at the time that there could be no survivors, and that Hoover made the right decision. Hoover had Helena pass word of the sinking by flashing light to a passing Army B-17, with the position, about 10 degrees South, 161 degrees East. The bomber acknowledged, but the message did not reach Halsey’s headquarters at Nouméa in time to start a search. Halsey relieved Hoover of command soon afterward, a decision that was criticized then and has been argued over since; Wylie defended Hoover’s judgment.

There were survivors. NHHC’s account of the loss gives about 115 men alive in the water after the explosion; the ONI narrative says about 120, about two-thirds of them wounded; the history of the destroyer The Sullivans says roughly 140. They clung to rafts and wreckage in the open sea with little water and no food, and over the following days exposure, wounds, exhaustion and sharks killed most of them. ONI records that a search plane sighted about sixty survivors at 1100 on 14 November, but a destroyer search on the 16th found nothing. Three men reached an island on the night of the 18th and were picked up by a Navy PBY flying boat on the 21st; rafts with ten men were sighted on the 19th and seven of them were rescued. In all, ten men were rescued, eight days after the sinking. The DANFS history of Juneau counts ten survivors of the 693 aboard when she was hit, plus O’Neill and the three corpsmen who had been in San Francisco. Cox gives 683 dead; other counts range up to 687.

The deaths of five brothers in one ship were national news, and the Navy afterward adopted a policy, in NHHC’s words, “discouraging family members from serving together in the same ship.” It was a policy, not a law, and it did not end the practice altogether. The story of the brothers and of Juneau’s other crew, and the 2018 discovery of the wreck by the research vessel Petrel at about 4,200 meters (13,800 feet), is told in detail on the special-units page USS Juneau and the Sullivan Brothers.

Hiei’s last day

At dawn on 13 November Hiei was north of Savo Island, her steering damaged and her turrets partly out of action, circling slowly with a few destroyers standing by. She was within easy range of Henderson Field. The airfield, which Abe had been sent to wreck, had not been touched. For the rest of the day Marine, Navy and Army aircraft struck her in relays. The ONI narrative, the NHHC chronology and Cox together describe the sequence: Marine torpedo planes and dive bombers from Henderson Field attacked in the morning; at about 1100 torpedo bombers from Enterprise’s Torpedo Squadron 10, flying to Guadalcanal from the carrier as it came north from Nouméa, put torpedoes into her; B-17s from Espiritu Santo dropped 68 500-pound bombs and claimed one hit and possibly more; and at about 1430 another strike of Marine and Navy planes from Henderson Field scored, by the chronology’s count, three torpedo and three bomb hits. Among the fliers was Lieutenant Harold “Swede” Larsen’s Torpedo Squadron 8 detachment, which H-Gram 012 mentions.

By evening Hiei was still afloat about five miles northwest of Savo. Her captain, Captain Masao Nishida, had ordered abandon ship, according to Cox after an incorrect report that her engines were damaged, and the escorting destroyers took off her crew. Accounts differ on what happened next. Cox states that Yamamoto ordered that she not be scuttled; other accounts have the destroyers torpedo her before leaving. She sank during the night of 13–14 November, the first Japanese battleship lost in the war. Cox puts her dead at about 300; other accounts give lower figures. Abe, wounded on her bridge, was removed from command after the battle, and Kondō took personal charge of the next attempt.

The loss of Hiei showed what the Cactus Air Force had become since August (its earlier story is in Chapter 3). A crippled battleship within 30 miles of an operating airfield in daylight could not survive. That lesson would decide the fate of Tanaka’s transports the next day.

Mikawa’s bombardment and the morning of the 14th

Abe’s failure did not stop the Japanese plan; it delayed it one day. Tanaka’s convoy, which had turned back on the 13th, was ordered south again. Mikawa’s cruisers came down the Slot on the night of 13–14 November, and this time there was no American surface force in Ironbottom Sound to meet them. The battered ships of TG 67.4 had withdrawn, and Halsey’s battleships, coming north with Enterprise, were still too far away.

