by John Prados
Hara’s misgivings were well-founded. While it is perfectly true that, with the New Georgia campaign in full swing and the Japanese operational tempo well understood, SOPAC could expect a Tokyo Express, Halsey’s preparations were based on Ultra intercept of an Eighth Fleet dispatch, enabling him to set the ambush. As on the night of the PT-109 incident, when SOPAC had put cruisers off Vella Gulf (on the other side of Kolombangara from previous Express missions), with PTs down near Blackett Strait, this time Halsey posted PT boats in the same place, with destroyers right inside the gulf. He instructed Vice Admiral Theodore S. Wilkinson, his operational leader, to put the forces in motion. The preparatory order went to Captain Frederick Moosbrugger the preceding day. The Japanese destroyers, still at Rabaul, left at 5:00 a.m. on August 6. Moosbrugger had had nearly a day’s notice. In any case, the American captain raised steam by noon and led a half dozen destroyers from the anchorage. They were in position by 10:00 p.m. Moosbrugger delivered on his instructions.
The standard practice with Ultra was to post an air scout in the vicinity of a predicted surface naval movement. While an air spotter tipped the enemy that the Allies were aware of their presence, more important was that it gave the Japanese an explanation other than codebreaking for why their actions were anticipated. Sure enough, as Sugiura’s destroyers passed Buka Island at 6:30 p.m., they saw a snooper and then overheard its contact report. Sugiura pressed on without altering course or speed. By 9:00 p.m. the Express neared its goal. Captain Sugiura arrived in Vella Gulf at his appointed hour.
Moosbrugger had put his ships close off the Kolombangara shore, difficult to see against the dense jungle foliage. The last-quarter moon, obscured by clouds, was due to set near 10:30. The Southern Cross constellation was already dropping beneath the horizon. There were occasional squalls. This darkness completed the frustration of Japanese lookouts. At 11:18 destroyer Dunlap made radar contact. Captain Moosbrugger decided it was a ghost signal, but fifteen minutes later the same ship acquired a real contact, almost due north, twelve miles distant. Craven this time confirmed the contact. Moosbrugger immediately ordered his ships to set for torpedo attack. By 11:37 the tin cans knew there were four Imperial Navy warships. The launch took place at 11:43, with the Japanese at 6,500 yards, before a single gun had spoken. Commander Gelzer L. Sims of the Maury sent her fish against Sugiura’s Hagikaze. These were classic night-destroyer tactics, the kind the Japanese had so often employed against American flotillas whose own tin cans were usually restricted by conforming to a battle line of heavy ships.
Captain Sugiura saw everything as nominal. His Tokyo Express had settled on a southeast course at thirty knots. At 11:30 p.m. he altered to south-southeast. On the Kawakaze, the second ship, Petty Officer Tokugawa Yoshio, an ammunition hoist operator, was grabbing some shut-eye, in common with his comrades. Aboard the other vessels only Hara Tameichi had any inkling of danger. His Shigure, an older tin can in need of a refit, had lagged behind. Seeing nothing but forbidding darkness, rather than speeding up Hara ordered the Shigure to battle stations and doubled his lookouts. At 11:42 a spotter on the Hagikaze reported dark shapes along the Kolombangara coast. American torpedoes were already in the water.
After that, bedlam. A Shigure lookout saw torpedo wakes. Captain Hara had hardly gone on the long-wave radio to warn of torpedoes when lookouts on Arashi and Kawakaze reported enemy ships. Hagikaze and Arashi heeled to port and Kawakaze to starboard, but none could avoid the deadly tin fish. The first two were hit amidships. On the Kawakaze the crew barely made it to battle stations before torpedoes struck. Petty Officer Tokugawa believed the first one hit the bow. Sailors claimed a PT boat had delivered it. She quickly began sinking. The Japanese later established that seven torpedoes struck home, including one no one noticed, right through Shigure’s rudder, apparently so encrusted with barnacles that the holed rudder was only slightly less efficient at turning the ship. It was Hara who put up the fight for the Japanese, and Shigure survived because she had lagged behind. Had she matched the speed of Sugiura’s other ships, the Shigure would have been in the torpedo water. Hara turned away under a cloud of smoke and made for Rabaul, joining cruiser Sendai, returning from a supply run to Buin.
