On July 20, 1944, the radio room of a Japanese submarine lying at Singapore sent Tokyo a full account of where the boat intended to be and when. Her captain was Takakazu Kinashi. Two years before, in the black water off Guadalcanal, he had put a spread of torpedoes into the American carrier Wasp and sunk her. The Navy had since handed him something it wanted more than a carrier.
The submarine I-29 had left Lorient in April with German cargo in her pressure hull. There was a Walter rocket motor of the kind bolted into the Messerschmitt 163, a Junkers turbojet of the kind hung under the 262, drawings and handbooks for both aircraft, twenty cipher machines, and a quantity of mercury and bauxite down in the bilges as ballast worth carrying home. She crossed two oceans with it. On July 14, she came up the Malacca Strait and berthed at Singapore, and her passengers went ashore.
One of them was Commander Eiichi Iwaya, a naval technical officer. On July 17, Iwaya boarded an aircraft for Tokyo with as much of the paper as a man can carry in his hands. The boat sailed for home on July 22. Four days later, the American submarine Sawfish found her in the Luzon Strait, and everything still aboard went down with her.
What reached Japan was not a rocket motor. It was an instruction manual, a set of general arrangement sketches, and some figures. Mitsubishi had to build the airplane out of that. The Japanese called it Shusui: short, tailless, wing-shaped, a rocket chamber in the back, and no undercarriage worth the name.
It existed to climb to 25,000 feet fast enough to shoot at a B-29 and then come down again as a glider. In the summer of 1944, no aeronautical problem in the Japanese Empire mattered more. It was also, on paper, the one program the Army and the Navy had agreed to run together. An accord between the two air services in August 1943 had set up a joint aeronautical technology committee.
The rocket interceptor belonged to both of them. The Navy would call it Shusui. The Army would call it the Ki-200. Same drawings, same manufacturer, same motor.
That arrangement deserves attention, because outside it, the two services shared nothing whatsoever. Each ran its own air headquarters under its own ministry. Each ran its own research establishment doing development work on new combat types: the Army Arsenal at Tachikawa and the Naval Air Depot at Yokosuka. Each kept what it learned.
They numbered their aircraft differently, numbered their engines differently, and issued guns of the same nominal bore firing ammunition that would not fit the other service’s weapon. Before the war, a purchasing mission staffed by officers of both services had gone to Germany together, inspected the industry together, and come home with a Heinkel, a Messerschmitt, a Junkers, and a set of Daimler-Benz engines. Then the Army bought a license for the inverted V-12 and gave it to Kawasaki. The Navy bought a license for the same engine and gave it to Aichi.
Japan built one German power plant on two production lines under two names. That was the pattern for anything foreign. When a captured aircraft or a set of drawings came in, whichever service had it insisted on studying the material first, handed it to industry afterwards, and then to one chosen firm, which kept it from the rest. The American survey team that went through the industry after the surrender found cases where both services had obtained the same foreign airplane and put it out to different companies for development, neither telling the other.
Nakajima was the largest aircraft manufacturer in the country, building better than a third of all combat airframes and close to a third of all aero engines. Inside Nakajima, there were two firms. Ota and Utsunomiya built for the Army, while Koizumi and Handa built for the Navy. The Hamamatsu engine works belonged to the Army alone, and the Omiya engine works to the Navy alone.
Only the Musashi engine plant did work for both. At Ota, they were turning out the Ki-84, the best fighter the Army ever put into squadron service. Ten miles away at Koizumi, in the same prefecture, under the same company name, the same firm was turning out Zeros. No pilot in either service would ever be issued the other’s airplane.
The survey put it flatly in its industry report. “No effective cooperation existed between the army and the navy,” the investigators wrote, “in spite of direction from the top of the government that there should be. ” Within any one company doing work for both services, two organizations existed side by side. One for army work, one for navy work.
The machine tools gave the game away. Between them, the two services owned or controlled most of the machine tools in Japan and leased them out to the aircraft firms as required. A lathe on the Army’s books could not be put on Navy work. A Navy jig could not be handed to an Army line.
It made no difference how urgently it was wanted or how long it stood idle. The government created the Ministry of Munitions in November 1943 to end exactly this. Its Air Ordnance Bureau was to gather planning, materials, and equipment into one pair of hands. What followed instead was that the Army and Navy kept their own supervisors inside the factories, held on to the plants that mattered, drew up their own programs for aircraft armament, components, and raw materials, and declined to give the ministry the details.
The ministry spent the rest of the war handling civilian goods. Both services went on building aircraft in their own workshops as well, and between them they finished the war as the sixth largest producer of combat aircraft in their own country. Saburo Endo ran that bureau. A lieutenant general who had spent the war arguing for standardization, he gave the Americans an answer afterwards that started with the two rival headquarters rather than with the bombing.