Mikawa’s force included the heavy cruisers Chōkai, Kinugasa, Maya and Suzuya, the light cruisers Isuzu and Tenryū, and destroyers. While Chōkai and Kinugasa covered, Suzuya and Maya shelled Henderson Field. Cox, in the H-Gram 012 attachment on the battleship action, states that they fired almost 1,000 rounds of 8-inch shell and destroyed two Wildcats and two SBD dive bombers and damaged fifteen other Wildcats. The ONI narrative gives the bombardment as lasting from about 0220 to 0340, with one SBD-3 and two F4F-4s destroyed and seventeen Wildcats damaged, and states that “the field was not seriously injured.”

The two sources disagree on what ended it. ONI, writing from the reports of the motor torpedo boats at Tulagi, says that six PT boats fired seventeen torpedoes and cut the bombardment short at 0340, claiming one or more hits. Cox writes that two PT boats tried to interfere and failed. Postwar Japanese records do not show torpedo damage to Mikawa’s ships that night, so the PT boats’ claims were probably mistaken, but their attack was the only American naval opposition in the Sound on the night of 13–14 November.

Mikawa’s ships headed northwest at dawn. At about 0715 Henderson Field launched six torpedo bombers, seven scout bombers and seven fighters after them; at about 0800 they set the heavy cruiser Kinugasa and light cruiser Isuzu afire, according to the NHHC chronology. Enterprise’s dawn search found the retreating cruisers next. ONI records that Lieutenant Gibson and Ensign Buchanan of Enterprise’s scouting and bombing squadrons dove on what they took to be a Nachi-class heavy cruiser at about 1015 and hit her, and that the ship “burst into flames.” The ship was Kinugasa. She flooded and listed, and the Enterprise attack group arrived at about 1115. Cox records that Kinugasa capsized and sank at 1122 with 511 of her crew. Chōkai, Maya and Isuzu were damaged; Maya lost 37 men when a crashing SBD struck her. The wartime ONI narrative reported the target as “Nachi” and claimed it sunk; postwar records identify the ship sunk as Kinugasa, a veteran of Savo Island.

The destruction of Tanaka’s transports, 14 November

At about 0830 on 14 November a B-17 found Tanaka’s convoy steaming down the Slot toward Guadalcanal: eleven transports in several columns, escorted by eleven or twelve destroyers, with intermittent fighter cover overhead. American pilots reported between nine and twelve transports. Captured documents, according to General Vandegrift as the ONI narrative records, suggested 30,000 to 35,000 men, a large exaggeration; other intelligence estimated about 4,600. What followed was the most destructive day the Cactus Air Force ever had.

From late morning until dusk, Marine, Navy and Army aircraft from Henderson Field and Enterprise attacked the convoy in a series of strikes. The ONI narrative lists them: seven Torpedo Squadron 10 TBFs, eighteen Marine scout bombers and twelve Marine fighters from 1220 to 1250, claiming three torpedo and five bomb hits; seventeen scout bombers, eight Marine torpedo bombers and eight fighters at about 1345; two flights of B-17s from Espiritu Santo beginning at 1500; a strike from Enterprise of eight scout bombers and twelve fighters, which at 1630 claimed six 1,000-pound bomb hits on five transports; and more Henderson Field strikes until after dark. Pilots landed at Henderson, rearmed and refueled, and flew again. Tanaka’s destroyers had no answer except antiaircraft fire and smoke, and the transports could only maneuver and steam on.

By nightfall, according to Cox, six transports had been sunk and a seventh, badly damaged, had turned back to the Shortlands with two destroyers. Tanaka’s destroyers pulled men from the water and from burning ships under air attack: Cox gives 1,562 rescued by the destroyers that went back with the damaged transport and 3,240 by four others overnight, and about 450 soldiers killed. Those thousands of men went back up the Slot; none of them reached Guadalcanal. Tanaka himself pressed on with the four surviving transports and his remaining destroyers. The NHHC chronology’s count at 1900, compiled from American reports, was four transports sunk, four dead in the water and burning, and four damaged and retreating, which shows how hard it was to keep track of the convoy from the air. Postwar accounts agree that seven of the eleven were lost at sea on the 14th.

The convoy’s ordeal was also the human cost of the Japanese navy’s decision to run transports into waters it did not control by day. The soldiers aboard, many of them from the 38th Division, which had fought at Hong Kong and in the Netherlands East Indies, had no means of fighting back. Those who reached Guadalcanal joined an army already short of food; the Japanese soldiers’ name for the island, Gatō (餓島), “Starvation Island,” dates from these weeks. Their experience ashore is part of Chapter 8.

