The navigator of the Taiho stood on the pilot house when the deck opened up beneath him. He watched his ship lurch over to port and go under stern first, and a year and a half later he told an American officer with a notebook exactly what he had seen. The Americans took his account, set it against what her designers told them in Tokyo, and decided they did not need to interview anyone else who had come off her. That was a strange conclusion for a technical mission to reach about the newest and most expensive aircraft carrier the Imperial Japanese Navy ever put to sea.
On June 19, 1944, the Taiho had been in commission for three months and eleven days. That morning she was doing the work she existed for. The trade wind in that part of the Pacific blows out of the east, and a carrier has to steam into the wind to get aircraft off the deck. This meant that Ozawa’s fleet, sitting west of the Marianas, could launch and keep closing on the Americans at the same time, while the Americans had to turn their backs on him to fly.
Ozawa had the wind, and he knew he had it. At about eight in the morning, the Taiho put nearly her whole air group into the sky, holding back sixteen fighters for combat air patrol. She was buttoned up, the crew at general quarters, and the whole force had spent the previous day on notice that American submarines were somewhere close. To understand what happened next, it helps to know the inside of the ship.
The Taiho was 850 feet long and carried an armored flight deck, three inches of it, which no Japanese carrier had ever had before. Her hangars, upper and lower, were closed structures with steel on all four sides in the British manner rather than the American, and both opened onto the forward elevator well. Her aviation gasoline sat in tanks in the bow, forward of the point where her torpedo protection ended, because that protection ran only the length of the engine and boiler rooms. Around those tanks, the designers had put a thin layer of fuel oil and nothing else.
Above them, there was no void, no empty space, no separating compartment of any kind between the top of the gasoline and the deck of the elevator pit. That pit floor sat below the water line when the ship was fully loaded. The men whose job it was to save her worked inside an organization that looked on paper like a serious one. The Navy Ministry ran a technical council on damage control and a standing battle report investigating committee whose entire purpose was to turn action reports into practice quickly.
Three schools taught the subject. Deck officers went to the naval navigation school for firefighting, flooding, counterbalancing, poison gas, and the handling of the wounded. Engineering officers went to the naval workshop and repair school or the naval engineering school for pumping, shoring, cabling, and machinery repair. Once a year, on the ministry’s order, a fleet commander exercised his whole fleet in damage control.
The war shortened all of it. The officer’s higher course went from a year to six months. The enlisted ordinary course from six months to four. The month an officer spent attached to the other two schools became a fortnight, and the enlisted man’s fortnight became a week.
Damage control duty on a fleet or squadron staff was never a post of its own; a man carried it on top of his regular work as a navigator or an engineer. Firefighting itself was taught at the Yokosuka Naval Station on four old ships tied up for the purpose. Students watched a gasoline fire put out with foam. Then they watched a second gasoline fire lit for the purpose and put out by shutting the compartment up, hosing the outside of it to keep it cool, and waiting for the fire to consume the air and die.
Both methods work. Only one of them can be used on a hangar deck with aircraft in it. A Japanese capital ship carried fire mains that could be split into three to five isolated sections in battle, sprinkling systems in the magazines and the hangars, carbon dioxide bottles to be carried to the fire, portable foam gear, and portable petrol-driven pumps in four sizes. In the flooding and pumping compartment, the only way a man could tell whether the sea was getting into a particular void was to put his ear to a vent valve and listen for escaping air, or to take the top off the tank and look.
The American paperwork on the same problem is dated, and the date is the point. On May 8, 1942, the Lexington was lost in the Coral Sea, not to the bombs and torpedoes that hit her, but to petrol vapor that found its way up out of her stowage and into the interior of the ship, going off an hour after the shooting stopped. The Bureau of Ships took that apart and issued recommendations for immediate application. Keep the layer of tanks inboard of the gasoline tanks empty and keep inert gas in them at all times.
Separate the port and starboard gasoline systems so that a sound one cannot pressurize a damaged one. On any suspicion of vapor after an attack, run exhaust ventilation early. Shut down every piece of electrical equipment liable to spark. Fifteen weeks later, the Enterprise took three bombs in four minutes off the Eastern Solomons, and her people had been scraping paint off the bulkheads for months on the theory that paint burns.
Her repair parties and their gear were scattered around the ship by design. The bureau’s own account of her performance credits it to preparation, to the study of what had happened to the Lexington and the Yorktown, and to regular drills. At about half past eight that morning on the Taiho, a torpedo fired by the submarine Albacore struck the starboard side forward, straight into the gasoline. The forward elevator, sitting flush with the flight deck, dropped about six and a half feet and jammed.
The floor of its pit tore open across a wide area. The ship went down by the head a foot and a half and slowed by exactly one knot. Nothing caught fire. Her crew laid planking over the open well, and the sixteen fighters held back for the patrol were flown off before nine.
