Chapter 19
* * * * * * * * * * * * * * * * * * *

Final Victory in New Guinea

According to the JCS directive of 12 March 1944,* MacArthur's immediate duty following RECKLESS was to "conduct such operations along the New Guinea coast and such other operations as may be feasible in preparation for support of the Palau operation and the assault on Mindanao." To execute this mission MacArthur had been assigned most of the U.S. Army forces in the SOPAC, forces which would have to be transported to the SWPA and integrated into its command structure. Although Nimitz was supposed to assist SWPA as practicable, POA's operations in the Marianas meant that SWPA would have no fast carrier support and would be unable to borrow amphibious shipping to augment its own insufficient number of assault vessels. No exact timing had been indicated by the JCS for the remaining phase of the New Guinea campaign, but it had been stipulated that both SWPA and POA would be in positions permitting the invasion of southern Mindanao by MacArthur on 15 November 1944.

As planning matured, it became evident that SWPA would have to complete its New Guinea campaign by early August or sacrifice in a move into the northern Halmaheras the advantage to be had from the strategic carrier cover planned by POA in support of its mid-September invasion of the Palaus. For SWPA to be delayed and POA to progress on schedule could mean that MacArthur would not lead the troops liberating the Philippines, if indeed the JCS did not cancel the invasion of those islands. MacArthur had no intention of allowing such a failure. As he remarked to Kenney, he intended to be ready to enter the Philippines on schedule even if he was "down to one canoe paddled by Douglas MacArthur and supported by one Taylor cub."1

*See above, p. 573.


The victory at Hollandia, particularly the destruction of the Japanese Fourth Air Army, permitted SWPA to accelerate the phasing of New Guinea operations specified in RENO IV, but other than a working draft of a RENO V plan which was never officially released, SWPA did not undertake a formal revision of its over-all strategy until 15 June--too late to include any New Guinea operations other than the final invasion of a Vogelkop target area. On 8 May, however, in reply to a request from the JCS for information, MacArthur indicated that SWPA intended to invade Wakde Island about 21 May, Biak about 1 June, and some site suitable for an airfield on the coast of the Vogelkop about August. Tentatively, SWPA planned to seize an airfield site in the northern Halmaheras about 15 September. Except the invasion of Noemfoor Island on 2 July, this selection of targets and timing would be generally followed. The draft of RENO V included an additional operation following the old concept of a contingent invasion of the Kai-Tanimbar Islands if necessitated by Japanese air strength in the Celebes and Netherlands East Indies. Such a maneuver was favored by the Allied Air Forces, and on 30 May, Kenney submitted plans to CINCSWPA for Operation JURYMAST, an airborne invasion of Selaroe Island in the Tanimbars, urging that the plan be implemented on or about 1-15 July without regard to Allied progress on Biak. He hoped to make this operation--designed to build a fighter field on Selaroe--completely airborne and air supplied for the first fourteen days in order to explore the practicability of an airborne invasion of Mindanao. But GHQ considered itself fully committed to the advance along the New Guinea axis, and refused to permit any diversion of effort.2

SWPA thus committed its entire effort to an advance up the New Guinea coast along an exceedingly narrow front. Its four remaining operations in New Guinea would advance the land-based bombers by successive occupations of minimum air-base areas, selected in positions lightly held by the Japanese. Air power would prepare the way for each invasion and would protect SWPA's flanks, increasingly vulnerable as the attack moved northward. SWPA experience had demonstrated that air power could perform such a mission. The only question was whether the execution of four operations in as many months with the limited amount of amphibious shipping and engineering forces available would allow SWPA to reach the point of departure for the Philippines within the time allotted.



Nearly a month before the ground troops had gone ashore at Hollandia, Kenney and Whitehead had perceived from low-level aerial photos that the area around Lake Sentani would not support an air base of the magnitude contemplated, and looking ninety-six miles farther up the New Guinea coast, they had picked Wakde-Sarmi as the next most likely air-base area. Although GHQ had been unwilling to modify RECKLESS, it had on 10 April issued a warning order instructing all forces to prepare plans to attack Wakde-Sarmi on not less than three days' notice following the establishment of two fighter groups at Hollandia and other necessary preparations of force.* 3

As usual in the swiftly moving SWPA campaigns, initial thinking about a Wakde-Sarmi operation was based on "meager to nonexistent" topographical information. Geographically, the area of interest was a 25-mile-long coastal strip, backed by swamps and dense jungles from two to six miles inland, lying between Sarmi and Toem villages, together with the offshore islands of Wakde, Masi Masi, and Koemamba. Of especial interest were the two Wakde Islands: Insoemoar, a generally flat island about 9,000 feet long by 3,000 feet wide, containing approximately 540 acres, and Insoemanai, a smaller island to the southwest. Insoemoar, commonly called Wakde, had been occupied by the Japanese in June 1942. Within a year they had cleared enough of the coconut plantation covering the island to build an excellent, coral-surfaced runway. Late in 1943 they started to build up their garrison on the adjoining mainland, and they had constructed an airfield at Sawar and cleared another strip at Maffin. Of these two, only the Sawar strip was operational in the spring of 1944. The general area was under the command of the Japanese Second Army, with headquarters at Manokwari, which, in turn, was under the Second Area Army. At Sarmi village the newly arrived 36th Division had established its command post, and within the area the division controlled a garrison consisting of parts of two infantry regiments, a tank company, and some 2,400 base and line-of-communications troops. Although the 222d Infantry Regiment was at Biak, the division had an estimated 6,500 mobile combat troops under its immediate command. The Allies estimated that the Japanese had another concentration of approximately 13,225 troops, including 7,150 mobile combat effectives, around the

*See above, pp. 609-10.


Map: Wakde-Sarmi Area
Wakde-Sarmi Area


Second Army command post at Manokwari. Utilizing the 85 to 100 barges believed available, they could, under favorable circumstances, move a reinforcing regiment from Manokwari to Wakde-Sarmi within a week. Interference by major Japanese naval units was believed unlikely.4

The Japanese air situation was likely to remain favorable to the Allies for some time, at least during May. On about 20 April--as soon as the Allied intention to land in New Guinea rather than in the Palaus or Mindanao became clear to the enemy--the 23d Air Flotilla headquarters had advanced to Sorong, and its strength, disposed on Vogelkop airdromes, was increased to 180 planes by reinforcements from Malaya, Japan, and Truk. In May it would receive seventy additional army aircraft from Manila. Despite these reinforcements, the 23d Air Flotilla was no longer the formidable organization it had been when its planes had raided Clark Field on 8 December 1941. The squadron which it received from Japan, for example, had been re-formed after heavy losses at Kwajalein. Remnants of the Fourth Air Army headquarters had withdrawn from Manado in the Celebes and, operating through the 7th Air Division at Amboina, were attempting to harass the Allies with night raids staged through Biak, Noemfoor, Sarmi, and Samate. Allied commanders estimated that the Japanese would have 282 fighters and 246 bombers available for use against a Wakde-Sarmi landing, but this number included 348 planes in the Philippines and Palaus which probably would remain where they were. It was assumed that Allied air attacks prior to the invasion would have reduced Japanese strength within 100 miles of the beachhead to 36 fighters and 24 bombers. Recognizing Allied air superiority, the Japanese would probably limit themselves to night attacks.5

SWPA warning instructions directed ALAMO headquarters to submit a coordinated air, naval, and ground plan for the Wakde-Sarmi operation--designated collectively as STRAIGHTLINE--before 22 April. Krueger, considering the probable enemy dispositions, decided to use one regimental combat team against Wakde and to use the remainder of the division from which it was taken against Sarmi. He would have preferred to employ an idle division, but the limitations imposed by shipping and by staging areas made it imperative that he use one of the divisions forward at Hollandia. Consequently, he selected the battle-hardened 41st Division, and designated the 163d RCT for the landing on Wakde. ADVON Fifth Air Force drew up the basic


tactical air force plan for STRAIGHTLINE, predicating its readiness for the operation upon the establishment of two fighter groups and an attack bomber group at Hollandia and an RAAF fighter wing and two RAAF light bomber squadrons at Aitape. The Allied Naval Forces allocated assault shipping for STRAIGHTLINE through D plus 10, at which time USASOS vessels would undertake the mission. These individual plans were consolidated at a commanders' conference on 27 April, and on the same day GHQ issued its formal operations instruction.6

During the week following the issuance of these orders, low-level oblique photography revealed that the Sarmi area was most unsuited for airdrome development. Whitehead, reporting observations of 17th Reconnaissance Squadron pilots, informed Kenney on 3 May that the "Sarmi areaÉ is fuller of Nips and supplies than a mangy dog is with fleas." Two days later, Whitehead, recalling that SWPA had selected the objective areas at Gloucester, Saidor, and Hollandia without adequate reconnaissance and with adverse results, warned that the entire Maffin-Sawar area was nothing more than a "mud-hole." He strongly recommended that STRAIGHTLINE be limited to capture of a mainland area opposite Wakde and the island itself. He would move the 348th Fighter Group there, and staging two heavy groups through the Wakde area he would neutralize Biak sufficiently to make possible its capture early in June. From the rate of progress that the Japanese were making with their airfields there, Whitehead was sure that Biak offered the best possibilities for airdrome development between Nadzab and Mindanao. MacArthur tentatively accepted these recommendations without delay, and so informed Krueger on 6 May. He sent key staff members to Finschhafen to discuss the changed maneuver, and on 9 May the conferees agreed that it would be practicable to invade Wakde on 17 May (D-day) and Biak on 27 May (Z-day). The selection of this D-day, however, was predicated upon the establishment at Hollandia of the air units previously specified by ADVON, and Z-day upon establishment of a fighter group on Wakde and at least two B-25 squadrons at either Hollandia or Wakde.7

Biak's geographical location in the approximate center of Geelvink Bay, together with its Japanese-built airfields, justified plans for its capture, but its topography would make it a difficult ground campaign. Being little more than an uplifted portion of the shallow seas thereabouts, the island is predominantly a narrow coastal shelf. This shelf,


which occupies the southern two-thirds of the island, is somewhat broken up by roughly parallel limestone ridges and cliffs which rise like steps to an inland plateau. Within these ridges and cliffs are countless caves, many of which are interconnected by fissures and tunnels. Vegetation is primarily rain forest on the coast; inland, scrub predominates. Access to Biak from the ocean is made difficult by fringing coral reefs which border most of the coast line, and the coast is otherwise made inaccessible by cliffs. There are breaks in the reef at Korim Bay on the east coast, at Wardo Bay on the west coast, and off Bosnek village on the southeast shore of the island. On its northwest, Biak is separated by a narrow channel from the mountainous Soepiori Island; off its southeastern coast are the islets of the Padaido group, two of which--Owi Island and Mios Woendi Atoll--would prove of interest to the Allies. Specific information regarding Biak, as late as 7 May, was scarce. The first good photographs of the island, taken on 17 April, showed that the Japanese had made Bosnek village their main supply and bivouac area. This village was connected by a fair motor road, running westward along the coast and skirting cliffs near Parai, with the three airdromes which the Japanese had been hurriedly building during April. By early May they had completed Mokmer and Sorido dromes, while Borokoe, lying between the other two, was still under construction. ALAMO headquarters believed that the main defense force on Biak would be the 222d Infantry Regiment, but adding engineers, AA troops, and service organizations it estimated that there would be from 2,575 to 7,255 Japanese troops on Biak by Z-day. Barring an interdiction of water movement, the Japanese possessed ability to move in additional men from Noemfoor Island, 72 nautical miles distant, or from Manokwari, only 123 miles away across Geelvink Bay. Landbased air opposition to an Allied landing on Biak would probably be about the same as for Wakde-Sarmi, with possibly some intensification of night raids since Biak was ringed on the west by Japanese Vogelkop airfields. The Allies had also learned that the Japanese were moving to Tawitawi in the southern Philippines two carrier divisions which, after about Z plus 14, could attack Biak.8

To effect the change in objectives from STRAIGHTLINE to STICKATNAUGHT-HORLICKS (Wakde-Biak), GHQ on 10 May issued an amendment to its operations instruction. The final ground campaign plan was a modification of STRAIGHTLINE, substantially as indicated by MacArthur on 6 May. The task force for


TORNADO, consisting of the 163d RCT plus the 218th Field Artillery Battalion and supporting troops and commanded by Brig. Gen. Jens A. Doe, would land near Arara on the coast opposite Wakde, emplace its artillery for close support, and would transfer across the two miles of water to make a landing on Wakde on D plus 1. With utmost speed its engineers would prepare facilities for an initial fighter group on the island, and then they would expand the airdrome to accommodate an additional fighter group, a night fighter flight, and a long-range reconnaissance squadron. The task force for HURRICANE, commanded by Maj. Gen. Horace H. Fuller and comprising the remainder of the 41st Division and supporting troops, would land at Bosnek on Z-day, seize a beachhead, and occupy the three Biak airdromes. It would quickly rehabilitate one of them to accommodate an initial fighter group and, as soon as practicable, an additional fighter group, a night fighter squadron, a reconnaissance group, a photo reconnaissance squadron, and a heavy bomber group. Krueger would arrange to reinforce TORNADO with the 158th RCT, effective on about 23 May.9