Lee and the battleships: the night of 14–15 November

Kondō, now in direct command, came south on 14 November with the surviving battleship of Abe’s division, Kirishima, the heavy cruisers Atago (his flagship) and Takao, the light cruisers Nagara and Sendai, and destroyers. Cox divides the fourteen ships into three groups: a bombardment group of Kirishima, Atago and Takao; a screening group of Nagara and six destroyers; and a sweeping group of Sendai and three destroyers, which was to clear the passage east of Savo. Their task was the one Abe had failed at: to shell Henderson Field during the night so that Tanaka’s four remaining transports could land their troops at dawn.

Halsey had taken a calculated risk. On the 13th he detached the new fast battleships Washington and South Dakota, with four destroyers, from the Enterprise group as Task Force 64 under Rear Admiral Willis A. Lee Jr. The destroyers, Walke, Benham, Preston and Gwin, came from different divisions and had never operated together. The force spent the 14th about fifty miles south-southwest of Guadalcanal, trying to stay out of sight of Japanese reconnaissance planes, and at 2249 by ONI’s clock passed between Savo and Florida Islands into Ironbottom Sound. Lee put the destroyers in the van, followed by Washington and South Dakota. Fast battleships were rarely risked in night fighting in waters so confined, with islands close aboard and torpedo-armed destroyers about.

Washington’s SG radar made contact east of Savo at about 2252, about nine miles off: Kondō’s sweeping group under Sendai. At 2316 (ONI: 0016) Lee ordered fire opened, and Washington fired 42 rounds of 16-inch shell at a range of about 18,500 yards in roughly three minutes. South Dakota joined. Sendai made smoke and turned away unharmed. Meanwhile the Japanese screening group under Nagara, coming around the southwest side of Savo, closed on the American destroyers in the van, which were silhouetted against the battleships’ gun flashes and illuminated by Japanese searchlights and flares.

The destroyer fight lasted about ten minutes. Walke opened fire at about 2322 on the destroyer Ayanami. Within minutes Japanese shells and torpedoes had wrecked all four American destroyers. Preston, hit by 6-inch shells from Nagara and then by heavy shells, sank at about 2332 (ONI: 0032) with Commander Max C. Stormes and 117 of her crew. Walke was torpedoed and hit by gunfire and sank stern first at about 2342; Commander Thomas E. Fraser and 80 of her crew were lost. Benham had her bow blown off by a torpedo and could barely steam. Gwin was hit in an engine room and lost six men. Cox’s figures for the four destroyers are 80, 117, none and six killed. Lee ordered the two damaged survivors to retire.

Then South Dakota failed. At about 2330 (ONI: 0033) a short circuit in a secondary-battery director feed, made worse by a circuit breaker that had been locked closed, cut all power aft. The ONI narrative says the turrets were without power for about three minutes and the trouble was isolated in a little over two, but her SG radar went off for several minutes more and her radio and fire control were disrupted for longer. Steaming without radar past the burning destroyers, she was silhouetted against the fires and lost touch with Washington. Lee, in Washington, no longer knew exactly where she was, which made him reluctant to fire at any radar contact that might be his own ship.

At about midnight (ONI: 0100) Atago’s searchlights found South Dakota at short range, and Kirishima, Atago and Takao opened on her. In the next minutes she was hit, by the count of the 1947 Bureau of Ships damage report cited by Robert Lundgren and Nathan Okun, 26 times: one 14-inch, eighteen 8-inch, six 6-inch and one 5-inch shell. Cox gives 27 hits from five Japanese ships. Most struck her superstructure; the radar plot was wrecked by a 6-inch shell, fires broke out topside, and her radio and fire control were knocked out. She lost 39 men killed (one of them a Marine) and 59 wounded, by Cox’s count; DANFS gives 38 and 60. The armor kept the shells out of her machinery and magazines, but she was no longer fit to fight. Her captain, Thomas L. Gatch, turned her away. In his report Lee described her condition as “deaf, dumb, blind, and impotent.”

