Arleigh Burke spent the last week of July 1943 at anchor in Purvis Bay off Tulagi with eight destroyers and almost nothing to do. What he did with the time was talk to their captains. He had a plan he wanted them to know by heart. Destroyers would go out alone without cruisers over them, in two short columns.
Whichever column was nearer the enemy when radar found him would fire torpedoes and turn away hard. The other column would keep its guns quiet, swing across the front of the enemy line, and open fire only after the torpedoes had arrived. No searchlights. Nothing on the short-range voice radio except orders.
None of that was new. It was Japanese doctrine written out in English. The Imperial Japanese Navy had built its whole night battle on those ideas in the years between the wars. Torpedoes first from the dark at ranges nobody else thought possible.
Guns when guns were needed and not before. Searchlights almost never, because a searchlight tells the other man where the light is coming from. Underneath all of it, lookouts. The Japanese picked men for night vision, trained them as a specialty, and handed them glasses with lenses the width of a saucer.
They were the best trained eyes at sea anywhere in the world, and there is no serious argument about that. They also had a torpedo nobody else could match. And for most of the year, the Americans did not know they had it. It ran on oxygen instead of compressed air, which meant that at night it left almost nothing on the surface to give it away.
And it reached out to distances American officers thought impossible for a weapon of that size. Ships kept coming apart in the dark with no visible cause. American staffs put those losses down to submarines nobody had detected and went on building their plans around a danger that was not the one killing them. A complete Japanese torpedo was not taken apart and understood until the middle of 1943, which was not long before Vella Gulf.
The men who worked this out were not staff theorists. Captain Kaju Sugiura, who would take four destroyers down into Vella Gulf on the 6th of August, had been chief instructor until shortly before at the Navy Torpedo School at Yokosuka. Commander Tameichi Hara, who would bring up the rear of that same line, had a hand in writing the tactical manual as the war began. Burke had not invented any of it.
What he was teaching in Purvis Bay was the enemy’s own method, taken apart across a year of losing to it, and put back together by men who had been on the receiving end of every part of it. On the 4th of August at Rabaul, Sugiura called his captains together and told them the last delivery run to Kolombangara had gone perfectly. Both the Navy and the Army were pleased, and they would do it again in two days. The same passage on the same schedule.
Hara sat through this and then said what he thought: that three identical runs through the same stretch of water was an invitation. The order stood anyway. Sugiura’s answer was that the way in had been chosen above him and could not be altered because the garrison at the far end was hard to reach by radio to arrange anything else. Behind that sat two more things.
The reports from the previous two runs had described them as successes, and they had been. And on the 1st of August, a squadron of American torpedo boats had tried to stop this same operation in the same water and had failed, which is what the staff at Rabaul had in front of them when they ordered the third run. Sugiura offered Hara the scouting position at the head of the column, unloaded and running clear. Hara turned it down because his engines were long overdue for a refit and could not hold much more than 30 knots.
So Sugiura took the point himself in Hagakaze, and then did something harder to explain. He loaded his own ship with troops as well. When the four destroyers weighed anchor at 5:00 in the morning on the 6th, not one of them was a fighting ship with clear decks. The Americans had spent a year finding out what was wrong with their equipment.
Their surface torpedo, the Mark 15, swam about 10 feet deeper than the setting on its dial, and its magnetic firing gear had a habit of doing nothing at all. Both faults were understood at the working level by now, whatever the Ordnance Bureau thought. Crews had switched the magnetic gear off, so the weapon had to strike a hole to matter. The warheads carried a stronger filling.
Somebody had bolted shields over the tubes to hide the flash of a launch, because a flash at night is a bearing, and a bearing is a firing solution. The last fix of all was made by one officer on his own authority. In the afternoon, on the way to the battle, Russell Crenshaw was executive officer of the destroyer Maury. Instructions from the destroyer command said to set torpedo depths at 5 feet and 9 feet in turn.
Crenshaw had worked out that if the fish were swimming deep, the 9-foot ones would pass under a destroyer. He put this to his captain, who agreed something might be wrong but would not go against a written instruction on a hunch. Crenshaw left the bridge, thought about it for an hour, went down to the torpedo battery himself, and supervised the setting of all 16 to 5 feet. Then he had the depth spindles withdrawn and locked and told two torpedomen by name to let nobody near them.
There was a reason the destroyers were going out on their own, and part of it was not doctrine at all. Twice in July, at Kula Gulf and then off Kolombangara, American squadrons built around cruisers had gone up the slot to stop these same supply missions, and twice they had failed to stop them and been hurt trying. Theodore Wilkinson took over the amphibious force in the middle of that sequence and found he had very few undamaged cruisers left to send anywhere. What he did have to hand was a striking force with nothing above it.