The Allied Naval Forces, directed to furnish transportation and protection for the invasion troops and to move light naval forces into the target areas, allocated its transport mission to TF 77, to be commanded by Rear Adm. William M. Fechteler, deputy commander of the Seventh Amphibious Force. TF 74 and TF 75, the two Seventh Fleet cruiser forces, would furnish fire support and naval cover for the operations. Limited virtually to its own resources, TF 77 would load the Biak-scheduled troops of the 41st Division at Hollandia and the 163d RCT at Aitape, utilizing in part the amphibious craft which it had begun to concentrate forward on orders from GHQ.10

Most of the duties allotted to the Allied Air Forces for the operation were passed on to ADVON Fifth Air Force. ADVON, in turn, directed its fighter and bomber commands to push forward the 49th and 475th Fighter Groups and the 3d and 345th Bombardment Groups into Hollandia. This force, employed under the 310th Bombardment Wing (M), would provide local defense, offer fighter cover to forward missions, and extend direct support and cover to the beachheads and naval convoys. The V Bomber Command would employ its three heavy groups from D minus 5 to Z-day in concentrated attacks in the Wakde-Biak area. Using SB-24's, with 1,000-pound bombs, it would destroy enemy ground defenses on the coast at Toem just after first light on


P-39 Airacobras, Makin
P-39 Airacobras, Makin


Refueling P-38 Lightning, Middelburg
Refueling P-38 Lightning, Middelburg


Lookout on New Caledonia
Lookout on New Caledonia


Portable Radar, Noemfoor
Portable Radar, Noemfoor


Dispersal, Espiritu Santo
Dispersal, Espiritu Santo


Antiaircraft Emplacement, Sansapor
Antiaircraft Emplacement, Sansapor


But, New Guinea
But, New Guinea


Wotje Atoll
Wotje Atoll


D-day. On Z-day, V Bomber Command heavies would strike the airfields at Manokwari, Moemi, Ransiki, and on Noemfoor. After Z-day, all heavy bombers would be ready to support ground operations. The 38th Group would stage its B-25's to Merauke on 25 May and during the three following days would neutralize the Japanese airfield at Nabire. Shorter-range units would continue to strike Japanese opposition at Aitape and Wewak.11

The Thirteenth Air Task Force, flying its two heavy bomber groups from the Admiralties under operational control of ADVON, would harass Woleai and Truk enough to hamper repair and reinforcement, at the same time devoting a part of its efforts to Wakde, Sarmi, and Biak. Beginning on 1 June and continuing through 22 June, it would have to exert maximum effort against Truk and Woleai in support of POA's Marianas operations. This force would also send daily searches up the north coast of New Guinea, and prior to H-hour on Z-day it would soften ground defenses at Biak. On the west flank, the RAAF Command, in addition to routine operations, would neutralize hostile air bases in the Vogelkop and Geelvink Bay areas west of and including Noemfoor and Nabire.12

Tentatively the Allied Air Forces planned to develop Biak as a major air base for both tactical and service units, but the more pressing problem would be to establish the tactical air garrisons required for immediate support as quickly as facilities could be rehabilitated. The 308th Bombardment Wing (H) , commanded by Col. David W. ("Photo") Hutchison, would serve as the air task force headquarters for both operations. It would send an advanced echelon to Wakde to call forward the 348th and 8th Fighter Groups, a flight of the 418th Night Fighter Squadron, and a PB4Y squadron. The main body of the 308th would move to Biak and establish the 49th and 35th Fighter Groups, the 421st Night Fighter Squadron, the 25th Photo Reconnaissance Squadron, the 17th, 82d, and 110th Reconnaissance Squadrons, and the 43d Bombardment Group. When the 308th left Wakde, command there would pass to the 310th Bombardment Wing.13

Construction of air facilities at Wakde and Biak would be the function of the ground task forces, but for the first time in SWPA the air task force commander would be permitted to make minor changes in construction directives to suit terrain conditions. GHQ's comprehensive definition of facilities for Wakde, issued on 17 May, included an initial fighter field suitable for expansion into a medium bombardment


field with an extra-length, 7,000-foot runway, sixty-four heavy bomber hardstands, and a surfaced parking area sufficient to stage ninety-six heavy bombers. The fighter field with limited dispersal for a fighter group should be ready by D plus 2, the augmented medium bomber field by D plus 14, and the heavy bomber parking by D plus 16. On Biak, GHQ directed the construction of one fighter field suitable for expansion to medium bomber specifications with a 7,000-foot runway and fifty-two heavy bomber hardstands. Another regulation heavy bomber airfield with seventy-three heavy bomber hardstands would be built. One usable runway with limited dispersal for a fighter group should be operational by Z plus 2; the fighter airdrome with dispersal for two fighter groups by Z plus 10; one 7,000-foot runway at the heavy bomber airdrome, together with limited dispersal for a heavy bomber group, would be suitable for use by Z plus 20; and it was expected that the whole project would be completed by Z plus 60. To supervise the construction and to make such minor field modifications as were now permitted, Whitehead arranged to send his air engineer, Col. Leland H. Hewitt, on detached service with the 308th Bombardment Wing.14

Wakde, Sarmi, and Biak had been attacked prior to the RECKLESS landings, but the first large daylight strikes on the three areas began on 28 April. Forty-seven B-24's from the 43d and 90th Groups flew from Nadzab to bomb Mokmer drome on Biak. The 90th Group, arriving first, encountered twelve fighters, shot three of them down, and destroyed an additional ten planes on the ground. The 43d Group, meeting only one interceptor, was credited with three grounded aircraft. Before dawn that morning, twelve SB-24's of the 63d Bombardment Squadron had attacked Wakde, and at the same time that the Liberators struck Mokmer twenty-five B-25's of the 38th Bombardment Group bombed and strafed Wakde, destroying five grounded planes and setting a number of fires. Simultaneously, also, twenty-three Mitchells of the 345th Group bombed and strafed Sawar airstrip, scoring direct hits on four more Japanese planes. About ten minutes later, thirty-two Liberators from the 22d Group bombed the Japanese headquarters at Sarmi village. Amidst all this activity, twelve B-25's of the 17th Reconnaissance Squadron swept coastal targets between Sarmi and Sawar. None of the covering fighters met any opposition during the day and no Fifth Air Force planes were lost in the extensive activity. That night seventeen SB-24's from the 63d and 868th Squadrons teamed up to


bomb Wakde. While the heavies rested on 29 April, twenty-seven B-25's raided Sawar and Sarmi, six B-25's bombed and strafed Wakde AA, and late that night six of the 63d Squadron's snoopers harassed Wakde in cooperation with a naval bombardment of coastal targets. Once again the fighters had an uneventful day.15

This was a good beginning toward the neutralization of both targets, but the weather closed in along the north coast of New Guinea on 30 April and persisted until 13 May, barring all but a few missions into Wakde-Sarmi. Fortunately, the weather, which thwarted Markham River valley-based planes, did not seriously hinder the Liberators of the XIII Bomber Command, now based in the Admiralties, and on 4 May they took over the task of neutralizing Biak. The first three missions, flown by the 5th Bombardment Group with fighter escort, were hotly contested by Japanese fighters. On 5 May one Liberator was lost and another had to crash-land at Hollandia. On 7 May the 49th Fighter Group's P-40's from Hollandia escorted planes from the 5th and 307th Groups on a raid against Mokmer, and on this mission the heavies destroyed two interceptors while the P-40's claimed eight planes shot down. Thereafter the Japanese did not attempt to intercept the bombers, although sixteen to eighteen fighters attempted to attack an F-7 on 15 May over Biak, only to lose seven of their number to the 7th Fighter Squadron's P-40's. The XIII Bomber Command--skipping only 10 May, when it bombed Truk, and 15 May, when it raided Woleai--bombed Biak daily from 4 to 19 May. After 7 May these missions were uneventfully successful, except for one B-24 shot down by AA on 12 May.16

As the weather improved, the Fifth Air Force returned to the job. Heavies and mediums had a successful day against stores and personnel targets at the coastal villages of Sarmi, Arara, and Sawar on 7 May. Wakde was heavily attacked on 11 May. Weather did not permit any large-scale air activity again until 13 May, but beginning on this day and continuing through the three succeeding days, the heavies and mediums loosed a heavy bomb tonnage on Wakde-Sawar. On 17 May, which was D-day at Wakde, Whitehead shifted V Bomber Command's three heavy groups to attacks on Biak.17

Meanwhile, the 380th Bombardment Group, flying from the Darwin area, had been carrying out what its personnel, committed to twelve- to seventeen-hour unescorted flights since early March, considered to be their private war against the Japanese. The group had continued its


string of missions against Noemfoor through 23-24 April and then switched its efforts to the Vogelkop airfields. In addition, the group sent B-24's to reconnoiter the Halmaheras and northwest New Guinea, missions which were usually intercepted, although not always persistently. During May the group reconnoitered Geelvink Bay on the 4th; bombed Penfoei airdrome in Timor on the 6th; reconnoitered Geelvink Bay and bombed Utarom strip on the 7th; reconnoitered the Halmaheras and bombed Jefman Island, losing a B-24 to interceptors, on the 8th; reconnoitered the Vogelkop and bombed Kai Island fields on the 9th; bombed Laha and Namlea airfields on Amboina and Boeroe Islands, respectively, on the 11th; hit Namber and Moemi airfields on the 13th; searched for shipping in Geelvink Bay on the 15th; and on the 16th, covered by a squadron of the 475th Group's P-38's from Hollandia, bombed Kamiri airdrome on Noemfoor. Taking advantage of fires set by the British Eastern Fleet's carrier strikes on Soerabaja, seven 531st Squadron B-24's bombed the dock installations there shortly after midnight on 17/18 May.18

By nightfall on 16 May the 310th Bombardment Wing (M) had a garrison in place at Hollandia, which fact enabled the Allied Air Forces to inform GHQ that they were ready. That evening the naval vessels carrying the TORNADO landing force set out from Hollandia, joined the naval escort from the Admiralties during the night, and by early morning the convoy was standing off the mainland beaches. Direct aerial support for the landing had actually begun the preceding morning when the 38th Bombardment Group had undertaken to detonate possible mines in the waters off the coast and surrounding Insoemoar and Insoemanai. Neither the TORNADO force nor the Navy had desired any heavy pre-invasion air bombardment of the landing beaches, but Whitehead had nevertheless scheduled a small snooper mission to bomb the coast between Sawar and Sarmi, west of the invasion beaches, at first light on D-day. Fighter cover from Hollandia arrived at 0715, the precise minute that the naval bombardment lifted and the 163d Infantry began an unopposed landing. During the day, the 163d established a defensive perimeter and concentrated at Toem, while one reinforced rifle company landed on Insoemanai. Flights of the 8th Squadron's A-20's, orbiting in relays until called in by the local air control, bombed and strafed Wakde. The next morning, following an aerial and artillery barrage, elements of the 163d landed on Wakde, where for the first time they met intense rifle, machine-gun, and mortar


Map: Wakde Islands
Wakde Islands


fire from well-entrenched Japanese troops who seemed to have been but slightly bothered by the heavy naval bombardment which preceded the landing or by the tons of aerial bombs which had battered the island during April and May. About 800 Japanese troops had dug themselves in so completely that only direct bomb hits on dispersed entrenchments and hand-to-hand combat could destroy them. During the day, A-20's were directed on close support missions against enemy bunkers and strongpoints by an airborne B-25 "control ship," and after a sharp fight, all organized resistance ended on Wakde at approximately 1700 hours on 18 May.19

A better employment of the full strength of the Japanese 36th Division--10,000 combat troops against 10,500 Allied soldiers--might have made it a different story. The division commander within the week preceding the landing had shrewdly estimated the size of the U.S. invasion force and had anticipated a landing at Wakde by moving an infantry company there on 16 May. Otherwise, the Japanese commander had failed to prosecute his advantage. Believing that the landing could not come before 18-19 May, he had failed to get two other infantry companies moved to Wakde as planned. He had, moreover, prepared his defenses to withstand an invasion either between Sawar and Sarmi or directly on Wakde and was thus unprepared for a landing at Arara. His best opportunity--the chance to attack the 163d Infantry on 18 May when it was divided between the mainland and Wakde--could not be exploited because he had previously sent most of his 224th Infantry Regiment overland to attack Hollandia, and he was unwilling to give battle until these troops had returned. By that time the 158th RCT would have disembarked at Toem on 21 May and the situation would never again favor the Japanese.20

The 836th Engineer Aviation Battalion landed on Wakde on 18 May, but because of sniper fire it did not begin repairs on the captured airstrip until next day. The engineers had it operational on 21 May when a C-47, first buzzing the field to get the bulldozers out of the way, landed at 1310 hours. There had been little difficulty in filling enough bomb craters to make the runway operational, but preparation of facilities specified by GHQ was another matter. Charged with support for Biak and with initiating heavy bomber strikes on the Palaus in preparation for POA's Marianas operations, Kenney had to have base facilities at Wakde for a fighter group and a Navy PB4Y squadron and staging facilities for a B-25 reconnaissance squadron, an additional


fighter group, and two heavy bomber groups. GHQ directed Krueger and Whitehead to plan a program which would complete facilities by 27 July, and after studying the terrain, Hewitt and the TORNADO force engineer agreed that the only solution was to prepare a 6,500-foot strip and several large undispersed parking areas. The presence of all the indicated units at any one time would necessitate parking aircraft close together, thereby presenting a lucrative target to a Japanese night raider, but Whitehead, realizing that this could mean losses in both personnel and planes, agreed to take the risk. By 27 May the facilities were ready for use substantially as outlined by Hewitt.21 These facilities, so hurriedly completed, were imperfect (a dip near the center of the coral-surfaced runway caused a roller-coaster effect and became soft in heavy rains), but they would improve with more effort.