A number of Japanese survived tribulations as great as those of Jack Kennedy. Chief Petty Officer Kawabata Shigeo of Kawakaze swam fifteen hours before reaching land. Friendly natives gave him coconuts and young shoots to eat. Petty Officer Tokugawa drifted two days until the current took him to Vella Lavella, where he found more than two hundred survivors of the other ships. Seaman Kawahara Jihei drifted about twenty hours and cut himself badly on coral as he beached at Vella Lavella. The U.S. patrol that captured him was guided by a Melanesian tribesman.
The Battle of Vella Gulf, as this action is known, marked the onset of Japan’s dark period. Suddenly a draw seemed the best that Japanese forces could accomplish. Amid the succession of inconclusive actions and actual defeats, Kusaka’s position collapsed. Come to witness that sorry end would be Baron Tomioka Sadatoshi, the erstwhile NGS planner. After a time commissioning a new cruiser for the Imperial Navy, Tomioka arrived in Rabaul as a staff aide to Admiral Kusaka. Only a year earlier Tomioka had been debating the merits of South Pacific offensives versus an invasion of Australia. Now he had to assist his admiral in the desperate defense of a Japanese bastion, its power ebbing.
This transformation boggles the mind. Until very recently Halsey’s SOPAC forces had not exceeded the Japanese. Indeed, for roughly the first half of Tomioka’s year it was Japan that had been superior. Intelligence made the difference. Not that Japan lacked for resources in this field. The Japanese had aerial reconnaissance; they set up their own network of coastwatchers; the radio traffic analysts of the Owada Group and the communications units—like the 1st and 8th at Rabaul—were very good. But there were marked differences in the two sides’ capabilities. The Japanese simply never devoted the weight of effort to intelligence that the Allies—the United States, Great Britain, Australia—all did. There were many thousands of officers and men involved with Allied activities. On the Japanese side the number was a fraction of that. While hard data are lacking, a reasonable estimate would put their personnel at a tenth to a quarter the size of the Allied intelligence force.
The professional spook in the Imperial Navy lacked standing. So did the Allied pros, but on their side wartime events, starting with Midway, ended any confusion over the value of their work. During Tomioka’s year, intelligence proved so central to enabling meager Allied forces to trump the enemy that by mid-1943 it had become integral to the entire enterprise. The Japanese tolerated intelligence but regarded the product more as demonstrating the dimension of obstacles a commander must overcome to achieve victory. At root this was different from the Allied concept, in which intel identified targets; then operating forces blasted them.
Japanese intelligence nevertheless employed identical principles. Documents captured in the South Pacific show that the Japanese graded information for accuracy (“undoubtedly reliable,” “probably reliable,” “authenticity is undetermined”), collected topographical and other information from natives and friendly residents, had an actual propaganda strategy, exploited captured documents, recognized the value of prisoners as information sources, closely followed Allied radio news broadcasts, and, of course, valued radio intelligence and aerial reconnaissance. Japanese instructions placed special emphasis on this data: shifts in strength and movements of adversary air units, status and equipment of airfields, movements and status of warships and supply forces, the state of signaling and broadcasting, and adversary unit identifications. Japanese Army documents also emphasized data on enemy airborne raiding forces (MacArthur would conduct a parachute assault against Nadzab in New Guinea). Any Allied intelligence officer would recognize these collection targets immediately.
Much as IGHQ reached “central agreements” on operations, and local commanders negotiated parallel arrangements for their regions, the Imperial Navy and Army mad
e formal agreements that assigned primary collection responsibility in given sectors to one service or the other. In the Solomons the Navy took the lead on intelligence.
The Japanese were careful statisticians. Captured documents showed they closely tracked attacks on their air bases in an effort to divine overall patterns. Ground observers recorded the numbers and types of aircraft in an attack or over an area, as well as the hour, altitude, and general technique of attacks. Halsey was thus gaming the Japanese system when he flung a 150-plane attack at Kolombangara but made his next amphibious landing on Vella Lavella. Kusaka expected SOPAC to invade the former.
Given the intelligence juggernaut that fueled Allied success, it is perplexing that their system fell short when it came to Japanese pullbacks. The Guadalcanal evacuation owed much to Allied tardiness in divining the enemy’s true intentions. This happened again in the summer of 1943. The view from Rabaul was that supplying such exposed outposts as Munda and Vila had become too costly. Admiral Kusaka had also begun to suspect that Halsey would invade Bougainville—hence the sudden effort to build up Buin and surrounding bases, including Shortland, which Southeast Area Fleet believed a specific SOPAC target. Kusaka determined to regroup his garrisons, relinquishing Munda, which the Americans captured on August 4. He also ordered that troop movements be made primarily by barge.
