Jungle, Hunger and Fever: The Army Takes Over and the Islanders’ War

Marines moving through thick jungle on Guadalcanal. Off the few trails, men cut their way through undergrowth with machetes and bayonets; the Army’s official history put the average march rate at about one mile an hour on jungle trails and half that off them. Signal Corps Archive from Ireland and United States; public domain, via Wikimedia Commons.

Guadalcanal · Chapter 8 of 9

Jungle, Hunger and Fever: The Army Takes Over and the Islanders’ War

From November 1942 to the end of January 1943 the campaign was decided as much by mosquitoes, rain and empty supply drums as by rifles, while thousands of Solomon Islanders scouted, carried and built for the Allies.

Nov 1942 – Jan 1943Lunga perimeter · Koli Point · Mount Austen · Kokumbona · Malaita~30 min read

By the first week of November 1942 the great crises of the Guadalcanal campaign had passed through the Lunga perimeter one after another: the landings, the disaster at Savo Island, the Tenaru, Edson’s Ridge, the battleship bombardment of 13–14 October and the Japanese Army’s failed October assault on Henderson Field. The naval battles of 12–15 November, told in Chapter 7, would shortly end Japan’s last serious attempt to reinforce the island in strength. What followed was less dramatic and in some ways harder: three months in which two exhausted armies faced each other across jungle ridges, and in which disease and hunger killed and disabled more men than bullets did.

This chapter covers that period: the island’s terrain and climate, daily life on the perimeter, the malaria epidemic, Koli Point and Lieutenant Colonel Evans F. Carlson’s “Long Patrol,” the Army’s takeover under Major General Alexander M. Patch and its fights on Mount Austen, the Gifu, the Galloping Horse and the Sea Horse. It then turns to the starvation of the Japanese 17th Army on what its soldiers called Gatō (餓島), “Starvation Island,” and to the Solomon Islanders, whose scouts, police, carriers and laborers were part of the campaign from first to last.

Times in this chapter are local Guadalcanal time (GMT+11) unless stated otherwise. Figures for sickness and deaths come from official histories written in 1949–1963 and from later scholarship; where they disagree, the competing numbers are given with their sources. The final evacuation of the Japanese garrison in February 1943 is told in Chapter 9.

The island itself: rain, kunai grass and rain forest

The U.S. Army’s official history, John Miller Jr.’s Guadalcanal: The First Offensive (1949), describes an island about ninety miles long and averaging more than twenty-five miles wide, with a forested mountain spine whose highest peaks rise to about 7,000 feet. Between the mountains and the sea on the north coast, from Aola Bay west to the Matanikau River, ran the strip of land that mattered to both armies: a flat, narrow plain lined along the beach with coconut plantations, broken by stretches of high, tough kunai grass, and crossed by rivers that Miller calls “generally deep and swift.” Behind the plain, the ground rose into grassy foothills and ridges and then into heavy tropical rain forest.

Miller called the Solomons “one of the world’s wettest areas.” Rainfall in some places exceeded 200 inches a year, and at Tulagi it averaged 164 inches annually between 1922 and 1942. The northwest monsoon, roughly November to March, brought almost daily rain; the dry season was dry only by comparison. Sea-level temperatures ran from 73 to 93 degrees Fahrenheit through each day, with high humidity. When the rains came, the rivers flooded, silt built up at their mouths, and the stagnant pools left behind became breeding grounds for Anopheles mosquitoes. The second half of the campaign, from November onward, was fought in the wet season.

The terrain shaped every operation that followed. Miller’s chapter on fighting on Guadalcanal reports that “most targets lay less than fifty yards from the infantrymen,” that commanders could seldom see or control their dispersed units, and that deep valleys blocked the signals of the lighter radios. The SCR-284, the most reliable infantry radio, weighed 110 pounds and took several men to carry. Beyond the road’s end, supplies moved by hand, by boat and by improvised cableway, and in the front line they were often thrown from foxhole to foxhole.

The Army even tried pack mules with the 97th Field Artillery Battalion, whose animals needed 1,500 pounds of oats and 2,600 pounds of hay hauled forward daily and managed only a mile an hour on the trails; Miller judged the experiment “unsuccessful.”

Life on the perimeter: food, flies, rot and no sleep

For the 1st Marine Division, which had landed on 7 August, the perimeter was home for four months. Seth Paridon’s account for the National WWII Museum describes how, after the transports withdrew following Savo Island, captured Japanese rice became the Marines’ main food until early November, with portions issued from the same pot day after day. Some men lost as much as forty pounds. Miller reports that the diet remained inadequate after the Army arrived: nearly every man lost weight, secondary anemia was common, and weakness caused many casualties on the march. The ration was C or K ration, a day’s C ration weighing “over five pounds”; fresh food was almost unknown apart from one Christmas shipment.