The orders that sent the force out were three pages long. Be underway from Tulagi at half past 11:00 in the morning. Be off the bottom of Vella Gulf at 10:00 at night. Sweep it.
Every surface contact inside the Gulf is an enemy. Destroyers are the target of choice, and barges come second. And if nothing has been found by 2:00 in the morning, go home down the slot at maximum speed. That gave the force a box of water and four hours to find something inside it.
Frederick Moosbrugger had taken over the striking force from Burke three days earlier. His admiral had told him to be wary of Japanese torpedoes and lean on long-range gunnery instead, and had also given him the freedom to ignore that advice. He ignored it. On the morning of the 6th, he brought all six commanding officers aboard his flagship and went through the plan and every branch of it until there was nothing left to ask.
His written account afterwards described the result as complete mutual understanding of all possible situations to be encountered. At half past 4:00 that afternoon, an American patrol aircraft found the Japanese column and reported its position, course, and speed. Sugiura watched it go into cloud, understood that he had been seen, worked out that he would have to fight before morning, and changed nothing whatever about the passage, the speed, or the timing, because none of those things were his to change. The quarter moon went down at 26 minutes past 10.
After that, there was no moon at all. A low overcast, rain squalls drifting through, a calm sea with almost no air moving, and visibility that came and went between 2,000 and 4,000 yards. An Australian coast watcher on Vella Lavella had already sent word of about when to expect company in the gulf. So the Americans were not searching so much as waiting in the right place.
They came up the eastern side, hugging the coast of Kolombangara, some of them within 4,000 yards of a beach nobody was certain was empty of Japanese. The mountain behind them did the rest. Against that black bulk, six darkened destroyers were nothing at all. They swept the mouth of Blackett Strait, found it clear, and turned north up the coast a little after 20 minutes past 11, two columns in line, 500 yards between ships.
The Japanese came in from the north end of the gulf at 30 knots. Harusame could not hold her station and had dropped back to something like a mile behind the ship ahead of her. Her navigator asked whether they should force the engines to catch up. Hara said no and left her where she was.
At 33 minutes past 11, the search radar in the American flagship picked something out of the weather clutter to the north, about 10 miles off. Within a couple of sweeps of the antenna, it had resolved into four distinct returns in a line, and the other ships had it too. The next eight minutes happened below decks. In the plotting rooms, men who could see nothing at all began building a firing solution out of numbers.
Bearings and ranges came off the radar screen every few seconds and went onto a plotting sheet. The plot gave back a course and a speed for a column of ships nobody aboard had laid eyes on. That answer went up a voice tube to the bridge and sideways to the torpedo director. The tubes trained out to a bearing on the black.
Moosbrugger brought his division around to shorten the range, and it fell fast, the two forces approaching each other at better than 50 knots between them. On the bridges, officers stood at the voice tubes and stared into nothing. Every gun in the division was trained on a target, and not one of them fired. A few men topside whispered that they thought they could see something out there.
Nobody was sure. For 20 months of war, a night action had begun when somebody saw something, and the Japanese had built a navy around winning that moment, and were still winning it night after night with better glasses and better lookouts than anyone else had. What they had no answer for was an enemy for whom seeing had stopped being the qualifying event. At 41 minutes past 11, at a range of about 6,000 yards, the order to fire went out over the voice radio.
Twenty-four torpedoes went into the water in a little over 60 seconds, three seconds apart, sliding out of their shields without a flash. Then the whole division put its rudder over together and ran, so that anything coming back would find three sterns. In the plotting room aboard Maury, a man started a stopwatch on the first weapon and called out each one as it should have crossed the enemy track. The first call went by unanswered.
The second brought an explosion on the quarter. The third, silence. The fourth, a column of fire on a different ship. Then another.
Then another. The reports coming down from topside stopped being about bearings and started being about the height of the flames, which one man put at 1,000 feet. Nobody in the American force had yet seen a Japanese ship. Almost nothing in that attack was original because it was the enemy’s own night doctrine, taken apart across a year of losing to it and handed back whole.
On the other side, those same four minutes had gone like this. Hara, who liked nothing about the night, had doubled his lookouts and had his guns and tubes trained out to port into the worst of the darkness. Lookouts in all four ships began calling in shapes they could not identify, and half of them took the shapes for torpedo boats. Then a lookout in Shigure called white water and dark hulls and several ships coming at them.