The Taiho had taken a torpedo in the worst possible place and shrugged. Underneath the planking, the elevator pit was filling with a mixture of seawater, fuel oil, and liquid aviation gasoline, and the pit opened into both hangars, and the hangars had steel on every side. The fumes went where the air went. Damage control saw the danger within minutes and moved on it.
They opened every door and hatch they could reach, betting that the ship’s own speed would drive a draft through her and take the vapor out over the side. What it did was carry the vapor into compartments that had been clean. An attempt was made to pump the free gasoline out of the pit and over the side, and it was, in the flat wording of the American report, bungled. So the order went out to start every ventilation set aboard, supply and exhaust together, and it was carried out.
The fumes now had a distribution system. The ship, the report says, was wreaking. What separated the two navies was a matter of stowage rather than nerve. Both fleets knew that aviation gasoline vapor was the thing that killed carriers.
The difference is what each of them was willing to put in the space around the tank. The Americans emptied those compartments and filled them with inert gas, and where they could, they displaced the fuel itself with seawater as it was drawn off, so that a tank never held a volume of air to explode. Water replacement was no secret to the Japanese. They did not fit it.
Their own aviators objected to it, and the aviators won that argument. The space around the gasoline stayed full of fuel oil. The Lexington burned in May 1942, and her lessons were on the Saratoga’s deck plans within a month and in the hands of the Enterprise’s repair parties by August. The Japanese Navy had a committee whose written purpose was to do that same job at that same speed.
What it no longer had by the middle of 1944 was margin. The carriers that would have carried the experience were on the bottom, and so were the chief petty officers who had fought fires on hangar decks at Coral Sea and in the Solomons, who would have known without being told what a hangar full of steel and vapor was. The Taiho commissioned on March 7, 1944, and went into action on June 19. Her own designers rated her to handle 52 aircraft, and her air group carried 64.
Nobody aboard her had ever done this before. At three in the afternoon, seven hours after a single torpedo had cost her one knot, the vapor found whatever it was waiting for. The navigator up on the pilot house put the center of the explosion at the forward elevator opening. The armored flight deck, the pride of the design, split down the middle.
Both sides of the hangar blew out. The armor did what armor does and turned the blast back down into the ship. The only men who got out of the engineering spaces came up out of number two fire room, climbing straight up through the wreckage. A few men from the crew spaces on the orlop deck got out through holes torn in the side of the hull.
The fire that started then never went down, because it was being fed from ruptured tanks. She was abandoned with her engine rooms dead and burning, and at four o’clock she rolled to port and went under. Compare the afternoon of August 24, 1942, on the Enterprise. First bomb at sixteen minutes to five.
Three hits and three near misses. Thirty-five men killed in a mess compartment by the first one alone. By sixteen minutes past six, the ship was making 24 knots, and her deck was ready to take aircraft. The first one landed at nineteen past.
Twenty-five came aboard before a steering casualty stopped the work. Ninety-five minutes. What lit the Taiho is not known, and it is not going to become known. A special committee drawn from every section of the Navy Ministry took the loss apart, because a modern carrier destroyed by one torpedo alarmed the operating fleet and the constructors equally.
Their report burned with the ministry in May 1945. The compartments where the ignition happened were forward and low, and the men who were in them did not come up. The report the Americans wrote afterwards lists the vapor, the open hatches, the running fans, and the live switchboards, and then observes that the fans provided innumerable sources of ignition, a sentence that names a category and no member of it. The fix the Navy Ministry chose after the Taiho was to encase every gasoline tank in the fleet in three feet of reinforced concrete.
That decision and the date on it is the whole story. The simpler fix had been talked down years earlier by aviators who did not want seawater near their fuel. Commander Hedgecock of the United States Navy did the investigation out of the hull, engineering, and gunnery sections of the Navy technical department in Tokyo, across a table from men who had spent their careers on the problem. What Hedgecock filed on February 4, 1946, opens by saying that against American and British standards, the Japanese organization, training, and equipment were all inferior, that real improvement had been made in all three during the war, and that good results were never obtained.
He also recorded that his was a negative report, in that nothing had been found which could be used to improve American procedure or training. The alterations the Japanese ordered after the Taiho went into a short list. Gasoline capacity on the surviving carriers was cut hard. Every gasoline tank was to be wrapped in three feet of reinforced concrete to soak up leakage or seal it in.
Large supply fans went into the forward bulkhead of every hangar. Canvas screens were issued to be rigged on the flight deck ahead of the forward elevator to act as wind scoops and force air below. No further Japanese carrier was ever designed. How many men died on the Taiho is not settled.
The American technical report says more than a thousand of her crew were lost and fewer than five hundred survived, almost all of them from topside spaces. Japanese records give it the other way round, with over a thousand recovered from the water by the destroyers and the boats of the cruiser Haguro, and twenty-eight officers and 632 men lost. An inquest into the carriers lost off the Marianas convened in Tokyo in July 1944.
Its file is not in either archive.