On the night of August 28, 1943, a Japanese officer named Kobayashi was carried along the north coast of New Guinea in a landing barge from Shio to Finchhafen. The craft made about eight knots, its helmsman steering by the dark line of the shore a few hundred yards away. The trip passed without incident, and Kobayashi counted that as luck rather than routine. He noted that the barge unit which carried him was attacked on the return run, and the barge commanding officer and his men were all lost.
The barge line was not an act of desperation by that summer. It was a considered answer to a problem Japan had already tested to destruction. In March 1943, off Lae, Allied aircraft caught a single Japanese convoy in the open. Eight transports and four destroyers went down in three days.
Through July and early August, in the waters around New Georgia, three more destroyers and a light cruiser were lost trying to move troops and rice at night. Anything large enough to be worth a torpedo or bomb was going to be found and killed within range of Allied airfields. So the traffic was handed to craft too small to be worth killing. A barge was not worth a bomber’s load, could not be torpedoed at all, and could be hidden under palm fronds in daylight.
It drew so little water that it could run inside reefs where a destroyer would tear its plating open. For most of a year, the barge line worked. The line was a chain of hiding places rather than a shipping lane. Along the north shore of New Britain, staging points ran west from Cape Hoskins through Talasea, Gava Island, Iboki, Natimo, and Cape Gloucester, then around the western tip to Sagsai and across the strait to Rooke Island.
The southern shore carried its own set at Kasmata, Arawe, and Cape Bushing. Trails were cut between them over the mountains and along the beaches so a stranded crew could walk to the next station. On May 6, 1943, the area army commander signed an order strengthening the sea transport base in the Cape Gloucester region and naming it as the supply base for all his forces in the Lae area. Garrison troops at these points were mainly shipping engineers, with a few medical men to look after troops passing through.
Distances were short by design. A crew left one hiding place after dark, ran for a set number of hours, and was under camouflage again before first light. Nobody crossed open water, and nobody kept a schedule an enemy could learn. In the Solomons, the same method ran barges down from Choiseul to the top of Vella Lavella, along the coast to the garrison at Vila, and up Kula Gulf into Bairoko.
It moved thousands of men and their rice while the American Navy was still looking for warships to fight. The standard craft was a steel box 47 feet long with a ramp at the forward end. Its Japanese name meant nothing more exotic than “large motorboat. ” It weighed about 21 tons, carried a 60-horsepower diesel, and made eight knots flat out with a following sea.
It could take 70 men, 10 tons of cargo, or one light tank. A crew of 12 worked it. Its draft was two feet seven inches, and that figure decided the shape of the next year and a half of fighting in those waters. Armament was whatever the unit could bolt down: two light machine guns at first, then 25mm cannon, then anything heavier that could be mounted, up to and including 40mm guns on the larger craft.
Steel plate went on around gun positions and along the sides, improvised in the field. A barge worked over by a unit that knew what it was doing was a small, low, armored gun platform that happened to be carrying rice, and it was almost invisible against a jungle shoreline at night. The thing that came hunting it was an 80-foot American motor boat with three aircraft engines: 12-cylinder Packards turning out 1,500 horsepower apiece, burning aviation gasoline, and carrying 3,000 gallons of it. The boat could make better than 40 knots with a clean bottom.
Everyone called them plywood boats, including their own crews. The United States Navy’s official history of the type says the hull was not plywood at all, but two layers of mahogany planking laid over laminated wood frames, with a sheet of aircraft fabric soaked in marine glue ironed between them. Plywood went into the cockpit and the joinery. The hull was a wooden shell built the way a fast racing motorboat is built.
What mattered was that bullets went through the American boat without stopping it. A wooden hole made by rifle fire does not flood the way a steel one does. Crews came home again and again with their boats punched full of holes and nobody aboard hurt. What could stop one was a hit in the fuel tanks, and when that happened, the entire boat went up in a single flash that lit the coastline for miles.
The American boat also carried four torpedoes, and that is where the arithmetic breaks. A torpedo runs at a set depth; six feet was about as shallow as the weapons of the day could be persuaded to go without running wild, and even then they were unreliable. The target drew two feet seven inches. The most powerful small warship for its size in the United States Navy, built around a torpedo battery meant for destroyers and cruisers, had arrived in waters where its main weapon could not be aimed at anything it was going to meet.