Each of them had a decisive battle it meant to win. The Navy expected its battle north of New Guinea in June, the Army expected its own in the Philippines in August. Until those two came, nothing else counted. Everything went into the peak, and the peak arrived, and both battles went the wrong way.
By then, the machinery had worn out and the shelves stood bare. By that stage, the quarrel over parts had stopped being a quarrel. Competing for the same steel, the same light alloy, and the same fittings, the two services came to physical violence over it. Armed patrols belonging to one went and carried off supplies set aside for the other.
That was not an incident but a practice, and it got worse the longer the war ran. It reached the sky over Tokyo. An American officer interrogating a staff officer of the 10th Flying Division in October 1945 worked through the capital’s air defense and eventually asked whether there had been a common fighter director center for the two air forces. The answer was that they operated independently, but there was liaison.
The Army held the air over the land. The Navy from Yokosuka, Kisarazu, and Atsugi covered its own immediate area. When an attack came, a telephone line ran between Army East District Headquarters and Yokosuka. Two fighter forces, one city, two chains of command, and by the middle of the year an aviation gasoline shortage severe enough that the Army division had pilots it could not put in the air.
That brings the matter back to the rocket fighter, and to the one part of it nobody has explained. Production of the airplane went two to one between the services. The rocket motor itself, the hard part, the part nobody in Japan had ever built, became the Army’s responsibility under the joint arrangement. The Navy developed a motor of its own regardless.
The propulsion unit was defeating everyone. Hydrogen peroxide and hydrazine hydrate in a thin steel chamber is unforgiving chemistry, and Japan was attempting it from a handbook. When bombing forced Mitsubishi to move the prototypes out of Nagoya to Yokosuka in March 1945, the motor came with them unsolved. From there, the Army and the Navy each set up a research effort of its own on the same failure.
In June 1945, each of them finished its first working prototype. Two motors, two teams, one airframe, one country, ten weeks from the end. The Army went further still and started design work at Tachikawa on an improved version of an airplane it did not yet have flying. The survey listed that program as the best example of joint Army and Navy working it could find anywhere in the Japanese aircraft industry.
It was not being generous. Why the Navy built the second motor is the question the record leaves open. The sequence is documented in detail, dates and all. No reason appears anywhere for the decision, from the men who took it or from the men who came afterwards to investigate.
There is no memorandum, and nobody sat either service down and asked. It survives in the American technical account as a bare fact of engineering history. There is one more thing about the boat. Nothing in the Japanese record suggests anyone in Tokyo ever asked how the Sawfish came to be in the Luzon Strait that particular week.
The signal Kinashi’s radio room sent on July 20, the one with a route and the dates in it, went into the fleet radio unit at Pearl Harbor. The American command had the boat’s own schedule inside days and passed it out to the submarines. Berlin had already cabled Tokyo a description of what was in her pressure hull, and the Americans read that too. Kinashi was not hunted.
He was met. That is why two Japanese services spent the last year of the war building two rocket motors from a handbook. A cargo that had survived the Atlantic, the Indian Ocean, and the South China Sea came off the board in the final four days of the voyage, taken by a capability the Imperial Navy never once accounted for. After that, the only way to have a rocket motor was to invent one.
The two air services could not agree to invent at once. There was dissent inside the program by then, and it had nothing to do with the enemy. The army arsenal at Tachikawa began drawing an improved version of an airplane that had not yet flown. Engineers who took a middle position argued that the sane course was to finish the machine already designed.
Out at Mitsubishi, the shop floor held a view of its own. The men building the prototypes did not believe the rocket motor would ever be made to work. They put no great vigor into the job, and the American investigators noticed it and wrote it down. The Army got its engine into an airframe first.
The machine stood at Kashiwa, northeast of Tokyo, with the motor installed and the flight already canceled because the fuel system leaked. Before any of that, the airframe itself had flown unpowered nine times while the engine work went on, and those flights changed the machine. The hydraulic system for the undercarriage came in for modification, and the slit at the aileron was narrowed, the ordinary business of getting an airplane right, carried out on a design that still had no working motor to put in it. The Navy flew on July 7, 1945, from the airfield at Yokosuka.
Everyone involved understood that Yokosuka was too small for this. The strip was short and the margin was nothing. To hold the weight down, they put aboard the smallest quantity of fuel the profile allowed. Takeoff was clean.
The climb was clean. At about 1,100 feet, the motor stopped. The pilot did the only things available to him: dumped what fuel he had and turned back toward the strip. He did not reach it.
The airplane went into a house near the field, was wrecked, and he died a short time afterwards. When they opened up the fuel system to find out what had happened, the answer turned out to be the small fuel load itself. Under hard acceleration in a steep climb, the surface of the liquid in the tank had moved off the outlet and the flow had stopped. They redesigned the whole system around that finding, moved the drain port and enlarged it, fitted a jet pump, and reissued the drawings.
Forty miles away at Kashiwa, the other service’s airplane was still standing on its own field with its own engine in it, waiting on a fuel line that would not hold.