What began as tactical maneuver soon became strategic necessity. Tokyo viewed the Solomons with increasingly jaundiced eyes. The Americans had just ejected Japan from the Aleutians—in the end without any intervention by Koga’s Combined Fleet. In various encounters during early August the emperor raked both Army and Navy chiefs over the coals, and complained to Prime Minister Tojo as well. The Allies had to be stopped somewhere. Imperial Headquarters initiated yet another strategic review. Planners decided that positions throughout the Pacific needed strengthening and the Solomons were draining capabilities. As an interim measure while Army-Navy discussions progressed, on August 13, Admiral Nagano issued NGS Directive No. 267, providing that the Solomons battle be waged by forces in place, which should withdraw to rear positions from late September. Meanwhile at the front, the tenor of operations quickly changed. Naval officers at Rabaul found themselves dispatched to convoy barges or to distract SOPAC while barges sneaked past the the Allies. General Sasaki and Admiral Ota made it to Kolombangara with many of their troops. The Americans reckoned they had eliminated about 2,400 Japanese in the seven-week Munda campaign. That represented a fraction of the Japanese force. It soon became clear that the Allies were not going to assault Kolombangara. On August 15 Halsey’s forces landed on Vella Lavella instead. The Japanese Army rejected any counterlanding. The Americans invaded more points on the island. Kusaka responded with air strikes. A transport was sunk off Guadalcanal and an LST at Vella Lavella, while a few other ships were damaged, but there was no halting SOPAC, which funneled 6,300 troops into Vella Lavella. Allied participation in combat reached a new level when New Zealand troops engaged there. The Japanese barges worked overtime to shuttle men to posts the Allies had yet to reach.
Bypassed, the 12,000-strong garrison at Kolombangara still needed recovery. To facilitate this, Kusaka decided to set up a barge station at Horaniu at the northeast end of Vella Lavella. Since the Allies on that island had stopped to form a perimeter and build an airfield, this remained possible. A small force loaded on barges at Rabaul. They were covered by a destroyer sortie. Rear Admiral Ijuin Matsuji of Destroyer Squadron 3 led the operation. Tall and gangly for a Japanese, with an optimistic disposition and wide-open eyes, the baron was a navigator of excellent reputation. Formerly master of the battleship Kongo, Ijuin had advocated reliance on barges. Intent on showing this would work, the baron gave his captains a free hand in making preparations. Only four destroyers could be used, but Admiral Ijuin picked the best tin cans at Rabaul, including Hara Tameichi’s Shigure, and the Hamakaze, equipped with radar. Ijuin’s force sailed before dawn on August 17.
American search planes discovered the baron’s ships in The Slot. Admiral Wilkinson detached a destroyer division from the Vella Lavella invasion flotilla’s screen and sent it after Ijuin in a high-speed chase. AIRSOLS also contributed a night attack by two flights of TBF Avengers. The torpedo planes failed to score, but they delayed Ijuin while Captain Thomas J. Ryan’s destroyers came up on him. A surface engagement took place around midnight. The Americans were silhouetted by a bright moon behind them. The fight was inconclusive. One Japanese destroyer, slightly damaged by near misses, suffered a few casualties; another sustained even lighter damage. Two of Captain Ryan’s tin cans had their prows battered by Japanese torpedoes. The barges sought refuge in a cove and continued to Horaniu the next night. Frustrations at home had reached such a point that when the emperor learned of the battle he erupted at Admiral Nagano, accusing the Navy’s destroyers of running away and leaving the Army troops to their fate. In its essentials, however, the mission had worked.
For weeks the barge chain continued regrouping Japanese forces. Tokyo Express runs supplemented them on certain key evacuations. More Army troops were dispatched by convoy to the Outer South Seas. The Japanese reinforced Rabaul with nearly a full infantry division. This place became a fortress in more than name. General Sugiyama conferred with Hirohito on the plans. Their September 11 conversation shows the depths to which the Empire had fallen. In a talk replete with references to the historic Emperor Meiji, Hirohito expressed himself openly. He would not “tolerate” another episode where the generals came back to report their soldiers had “fought bravely, then died of starvation.” Supplying Rabaul lay at the heart of the matter. Why defend the place at all? Sugiyama observed, “Rabaul is vital to the Navy and they have asked us to hold it somehow.” General and emperor explored the implications of the initiative while Hirohito also harped on New Guinea, another of his sore points. But everything came back to the fortress. “If we lose Rabaul,” Sugiyama admitted, “we will lose all mobility.”