The old October perimeter enclosed less than thirty square miles, and Miller notes that it had no real rest areas: a man relieved from the line went back to ground still exposed to Japanese shelling and nightly bombing. There was not enough water for bathing or laundry, and men who bathed in the rivers picked up fungus infections—the “jungle rot” of so many memoirs. Flies, drawn by unburied enemy dead, gathered so thickly that, in Miller’s description, men eating in the open had to brush them off their food with one hand while eating with the other.

Sleep was the other casualty. Japanese ships shelled the airfield area at night, aircraft came over to drop a few bombs and keep the defenders awake, and patrols and infiltrators moved after dark. Miller records that “war neuroses” were diagnosed in 110 men during October, when bombardments were heaviest, and in only 13 in November, after the naval victory reduced the shelling. The diagnostic categories of 1942 under-recorded combat stress, so the numbers show a trend rather than a true count; the bombardment of 13–14 October (Chapter 6) left far more men shaken than wounded.

Work was constant. Supplies had to be lightered from ships to the beaches, trucked inland and dispersed into dumps against air attack, and there were never enough trucks. Miller writes that “combat troops as well as native laborers were forced to handle cargo,” a job for which the riflemen “showed a marked lack of enthusiasm”; Patch himself reported that combat troops were “apathetic toward labor.” In late November, Miller adds, 91 ships with 180,000 tons of cargo waited to be unloaded at the rear base at Noumea.

The perimeter itself changed shape through the campaign, from the narrow beachhead of August to the wide front of February. The slider below follows it month by month.

Malaria, dengue and dysentery: the enemy nobody planned for

The planners of Operation Watchtower knew Guadalcanal was malarious. Miller’s description of the island lists malaria and dengue among its expected hazards and predicts that fungus infections and sores would affect all troops. What the planners did not provide was the means to prevent it. Miller records that no malaria control personnel were allowed on Guadalcanal until mid-November 1942, that aerosol dispensers and insect repellent remained scarce for nearly a year, that quinine was scarce, and that mosquito control and “malaria discipline” were lax.

The results appear in the figures Miller drew from the history of U.S. Army Forces in the South Pacific Area. For units of all services on Guadalcanal, the malaria rate per 1,000 men per year rose from 14 in August to 1,664 in October and 1,781 in November, before falling to 972 in December and rising again to 1,169 in January 1943. A rate above 1,000 means that, if it continued for a year, the average man would be admitted more than once. The September figure is not given in the source.

For the 1st Marine Division the totals were worse than any battle. Miller reports that up to 10 December 1942, of 10,635 casualties in the division, only 1,472 resulted from gunshot wounds, while 5,749 were malaria cases; 3,283 Marines were hospitalized with malaria in November alone. The Marine Corps’ own official history, Hough, Ludwig and Shaw’s Pearl Harbor to Guadalcanal (1958), counts 8,580 Marines who fell ill with malaria or other tropical diseases and warns that these cannot be cleanly separated from other totals because many men were admitted more than once. When Navy doctors examined one regiment on 7–8 December, they found 34 percent unfit for combat-related duty—and that was after 400 fresh replacements had joined it. Front-line units, the same history records, worked by a rule of thumb that a man without a temperature above 103 degrees was not excused patrols.

The Army’s medical historians, writing later, reached a blunter conclusion. In the chapter “Statistics of Malaria” in the Army Medical Department’s Internal Medicine in World War II, volume II, Infectious Diseases, Fred H. Mowrey wrote that the 1st Marine Division “ceased to be an effective combat unit for many months” because of the hospitalization for malaria of approximately 80 percent of the command—a figure that includes the relapses that followed the division to Australia. The Paridon article puts the proportion another way: tropical diseases disabled nearly two-thirds of the division’s disabled men, enemy action one-third.

The Army units that followed fared little better. Miller reports that Army admissions for malaria averaged 420 per 1,000 men per year from 1 November 1942 to 13 February 1943, and that nearly 65 percent of ineffective Army troops were out of action from disease against about 25 percent wounded. Mowrey adds that the Americal arrived with few antimalarial measures and later saw weekly rates of 2,500 per 1,000 per year, and that about 46 percent of the 25th Division had at least one attack during 1943. Malaria kept coming after the fighting ended: Mowrey counts 10,206 new Army cases on the island from March to December 1943.

The suppressive drug was atabrine (quinacrine), taken as a daily tablet. Miller writes that atabrine treatment “had been inaugurated but had not halted the spread of the disease,” partly because “many men swallowed atabrine tablets reluctantly if at all.” He lists the beliefs behind the reluctance: that it was poisonous, that it caused sexual impotence, or that it stained the skin permanently—all false, though the drug did turn skin yellow while it was being taken. Early doses were also heavy enough to make some men sick, which fed the rumors.