By then the torpedoes had been running for three minutes. Hara put his rudder hard over and got a salvo away down the reciprocal bearing, and while his own weapons were still hitting the water, the three ships ahead of him came apart one after another within seconds of each other. One American torpedo went clean through Shigure’s rudder without going off. He did not find out that it had happened until somebody looked at the rudder afterwards.
The second American division, which had crossed the front of the Japanese line while all this was going on, then opened up with 5-inch guns on ships that were already burning, and the thing turned into a demolition. There was one more difficulty about looking, and it belonged to the winners, because running away from their own torpedoes had put the burning ships almost dead behind them, inside an arc of about 30 degrees either side of the wake in which the search radar showed nothing whatever for the first stretch of the shooting. The men who had just destroyed three ships could not count them. Moosbrugger brought his own division back around to the south and gave the order to engage with guns.
Aboard Maury, the search picture began to fade the moment her first salvo went out, the picture breaking up under the shock of her own 5-inch mounts. Crenshaw wanted the radar more than he wanted a few extra hits. He called a ceasefire over the internal telephone circuit and explained himself up the voice tube to his captain afterwards. Within a few moments, Maury was the only ship in her division with a search set still working.
Hara laid smoke and took Shigure away to the northwest into the rain. There his crew reloaded the torpedo tubes, which took them 23 minutes. Then he turned back south towards the burning. He had been asking the other three ships for their situation and getting nothing back.
At 10 minutes past midnight, the magazine in Arashio went up. Somewhere above, Hara could hear the engine of an American night patrol aircraft. From those two things, he worked out that his consorts were being bombed from the air. He also believed that his own salvo had sunk an American destroyer.
Neither of those things had happened. At a quarter past midnight, he turned for Rabaul. It was finished before half past midnight. Sugiura’s flagship went last under fire from everything the Americans had.
Afterwards, there were hundreds of men in the water and a sheet of burning oil across it. The smell of gasoline off the deck cargo was strong enough to hurt the eyes. The American ships moved through it for the better part of an hour, offering lines and floats. Almost nobody took one.
The men in the water could be heard talking and chanting together. When a ship came near a group, a whistle would sound and the chanting would stop, and they would swim away from it. The only injuries in the whole American force were a hand hurt on one ship and second-degree burns to two men fixing a feed pump on another in an engine room in the dark on the way home. How many Japanese died is not settled and probably never will be.
The figure usually given is a little over 1,200, the larger part of it soldiers rather than ships’ crew. Some accounts run well above that. Nobody can do better because no war diary survives from any of the ships that went down, and the officers who could have settled the arithmetic went into the water with them. Around 300 men reached Vella Lavella and were taken off later.
The American paperwork on the night is otherwise in good agreement with itself and with what was found afterwards, apart from one thing: the men watching the two biggest ships burn took them for cruisers. What the Japanese did about it afterwards is the interesting part. They stopped using Vella Gulf and Blackett Strait for the run to Vila. Inside a week, the conclusion at Rabaul was that Kolombangara could not be held, and the work of getting the garrison off it began.
Sugiura himself came out of the water alive after 30 hours, walked in the jungle for a week, was picked up and went back to Rabaul in disgrace. Shigure got home. What waited for her there was a version of the night in which the only ship to come back was the ship that had run. Members of her crew ended up in fistfights ashore defending her against it.
What did not happen was any change in how that navy fought at night. The doctrine stayed as it was. The lookouts stayed the best in the world. The lesson available that night was about how a night action now started, and the lesson taken was about a stretch of water.
It did not cost them everything either. Two months later off Vella Lavella, three American destroyers with radar and plotting rooms and the new doctrine ran into nine Japanese ships and came off worse, losing one of their own and having the front end blown off another. American destroyers fought their opposite numbers five more times after Vella Gulf before they won another action like it. Russell Crenshaw was given the Legion of Merit a year afterwards for the business with the depth spindles.
That leaves one book. Tameichi Hara wrote his memoir after the war, and it appeared in 1961. It is the only sustained account of that night from inside the Japanese line, running from the conference at Rabaul through to the water afterwards. And it exists because his engines were tired and he was too far back in the column to be killed.
He was not grudging about any of it. He wrote that the Americans had done everything correctly, that they had ambushed his force brilliantly, and that never before had he seen such marksmanship. Every ship in that line is on the bottom of one sea or another now, his own included, sunk two seas away and a year and a half later by an American submarine.
The man who could not hold his station is the reason anyone knows what the night looked like from the other end of those torpedoes.