The ground echelon of the 348th Fighter Group, which had been awaiting a call forward at Hollandia, moved into Wakde on 22 May. Finding the island crammed with service troops and supplies, the 348th pitched its camp on Insoemanai and started a ferry across the 300 yards of water separating the islands. On 26 May the group received its P-47's. The air echelon of the 17th Reconnaissance Squadron, a unit which Whitehead wanted forward to patrol Biak at dawn and dusk, flew to Wakde on 25 May, leaving its ground crews behind at Finschhafen. On 26 May, seven PB4Y's of VB-115 flew to Wakde, and on the next day this squadron made the first regular air reconnaissance of southern Mindanao since early 1942. On 28 May the 421st Night Fighter Squadron sent a flight of P-70's forward to complete the Wakde air garrison. The 8th Fighter Group, displaced at Wakde by a decision to use its space for staging, was held for movement to Biak. Command at Wakde, assumed by a detachment of the 308th Bombardment Wing shortly after its arrival on 19 May, would pass to the 310th Bombardment Wing on 29 May, and early in June the 85th Air Defense Wing would arrive to exercise the local command function at Wakde under the 310th Wing.22

Wakde, though operational as an air base, would remain an unpleasant and hazardous locality for some time. The 303d Airdrome Squadron, one of the first units at Wakde, killed five Japanese officers and fifty-one Japanese soldiers in suicide charges on the nights of 19 and 20 May. Air Corps troops found the island too small for comfort, and the stench of Japanese corpses left unburied in the press of other


work, as well as the swarms of flies which bred on them, promised an epidemic of disease. Liberal use of DDT in air-spray solution--a new technique in SWPA--and other sanitary measures soon alleviated the hazard. The accumulation of supplies and especially of aviation gasoline and bombs--in the hurry to repair operating facilities the engineers had no time to build dispersed fuel and ordnance dumps--made it such a potential powder keg that on 5 June all engineers, other than those required to keep the runway usable, were ordered to the construction of dumps.23

In the circumstances, fighter defense was a matter of the most vital importance, especially in view of the slowness of the ground forces in building up an adequate gun and searchlight defense. As at Aitape, Detachment H, Fighter Wing landed on D-day with echelons of the 1st Fighter Control Squadron and Company A, 583d Signal AW Battalion. An assault fighter control center opened immediately, and on 23 May a temporary control center on Insoemanai replaced it. Company B, 574th Signal AW Battalion soon supplemented the radar network and on 24 May opened a GCI radar on Insoemanai. Early in June the 421st Night Fighter Squadron flew in five new P-61 night fighters to replace obsolescent P-70's. Not until 27 June were the permanent installations of the 32d Fighter Sector established. But once again the Japanese air forces had failed to exploit a golden opportunity; they made no air attacks against Wakde until 27 May, and during June, devoting most of their effort to less lucrative attacks on Biak, they sent only eight listless night attacks against Wakde. Even these small raids were particularly destructive; two raiders, making the most successful attack of the series on the night of 5/6 June, killed five men and wounded four others and destroyed six planes and damaged eighty others.24

While the Fifth Air Force was perfecting its Wakde establishment, the ground campaign continued on the mainland, supported initially by the 49th's P-40's from Hollandia. On 20 May these P-40's were credited with yeoman success in assisting the ground troops to break up a Japanese counterattack along the Tor River. After a few weeks of hard fighting the campaign was not pressed; and ALAMO headquarters was content to move the 6th and 31st Divisions through the area in succession, using them in patrols while they were awaiting employment at Sansapor and Morotai. Glide-bombing P-47's of the 348th Fighter Group from Wakde furnished most of the air support required during


the lagging campaign. Results of these strikes were hard to assess, but one notable success--destruction of bridges across the Orai River by the 341st Fighter Squadron on 12-14 June--was believed to have hastened the withdrawal of Japanese artillery in forward areas, thus lightening the work for American infantrymen. Virtual collapse of Japanese resistance by 20 July curtailed the need for air activity, but sporadic strikes continued even beyond the official termination of the operation on 2 September.25


Meanwhile, the Biak campaign had been mounted on schedule. Whitehead, joining his three heavy groups to the two of XIII Bomber Command, sent ninety-nine heavy bombers over Biak on 17 May for an hour of "excellent" to "superior" bombing of Bosnek, Sorido, and Mokmer. Each day thereafter, with the exception of 21 May when it raided Truk, XIII Bomber Command bombed Biak, accumulating a total of 185 sorties against that target between 4 May and Z-day.26 Fifth Air Force heavies also devoted most of their effort to Biak, but handicapped by adverse weather and the long trip from Nadzab, they had managed a total of only 139 sorties prior to Z-day. In addition, light bombers of the 8th Squadron made low-level strikes against Sorido drome and near-by barge shipping on 20 and 22 May, and on 23 May they attacked targets of opportunity west of Mokmer drome.27 All of these bombing missions were escorted by 49th Fighter Group planes. Bored by a lack of aerial opposition, fourteen P-38's of the 9th Squadron, returning from an abortive escort mission on 23 May, strafed targets on Biak, silencing an AA position near Mokmer, burning four barges offshore, and killing about fifteen enemy troops who tried to swim ashore.28

Neutralization of other Japanese airdromes in Geelvink Bay and on the Vogelkop, made doubly important by Japanese reinforcements, was the assigned function of the 380th Bombardment Group. On 19 May, thirteen B-24's of the 529th and 530th Squadrons flew from northwest Australia to bomb Manokwari airdrome. After a great amount of confusion in arranging the escort, Col. Donald R. Hutchinson finally sent out the 9th and 431st Squadrons from Hollandia to fly cover. The 380th's Liberators were intercepted by six Tojos, one of which was shot down by the bombers and four others of which were destroyed by the 9th Squadron's P-38's. The 431st Squadron met four


Japanese fighters and destroyed one of them. The 380th Group, carrying out its maximum effort in support of Biak, made two more strikes against Manokwari. On 21 May, one squadron made rendezvous with 433d Squadron P-38's to attack the primary target, but the other three Liberator squadrons, failing to meet their escort, attacked Utarom and Moemi airstrips. By another miscalculation on the 23d, the 9th Squadron's P-38's were dispatched to cover the 380th on a strike against Sagan, only to receive a much-delayed radio canceling the B-24 raid. On 26 May the 8th Squadron's P-38's joined twenty-one heavies from Australia for a successful strike on Manokwari. On 27 May, using two squadrons and ten B-24's, the 380th bombed Babo airdrome. The effort in general had been disappointing, for the group had managed only one really good strike, that against Manokwari on the 26th. Moreover, in view of the continuing confusion over fighter cover, Whitehead would soon pronounce such coordinated missions to be impossible.29

The 8th Bombardment Squadron, flying its fifteen A-20's from Hollandia, operated more satisfactorily. On 19 May, twelve of these A-20's, sweeping in over Manokwari harbor in an early morning raid, destroyed or badly damaged seven or eight vessels, including one of 1,000 to 1,500 tons and another of 2,000 tons. Still having some ammunition left, the A-20's continued to their secondary target, Kamiri drome on Noemfoor, where they destroyed at least four grounded planes, heavily damaged five others, and strafed about 100 Japanese workers on the strip. P-38's of the 432d and 433d Squadrons covered the light bombers, and the latter squadron shot down a Rufe floatplane. After being weathered out on 20 May, twelve A-20's virtually ignored Japanese gunners to make four bombing and strafing passes over Kamiri on the next day, destroying seven to eight grounded planes. The next two days the light bombers were unable to reach their primary targets and attacked Biak, but on 24 May they swept Namber and Kamiri dromes, destroying ten additional planes on the ground.30 These raids were singularly effective in reducing the Japanese air potential and in breaking up shipping, but the single light bomber squadron lacked the bomb-carrying capacity to crater Japanese airfields on Noemfoor.

ADVON, although unable to secure base facilities for the 345th Bombardment Group at Hollandia as it wished, was successful in staging the air echelon of the 38th Group to Merauke where, operating at extended range, its B-25's could attack some of the Vogelkop dromes. This movement was accomplished on 23-24 May, and it being


Schouten Islands


evident that Nabire was no longer a profitable target, the group was sent out to raid Babo on 26 May, only to find the area blanketed by weather and to be turned back to attack Dobo township in the Aroe Islands. The next day, Z-day at Biak, the group got through to the Vogelkop and, dividing into two-squadron echelons, attacked Babo and near-by Otawiri dromes at minimum altitude. The 71st and 822d Squadrons found only a few enemy planes at Babo but destroyed a Betty; the 823d and 405th Squadrons found Otawiri still under construction.31

The plan had been to use the heavy bombers of the Fifth Air Force against Vogelkop and Noemfoor airfields on Z-day, but Whitehead, anticipating strong enemy opposition to the landings, decided to employ them on Biak in supplement to the scheduled efforts of XIII Bomber Command. Ground and naval commanders had been somewhat reluctant to accept Whitehead's plan to use the Admiralty-based B-24's against the invasion beaches, but ALAMO headquarters, prompted by GHQ, finally authorized the attacks.32 As soon after first light as it was possible to identify the target on 27 May, twelve B-24's, drawn from the 63d and 868th Squadrons, flew in from Los Negros to bomb the Bosnek defenses. A second wave of twenty-five B-24's appeared between 0700 and 0704 and dropped 234 X 500-pound bombs along the beaches which, at the moment, were also under naval fire. XIII Bomber Command had scheduled forty-eight heavies for the mission, but one had crashed and exploded on the runway while taking off and had prevented the remaining planes from taking off. Another B-24 had crashed in the ocean en route.33 Between 1103 and 1150 hours, seventy-seven of the Fifth's Liberators bombed targets in the airdrome area.34

Including this effort, the Allied Air Forces in one month of sustained activity had dropped 2,260 tons of bombs on Biak defenses--899 tons by V Bomber Command and 1,361 tons by XIII Bomber Command. Study by a bomb-evaluation section of V Bomber Command and Japanese testimony soon revealed the extent of the damage caused by this aerial attack. Approximately 90 per cent of the Japanese supply dumps south of the airdromes and surrounding Bosnek and Mokmer villages had been destroyed. Bosnek village itself was practically devastated, and P/W's indicated that the main body of the 222d Infantry had evacuated Bosnek on 23 May because of the severity of the air strikes. Leaving only one company at Bosnek, the Japanese had


withdrawn to the airdrome area, despite their doctrine that "the enemy must be annihilated on the beaches." Coastal defense guns which might have imposed serious casualties on the Allied landing forces had been knocked out before they could fire. ALAMO credited the immediate pre-landing aerial bombardment with destruction of one 5-inch and two 3-inch naval dual-purpose guns located halfway between the two jetties at Bosnek, although the naval barrage which converged on the same area made definite assessment of the destroying agency difficult.35 "My hat is off to the Air Corps," signaled General Fuller after he had witnessed the virtually unopposed landing of his invasion troops.36

As Fuller's comment suggests, the establishment of a beachhead at Bosnek proved not to be very difficult. The first echelon of the 41st Division had departed Hollandia on the evening of 25 May. Naval covering forces joined the amphibious fleet next morning, and steaming directly toward Biak, with cover by fighter patrols from Hollandia and Wakde during the daylight hours of the 26th, the invasion fleet arrived off Bosnek on the early morning of 27 May. Following the naval and air bombardment, landing troops started ashore at 0715. Once ashore, the 186th Infantry quickly established a beachhead, and the 162d Infantry, second in the debarking column of regiments, pushed speedily toward Mokmer village.37 Air cover was instituted shortly after first light by four 17th Reconnaissance Squadron B-25's, but the next two successive fighter patrols were blocked off by weather and scheduled air cover was not in place again until 1100. Thereafter, fighters rotated on patrol, encountering no enemy airborne opposition (although friendly AA sent up occasional bursts of flak) until late in the afternoon.38 Four squadrons of the 3d Group's A-20's, orbiting off Biak, furnished direct support on call, and the fighters, as they were relieved from patrol, strafed targets specified by the ground controller.39 Except for one A-20 shot down by enemy AA with the loss of its gunner, there were no aerial casualties during Z-day.