The Japanese searchlights also gave Washington what she needed. With South Dakota now clearly identified, Captain Glenn B. Davis opened fire on Kirishima at 8,400 yards. Over the next seven minutes Washington fired 75 rounds of 16-inch and 107 of 5-inch ammunition. The ONI narrative recorded about eight 16-inch hits, the figure in Lee’s report. Lundgren’s hit-by-hit analysis of Japanese damage records, published by NavWeaps, identifies twenty major-caliber hits, seven of them below the waterline where American observers would have counted them as misses, and about seventeen 5-inch hits; NHHC gives at least nine and possibly up to twenty 16-inch hits. The shells wrecked Kirishima’s forward turrets, started fires and flooding, and jammed her rudder so that she circled out of control.

“Radar has forced the Captain or OTC to base a greater part of his actions on what he is told rather than what he can see.”

— Captain Glenn B. Davis, commanding USS Washington, in his report on the action, as quoted in NHHC H-Gram 012

At about 0025, Cox writes, Kondō ordered his ships to converge on the American battleships; finding only Washington, which turned northwest and drew the Japanese away from South Dakota, he ordered a general withdrawal. Japanese destroyers fired torpedoes at Washington as they retired; the ONI narrative records seventeen torpedo sightings between 0148 and 0219 by its clock, four or five of them close, and none hit. Lee turned south to avoid a torpedo trap and retired, and Washington found South Dakota at the morning rendezvous. Washington had not been hit at all.

Kirishima did not survive the night. Lundgren cites the Japanese report that she went down at 0125 Tokyo time, 0325 local, about eleven miles west of Savo Island, after capsizing to port; whether she was deliberately scuttled is disputed, and Lundgren writes that “there is no real evidence of any scuttling order.” The destroyers Asagumo and Teruzuki rescued 1,100 of her crew, by the count Lundgren cites. Ayanami, set afire by the American destroyers and by Washington’s 5-inch guns, was abandoned and scuttled. Benham, her bow gone and her hull cracking, was abandoned on the afternoon of the 15th and sunk by Gwin. Cox gives the American loss for the night as 242 killed and 142 wounded; every commanding officer of the six American ships received the Navy Cross, two of them posthumously.

On the beach at Tassafaronga, 15 November

Tanaka had held his four surviving transports northwest of Savo while the battleships fought. With the Americans gone, the decision was taken not to unload in the normal way but to run the transports aground at Tassafaronga, west of the American perimeter, so that troops and some cargo could get ashore before daylight brought the aircraft back. At about 0400 on 15 November Hirokawa Maru, Kinugawa Maru, Yamaura Maru and Yamazuki Maru grounded on the beaches between Tassafaronga and Doma Cove. The troops went ashore.

From shortly before 0600 Marine and Army artillery on the perimeter and aircraft from Henderson Field attacked the beached ships, and later that morning the destroyer Meade came over from Tulagi and fired several hundred 5-inch rounds into the transports and the supply dumps along the shore. By midday the four ships were burning wrecks. Most of the cargo still aboard, and much of what had been landed, was destroyed. The hulks lay on the beach for the rest of the war and after it; the remains of Kinugawa Maru were still visible decades later.

Cox gives the result in a sentence: only about 2,000 Japanese troops got ashore, with very little ammunition and food. The Australian historian Mike Dunne gives the same figure of 2,000. The 38th Division’s heavy equipment, its artillery ammunition and the rice that might have fed the 17th Army through December were on the bottom of the Slot or burning on the beach.

What the November battles decided

Counted in ships, the four days were close to even. Across both nights the Americans lost two light cruisers and seven destroyers (four on the first night, three on the second), with a battleship, two heavy cruisers, a light cruiser and several destroyers damaged; NHHC’s historical summary gives 35 American aircraft lost. The same summary lists Japanese losses as two battleships, one heavy cruiser, three destroyers and eleven transports, the last including the four beached on the 15th, and 64 aircraft. Cox’s totals of killed, 1,732 Americans against about 1,900 Japanese personnel, do not include the soldiers lost in the transports; other estimates of Japanese dead are higher.

Measured by what each side was trying to do, the result was not even. The Japanese plan required Henderson Field to be put out of action and a division landed with its supplies. Neither happened. The first night action, costly and confused as it was, turned Abe back before he could fire a shell at the airfield, and his flagship was then sunk by the planes he had been sent to destroy. Mikawa’s bombardment did little damage, and the airfield’s aircraft sank Kinugasa and most of the convoy the next day. Lee’s battleships turned back Kondō’s second bombardment attempt and sank Kirishima. The convoy’s destruction meant that the 17th Army entered December weaker relative to the Americans than it had been in October, and hungrier.