The Navy’s own history states the position without softening it: the barges were too small and too shallow to be worth a torpedo, and they carried heavier guns than the boats sent to hunt them. That second half took some weeks to believe. The first contact at Rendova came on July 21, 1943. On the night of the 26th, three American boats found six large barges and opened on them at short range with .
50 caliber machine guns and 20mm cannon. The fire bounced off the steel and away into the dark, and the six kept going. Then the American guns ran dry, all of them at once, and while the crews changed belts and drums, the barges turned toward them and opened fire. The American boats took a great many hits that night and were lucky to lose nobody.
Out of that engagement came the first American counter, which was procedural rather than material: keep the barge under fire without a break, never let all guns empty together, and have a second boat cover the first while it pulls away. Boat captains worked those three rules out in the Solomons in the last week of July 1943, and they owed every one of them to six enemy crews who had found the weak point first. The man who put the problem on paper was Lieutenant Commander Robert Kelly, who had brought a squadron of boats to Rendova at the end of June and taken over the boats at Lever Harbor on July 26. Late in August, he sat down and set out what his boats were up against.
Marine observers ashore at Enogai had counted 28 barges going in and out of Bairoko across six days. On three of those nights, his boats had attacked 17 of them and achieved almost nothing, because the return fire from the barges and shore batteries was so heavy that they could not get into a range where their guns would tell. Armored craft carrying 40mm guns were escorting the medium ones, which had nothing but machine guns and light cannon. To sink a single barge, Kelly wrote, a boat had to come in to inside 100 yards and put more than 1,000 rounds of .
50 caliber and 500 of 20mm into it. That meant lying almost stopped for about ten minutes inside the effective range of both the barges and the guns on the beach. Kelly’s phrase for what that amounted to was “sacrificing the boat. ”
Then he set out the part that turns the picture around.
Without illumination, his crews could not see barges hugging a dark shore. His own boats, though, showed up from the beach, silhouetted against the horizon in the moonlight. In the standard telling, the hunters are wooden and fast, and the prey is a slow steel box full of rice. In the paperwork of the men doing the hunting through the second half of 1943, it was the barge crews who saw the other side first.
Kelly’s conclusion was that his boats were the wrong instrument for the job. The practical answer, he wrote, was to build armored craft of America’s own and fight the enemy with his own kind of boat, or else put guns ashore along the barge lanes and shoot the traffic from the land. His superiors agreed with him only in part. The endorsement that went forward with the August action reports conceded that using motor torpedo boats as barge destroyers left a great deal to be desired, and that doing it in daylight or bright moonlight was expensive beyond what was acceptable.
It declined to accept that the boats were useless at it under all conditions, and it named the fix, which was neither of the two things Kelly had asked for. The boats would be given a heavier gun. Three of them were to have their torpedo tubes and depth charges taken out altogether, and the weight saved would go into an automatic cannon and armor. A torpedo boat carrying four torpedoes and their tubes has no margin left.
Take the torpedoes off, and the boat has enough spare tonnage for a real gun, and a real gun is what kills an armored barge. The gun the Navy settled on in the Solomons came out of an airplane. A single-shot anti-tank piece was tried first and turned down for its rate of fire. The automatic 37mm cannon fitted to Army fighter aircraft was then made available and went onto the forward deck as a standard installation.
Three boats in the Solomons had everything stripped off them except the heavy machine guns and took two 40mm mounts each, one forward and one aft, with four more twin . 50s besides. In New Guinea, that heavier gun became the rule, and boats carrying it had authority to go out with two torpedoes instead of four, or with none at all. By the end of the year, the fastest torpedo boat in the world was going to sea as a gunboat.
The other half of the answer came from the air. Patrol planes painted black were assigned to work with the boats at night. Their job was to find the barges, tell the boats where they were, and drop flares over them. A flare changes everything for a crew that has spent the whole war depending on not being seen.
It puts the barge in a lit circle on black water, with an armed boat somewhere outside it in the dark, and it takes the crew’s attention off their gunnery at the moment they need it most. The effect ashore showed up before the barge crews could have measured it themselves. In a letter to the American boat commanders, an Australian corps commander in New Guinea set out what his own intelligence people were seeing on the other side of the hill. Japanese diaries taken from the dead described a shortage of rations and daily working parties sent out to gather native food.
A prisoner stated that three days of rice now had to last nine. In one area, the enemy had killed and eaten his pack animals. Reinforcements had reached the rear areas and could not be pushed forward because there was no way to feed larger numbers at the front. The line about the artillery is the one worth sitting with.