First indications were that seizure of the Biak airdromes also would be easy, but the Japanese soon revealed that they, no less than Whitehead, considered the Biak operation to be "the decisive step on the road to the Philippines."40 Alerted by a search plane which had sighted the Allied convoys on Z minus 1, the 23d Air Flotilla had immediately started moving reinforcements in from Davao.41 The first Japanese airborne opposition, however, did not appear until 1630 on Z-day, when the 342d Squadron met eight enemy fighters flying in on the


deck about ten miles east of Bosnek. The patrol promptly shot down five of these planes but lost one P-47. Two hours later, five other Japanese planes, taking advantage of the temporary absence of Allied fighters, attempted a raid on the beachhead only to be shot down by AA.42 On the next day, the reinforced Japanese regiment at Biak, commanded by the exceptionally able Lt. Col. Naoyukie Kuzume, revealed that the Japanese meant to fight on the ground as well as in the air. Although ordered to defend the island at the water's edge, Kuzume, left little choice by the Allied bombing but to withdraw from Bosnek, had disposed the bulk of his force in the airdromes area and had prepared a perfect ambush in the cliffs high up over the Bosnek-Mokmer road. Permitting his advance guard to retreat before the 162d Infantry, he loosed a mortar barrage as the American troops were reaching Mokmer village and simultaneously erected a road block behind the Americans at Parai. Faced with counterattacks supported by tanks on 29 May, the 162d had to fight its way back through the road block and establish a defensive position near Ibdi village, halfway back to Bosnek. Kuzume then strengthened his defense at the Parai defile, completely blocking the easiest route to the airdromes. Fuller could only ask for the remainder of the 41st Division, especially the 163d RCT, so that he could undertake a difficult flanking attack up across the desiccated plateau back of the coastal shelf and envelop the Japanese positions by entering the airdromes from the northeast. At the same time, another regiment would attempt to reduce the enemy position at Parai defile and a third regiment would hold the beachhead. Krueger immediately ordered the 163d RCT dispatched from Wakde so as to arrive at Biak by 31 May and asked the Fifth Air Force to fly the 503d Parachute Infantry from Oro Bay to Hollandia to be ready for an emergency drop.43

Capture of the airdromes would obviously take time and would require more air support than the 310th Bombardment Wing could furnish. At Fuller's request, thirty-one Liberators, armed with 1,000-pound demos, were sent against AA positions at the east end of Mokmer strip on 29 May. They knocked out a four-piece, 120-mm. gun battery so emplaced that it could be used either as AA or to cover water approaches to the airdromes.44 Hearing about the Japanese tanks, Whitehead also sent six B-25H's of the 398th Bombardment Squadron up to Wakde that day in the hope that the 75-mm. nose cannons (so far a disappointment in combat) might finally prove useful. Finding the tanks already destroyed by the ground forces, the B-25H's hit gun


positions; but being forced to fly a steady course to sight their cannon, the B-25H's provided a good target and lost one of their number to AA.45 Light and medium bombers of the 3d Group and 17th Reconnaissance Squadron continued to give direct support on call, a business which was endangered by friendly AA gunners made "trigger happy" by Japanese activity. On 28 May the gunners had shot down a B-25 which had been cleared to drop pictures on the beachhead.46 Bad flying weather over Biak turned back all missions on 30 May but also gave the ground forces respite from hostile air attacks until late that evening.

Having received its reinforcements and regrouped, the HURRICANE force began its offensive on 1 June, sending the 186th Regiment up over the plateau and moving the 162d toward Parai. By 7 June the 186th had broken through the coral cliffs northeast of Mokmer and had captured the strip; its position, however, was precarious from lack of supplies, general exhaustion, and harassing Japanese artillery fire from ridge positions back of the strip. Having made amphibious landings west of the Parai defile, forces of the 162d RCT pushed through the Japanese barrier from both sides and opened the Mokmer-Bosnek road on 9 June. The Fifth Air Force had wanted to use six squadrons of B-24's a day to blast a path for the ground forces on Biak, but weather and more pressing targets prevented it. One mission of thirty-six Liberators bombed Mokmer artillery positions on 1 June, but the formation had so much difficulty finding a poorly identified target that one plane, running low on gas, crashed while attempting to land at Wakde. Weather prevented most of the heavy flights from reaching Biak on 3-4 June; from 4 to 7 June, they were largely committed to a futile search for Japanese naval units; and on 8 June, the Nadzab-based B-24's were staged to Hollandia for a mission to the Palaus. Such heavy strikes against Biak targets as were made proved only partially successful and were unduly expensive; weather-aborted missions used up effort and, in addition, the heavy missions from Nadzab had to arrive over Biak about noon, a time at which the Japanese soon prudently took cover.47 When it became necessary to commit the heavies to operations over the Palaus, ADVON supplemented the strength of the 310th Bombardment Wing by staging B-25 missions from Nadzab through Hollandia to Biak.48

While General Fuller's forces worked toward the airdromes, the Japanese Combined Fleet undertook to move 2,500 troops of the 2d Amphibious Brigade from Zamboanga to Biak. An escort of


battleships and cruisers sailed from Davao on 2 June, but alarmed by submarine and search-plane sightings and erroneous reports of a powerful Allied naval force northwest of New Guinea, the battleships soon turned back. Two cruisers, six destroyers, and the troop transports continued to Sorong, where after some debate, the Japanese decided to use the six destroyers for a fast run into Biak. Three of the destroyers were loaded with 200 men each, and covered by the three other destroyers they left Sorong early on 8 June.49 Special intelligence and sightings had kept Allied forces well posted. While TF 74 and TF 75 went on alert, Whitehead on 4 June rushed a large concentration of aircraft to Wakde, but when planes of the 63d Bombardment Squadron located the main body of the enemy fleet steaming back north that night and sank a troop transport, the alert was relaxed. During the day a large formation of Japanese planes, sacrificing four of their number to P-47's, had holed the cruiser Nashville; after the alert, our naval forces moved back to safer waters at Hollandia.50 But continued reports, resulting from attacks by the 3d Group on Japanese barge concentrations at Manokwari, indicated that mischief was still afoot, and on 6 June, B-24's of the 380th Group located the Japanese cruisers hovering off Waigeo Island.51 The subsequent attack proved unsuccessful.

When the Japanese destroyers began their run on 8 June, luck seemed to be with them. The Allied Air Forces, committed to heavy bomber attacks on the Palaus prior to Nimitz' move into the Marianas, had released as much forward staging space as possible. Moreover, the 5/6 June night raid on Wakde had damaged a great number of Allied aircraft, and all that the 310th Bombardment Wing had available for attack was the 17th Reconnaissance Squadron. Covered by P-38's, ten of the 17th's B-25's met the Japanese destroyers near Amsterdam Island at 1250 hours on 8 June. The AAF pilots, believing that they had encountered two cruisers (actually oversized destroyers) and four destroyers, sank the troop-laden Harusame and damaged three other vessels in a blazing low-level attack. Three crews, including that of the squadron commander, Maj. William G. Tennille, Jr., were shot down, and the remaining B-25's were so badly damaged by AA that the whole squadron had to be sent back to Finschhafen to re-form. P-38's of the 432d Squadron shot down three of the small force of escorting Japanese fighters.52 The Japanese force turned back temporarily but soon resumed a course for Biak, arriving off Korim Bay at


2230 hours. Just as they were preparing to unload, they were attacked by Allied destroyers, which had rushed back from Hollandia, and then pursued to the vicinity of the Mapia Islands in a running gun fight without damage to either side. The Japanese made preparations for another attempt to run in reinforcements, this time with escort from the super-battleships Yamato and Musashi, but the American attack in the Marianas forced cancellation of all such efforts to aid Biak.53 With the possible exception of the night of 12 June, when the enemy seems to have slipped approximately 218 men in from Manokwari on luggers, probably via Noemfoor, the Japanese had been unable to reinforce Biak.54

Enemy air operations over Biak, stepped up somewhat during the period of attempted reinforcement, reached a total of twenty-two attacks by some seventy-one planes to the end of June.55 As is usual in military operations, when one phase of an effort is retarded, others are thereby delayed. Thus, the establishment of adequate air warning and control facilities on Biali was thrown out of kilter by the slow ground campaign. Z-day echelons had set up an assault fighter control center west of Bosnek without delay, but the Japanese counterattack forced the control units to retire to Bosnek. The small perimeter available around Bosnek village circumscribed the siting of radars, and not even the GCI station, which went on the air on 31 May, could operate with efficiency. As rapidly as Owi and Woendi Islands were taken, radar stations were moved there to broaden the base of the air warning system, and on 26 June the GCI, now moved from Bosnek to Tamao, began operation under more favorable circumstances. As soon as Mokmer drome was secured, V Fighter Command's new airborne control center was flown in to relieve the assault fighter control center at Bosnek. This improved control and warning system, aided by the arrival of a detachment of 421st Night Fighter Squadron P-61's (one of them trailed and shot down a Dinah bomber over Japen Island on 7 July to score the first P-61 victory in the SCVPA) , made Japanese raids against Biak-Owi increasingly hazardous, but the permanent installations of the 33d Fighter Sector, delayed by a shipping impasse, would not be operational until the middle of August.56

The Japanese had built up raiding air garrisons at Jefman and Samate airdromes, at the north tip of the Vogelkop near Sorong, and at Babo, Sagan, and Otawiri, on McCluer (Reraoe) Gulf. They could stage these planes and others from the Netherlands East Indies through


Manokwari, Ransiki, Moemi, Waren, Noemfoor, and Nabire to attack Biak. A naval petty officer, shot down over Biak on 2 June, had been flown to Moemi to pick up a plane which had been ferried down from Davao the day before. Initially, Whitehead had relied upon the 380th Bombardment Group and the stripped-down 38th Bombardment Group, flying from Merauke, to keep Japanese air strength in the Vogelkop beaten down. At his request the 380th had flown to Jefman on 28 May, only to be weathered out; another scheduled strike on Jefman had been canceled when the 380th on 6 June attacked the Japanese cruisers; and by 7 June the enemy air strength in the area had so increased that Kenney did not care to send the unescorted 380th there again. The 38th Group efforts against Babo had been only moderately successful; it managed the long flights there on 27, 29, and 31 May, but during June, weather prevented any more missions to Babo. Engine failures caused by long cruising on lean gas mixtures, the expense of maintaining the Mitchells at Merauke by air resupply, and ordnance difficulties reduced the effectiveness of the 38th Group. Babo, however, was too close to Allied bases at Hollandia to continue long as a major Japanese base. On 3 June, P-38's, supposed to cover a weather-thwarted B-25 strike, virtually eliminated the Japanese air complement at Babo, but lost their squadron commander, Lt. Col. David A. Campbell. As quickly as the situation on Biak permitted, Whitehead committed the 3d Bombardment Group almost entirely to neutralization strikes on Geelvink Bay fields, and beginning on 5 June the group also undertook a continued neutralization of Babo, so releasing the 38th Group from its stint at Merauke.57

The Japanese air effort against Biak was sharply curtailed after 13 June, when the 23d Air Flotilla's potential reinforcements started moving to the Central Pacific front. Eradication of the Japanese strength remaining at Jefman-Samate would be a difficult matter, involving the longest B-25 flights yet undertaken in SWPA and prospects of a vigorous interception. Kenney, however, considered the area to be an anchor position in the Japanese defense line, and ADVON at Nadzab was making plans to neutralize it as soon as the commitments to Nimitz and the installations to turret and bomb-bay tanks in the B-25's permitted.58

On 14 June these requisites had been accomplished, and the Fifth Air Force began staging the 38th and 345th Bombardment Groups, together with P-38's of the 8th Fighter Group, into Hollandia. The


latter group and the 475th Fighter Group moved their P-38's up to Wakde early on 16 June, refueled them, and took off to cover forty-one B-25's from Hollandia which reached Jefman drome at 1255. The Mitchells went in over the field in an exceptionally low line-abreast sweep, strafing and dropping 100-pound parafrags on Japanese fighters which were either trying to take off or were parked in ready positions. Leaving fires and billowing smoke so thick that the last pilots had to fly on instruments, the wing continued across the four-mile channel separating Jefman Island from Samate drome on the mainland and repeated its performance. Surprised Japanese defenses managed to shoot down only one plane. Crews of the 38th Group, comparing their activities to a day at a county fair, claimed eleven enemy planes shot down. The 345th Group, flying second, claimed only one Hamp. Photos showed that fourteen grounded planes had been destroyed or badly damaged. The covering P-38's claimed twenty-five other definite victories, at a loss of one pilot.59 Whitehead intended to use the B-25's against Babo on the following day, but they had reported so many lucrative shipping targets around Sorong that he sent thirty-five Mitchells and sixty-eight P-38's back there. On this day the 345th Group attacked the two seemingly deserted airfields, while the 38th Group claimed at least eight vessels sunk and several others set on fire. Lightning pilots saw only four airborne Japanese planes and shot down one of them. A few Japanese fighters evidently remained around Sorong for several days, because two of them shot down a 38th Group B-25 during a shipping sweep on 22 June, but finally admitting the defeat of all organized Japanese air strength in New Guinea, the 23d Air Flotilla headquarters retreated from Sorong to Amboina on 28 June.60