Malaria was not the only fever. Dengue, spread by a different mosquito, struck without warning: Brigadier General William H. Rupertus, commanding at Koli Point, fell ill with dengue on 8 November and had to hand over to Brigadier General Edmund B. Sebree. Gastroenteritis struck nearly 500 Marines in August and September, though Miller notes only 12 cases in December once sanitation improved. Of the 2nd Raider Battalion’s 225 men evacuated sick during the Long Patrol, the Marine Corps’ Raider history lists malaria, dysentery and ringworm as the main causes.

Koli Point and the Long Patrol, November 1942

While the naval crisis of November built, the Marines fought east of the perimeter. Division headquarters expected a Japanese landing near Koli Point, and on 2 November Lieutenant Colonel Herman H. Hanneken’s 2nd Battalion, 7th Marines, took up positions near Tetere, east of the Metapona River. That night, about 2200, Japanese ships—Miller lists a cruiser, a transport and three destroyers—landed troops of the 230th Infantry at Gavaga Creek, about a mile from Hanneken. Miller gives their number as about 1,500, ordered by Hyakutake to join the survivors of Colonel Toshinari Shōji’s force retreating from the October battle and to begin an airfield. A footnote records the dissent of Major General Ito, who told postwar interrogators that the ships landed only supplies.

Hanneken’s radio failed; his report reached division headquarters only at 1445 on 3 November, by which time he had fallen back across the Nalimbiu. Lieutenant Colonel Lewis B. Puller’s 1st Battalion, 7th Marines, the Army’s 164th Infantry and naval gunfire followed. From 9 to 11 November the Japanese were trapped in a pocket at Gavaga Creek; Puller was wounded but stayed in command, and by 12 November the pocket was cleared. Miller counts about 40 Americans killed and 120 wounded, and about 450 Japanese dead, with fifty collapsible landing boats captured. Many of Shōji’s men had already slipped out to the south.

Those escaping Japanese were the quarry of the most unusual operation of the campaign. On 4 November two companies of the 2nd Raider Battalion landed at Aola Bay, about forty miles east of Lunga Point, with the 147th Infantry and Seabees sent to build an airfield there. Their commander, Lieutenant Colonel Evans F. Carlson, was ordered to march west and intercept Shōji. Major Jon T. Hoffman’s Marine Corps history From Makin to Bougainville: Marine Raiders in the Pacific War (1995) describes the column setting out on 6 November with Islander scouts leading. Among them was Sergeant Major Jacob Vouza, whom Hoffman calls “already a hero in the campaign” for his survival and warning before the Tenaru in August (see Chapter 4).

The patrol, followed day by day below, lasted until 4 December. After the first four days of canned food, rations were mostly tea, rice, raisins and bacon, landed by Higgins boats and carried inland by Islander porters. Holger Droessler’s 2026 study of Islander labor in Modern American History puts the number of scouts working with Carlson’s Raiders at nearly 100 at the peak, paid at first in food and tobacco. Miller credits the Raiders with about 150 miles of marching outside American lines and twelve separate actions.

The battalion reported 488 Japanese killed for 16 Marines dead and 18 wounded. Miller, writing in 1949, gives “over 400” enemy killed and 17 Raiders killed; the 488 figure is the battalion’s own claim, repeated by the Marine Corps histories, and like most kill counts in jungle fighting it could not be checked. The non-battle cost was larger: 225 Raiders were evacuated sick, and Hoffman notes that Companies C and F, which had landed with 133 officers and men each, came back into the perimeter with 57 between them.

The Army takes over: Patch, the Americal and the relief of the Marines

The U.S. Army had been on Guadalcanal since 13 October, when the 164th Infantry landed in time to fight in the October battle for the airfield. Its parent formation was unusual. The Americal Division had been assembled on New Caledonia in 1942 from National Guard regiments under Major General Alexander M. Patch; its name, suggested by a private first class, joined “American” and “New Caledonia.” The 182nd Infantry followed the 164th in November, and the 132nd Infantry landed on 8 December. On 11 December, Miller notes, the division counted 13,469 men and was more than 3,100 short of full strength.

Lieutenant General Millard F. Harmon, commanding U.S. Army Forces in the South Pacific Area, chose Patch to direct operations on the island, and on 9 December 1942 command ashore passed from Major General Alexander A. Vandegrift to Patch. Neither official history gives a time of day. Patch’s mission, as Miller puts it, was to “eliminate all Japanese forces” on Guadalcanal. The same day, three ships carrying the 5th Marines sailed for Australia; division headquarters, the 1st Marines and finally the 7th Marines followed through the rest of the month.