So the redeployment proceeded. The 3,400 Japanese at Rekata Bay returned aboard destroyers. Barges safely removed Japanese coastwatcher posts on Santa Isabel and Gizo. Troop strength at Vila and Choiseul was thinned out. At least one more Tokyo Express ran to the big island. Tip Merrill’s cruisers made nightly forays up The Slot from September 12 in an effort to interrupt the barge traffic. One October night, Merrill thought he had nabbed a convoy and lit the sky with star shells to help his tin cans shoot—but results proved illusory. One officer concluded that the destroyers’ five-inch guns could not track fast enough to follow barge maneuvers at close range, while the fuses on 40mm cannon shells were so sensitive they detonated prematurely. More than fifty PT boats patrolled constantly and slugged it out with the barges, but often got as good as they gave. On one occasion they claimed sinking twenty bargeloads of the enemy. The eye-opener came when the PT force started modifying boats, removing torpedoes and rearming them as gunboats.
For Allied sailors and airmen it seemed Japanese determination made them unstoppable. Day after day they went back to the same targets, blowing them to hell, but the enemy always came back—in numbers and with guns. Frustration eroded morale to a degree. One day at Espíritu the men were given a boost by the visiting first lady of the United States, Eleanor Roosevelt. Rear Admiral Aaron S. “Tip” Merrill’s cruiser force happened to be there, revictualing. Sailors on liberty attended the rally, and others saw Roosevelt as she toured the base. James Fahey of the Montpelier noted that Mrs. Roosevelt was the first woman he had seen in nearly ten months. Lieutenant Commander Richard Milhous Nixon—another future president of the United States—saw Mrs. Roosevelt’s jeep convoy as it moved between stops that September day. Nixon was a staff officer with the rear area forces. He too recalled the morale boost, and decades later remembered the first lady’s visit as one of the most memorable moments of his war service.
The completion of the Vella Lavella airfield on September 24 strengthened Halsey’s vise but did not stop the Japanese withdrawal. A couple nights lat
er an I-boat narrowly missed light cruiser Columbia when she illuminated herself by opening up at a beach fire the Japanese had set. The major pullout from Kolombangara began on September 27. Wrapped up early the next month, only sixty-six men of the garrison were lost in the withdrawal. Admiral Halsey would claim that over a three-month period SOPAC forces sank 598 barges and seriously damaged another 670. These figures are hard to square with evident Japanese success.
Kusaka achieved his goals despite SOPAC intelligence, Halsey’s knowledge that the movements were under way, and the strenuous Allied efforts to blockade Kolombangara by sea and air. The SOPAC leader claimed three or four thousand Japanese gunned down or drowned in the barges alone, but only about 1,000 of the 15,000 troops on Kolombangara and the surrounding posts ended up in Halsey’s trap. The rest escaped. Captain Yamashiro Katsumori of PT-109 fame led the final Express mission to the island on the night of October 2–3. Had Japanese offensives been conducted as meticulously as their evacuations, the Allied Powers in the South Pacific might truly have been driven onto the ropes. Kusaka’s bombers raided Guadalcanal, the Russells, Munda, Vella Lavella, and once even as far as Espíritu Santo, but accomplished no more than harassment.
In Tokyo a broad strategic review adopted a fresh approach. Japan would defend a restricted inner perimeter and adopt hugely expanded war production goals. But the Navy and Army differed on details yet again—and neither held fast to the plan’s logic. Admiral Nagano plumped for the inner perimeter but, in accordance with Combined Fleet commander Koga’s battle zone concept, held that opportunities for decisive action must be sought outside the perimeter. General Sugiyama favored holding on to what Japan already had, to gain time to build up the new defenses. Both supported more than doubling the existing rate of aircraft manufacture despite the fact that raw materials imports were already significantly below 1942 levels—and promised to diminish further as more merchant ships were requisitioned for war service. Even assuming the expanded production, Admiral Nagano refused to assure success. These issues were aired at an imperial conference with Hirohito on September 30. The IGHQ directive that followed sanctioned the new strategy without comment on its lack of realism. Rabaul lay hundreds of miles beyond the approved perimeter.