Back on Biak, General Fuller's troops had established tenuous control over Mokmer strip by 9 June, and aviation engineers began on the next day attempts to make it ready for our fighters. But it proved impossible to get heavy equipment over the road from Bosnek, and enemy fire forced suspension of work on the 13th. The infantry's attempt to clean up the Japanese pockets in the caves back of Mokmer rapidly exhausted the available forces, and even though Fuller put Air Corps troops into perimeter defenses at Bosnek, he lacked the strength to finish off the Japanese as rapidly as desired. As the delay in the rehabilitation of Mokmer ate up the schedule for the completion of the New Guinea campaign, MacArthur was growing restive. When Fuller, citing rumors of Japanese reinforcements and the fatigue of his men, asked


for an additional regiment on 13 June, Krueger arranged to move forward the 34th Infantry and directed Eichelberger to take command at Biak. Eichelberger assumed command on 15 June, and on the 19th launched an attack with the fresh 34th Infantry, which captured Boroltoe and Sorido dromes on the next day. The action also relieved Japanese pressure on Mokmer strip, and Eichelberger immediately informed Hutchison that he could move Air Corps units there as soon as he desired.61

ALAMO, anticipating the delay at Mokmer, had queried the HURRICANE force regarding alternate airdrome sites on 30 May, and by 8 June, Hewitt and the task force engineer had discovered a site for a regulation heavy bomber airfield on near-by Owi Island. With parts of two battalions of aviation engineers at work, stripping started at noon on 9 June, and work progressed so favorably that some of the P-38's returning from Sorong on 17 June were able to land. Colonel Hutchison considered the strip ready for two groups of fighters in an emergency, but preferring to give the engineers as much uninterrupted time as possible, he did not call forward the airplanes of the first two squadrons of the 8th Fighter Group until 21 and 23 June. The third squadron came in on 1 July. A detachment of the 421st Night Fighter Squadron and Airacobras--invaluable for close support--of the 82d Reconnaissance Squadron moved to Owi on 28 June. Mokmer received the air echelons of the 49th Fighter Group between 21 and 25 June, and by 2 July the rejuvenated 17th Reconnaissance Squadron had settled down there. The air garrison had now reached the minimum which Whitehead considered necessary before the Fifth Air Force could undertake support of the invasion of Noemfoor.62

Old-timers in the Fifth Air Force had long regaled new arrivals with tales of how "rough" it had been at Port Moresby, but the ground echelons of the first units moving into Biak calculated that they had an effective rebuttal of all such tall stories. An advanced detachment of the 308th Bombardment Wing headquarters had arrived at Bosnek on 30 May in time to experience the early bombings; it was fairly well inured to the local hardships by the time that the remainder of the headquarters arrived on 23 June. The ground echelon of the 49th Group had arrived on 5 June, and since space was limited it had been quartered in a campsite approximately 250 yards square. Except for furnishing labor to unload LST's, the group sat there in an idleness which the men, little knowing the precarious ground and shipping situation, considered


"purposeless, useless, and complete." Into this jammed camp area at about 0130 hours on 12 June, a single Japanese plane dropped four bombs with deadly effect, killing nineteen men and injuring twenty-nine. On 20 June the group moved to a new campsite near Mokmer drome and began preparations to receive its planes. Ground personnel of the 8th Fighter Group, having originally sailed for Wakde only to be held aboard ship for a month, finally debarked at Bosnek on 12-14 June and, primed with stories of Japanese atrocities, spent a wakeful series of nights. Uneasy guards, seeing faint lights moving along the ledges above the bivouac, alerted the whole camp on the night of 15/16 June, only to discover that the supposed Japanese infiltrators were fireflies. The group ruefully dubbed this incident the "Battle of Firefly Hill." On 18 June the group's personnel and equipment were ferried to Owi, where a permanent camp was pitched. Just as these first units were getting settled on Biak and Owi an epidemic of tropical fever--later diagnosed as mite-borne scrub typhus--broke out, and by 28 July there had been 202 cases of this debilitating and, in some cases, fatal disease among the Air Corps troops. Clearing and burning of mite-infested areas and rigidly enforced wearing of clothing impregnated in a dimethyl phthalate solution proved an effective prophylactic measure against new infections, but by this time the troops had begun to wonder whether there might not be some truth to the rumor that "Owi" in the native tongue meant "Island of Death."63

On Biak, as the 34th Infantry attacked westward, the 162d and 186th turned into the "West Caves," a maze of sump holes, caves, and interconnecting fissures in which Kuzume had centered the remnants of his regiment. The American infantry slowly liquidated this pocket, using gasoline and dynamite to flush the Japanese from their hiding places and calling occasionally for aid from the air forces, as when twelve B-25's of the 38th Group on 24 June made a "bull's-eye" attack with 1,000-pound bombs.64 The West Caves and other ridge pockets of resistance having been cleaned up by 29 June, Eichelberger returned to Hollandia. During early July the infantry localized remaining organized resistance in the Ibdi pocket, just north of Mokmer village, and following a heavy artillery barrage and Liberator bombardment, the 163d Infantry wiped out the pocket on 22 July. Patrols continued to hunt down bands of Japanese, badly demoralized and reduced to cannibalism, until the official termination of the Biak campaign on 20 August and even thereafter sporadically.65


While the Fifth Air Force had been able to locate the minimum air garrison needed on Biak-Owi to support the invasion of Noemfoor, the early ground delays would have a cumulatively delaying effect on the comprehensive base-development program and the air force deployment dependent upon it. The Navy had planned to start moving Liberty ships to Biak by Z plus 8, but delays in the capture of Mokmer strip, coupled with the vigorous Japanese air attacks, caused TF 77 to refuse to leave cargo vessels at Biak even for one daylight period. Few such large vessels could run to Biak until 21 June, and consequently some 30,000 to 40,000 tons of shipping had been backlogged by that time. This backlog would have to be sandwiched in with current resupply shipments and taken in over the beaches at Biak at a rate of about 2,000 tons a day, a rate which could not be appreciably increased until docks were built. ALAMO estimated that the backlog could not be worked off until late September.66

On 20 June, GHQ issued a revised construction authority for Biak. Four airdromes would be constructed: one regular heavy bomber drome with 180 hardstands on Owi; one heavy bomber drome (less a parallel runway) with 115 hardstands at Mokmer; one heavy bomber drome (less a parallel runway) with 115 hardstands at Sorido; and one fighter drome with 94 hardstands or equivalent aprons for two troop carrier groups at Borokoe. This program was broken down into monthly completion objectives, and it was expected that it would be completed by 15 September, at which time construction of air depot facilities could begin. After a personal observation and conferences with HURRICANE and 308th Bombardment Wing engineers, Whitehead projected his own needs in relation to scheduled post-Noemfoor operations. By 15 July he needed one 7,000-foot runway and 136 hardstands at Owi and another 7,000-foot runway and 142 hardstands in the Mokmer-Sorido-Borokoe area. By 25 July he required two 7,000-foot runways and 189 hardstands at Owi and both a 7,000-foot and a 5,000-foot runway and 244 hardstands in the Mokmer-Sorido-Borokoe area. By 5 August he wanted an additional taxiway and 282 hardstands at Owi and an extension of the shorter runway in the Mokmer-Sorido-Borokoe area to 5,500 feet together with 238 hardstands in the area. He was quite optimistic that the engineers, by concentrating effort on three airdromes and leaving Sorido until last, could meet these target dates. The Allied Air Forces forwarded Whitehead's request to GHQ, which permitted the extension of Borokoe to 5,500 feet but rescinded


all previously announced target dates and left this matter up to the air forces and the construction agency.67

The engineers cooperated with Whitehead's program to the best of their ability, but a variety of factors, including some worn equipment and an average delay of three to four weeks in the arrival of engineer units, caused construction lag behind the GHQ schedule by as much as fourteen days. Mature investigation revealed that Sorido could not be developed into a heavy bomber field without almost prohibitive labor; and, after 4,000 feet of the strip had been repaired enough to permit troop carrier use, GHQ curtailed its development and expanded the other three airdromes to make up for the loss of hardstanding capacity at Sorido. Tactical facilities would be practically finished by 1 October, and at that date work could begin on the Biak air depot.68

The same shipping bottleneck which impeded construction hampered the movement of air units and supplies to Biak in preparation for invasion in the Vogelkop. Air movements, however, utilizing troop carrier C-47's and bombers, enabled the Fifth Air Force to get stripped-down units into Biak and Owi despite the shipping tangle. Between 11 and 20 July, for example, the 22d and 345th Bombardment Groups devoted almost all their effort to hauling cargoes from Nadzab to Owi. By air transport the flight echelons of the 25th Photo Reconnaissance Squadron, 82d Reconnaissance Squadron, 475th Fighter Group, 43d Bombardment Group, 345th Bombardment Group, part of the 22d Bombardment Group, and such service organizations as were suitable for air movement were moved forward during July.69 Additionally, P-40's of the RAAF 78 Wing were flown from Hollandia to Biak, where they were held in readiness to move to Noemfoor. To extend the SWPA search pattern, PB4Y's of VB-115 moved to Owi on 15 July, and Catalinas of VP-34 and VP-52, based on the tender Tangier, took up station in Woendi Atoll on 15-16 July. Catalinas of the 2d Emergency Rescue Squadron, just arriving from the United States, began operating from Mokmer late in July.70 During the last week in July, advanced echelons of the Fifth Air Force, V Bomber Command, and V Fighter Command headquarters arrived at Owi. The Fifth Air Force formally transferred its command post from Nadzab to Owi on 10 August, relieving the 308th Bombardment Wing for a period of badly needed rest at Hollandia.71

Movement forward of the air echelons of so many tactical organizations without their full complement of ground crews and maintenance


equipment met the immediate tactical situation but contributed to one of the most confused logistical situations ever encountered by the Fifth Air Force. Proper maintenance was impossible, and by 3 August some ninety-one aircraft were grounded at Biak-Owi for want of repairs, reducing, in effect, by one full tactical group the potential striking force based forward.72 This unfavorable condition, moreover, was not limited to Biak-Owi. The 90th Bombardment Group, to take but one example, had set up an air echelon at Wakde on 22 June, a part of the ground echelon was aboard ship at Hollandia awaiting movement to Biak during most of July, while the rest of the ground echelon was at Nadzab, the official group station.73 Concentration of these scattered echelons at Biak and Owi would be one of the gravest problems confronting the Fifth Air Force during August and early September.

The Far East Air Forces

During the three months following the JCS announcement of plans for liquidation of SOPAC forces, SWPA had been so busy with its New Guinea campaign that none but extemporized command arrangements had been practicable. These improvisations, however, suited a climate of opinion in SWPA that placed military operations ahead of the niceties of administration.