The 1st Marine Division’s combat cost from 7 August to December, as tabulated in the Marine official history, was 605 killed in action, 45 died of wounds, 31 missing and presumed dead, and 1,278 wounded—1,959 casualties from enemy action, a fraction of its losses to disease. Miller records more than 600 Marine dead and 691 dead among other U.S. units on the island by 10 December, and more than 2,100 sick and wounded already evacuated. Men who had arrived in August walked to the beach thinner, yellowed by atabrine and in many cases still carrying malaria parasites that would relapse in Australia.

Reinforcements came in fast. The 25th Infantry Division, a Regular Army division from Hawaii under Major General J. Lawton Collins, was sent directly to Guadalcanal without reloading at Noumea—a risk Harmon accepted. Its 35th Regimental Combat Team landed at Beach Red on 17 December, the 27th on 1 January 1943 and the 161st on 4 January. The 2nd Marine Division, whose 2nd and 8th Marines were already on the island, brought in its headquarters and the 6th Marines from New Zealand on 4 January. Its commander, Major General John Marston, stayed in New Zealand because he was senior to Patch.

On 2 January 1943 Harmon activated XIV Corps under Patch, with the Americal and 25th Divisions and the 2nd Marine Division attached; Sebree took over the Americal. The corps was a skeleton—on 1 February its headquarters still had only eleven officers and two enlisted men, and it had no corps artillery or service troops of its own—but by 7 January it controlled 50,078 men: 16,196 in the Americal, 12,629 in the 25th Division, 14,733 in the 2nd Marine Division, 1,607 in air units and 4,913 in naval units. Against them, Miller estimates from Japanese sources, stood about 25,000 Japanese, down from a peak of about 30,000 in November; American intelligence at the time put the figure between 9,100 and 16,000.

Mount Austen and the Gifu, December 1942 – January 1943

Mount Austen had been a named objective since before the August landings, and Harmon argued that Henderson Field would never be safe while the Japanese held it. Miller’s chapter on the December offensive describes it not as a single peak but as a spur of the main range jutting north between the Matanikau and Lunga Rivers, its 1,514-foot summit about six miles southwest of the airfield. Dense jungle covered the summit and its steep, rocky ridges; grassy open patches lay on the foothills; aerial photographs showed little of either. Hill 27, a rocky 920-foot mound, stood southwest of the summit; Hill 31, a grassy rise about 750 yards to its north, looked out over Lunga Point.

Patch planned to take Mount Austen first, then attack west with two divisions while a third held the airfields. On 16 December he ordered the 132nd Infantry, a regiment with no combat experience that had landed eight days earlier, to occupy the mountain. The Japanese line between Hills 31 and 27 was held by understrength battalions of the 124th and 228th Infantry and the 10th Mountain Artillery under Colonel Oka, men weakened, in Miller’s words, by battle weariness, malnutrition and disease, and supplied almost entirely by hand-carriers.

At its center was the position the Japanese named the Gifu, after the prefecture in Japan. Miller describes about forty-five interconnected pillboxes in a horseshoe-shaped line (Americal Division sources estimated seventy-five), built of logs dug into the ground and revetted with earth, roofs three logs thick and walls two, rising less than three feet above the ground and hidden under earth and foliage. Each held at least one, sometimes two machine guns and two or three riflemen; the garrison was about 500 men.

The 132nd’s 3rd Battalion moved up on 17 December. Two days later its commander, Lieutenant Colonel William C. Wright, was hit by machine-gun fire at about 0930 and bled to death shortly after noon; Lieutenant Colonel Louis L. Franco took over. The battalion spent the following week finding the Japanese line by running into it; on Christmas Day alone its three rifle companies lost twelve killed and sixteen wounded for no gain. Between 19 and 30 December the regiment’s two committed battalions lost 34 killed, 129 wounded and 19 missing, and 131 men were evacuated sick. By 28 December their combined effective strength was 1,541.

The 57th Engineer Battalion pushed a jeep track, Wright Road, up to Hill 35; beyond it, Miller writes, supplies went forward on the backs of Islander bearers whom the Americans called the “Cannibal Battalion,” a nickname that tells more about American attitudes than about the men who carried the loads. On 29 December a patrol found a clear route to Hill 27 that avoided the Gifu’s eastern bulge, and Harmon, Patch, Collins and Sebree agreed to keep attacking. Colonel Alexander M. George took command of the regiment on 1 January from Colonel LeRoy Nelson, who had asked to be relieved because of illness. On 2 January Lieutenant Colonel George F. Ferry’s 2nd Battalion took Hill 27 at about 0907 without firing a shot, then held it against six counterattacks that afternoon, losing eight killed and seventy wounded in forty minutes of the fighting.