In order to protect his barges, the corps commander wrote, the enemy had been forced to spread his field guns along more than 50 miles of coastline. Those guns had been brought to New Guinea to shell Australian infantry. They were now parked at intervals along a beach, pointed out to sea, waiting for wooden motorboats that might or might not come that night. None of this stopped the traffic.
Through the autumn and into the new year, the barges kept running, and the crews kept fighting back and kept scoring. Boats were hit by machine gun fire from the beach and from the barges. Boats grounded on unmapped reefs. Boats burned.
Every night the barge units filed their claims, and the claims went up the chain to the army headquarters that was keeping the books. In March 1944, the 17th Army wrote its monthly transport return for the Solomons. That document is the single best thing anyone has for what the barge campaign looked like from inside it. It states that a line of communications built on landing barges appears to offer great capacity, but that constant interference from enemy aircraft and enemy motorboats, with occasional attacks by destroyers working with aircraft that dropped flares, had pushed all traffic into the hours of darkness and by preference into moonless nights.
The report notes that the method also needed suitable landing spots and hiding places, and that the capacity available in theory was therefore very difficult to realize. It states that in spite of all of it, the transport had been pushed through anyway. Then it gives the two numbers the men writing it were using to judge their own situation. Since the first of February, they had sunk more than ten of the enemy’s motorboats.
In the same period, they had lost more than 20 barges, large and small, which the report describes as 90% of the total number of barges possessed by the army. The second number is correct. The first is wrong by a factor of ten. In February and March 1944, across the whole of the Solomons, the Americans lost a single motor torpedo boat to enemy action.
Here is where the surviving paper runs out. No Japanese barge captain wrote a memoir. No interrogation exists of a shipping engineer who ran that line and came through it. What survives is a diary taken off a dead man, an army logistics return picked up in the field, and the reports of the men on the other side of the gunfire.
Not one of those says what the crews themselves believed they were achieving. The comparison cuts one way as well. Wherever American ground troops moved into positions the Japanese had held, they kept finding wrecked barges on beaches and reefs that no boat crew had claimed. So the American figures were running low rather than high.
There is no reason to think the 17th Army was inflating its own losses. The report was right about its own side and wrong about the other one. A force that cannot count what it is fighting has no way of knowing when to stop. One boat had a number and a night.
She was the 251. On the night of February 25, 1944, she went down to Gazelle Harbor on the Bougainville coast in company with the 252. The stretch of shore they were working, from the Magine Islands past the mouths of the Jaba River and on to Motupena Point, was held by the Japanese along almost its whole length, and the barges ran it every night hard against the beach. Early in the evening, the two boats found one barge and sank it on the second firing run.
Rifle fire came out of the barge and machine gun fire came off the beach, and neither did any damage. At ten minutes to two in the morning, the boats made contact with the patrol plane working that sector and asked for light. Three barges appeared in the flares 300 yards off the beach south of the river. Both boats ran in and opened fire.
The plane strafed and dropped two bombs alongside, while mortars and machine guns on shore tried to push the boats back out. One barge drove itself up the beach. Two ran for open water. The boats came in to 75 yards and watched one go under, then had to break off because the guns on both boats had jammed.
Half an hour later, a mile and a half further south, they found six more barges in two groups running within 100 yards of the sand. They shot up all of them. A large barge came up astern of the 251 and was fired on with no result anybody could see. More flares went down.
Another barge was worked over between the flares and a signal light burning on the beach. Then the 252 turned on a barge that appeared 50 yards off her port side. At the end of that run, the barge was going down by the stern, and the 251 was hard aground on a reef. She could not be moved.
The 252 put herself on the same reef trying to help and managed to back off. Two more boats came down at quarter to five and spent over an hour feeling their way in with the sounding line to get a tow line across. The 252 had worked in to within 75 yards when a gun on shore started shooting at the stranded boat. The first round went over.
The second went under. The third hit her, and 3,000 gallons of aviation gasoline did the rest. Three men were pulled out of the water. When the light came, the boats and the aircraft went over the whole area and found nobody else.
Two officers and 11 men were gone. Back at base that morning, somebody found an unexploded 57mm shell sitting in the warhead of one of the other boats’ torpedoes. One more, and it cost nothing. On the 26th of February, 1944, the United States Navy lost one motor torpedo boat in the Solomon Islands.
The other patrol section that had worked over that night lay beached a short distance up the same reef, and the light came up on all three of them together.