The American side drew the opposite conclusion. NHHC’s summary notes that after November the United States could resupply Guadalcanal more or less at will, while the Japanese navy stopped committing battleships, cruisers and carriers to Guadalcanal waters and limited itself to destroyer and submarine supply runs. President Roosevelt said that “it would seem that the turning point in this war has at last been reached.” That judgment proved premature for the war as a whole but was close to right for the campaign. The Army official historian, John Miller Jr., titled his chapter on these weeks “Decision at Sea.”

Tassafaronga: the night of the Long Lances, 30 November–1 December

With transports ruled out, the Japanese turned to the destroyers of Tanaka’s squadron and a new method of supply. Instead of landing cargo by boat, which kept the ships stopped off the beach for hours, the destroyers would carry food and medicine sealed in steel drums, roped together in long strings. Off the coast the drums would be pushed overboard, and men on shore would haul the lines in with boats or by swimming. To make room, the destroyers taking part left ashore their torpedo reloads and part of their ammunition. The first run of this kind was set for the night of 30 November.

Halsey had formed a new cruiser force, Task Force 67, at Espiritu Santo to stop exactly this. Its first commander, Rear Admiral Thomas C. Kinkaid, wrote a plan that drew on the lessons of August and November: all ships with SG radar, destroyers to scout ahead and attack with torpedoes first on radar contact, and cruisers to open fire only after the torpedoes had had time to hit. Kinkaid was transferred on 28 November, and Rear Admiral Carleton H. Wright took command two days before the battle. He kept Kinkaid’s plan but brought the destroyers closer to the cruisers. His force, according to Cox in H-013-1, comprised the destroyers Fletcher, Perkins, Maury and Drayton in the van; the heavy cruisers Minneapolis (flagship), New Orleans and Pensacola; the light cruiser Honolulu under Rear Admiral Mahlon Tisdale; the heavy cruiser Northampton; and the destroyers Lamson and Lardner in the rear.

Tanaka had eight destroyers, flagship Naganami, with Takanami ahead as a picket. Six of them were loaded with drums. At 2306 Minneapolis’s SG radar detected the Japanese at 23,000 yards off Cape Esperance, steaming southeast along the coast toward Tassafaronga. At 2315, with the range down to about 7,000 yards, Commander William M. Cole in Fletcher, whose radar had the targets clearly, asked permission to fire torpedoes. Wright took about two minutes to reply and then refused; after a second delay of about two minutes, at 2320, he ordered the destroyers to fire. Cox calls it “five minutes of indecision.” Fletcher fired ten torpedoes, Perkins eight, and Maury and Drayton three between them, but the targets were now passing on an opposite course and the shots were poor. Almost at once Wright ordered the cruisers to open fire, before the torpedoes could reach their targets.

The gunfire fell on Takanami, the nearest ship, which all five cruisers engaged; she was wrecked and burning, and only 33 of her 244 men eventually reached Guadalcanal, according to Cox. Her fire also showed the Japanese exactly where the American cruisers were, by their gun flashes. Tanaka had already cancelled the unloading at 2312 when Takanami reported the Americans. His destroyers, many still with drums on deck, turned and fired. Cox writes that they fired 44 Type 93 “Long Lance” torpedoes in a few minutes.

At about 2327 two torpedoes struck Minneapolis; one blew off her bow, the other hit No. 2 fireroom. New Orleans, next astern, steamed into a torpedo that set off her forward magazines and severed the bow forward of No. 2 turret. Pensacola, turning to avoid the two damaged ships ahead, was hit under the mainmast at 2339 by ONI’s timing; burning fuel oil rose to the maintop, killed everyone in her after control station and burned for four hours, while about 150 8-inch rounds in No. 3 turret exploded one after another. Honolulu maneuvered at high speed and escaped. Northampton, last of the cruisers, was hit at 2348 by two torpedoes, so close together, ONI says, that many aboard felt only one explosion. She took on a heavy list and caught fire. At 0130 all hands except a salvage party were ordered over the side; at 0240 Captain Willard A. Kitts and the salvage party left her, and just after 0300 she rolled over and sank stern first.