In twenty-two days on Mount Austen the 132nd Infantry lost 112 killed, 268 wounded and 3 missing and estimated it had killed 400 to 500 Japanese. Malaria, dysentery and battle fatigue had worn it out, and the 2nd Battalion, 35th Infantry, of the 25th Division relieved it. The Gifu itself was still there.

XIV Corps’ January offensive: the Galloping Horse, the Sea Horse and Kokumbona

Patch’s plan for January, set out in a letter of 5 January, gave the 25th Division the inland flank: relieve the 132nd on Mount Austen and seize a line about 3,000 yards west of it. The 2nd Marine Division would attack west along the coast from the Point Cruz–Hill 66 line, keeping contact with the 25th Division’s right. The inland ground was dominated by two hill masses named, Miller explains, for their appearance in aerial photographs: the Galloping Horse, a complex of open, grass-covered hills numbered 50 to 57 that rose to about 900 feet and dominated the Point Cruz area, and the Sea Horse, Hills 43 and 44, about 1,500 yards northwest of Hill 27 across the Matanikau gorge.

The attack began at 0635 on 10 January. On the Galloping Horse, Colonel William A. McCulloch’s 27th Infantry made what Miller calls spectacular gains over open ground: the 1st Battalion took Hill 57 on the first day and the 3rd Battalion Hills 50, 51 and 52. The day was hot and sunny, and little water reached the forward troops. On 11 January the 3rd Battalion attacked Hill 53—the horse’s head—with only what remained in its canteens and was stopped by mortar and machine-gun fire. On 12 January the 2nd Battalion went in; Lieutenant Weldon Sims was killed on the ridge that took his name.

On 13 January Captain Charles W. Davis, the battalion’s executive officer, led a close-range rush against the Japanese machine-gun position on Sims Ridge that had pinned down part of the battalion, and the 2nd Battalion took Hill 53 and reached the division objective. A sudden cloudburst that day gave the men enough water to go on. Davis received the Medal of Honor under War Department General Orders No. 40 of 17 July 1943. The 2nd Battalion lost two officers and twenty-nine enlisted men killed between 10 and 13 January and estimated it had killed about 170 Japanese of the 228th and 230th Infantry. For January as a whole, the 27th Infantry lost 74 killed and 226 wounded, mostly on the Galloping Horse.

On the Sea Horse, Colonel Robert B. McClure’s 35th Infantry sent Lieutenant Colonel William J. Mullen Jr.’s 3rd Battalion on a wide envelopment cut through the jungle. Its K Company drove the Japanese off Hill 43 on 11 January, and the battalion reached Hill 44 by 1831. By nightfall the Gifu was surrounded on east and west. Supply was the 35th’s great difficulty: about 300 Islander bearers carried loads forward from the end of Wright Road but could not make the round trip in a day. B-17s dropped supplies—one released 7,000 pounds on 13 January—but Miller reports that only about 15 percent of the ammunition dropped was usable and nearly all the water cans burst. Ground supply resumed on 17 January when a supply boat line reached the foot of Hill 50. The 1st and 3rd Battalions reported 144 and 414 Japanese killed and 17 prisoners.

The Gifu was reduced in the week that followed. Captain John M. Burden of XIV Corps intelligence broadcast surrender appeals by loudspeaker on 15 and 17 January; no one came out. On the 17th forty-nine howitzers fired more than 1,700 rounds into the position, with less effect than hoped. Miller reports that Colonel Oka himself apparently left around 14 January and sent orders for the garrison to withdraw, but that Major Inagaki’s men stayed. On 22 January a Marine light tank, one of three made available the day before and the one that reached the crest of Mount Austen, broke into the Gifu line and destroyed several pillboxes. At 0230 on 23 January about a hundred Japanese led by Inagaki attacked the American lines; the Americans found eighty-five bodies afterward, his among them. Lieutenant Colonel Stanley R. Larsen’s 2nd Battalion, 35th Infantry, cleared the Gifu that day.

Counts of the Japanese dead on Mount Austen do not agree. The 2nd Battalion reported 518 killed in the Gifu; McClure reported about 1,100 across Mount Austen and the Sea Horse; the 25th Division’s operations report gives 431, excluding those killed by artillery. Most of the Gifu’s defenders, by the testimony of Japanese survivors summarized in Miller, were already sick and starving when the last attacks began.