Since 1942 the ranking air headquarters in SWPA had been the Allied Air Forces, SWPA, which exercised operational control over the Fifth Air Force, RAAF Command and attached Netherlands East Indies units, and Aircraft Seventh Fleet. This headquarters, once heavily staffed with Australians, had become more and more an American body as the Fifth Air Force had dwarfed the RAAF Command in size, and by June 1944 the Allied Air Forces staff was, with the exception of the Directorate of Intelligence, the corresponding staff of Rear Echelon, Fifth Air Force. As Kenney was fond of saying, he and his staff merely had two hats, one "Allied Air Forces" for dealing with Allied or naval units and another "Fifth Air Force" for controlling AAF organizations. With the Thirteenth Air Force assigned to SWPA, it was no longer feasible for Rear Echelon, Fifth Air Force to serve in effect as the supreme air headquarters in SWPA and thus control another American air force. Kenney's first intention was to set up a new USAAF headquarters in SWPA--MacArthur wanted to call it the "First Air Army"--which would absorb the personnel and functions of Rear Echelon, Fifth Air Force and in effect the Allied Air Forces. Thus


the "First Air Army" staff would exercise administrative control over the Fifth and Thirteenth Air Forces and, in the name of the Allied Air Forces, could control operations of these air forces as well as those of the RAAF Command, Aircraft Seventh Fleet, and the composite Marine and Navy air command to be inherited from SOPAC. The new headquarters would establish and control directly the logistic and training functions common to both the Fifth and Thirteenth Air Forces. It would gradually undertake to control the operations of its AAF subordinates directly, because Kenney foresaw that the operational importance of the Allied Air Forces would decline as USAAF units redeployed from Europe and as operations progressed northward from the Australian base.74

Organization of the "First Air Army" would take care of the Thirteenth Air Force, but some other arrangement would be required under the Allied Air Forces framework to control Marine and Navy air units to be taken over when SWPA assumed responsibility for operations west of the 159th meridian. After intertheater conferences, SOPAC, utilizing the latitude permitted the Navy in the organization of task forces, established a new headquarters, designated Aircraft Northern Solomons (AIRNORSOLS), under Maj. Gen. Ralph J. Mitchell, USMC, with headquarters at Torokina. Effective with the organization of AIRNORSOLS on 15 June, the Allied Air Forces assumed operational control and directed Mitchell to support the operations of the U.S. Army XIV Corps along the New Ireland-Solomons axis. Mitchell was seemingly none too pleased by this mission against 70,000 by-passed Japanese, but Kenney assured him that he was not going to be "kicking a corpse around."75

The question of a new USAAF headquarters for SWPA was not of such easy solution. Kenney incorporated his ideas on the subject in a recommendation to CINCSWPA on 5 May, and SWPA, adding a request for an additional Army headquarters as well, passed the recommendation on to the War Department six days later. On 15 May the War Department readily granted the request for the additional Army headquarters but, pending study by AAF Headquarters, permitted the activation of a new air headquarters only on a provisional basis from personnel available in Brisbane. The AAF immediately objected to such an unusual name as "First Air Army," suggesting instead the name "USAAF in the Far East." AAF reaction to Kenney's request for 1,304 officers and men for the new headquarters was also skeptical,


especially since Kenney had earlier claimed that consolidation of two air forces would result in personnel savings and had just persuaded the AAF to authorize a fourth fighter squadron for the 348th Group on that assumption. The Troop Basis Division of AC/AS, Operations, Commitments, and Requirements first recommended that no additional personnel be authorized, but it soon relented and offered a standard air force headquarters and headquarters squadron, or an aggregate of 727 officers and men.76

After waiting as long as possible for Washington to approve a personnel authorization, USAFFE (the SWPA administrative headquarters) cut orders on 14 June 1944 announcing organization of a command designated as the Far East Air Forces (Provisional), effective the next day. The old Far East Air Force, as the predecessor of the Fifth Air Force, had fought in the Philippines early in the war,* and SWPA preferred a plural version of the same designation rather than that suggested by the AAF both for sentimental reasons and because an abbreviation of the suggested designation might be confused with USAFFE (U.S. Army Forces in the Far East). Kenney, utilizing personnel of the Brisbane headquarters, formed FEAF (P) on 15 June; it would continue as a provisional organization until 5 August, when receipt of an air force T/O&E permitted activation of a Headquarters and Headquarters Squadron, FEN. On 15 June, Whitehead and Streett assumed command of the Fifth and Thirteenth Air Forces, respectively, opening their command posts at Nadzab and Los Negros. They disbanded the old ADVON Fifth Air Force and the Thirteenth Air Task Force and absorbed the assigned missions. Since only part of the Thirteenth Air Force units were in the Admiralties, Streett opened a rear echelon of his headquarters at Guadalcanal to supervise rearward units, which, pending movement forward, would support ground operations in New Britain and assist COMAIRNORSOLS. FEAF (P), assigned for operational control to the Allied Air Forces, assumed command of the Fifth and Thirteenth Air Forces, following a pattern of command and control which was not fundamentally new in SWPA. Internally, the FEAF, Fifth Air Force, and Thirteenth Air Force headquarters were organized along conventional lines with the familiar A-1, A-2, A-3, etc., and special staff sections and would continue so organized for the remainder of the war period.77

This reorganization so far was little more than a paper transaction

*See Vol. I, p. 176 ff.


formalizing a command structure already existing, albeit with redesignations of headquarters. The only significant innovations were the command arrangements announced for control of logistic and training functions common to both the Fifth and Thirteenth Air Forces. The fourth echelon supply and maintenance function was assigned to the Far East Air Force Service Command (Provisional), organized on 15 June under command of Brig. Gen. Paul H. Prentiss. This command would continue provisional until regularly activated with an aggregate of 1,029 personnel on 18 August. Assigned to FEASC (P) were the Brisbane detachment of the V Air Force Service Command headquarters, the headquarters squadrons of the IV and V Air Service Area Commands, eight air depot groups and their subordinate units, aircraft assembly and overhaul squadrons, and miscellaneous supporting units. In brief, FEASC would receive aircraft and Air Corps technical supplies from the United States for modification, assembly, and delivery to the back doors of the fighting air forces. FEASC depots, assigned to the two air service area commands, would provide fourth echelon maintenance and supply for the combat air forces, and the duplicate area command headquarters--IV ASAC and V ASAC--would be located initially at Finschhafen and Townsville but would leapfrog forward to keep pace with the fighting. Both the V and XIII Air Force Service Command headquarters would be continued because Kenney considered them needed for logistical planning, for control of requisitions, distribution, and maintenance on combat bases, and to supervise and administer the third echelon functions of their assigned service groups. The staff of IV ASAC had been, in effect, functioning as the service command of ADVON Fifth Air Force at Nadzab; in the reorganization this headquarters was redesignated as the headquarters of V Air Force Service Command, and a new staff for IV ASAC was built up to take control at Finschhafen.78

Theater training and indoctrination of replacement aircrews became the responsibility of a new Far East Air Force Combat Replacement and Training Center (Provisional), organized on 15 June under command of Col. Carl A. Brandt with station at Port Moresby. To this FEAFCRTC (P) were assigned the 8th Service Group, the V Bomber and V Fighter Command Replacement Centers, and miscellaneous service units. Thereafter all incoming aircrews reported to Port Moresby, where they received a course of instruction and flew a series of indoctrination missions against by-passed Japanese positions. The


CRTC (P) moved to Nadzab on 4 September, where on 16 October it was reorganized on the basis of a newly activated service group, the 360th Service Group. This organization, functioning under policies established by FEAF's A-3, would train and distribute all incoming airmen until the end of the war. Reception and distribution of incoming ground replacements for the air forces remained the duty of the 91st Replacement Battalion (AAF), operating directly under the supervision of FEAF's A-1.79

The skeleton framework of the new organization was thus closely akin to the SWPA air organization which had grown up since 1942. Fundamentally, command relationships would continue to be inextricably connected with the personalities of the generals commanding. Kenney's prestige as the senior air commander in SWPA was high in all SWPA councils, and MacArthur, observing that he had "found that it takes an aviator to run aviators," left air matters within the theater generally to Kenney, who never forgot that the august CINCSWPA--Kenney always referred to MacArthur as "the General"--was in name and in fact the repository of all military authority in the SWPA. Kenney's air autonomy did not extend outside the theater, chiefly because no other theater commander was as jealous of his prerogatives as MacArthur, and communications between FEAF and AAF Headquarters in Washington, with the exception of personal correspondence between Kenney and Arnold, always passed through either GHQ or USAFFE, often receiving far more than routine attention. Thus early in July 1945, GHQ would refuse to pass officially to Kenney a request from Arnold that he be advised of FEAF's plans for supporting an invasion of Japan.80

Shortly after taking command of FEAF, Kenney informed Whitehead and Streett that his basic operational principle would be to insure flexibility in tactics, and he encouraged them to visit and correspond with each other and merely advise him of their decisions. While he planned that all Allied Air Forces entities would ordinarily operate in well-defined areas, their subordinate units might be switched about to meet tactical situations and each entity might cross its boundary lines if the commander concerned were properly notified. Since most of the original FEAF personnel had been members of the Fifth Air Force, there existed from the first a rapport between their headquarters which did not extend to the Thirteenth Air Force. The chiefs of staff of FEAF and of the Fifth Air Force kept up a steady exchange of


memoranda during most of 1944, settling many matters on an informal basis. Kenney, moreover, had a high admiration for Whitehead, whom he considered "an essential member of a winning team which is producing maximum results with a minimum cost of personnel and equipment."81 The two generals exchanged letters, often daily, when their command posts were separated.

The Fifth Air Force, because of its size rather than favoritism, would be the assault air force in most of the operations between June 1944 and August 1945, while the Thirteenth, smaller in size and a much tighter administrative unit than the sprawling Fifth Air Force, generally would be cast in a supporting role. In this respect, the Fifth Air Force would generally work closely with the ALAMO Force and its successor, the Sixth Army, and the Thirteenth would work with the new Eighth Army in consolidation operations. Predominance of Fifth Air Force-indoctrinated personnel on the FEAF staff provoked some complaints from the Thirteenth Air Force, but although Kenney would have preferred to draw staff officers from both air forces for the new headquarters, the Thirteenth had actually been so short of staff when it was transferred to SWPA (it had always depended upon the Navy for operational direction and intelligence) that Kenney was hard pressed to fill it up, much less take away its key staff members. The supporting role of the Thirteenth also occasioned "some chaffing on the part of the staff," and Streett in January 1945 went so far as to reveal a personal opinion that the Thirteenth should be inactivated and its units absorbed by the Fifth. Plans for the employment of FEAF against Japan made this suggestion impracticable, and not only would the Thirteenth be needed but the Seventh Air Force would be assigned to FEAF on 14 July 1945 to support the planned invasion of Japan.82

In general, FEAF would develop substantially as had been visualized, assuming more and more of the operational attributes of the Allied Air Forces until by September 1945 it was ready for the liquidation of the latter agency, a change planned to coincide with a redefinition of MacArthur's theater preparatory to the invasion of Japan but which actually would take place after V-J Day. This, however, will be a subject for later attention. To return to June 1944 and the New Guinea campaign, it may be observed that the reorganization had been accomplished smoothly and with no delay to the Allied war effort.



Neither RENO IV nor MacArthur's radio to the JCS on 8 May had contemplated the seizure of an objective in Geelvink Bay other than Biak, but by mid-May air planners were agitating their need for Noemfoor Island. Whitehead urged that such an operation could be undertaken as soon as the Fifth Air Force had two fighter groups operating at Biak, or about 15 June. By this time the RAAF engineers would be completing their work at Tadji and could be released to construct fields on Noemfoor needed for the 10 Operational Group and a few Fifth Air Force units. An air garrison on Noemfoor would facilitate fighter escort for bomber strikes on the Halmaheras, could maintain the neutralization of Vogelkop airfields, could break up Japanese efforts to reinforce Biak from Manokwari, and would also be of value in case the Japanese navy, observed to be effecting a concentration around Tawitawi, attempted to raid Biak.83

Geography and enemy efforts had fitted Noemfoor for exploitation. It lies near the northwestern limit of Geelvink Bay, eighty-one miles west of Biak and forty-five miles east of Manokwari. Roughly elliptical in shape upon axes of about fourteen miles north-south and thirteen miles east-west, Noemfoor's low coral-limestone hills are predominantly timber covered. The northern half of the island is low and flat. A fringe of coral reef almost completely surrounds the island, allowing access to only a few landing beaches. In November 1943, the Japanese, seeking to speed development of airfields, had impressed some 3,000 Javanese men, women, and children for labor on Noemfoor, where all but 403 of the mistreated slaves were to die before liberation. Three partly completed airdromes had been built: Kornasoren, on the north-central coast, with a partially cleared 5,000-foot strip; Kamiri, on the northwest coast near Kornasoren, with a 5,000-foot strip and seemingly extensive parking areas; and Namber, one-half mile inland in the southwest part of the island, with a 4,000-foot runway and limited dispersals. Other than a good path running around the perimeter of the island there was no well-developed system of communications.84

Allied prognostication as to the number of defending troops likely to be encountered on Noemfoor was hampered by the closeness of the island--two hours by destroyer or eight hours by barge--to Manokwari. On 21 June, ALAMO estimated that 2,850 enemy troops, including 1,600 combat troops, were then on Noemfoor, and that the Japanese


would be able to move an additional combat battalion there prior to an Allied landing. Estimates of air resistance for Noemfoor were about the same as for Biak. Within 800 miles the Japanese held at least forty-nine operational airdromes, and within 200 miles they had seven advanced staging fields. FEAF intelligence expected a maximum enemy air strength of 554 aircraft, predominantly fighters, to be within a radius of 600 miles of Noemfoor. The concentration of Japanese warships at Tawitawi was first believed a threat to Noemfoor, but before the target date these ships would have sortied for the Marianas. Whitehead, convinced that the Japanese had been reinforcing Noemfoor steadily, felt that they would "fight to beat hell."85

Map: Noemfoor Island
Noemfoor Island

GHQ planners had been in no hurry to commit themselves on a Noemfoor invasion, obviously preferring to await developments at Biak, but on 5 June, MacArthur indicated that it might be necessary to


use one regiment of infantry against Noemfoor in order to consolidate Biak. A preliminary GHQ plan contemplated use of this regiment and two engineer aviation battalions to seize and develop Kamiri drome for a fighter group, beginning on 22 June. The Allied Air Forces staff, while realizing that seven miles of road would be required to exploit Namber, urged that all three of the airdromes would be needed. The air forces were facing a strong air concentration, and their attacks on Amboina, Jefman, and Halmahera bases would need a minimum installation on Noemfoor to serve air units which otherwise would have to be less economically placed on Biak. GHQ accepted the concept of a Noemfoor operation designed to secure all airdromes and issued a warning instruction on 14 June, designating the 158th RCT, augmented with combat and service units, as the invading force. The target date would depend upon the establishment of fighter units at Biak.86

Discussions, including the highly controversial determination of a target date, now passed to ALAMO's Finschhafen headquarters. MacArthur had maintained that Noemfoor could be invaded between 22 and 25 June, but most of his staff seemed to believe it impossible before mid-July, a time which also jibed with Seventh Fleet thinking. No one, however, was particularly anxious to inform MacArthur that there would have to be delay. Kenney was exceedingly anxious to get two fighter groups to Biak in time to permit an invasion on 25 June--so anxious, indeed, that he would have been willing to see the attack launched with only one fighter group in place. A preliminary conference at Finschhafen on 16 June, however, estimated that Biak air facilities could not permit the invasion before 30 June, and at a conference between Krueger, Whitehead, and Fechteler on 20 June, it was decided that 2 July would be the best target date. By then the task force, scheduled to have completed concentration at Toem by 26 June, would have held a landing rehearsal, the engineers would have completed a parallel taxiway on Owi desired by Whitehead, and additional LCM's and LCT's would have moved forward. MacArthur immediately approved this target date.87

GHQ had already released its formal operations instructions on 17 June, and the naval, ground, and air plans for TABLETENNIS soon followed. The mission of the Allied Naval Forces was the usual transportation and supporting function.* Reef conditions off Kamiri drome,

*TF 74 and TF 75 would furnish cover and fire support, while TF 77, commanded by Admiral Fechteler, would embark troops at Toem and proceed so as to arrive off Noemfoor in time for H-hour (1800K) on 2 July.