Fletcher and Drayton came back to pick up Northampton’s crew. The ONI narrative of Tassafaronga records that they rescued all 57 officers and 716 men in the water; Cox credits Fletcher alone with 646. Fewer than three weeks before, as Wylie wrote, Fletcher had steamed past Juneau’s wreckage because there seemed to be no one to save. Minneapolis, New Orleans and Pensacola crept into Tulagi, where the tug Bobolink and PT boats of the Tulagi squadron helped bring Minneapolis in. All three were out of action for more than a year. The PT boats also carried survivors into Tulagi, among them the boat in the photograph below, PT-109, which John F. Kennedy would command the following year.

American losses at Tassafaronga were heavy for a single short action. Cox gives 395 killed and 153 wounded; the ONI narrative counts 19 officers and 398 enlisted men lost, 417 in all. By ship, Cox gives 37 killed in Minneapolis, 183 in New Orleans, 125 in Pensacola and 50 in Northampton. The Japanese lost one destroyer, Takanami. The wartime ONI narrative recorded CINCPAC’s belief that four Japanese destroyers had been sunk, revised later to two; in fact there was one. Cox calls the Japanese attack “arguably the most successful surface torpedo attack in history” and the battle “one of the most ignominious defeats in U.S. Navy history.” American commanders at first attributed the hits to submarines, because they did not believe any surface torpedo could hit at the ranges involved.

Yet Tanaka’s mission failed too. His destroyers had turned to fight with drums still aboard and left without delivering their supplies; Cox concludes that, technically, Task Force 67 succeeded, since no supplies from that run reached Guadalcanal. The drum runs that followed in December delivered only a fraction of what was dropped: many drums broke loose, sank or were shot up by aircraft at dawn before the men ashore could haul them in. Through December the Tulagi PT boats attacked the runs night after night, and on 12 December they torpedoed and sank the new destroyer Teruzuki, the ship that had helped rescue Kirishima’s crew a month earlier. The story of those runs, and the Japanese destroyermen who made them, is the subject of the collection’s Tokyo Express record.

For the U.S. Navy, Tassafaronga was the last of the big cruiser battles in Ironbottom Sound. Cox traces lessons from it into 1943: Cole’s experience shaped Captain Arleigh Burke’s standing order that destroyers were “to attack the enemy on first contact without awaiting orders from task force commander,” and the tactics Commander Frederick Moosbrugger used at Vella Gulf in August 1943. Wylie of Fletcher and Tisdale of Honolulu helped develop the Combat Information Center, which turned radar data into a picture a commander could use. Halsey, however, awarded the Navy Cross to Wright, Tisdale and the five cruiser captains, and none to the destroyer captains whose torpedoes had been held back.

Myths, the record and how the battles have been told

The first accounts were the ONI Combat Narratives, written in 1943 and 1944 for officers of the Navy and classified at the time; the November volume was based on action reports and the TBS log of the first night, which it reproduced as an appendix. They are vivid and close to the events, and their times and ship movements remain valuable, but their claims of Japanese losses were wrong in almost every case. Samuel Eliot Morison’s The Struggle for Guadalcanal (volume V of his History of United States Naval Operations in World War II, 1949) used Japanese sources available after the war and corrected many of those claims; Morison was blunt about the first night, which he compared to a barroom brawl after the lights had been shot out. Richard B. Frank’s Guadalcanal (1990) drew on Japanese official histories and remains the standard single-volume account, and James D. Hornfischer’s Neptune’s Inferno (2011) retold the naval campaign through the experience of the men in the ships. Robert Lundgren’s damage analyses, published by NavWeaps, reconstructed hits on Kirishima and South Dakota from Japanese and American records. Samuel J. Cox’s H-Grams for the Naval History and Heritage Command (2017) gave the Navy’s current summary, with casualty tables ship by ship.

The places where the battles happened are still named for them. The sailors who fought there called the waters between Savo, Florida and Guadalcanal “Ironbottom Sound” for the ships on its floor; American and Japanese wrecks from November 1942 lie there within a few miles of one another, and several, including Atlanta, have been surveyed and dived. Juneau, found in 2018, lies far to the southeast, where I-26 caught her. Kinugawa Maru rusts on the beach west of Honiara. Solomon Islanders who live along that coast have grown up with the hulks and with the stories of the nights when the sky over the Sound lit up; the memorials, the wrecks and the islanders’ own memory of the war are the subject of Chapter 9.