On the coast, the 2nd Marine Division attacked west from the Hill 66–Point Cruz line on 13 January, the 8th Marines using tanks and a flamethrower on the 15th. By 18 January the Americans held a continuous line from Hill 53 to the sea. The 2nd Marines, in the Solomons since the August landings, then sailed for New Zealand.

XIV Corps’ second offensive opened on 22 January. A composite Army–Marine force of the 6th Marines, the 182nd and 147th Infantry and divisional artillery, later called the CAM Division, attacked along the coast while the 25th Division swung inland behind the Japanese. On 23 January two columns of the 27th Infantry’s 1st Battalion entered Kokumbona, the former Japanese headquarters area, and joined there by 1510; the 6th Marines took Hill 92 and destroyed three 150-mm guns. By 25–26 January the Americans had reached the Poha River, whose several channels their map showed as a single stream. Miller notes that the retreating Japanese made none of the night attacks that had marked earlier fighting—they no longer had the strength. Where supplies had to follow the troops into the hills, Islanders carried them: Miller records that bearers took food forward to a blocking force near Vurai and carried supplies from Hill 53 for the 161st Infantry.

Gatō: the starving island

Japanese soldiers abbreviated Guadalcanal as Ga-tō (ガ島), “Ga Island.” Written with a different character for ga—餓, meaning starvation—the same sound became 餓島, the island of starvation. The pun records what happened to the 17th Army after its supply lines were cut. Miller reports that by December rations had fallen to less than one-third of the regular daily allowance, and that front-line troops were “often reduced to eating coconuts, grass, roots, ferns, bamboo sprouts” and wild potatoes. Stealing food was common. Probably every Japanese soldier on the island had malaria. Miller also records “apparent instances” of cannibalism on Mount Austen; he gives no detail, and this chapter adds none.

The Imperial Navy tried to feed the army by destroyer. From 16 November to 9 February the “Tokyo Express” made about eleven supply runs and, by Miller’s count, lost ten destroyers sunk and nineteen damaged in the process (see the Tokyo Express record and Chapter 7). Destroyers dropped drums of rice and supplies overboard off the coast for troops ashore to haul in, rather than stopping to unload under air attack. More than 20,000 drums were delivered; less than 30 percent of their contents was recovered. On the night of 3 December, the Marine official history notes, about 1,500 drums were dropped and only about 300 recovered. Former officers told American interviewers that the army also lost all but about 10 percent of its ammunition.

Japanese testimony fills in what the figures mean. Sadao Suzuki, interviewed by the Japan Times in 2015 at age 97, landed on 27 August 1942 and left on a destroyer on 7 February 1943. He recalled that there were not even frogs, snakes or mice left to eat, that men killed one another over food, and that of the 1,885 soldiers who landed with him, 1,485 died, many of them from hunger. Miller records that a Japanese officer was reported to have sorted his men into three groups: those who could move and fight, those who could fight only from emplacements, and those who could not fight at all; and that when the Japanese retreated, medical staff apparently took only patients who could walk. Americans found unwounded dead at abandoned Japanese hospital sites.

The 17th Army’s commander, Lieutenant General Harukichi Hyakutake, and his staff stayed on the island until February. In Tokyo, Imperial General Headquarters’ senior leaders agreed on withdrawal on 26 December, and the Emperor formally approved it on 31 December 1942, according to the chronology drawn from Frank, Hayashi and other sources in the English Wikipedia article on Operation Ke (used here as a pointer). The same summary, citing Frank, reports that by early December the army was losing about fifty men a day to malnutrition, disease and Allied attacks, and that of roughly 30,000 army troops delivered to the island, about 20,000 were alive in December and only about 12,000 fit for combat.

How many Japanese died of hunger and disease rather than combat cannot be known exactly, and the estimates differ widely depending on what is counted and when. Miller, citing enemy figures, reports that of 21,500 Japanese casualties, 9,000 died of disease—malaria, malnutrition, beriberi and dysentery. The Marine Corps official history gives about 14,800 killed or missing and 9,000 dead of wounds and disease, without separating the two. The figure most often cited in Japanese summaries, presented on the Japanese-language Wikipedia as an estimate, is that of 31,404 army personnel landed, about 20,000 died, roughly 5,000 in combat and some 15,000 from starvation and disease. The Japan Times, citing an official estimate, reports about 20,000 dead, three-quarters from non-combat causes; the English Wikipedia summary, drawing on Frank, gives 19,200 dead including 8,500 killed in combat. What all the estimates share is that disease and hunger killed at least as many Japanese soldiers as American weapons did, and probably more.

Miller judged the Japanese soldier “a brave, resolute, and often skillful soldier” who nevertheless “could be soundly beaten.” What beat the 17th Army in the end was less any single battle than the sea and air blockade that left it on the far side of a supply line it could not hold. The evacuation that saved about 10,600 of its men is the subject of Chapter 9.