By PBY, Hermit Island
By PBY, Hermit Island

By Submarine, off Los Negros
By Submarine, off Los Negros

P-47 Pilot, Noemfoor
P-47 Pilot, Noemfoor


Advance Strip, New Guinea
Advance Strip, New Guinea


Airtrain to Finschhafen
Airtrain to Finschhafen


Cargo Loading, Stirling Island
Cargo Loading, Stirling Island


Supply-Drop, Bougainville
Supply-Drop, Bougainville


Noemfoor: Paratroop Landing
Noemfoor: Paratroop Landing


the desired landing area, necessitated more than usual attention. Air photos showed no depth of water over the continuous reef about 450 yards offshore, but a team of ALAMO Scouts reconnoitered on the night of 22/23 June and found four feet of water in a pronounced break about 400 yards long off the southwest end of Kamiri strip. LCI's, LCT's, and LCM's might be able to get across the reef there (the ALAMO Scouts had been uncertain as to the state of the tide when they sounded), but LST's would have to off-load the assault troops and supplies into LVT's and DUKW's. This would take time, and the Japanese garrison, probably alerted by the party of scouts, might well concentrate their fire on the narrow boat lane and inflict Tarawa-like damages before the 158th RCT could get ashore. Thus it was vital that naval gunfire and aerial bombing paralyze the enemy defenses immediately prior to H-hour.88

ALAMO organized the CYCLONE Task Force under command of Brig. Gen. Edwin D. Patrick and charged it to land at Kamiri, seize the airdrome area, and subsequently occupy all of Noemfoor. The total combat force numbered only 7,415 men, and both Krueger and Whitehead, while realizing the shipping limitations, were skeptical that so few combat troops could accomplish the mission with any degree of speed. There being no really valid information as to enemy strength, Krueger committed the 503d Parachute Infantry Regiment at Hollandia as the task force reserve and alerted the 34th Infantry at Biak. CYCLONE engineers were expected to prepare initially one 5,000-foot runway capable of extension into a fighter-medium bomber field if terrain permitted, one 6,000-foot runway for expansion into a heavy bomber field, and one additional fighter-medium bomber field. Large undispersed parking aprons might be prepared initially, but eventually 280 hardstands were contemplated. Of these facilities, one usable runway with limited dispersal for 75 fighters and 8 night fighters would be ready by D plus 3; limited dispersals for 75 additional fighters by D plus 10;one additional runway and limited dispersals for 128 light bombers and 75 additional fighters by D plus 28; and the whole program would be completed by D plus 66.89

At Brisbane on 26 June, an ALAMO representative presented the combined TABLETENNIS plan to MacArthur's staff conference. MacArthur approved it, seemingly without his usual enthusiasm. Kenney's side comments were somewhat skeptical. "If it were not for the confidence that I have in your flattening the defenses before the


infantry gets in," he wrote Whitehead, "I would be willing to bet that the show would be a flop, but having a lot of faith in the thousand pound bomb and reading the continuous stream of reports of the Gloucesterizing going on I am not worried about it at all."90 TABLETENNIS was a small operation, but if the Japanese defenders remained able to fight they could make it costly both in men and time for the Allies.

Much of the aerial preparation for Noemfoor had already been accomplished in support of Biak. As early as 11 June, Whitehead had directed Hutchinson to use the 3d Bombardment Group against Manokwari in force approximately equal to that against Noemfoor so as to confuse the Japanese command as to the next Allied objective. He especially enjoined him to continue the "intelligent and aggressive" strikes against barge and lugger concentrations at Manokwari, both to reduce their ability to ship reinforcements to Biak and to deny them any opportunity to build up Noemfoor. During June, the 310th Wing claimed destruction of 107 Japanese vessels, mostly barges and luggers but including some twenty-four freighters. The 3d Group's "Grim Reapers," in their busiest operational month overseas, claimed 74 of the 107 vessels. With such a splendid record to indicate what A-20's up front could do, delay in getting a second A-20 group to Hollandia was doubly bitter to Whitehead.91

Similarly, the campaign against the Vogelkop airfields would assist an engagement at Noemfoor. Following the eradication of Japanese air units at Jefman and Samate,* the Fifth Air Force had only to keep the Vogelkop strips sufficiently cratered to prevent their use by sneak raiders. During the latter part of June, the 3d Group attacked Babo, Moemi, and Waren dromes, and the 90th Group's B-24's, staged into Wakde on 22 June, raided Jefman, Samate, Ransiki, and Moemi. The 38th Group, staging B-25's through Hollandia, also hit Manokwari and Ransiki on 26 and 27 June. Neutralization raids against Babo, designed to cover the heavy bombers at Wakde, were made by the 380th Group on 27 and 29 June. Since there were no Japanese interceptions over the Vogelkop after 22 June, fighter escorts strafed targets of opportunity for want of better to do. By 1 July such hostile air power as remained in New Guinea was generally debilitated. Jefman seemed abandoned, there was little activity at Samate, Babo's strips were cratered, Waren was unserviceable, and Moemi was used but sparingly. The Allied Air Forces estimated that there were approximately fifty-

*See above, pp. 639-41.


six enemy planes in northwestern New Guinea, of which probably no more than twelve were actually serviceable.92 Only the weather and the long distance from Nadzab hindered the neutralization of Noemfoor.

Sustained air attacks against the island began on 20 June with a four-squadron Liberator attack on Kamiri drome. Nadzab-based B-24's returned to the target next day, but bad weather, which would linger in the Markham valley for the remainder of the month, then closed in, and the 22d and 43d Groups were able to reach Noemfoor only on 25 and 26 June. Except for 23 June, when weather completely shrouded Noemfoor, the A-20's of the 3d Group and dive-bombing fighters from the 49th, 348th, and 475th Groups continued the assault; but these planes could not deliver the bomb tonnage necessary to saturate the beachhead defenses, and Whitehead, after moving his advanced command post to Hollandia on 28 June, called for Thirteenth Air Force support. On 30 June the 5th and 307th Groups, joined by the 90th Group at Hollandia and assisted by a miscellany of shorter-range forward units, delivered 159.5 tons of bombs to Noemfoor. The next day a break in Markham valley weather permitted all five FEAF heavy bombardment groups, plus the miscellaneous units, to drop 220.6 tons of bombs on the island. Between 20 June and 1 July, FEAF planes had thus deposited 801 tons of bombs on Noemfoor, mostly on the defenses in the Kamiri area. These missions proved uneventful, for the Japanese garrison elected to save its ammunition to withstand invasion, permitting the island to become an undefended target over which Allied aircraft were free to bomb and strafe from any level. Thus on 26 June a 403d Squadron formation, after five runs over Namber looking for an opening in the clouds, descended to 3,000 feet to bomb. After such experiences, one intelligence officer wondered whether a "milk-run" to Noemfoor was actually "an engagement with the enemy."93

The naval convoys, covered by Wakde and Biak fighters, had begun leaving Toem on 29 June, and by about 0630 on 2 July the landing and fire support ships were standing off Kamiri. Just before the sun rose three cruisers opened fire, and a few minutes later destroyers and three rocket-launching LCI's joined. Four other destroyers off Namber and Kornasoren began a simultaneous barrage. Between H minus 80 and H-hour the ships off Kamiri had fired two and one-half times as much ammunition as normally required to neutralize such an area. Promptly at 0731 the first B-24's appeared, and by 0747 the last of the sixteen


Liberators had dropped their cargoes of 250-pound instantaneous demolition bombs precisely on Japanese ridge defenses and personnel areas overlooking the beachhead. Admiral Fechteler described their work as "the best example of coordination and timing yet achieved in the SWPA."94 Proceeding ashore successfully in LVT's, the assault wave of the 158th RCT found that the Japanese had abandoned their beach defenses. The first defensive troops encountered, near the center of Kamiri strip, had been so badly stunned by the aerial and naval bombardment that they could offer little resistance. Throwing a perimeter defense around the airdrome, the 158th spent the rest of D-day cleaning huddles of dazed Japanese troops out of the caves east of the airdrome. A captured Japanese lieutenant and abandoned documents revealed that the Japanese commander had observed the soundings of the ALAMO Scouts and had anticipated an Allied landing at Kamiri early in July. He had prepared defenses and registered weapons so as to destroy the landing forces on the reef, but his forces, their morale and combat efficiency already sapped by aerial bombings, collapsed under the immediate pre-invasion air and naval bombardment.95 The potentially hazardous landing at Noemfoor was thus accomplished practically without opposition.

The B-24's from the 90th Group represented only a part of the D-day aerial support. A 421st Night Fighter P-61 had taken station over the beachhead at 0630, and another provided last-light cover. Between times, the 8th, 348th, and 475th Groups covered the beachhead with 161 fighter patrol sorties. Encountering no opposition, they strafed such targets as the naval air controller designated. Four A-20 squadrons each sent two missions to orbit off Noemfoor until directed to ground targets, and the 17th Reconnaissance Squadron provided three similar B-25 missions. These support aircraft silenced mortar positions, knocked out a few automatic weapons, and strafed small parties of enemy troops attempting to reach the combat area. In general, however, few worth-while targets appeared, and one A-20 flight was sent home for lack of a suitable objective. At noon, forty-four B-24's from Nadzab bombed Kornasoren airdrome.96

The ground fighting on Noemfoor had progressed beyond all expectations on D-day, but Patrick, his early intent to ask for the 503d Parachute Infantry strengthened by erroneous P/W reports that there were 3,500 to 4,500 Japanese troops on Noemfoor, requested reinforcements by air beginning on 3 July. By the evening of D plus 2, he


had correctly assessed enemy opposition at not more than 1,500 men, but he still needed the additional force for expanding operations.97 The 317th Troop Carrier Group had been concentrated at Hollandia, and on the mornings of 3 and 4 July its C-47's dropped 1,424 parachutists on Kamiri strip. Both missions were marred by high injury rates--9.74 per cent on the 3d and 8.17 per cent on the 4th. On the former day, a smoke screen laid by A-20's and B-25's to mask the drop zone from sniper fire drifted over the strip, with the result that many of the parachutists, missing the strip, landed among debris and parked vehicles on either side of it. On the second day the C-47's released the jumpers properly and most of them landed in the drop area, but by this time the engineers had begun compacting the strip and there were more fractures than on the previous morning. A ground forces board subsequently concluded that an airstrip was unsuited for paradrops and, with the exception of a few planes which had dropped below the prescribed 400 feet, absolved the 317th Troop Carrier Group of blame for the casualties.98 Because of the high injury rate, a third battalion scheduled to be dropped was flown to Biak and thence moved to Namber by LCI's on 11 July.99

Meanwhile, the 158th RCT had occupied Kornasoren drome on 4 July, and the next day a battalion moving southward from Kamiri broke up a Japanese counterattack to end organized resistance. On 6 July, by means of a shore-to-shore landing, a battalion of the 158th secured Namber. The same day a platoon of paratroopers seized Manim Island, desired by the air forces as a radar site. After 7 July patrols pushed remnants of the enemy, soon reduced by a lack of food to the most loathsome and promiscuous cannibalism ever noted in SWPA, to the interior of the island, where by the end of August they had been surrounded and destroyed. Other than routine fighter patrols, the CYCLONE force required virtually no air support after a few strafing sorties on D plus 2.100