The Islanders’ war: scouts, police and carriers

The Solomon Islands were a British protectorate in 1942, and Guadalcanal had been home to its people for thousands of years before any of the armies arrived. When the Japanese occupied Tulagi in May 1942 and began the airfield on the Lunga plain in July, Islanders faced the war on their own land. Anna Annie Kwai’s Solomon Islanders in World War II: An Indigenous Perspective (ANU Press, 2017) puts its central argument plainly.

Islanders were not bystanders in the war but active participants.

— Anna Annie Kwai, Solomon Islanders in World War II: An Indigenous Perspective (ANU Press, 2017), ch. 2

The armed core was the British Solomon Islands Protectorate Defence Force and the protectorate police. Kwai records that Major Vivian Fox-Strangways trained 112 Islanders at Tanambogo before the government base was evacuated in May 1942, with old and insufficient weapons. Clive Moore’s entry on the war in the Solomon Islands Encyclopaedia gives the Defence Force a peak strength of about 800, including most of the prewar constabulary, and Droessler adds that 32 of its members died by war’s end and that 120 policemen served on Guadalcanal. Their work was scouting: slipping through Japanese lines, reporting troop movements, guiding American patrols and bringing back prisoners and documents.

Their organizer on Guadalcanal was Captain Martin Clemens, the protectorate’s district officer, who had stayed on the island after the Japanese arrived and walked into the Marine lines with his police in mid-August (see Chapter 3). Droessler, drawing on Clemens’s records, reports that by November Clemens had organized his scouts into four sections of about seventy-five men each, with sixty reserves and carriers per section; more than 300 scouts were at work by mid-November, none of whom had yet been paid. Payment began on 23 November 1942. Kwai records that when Clemens left Aola in May 1942 he put Andrew Langabaea in charge of the scouting network around it, and Moore names Goratara, Chak and Deke Gamu, former police, and the government employees Daniel Pule and Silas Sitai among the men who served and sheltered the coastwatchers.

Interviewed in the 1970s, Vouza summed up the Islanders’ part in words Droessler took as the title of his 2026 article.

But we all won that fight, and it was all because of what we did.

— Jacob Vouza, 1970s interview, quoted in Holger Droessler, “‘We All Won That Fight’: U.S. Empire and Military Labor in the Battle of Guadalcanal,” Modern American History (2026)

The Solomon Islands Labour Corps

As American forces grew, the demand for labor outran anything scouts and volunteers could supply. On 30 November 1942 the Resident Commissioner, William Marchant, approved the recruitment of a labor force, and the Solomon Islands Labour Corps was proclaimed. Moore records an initial strength of about 1,450—about 1,100 at Lunga on Guadalcanal and 350 at Tulagi—under Charles Widdy, with officers drawn from prewar plantation staff. Kwai describes the first recruits arriving from Malaita in groups of twenty-five. By 1944 the corps had enlisted 3,710 men, most from Malaita, southern Guadalcanal and San Cristobal.

The work was the work of a base: unloading ships, building and surfacing airfields, draining swamps and spraying ponds against malaria, carrying ammunition, food and stretchers to the front. Droessler reports that carriers hauled loads of about forty pounds and rarely had more than two rest days a month. On Mount Austen and the Sea Horse, as Miller’s figures for the 35th Infantry show, they were the supply line. They were also within reach of the enemy: on 26 January 1943 a Japanese bombing raid on the Lunga labor camp killed eleven Islanders and wounded nine, according to the official count, though the laborer Isaac Gafu later recalled about sixty deaths. More than seventeen Labour Corps recruits died during the war, Droessler reports.

Wages became a point of conflict almost at once. British officials, Droessler shows, wanted to hold pay to prewar plantation levels and feared postwar inflation; Widdy and other supervisors complained that the Americans, who paid and gave more freely, were “spoiling the natives.” On 9 January 1943, 130 Guadalcanal carriers struck for higher wages and their leader was arrested; in mid-March 1943 five of the corps’ six sections walked out over unpaid wages. Gafu remembered the other side of the experience: in a 1980s interview he said, “We really enjoyed working for the Americans.” Laborers traded for American clothing and goods and stored them in trade boxes to take home.

The Islanders also met African American servicemen in the service and labor units that arrived in the Solomons as the base expanded. Droessler notes that Islanders traded souvenirs for uniforms with Black soldiers and that the sight of Black Americans holding responsibility and respect undercut the racial order of the protectorate, even as some of those soldiers shared the period’s stereotypes about Pacific Islanders. Which specific Black units served on Guadalcanal during the campaign itself, as opposed to the base period that followed, is not established in the sources consulted for this chapter.

The coastwatchers’ Islander networks and the rescues

The coastwatching organization, code-named “Ferdinand” by the Royal Australian Navy’s Commander Eric Feldt after the peaceable bull of the children’s story, is described in Chapter 3 and on the coastwatchers special-units page. The Australian Department of Veterans’ Affairs’ Anzac Portal history counts more than 600 coastwatchers in the Pacific, of whom 38 died, and notes a detail that is easy to pass over: each station needed twelve to sixteen carriers, usually local men, to move its heavy teleradio, batteries, charging engine and fuel. Every famous European coastwatcher on Guadalcanal, Bougainville or New Georgia depended on Islanders who carried, cooked, scouted, guarded and kept silent.

On New Georgia, Captain Donald G. Kennedy of the Defence Force ran station ZGJ5 at Seghe (Segi), deep in the Japanese-held central Solomons. Kwai describes “Kennedy’s Army” of twenty-eight scouts led by Bill Bennett, an Islander born in the New Georgia area who also served as a wireless operator; Bennett claimed more than 100 Japanese killed and 82 captured without loss to his own men.

Rescues were where the Islanders’ role was most direct. Henry I. Shaw Jr.’s Marine Corps history First Offensive (1992) credits the coastwatching network with rescuing and sheltering 118 Allied pilots during the campaign. Kwai, citing Feldt’s The Coast Watchers, puts the total over the war at no fewer than 321 Allied airmen and 280 U.S. sailors rescued from behind enemy lines, and notes that “the bulk of the rescue efforts themselves were carried out by Solomon Islanders”—finding airmen and shipwrecked sailors and bringing them to the coastwatchers. The best-known rescue came later, in August 1943, when the scouts Biuku Gasa and Eroni Kumana found Lieutenant John F. Kennedy and the survivors of PT-109. Moore notes that Islanders also held prisoners and ran collection points for weapons.

Admiral William F. Halsey’s often-repeated tribute—quoted by Shaw as “the Coastwatchers saved Guadalcanal, and Guadalcanal saved the Pacific,” and by Droessler with “South Pacific”—is usually read as praise for a handful of European officers; Kwai and Droessler point out that most of the network’s members were Solomon Islanders, whose own accounts were long left out of the histories.

Villages, wages and Maasina Rule

The war also reached the villages. Moore records that protectorate and mission health services collapsed, food ran short, new diseases came with the foreign armies, and villagers were told to plant hidden gardens. Planters left their laborers to fend for themselves; more than 2,000 had to be shipped home. Droessler reports that the population of south Guadalcanal fell by 14 percent. Men left for paid labor and, Kwai writes, women took on new roles in village leadership; on Malaita, Kwara’ae women marched to government headquarters to demand the return of men recruited for the Labour Corps. Moore names only a few Islanders who collaborated with the occupiers and records that most sheltered the coastwatchers despite Japanese pressure.

The Labour Corps had a political afterlife. According to the Global Nonviolent Action Database at Swarthmore College, the movement that became Maasina Rule (“brotherhood rule”) began on Malaita in 1943 or 1944; its founders, Jonathan Fifi’i, Nori and Aliki Nono’oohimae, had all served in the Labour Corps. Its demands included Islander-led development and self-government, reform of the labor contract system and plantation wages raised toward the European rate. By 1946 it had spread to Guadalcanal and other islands.

The British response was arrest: 253 leaders and associates in 1947, 23 of them convicted, and thousands more in 1948–49 for refusing a census and a new tax. Droessler notes that many wartime Allied workers, Jacob Vouza among them, took leading roles. The movement dissolved after 1952, when the first island-wide council was established on Malaita. Moore judges the Islanders’ wartime experience “crucial for the development of Maasina Rule,” Solomon Islands became independent in 1978, and Maasina Rule is widely regarded as an early expression of the demand for self-government that led there. In 1968, Moore records, the British government paid $953,282 to the protectorate government to settle war-related claims.

Myths and the record

What the long middle of the campaign meant

Miller concluded that American tactical principles “were basically sound and sufficiently flexible” for the Solomons, but that they had to be learned on the ground: envelop rather than attack frontally, bypass strongpoints, dig in by mid-afternoon, keep water moving forward and enforce malaria discipline. The 1st Marine Division left Guadalcanal not defeated but medically broken; the Army divisions that finished the campaign carried its lessons into the rest of the Pacific war.

For the Japanese survivors, the name Gatō carried a memory less of battle than of hunger. For Solomon Islanders, the campaign began a longer change that ran through wages, wartime friendships, a political movement and eventually independence. The final chapter follows the evacuation, the campaign’s total cost and the ways all three peoples have remembered it.