During the four weeks in July that Noemfoor was the most advanced Allied base, enemy air attacks amounted to nine sorties in five raids, which, if credence is given to a P/W shot down at Noemfoor, seem to have originated at Ceram bases and to have staged through Moemi. Neither the first raid, made by a single bomber at 2150 hours on 4 July, nor the following efforts were effective. Aircraft warning and control functions at Noemfoor were performed by Detachment G, Fighter Wing, utilizing the 35th Fighter Control Squadron and


operating units of the 565th and 569th Signal AW Battalions. They opened an assault control center on D-day, and after encountering shipping delays and various unloading accidents they finally completed the permanent installations of the 34th Fighter Sector on 20 August.101

Construction of initial air facilities on Noemfoor had to be timed in relation to the Vogelkop operation scheduled for 30 July. Fortunately, construction could begin as soon as the engineers were ashore. The Japanese strips, unfortunately, were of little use. Kamiri was poorly surfaced with sand and clay and, instead of supposedly "extensive dispersals," had no more than ten hardstands. Kornasoren was "only a location." Namber strip was suitable for transports, but its utilization would require a supply route overland from Kamiri and construction of a standard airfield would be complicated by heavy standing timber and rugged terrain. Work began at Kamiri on D-day, when the 27th Engineer Combat Battalion dragged lengths of Japanese railway irons behind trucks to smooth ruts and used abandoned rollers to begin compacting the strip. By 5 July the 1874th Engineer Aviation Battalion started 24-hour construction, and after a coral surfacing coat had been laid the strip was opened on 16 July for transport aircraft. Work had not begun at Kornasoren when, without warning on 14 July, GHQ indicated that air plans to begin raids against the Halmaheras would require a serviceable strip and parking for fifty P-38's there by 25 July--this without slowing work at Kamiri--but the RAAF 62 Construction Wing, service units, and combat troops working together met the deadline. During August the Fifth Air Force and ALAMO would agree to forego a fighter-medium bomber field at Namber, reasoning that the engineering effort could better exploit a limited expansion at Kamiri and a large expansion at Kornasoren.102 But, after the middle of August, the base-construction program on Noemfoor became a factor in immediate pre-Philippine operations and will be discussed in a later volume.

The air garrison, moved to Noemfoor before early August, was by no means as extensive as the Fifth Air Force had scheduled, because of a shipping jam even more aggravated than that at Biak. With the worst unloading conditions ever encountered in SWPA, only about 4,940 tons of shipping could be debarked between 2 and 15 July. Naval demolition parties had blasted a slot through the barrier reef off Kamiri and engineers had built an LST jetty there by 13 July, but even with attainment of a maximum daily unloading capacity of 1,500


tons ALAMO predicted that the shipping backlog could not be cleared before the end of August. This tie-up both delayed the movement of air force units and hindered construction on Noemfoor. By 12 August air force movements into Noemfoor were 69 per cent behind schedule for troops, 76 per cent for vehicles, and 66 per cent for other impedimenta. An advanced detachment of 10 Operational Group went ashore on 4 July, and on 21-22 July, P-40's of the 78 Wing flew to Kamiri. Waterborne echelons of the RAAF 22 and 30 Squadrons (77 Wing) were to have reached Noemfoor before 12 July, but these shipments could not be debarked until 24 July and 14 August and their Bostons and Beaufighters could not be accommodated at Kamiri before the middle of August. The American garrison was even more tardy, and when the advanced detachment of the 309th Bombardment Wing reached Kornasoren on 28 July, two days before D-day at Sansapor, it found only a detachment of the 419th Night Fighter Squadron in place. Failing to get transportation from Saidor for the 58th Fighter Group, Whitehead flew the air echelon of the 35th Group, which had been biding its time at Nadzab since early June, to Kornasoren in time to cover Sansapor.103 Fortunately, the landing at Sansapor did not depend too heavily upon the build-up at Noemfoor.


Both RENO IV and the new RENO V plan had assumed that it would be necessary to establish an advanced air base midway between Geelvink Bay and the Halmaheras. An air garrison at such a base--located either on Waigeo Island or the coast of the Vogelkop--would assist in neutralizing the Halmaheras, cover the convoys and the invasion beaches there, and interdict Japanese air forces based on the left flank of the Allied movement. To assure completion of air bases and the orderly installation of air units in time to assist an invasion of the Halmaheras (tentatively set for 15 September), the Vogelkop operation would have to begin about 1 August.104

Allied planners soon discovered that it was one thing to recognize the strategic utility of a Vogelkop base and quite another to specify its exact location, and GHQ, lacking even general information as to what areas might be profitably explored by ground infiltration parties, overloaded the Fifth Air Force's 91st Photo Reconnaissance Wing with requests for aerial photos. Weather and lack of staging space for photo planes at Wakde and Hollandia proved a hindrance, and the


evidence gained was largely negative. Waigeo and Ayu Asia Islands clearly possessed no possible air-base sites. A reconnaissance party sent out by submarine returned the first of July with word that Cape Sansapor offered a good beach, though no very acceptable airstrip site. But Whitehead's air engineer, who had reconnoitered the area in a low-flying B-25, was favorably impressed with two potential sites on the mainland and with the possibility of building a fighter strip on offshore Middelburg Island, which a captured Japanese document had also indicated as a likely site. Another party, headed by Lt. Col. H. G. Woodbury, who would be the air engineer on the project, landed from a PT boat on the night of 14 July and explored the terrain between Capes Sansapor and Opmarai. His reconnaissance confirmed the possibility of building airfields somewhere in the region, but a final engineers' conference at Hollandia agreed that selection of specific sites would have to await Allied capture of the whole area.105

Map: The Sansapor Area
The Sansapor Area

The region of interest to SWPA was a roughly rectangular strip of coastal plain between Cape Opmarai and Cape Sansapor, about three-fourths of a mile wide and two miles long, lying fifteen miles west of the Cape of Good Hope, northernmost point in New Guinea. Here the coastal flat is heavily forested, swampy in places, and intersected by several small rivers. About three miles inland the Tamrau


Mountains parallel the coast. Beaches between the two capes are generally reef-free. Amsterdam and Middelburg (the Mios Soe Islands) lie five and three miles off the coast, respectively.106

The Japanese were believed to have no more troops in the immediate area than small LOC detachments at Sansapor and Mar villages. Sixty miles along the coast beyond Sansapor, however, they had an estimated 8,000 combat, base defense, and LOC troops at Sorong, and a greater distance to the east approximately 13,023 troops in the concentration around Manoltwari. ALAMO estimated that the enemy might, under the most favorable circumstances, begin to move 200 to 300 troops a day by barge into the target area, beginning on D plus 2 from Sorong and D plus 3 from Manokwari. Heavier troop concentrations in the Halmaheras and southern Philippines could be drawn upon, but it was not believed that the Japanese would risk combat ships for their transport.107 In the air the Japanese possessed prospects. During June they had increased their SWPA air strength by an estimated 150 planes, adding strength especially to the Halmahera and Amboina-Boeroe-Ceram airfields, all within 400 miles of Sansapor. Including southern Mindanao and the Palaus, the enemy had a total of 850 aircraft within probable range of Sansapor, and in the Philippines they had an estimated additional 1,157 planes. The Allied Air Forces, however, did not believe that the Japanese would consider an operation at Sansapor as vital to their defenses, and since Sansapor was closer to Noemfoor than to the most advanced Japanese air bases then in use, Brig. Gen. Paul B. Wurtsmith of V Fighter Command assumed that fighter cover there would be "a cinch compared to the last three operations."108

GHQ, after a radio warning instruction on 21 June, issued its operations instructions for the GLOBETROTTER operation on 30 June, setting the target date for 30 July. ALAMO set up the TYPHOON Task Force under command of Maj. Gen. Franklin C. Sibert and assigned it the 6th Infantry Division (less the 20th RCT in reserve at Toem), a heavy complement of AA units, and the equivalent of four and one-half engineer battalions. The TYPHOON force, transported by TF 77 from Toem, would land in the vicinity of Cape Opmarai at 0700 on 30 July, capture a beachhead, and subsequently extend its control to Cape Sansapor, Middelburg, and Amsterdam.109 The air plan, which assigned the major task to the Fifth Air Force and put the Thirteenth in reserve, was distinguished chiefly by a purpose to draw


on the latter air force for the air garrison to be established in the Sansapor region. Kenney, who was anxious to force the issue of moving certain Thirteenth Air Force units out of South Pacific bases, slated for Sansapor the 18th and 347th Fighter Groups (P-38), the five B-25 squadrons of the 42d Bombardment Group, and half of the 419th Night Fighter Squadron. The headquarters mechanism for control of the garrison would be another Thirteenth Air Task Force, commanded by Brig. Gen. Earl W. Barnes and utilizing personnel of his XIII Fighter Command headquarters. This air task force, functioning under operational control of the Fifth Air Force, would move into Sansapor and call forward its air units as soon as bases were ready. Because of uncertainties as to the terrain, GHQ was persuaded to allow Barnes to make the final selection of air-base sites and establish the priorities for construction of air facilities.110 "I intend to follow up this advantage," Kenney wrote Whitehead, "and some day we may get fields built our way in places we select."111

Aerial preparations for Sansapor were hampered by a generally unfavorable base situation and the confusion incident to the forward movement of many Fifth Air Force bombardment groups. During July the total weight of bombs dropped by the Fifth Air Force declined to 2,744 tons, less than half the amount dropped in April, but most of this activity was confined to attacks against Japanese airfields on the Vogelkop. Early in July small forces of heavy bombers raided Sorong. Throughout the month all types of bombers kept the Babo area airdromes under attack, dropping 744.9 tons of bombs (the heaviest Fifth Air Force concentration during July) on these targets. Light and medium bombers attacked barge shipping and raided the Japanese supply center and barge construction yard at Kokas village, on the south shore of McCluer Gulf. Other attacks, including missions weathered out elsewhere, bombed enemy concentrations and airdromes in the vicinity of Manokwari. RAAF P-40's and V Fighter Command planes flew a few sweeps over the Vogelkop, and the latter usually strafed targets of opportunity when returning from uneventful escort missions. The 418th Night Fighter Squadron, using B-25's for want of regulation night fighters, made night intruder missions over the Vogelkop fields. Most of these missions were uneventful, but on occasions hostile AA, seemingly improving in marksmanship and increasing in volume, shot down Allied planes.112

Only the Halmaheras promised to provide an exciting target. Photos


taken by the 26th Photo Squadron on 22 July showed 128 planes dispersed at Galela, Lolobata, and Miti dromes. While the number of aircraft was only slightly larger than that observed on 13 July, the Japanese were attempting to build new fields at Kaoe, Hatetabako, and Laboeha, obviously to accommodate a larger air garrison. Whitehead sensed that this was intended as a defensive garrison, but he conceded that "maybe the Nip is fooling us."113 On 24 July, the air bases at Biak, Owi, and Noemfoor now being ready to receive the concentration of units needed for raids on the Halmaheras, Whitehead ordered an attack. A front over the target area delayed the initial mission until 27 July, but early that morning Hutchison sent out the largest coordinated bomber formation employed in the SWPA since Hollandia. The 90th Group and two squadrons of the 22d Group took off at Wakde for rendezvous with the other two 22d Group squadrons from Owi. Over Japen Island the heavy bomber wing assembled with the 43d Group, which had taken off from Owi, in the lead. Joined by four P-38 squadrons, the wing flew directly to the coast of the Halmaheras, where the groups divided to drop their 20-pound frag clusters on the dispersal areas at Lolobata and Miti airdromes. In all, fifty-two Liberators of the 90th and 43d Groups reached Lolobata, and twenty-eight Liberators of the 22d Group bombed Miti. Breaking away from these targets, the groups, with only two B-24's slightly damaged by AA, left at least ten enemy planes destroyed at Lolobata and seven more at Miti.114

That same morning the 38th and 345th Groups had launched a force of B-25's from Mokmer airfield, had contacted two squadrons of P-38's above Ajoe Island, and had thence flown directly to Galela airdrome. Swooping down to treetop height, twenty-four Mitchells of the 345th Group caught the Japanese completely by surprise. Two minutes later, twenty Mitchells of the 38th Group repeated the low-level attack. By the time they had cleared the area, the Mitchells had toggled 402 × 100-pound parademo bombs into the runways, dispersals, and other targets of opportunity and had expended 68,790 × .50-cal. and 29,375 × .30-cal. ammunition in strafing. With only three planes slightly damaged by AA, the Mitchells had destroyed at least ten parked planes.115

The 431st and 432d Squadrons from Mokmer and the 36th and 80th Squadrons from Owi escorted the heavy bombers. The 36th, covering Miti, chased and shot down three uneager Oscars, while a fourth, whose pilot was believed to have been inexperienced and to have


looked over his shoulder too long, crashed into the ocean. The 432d sighted five Zekes and a Hamp at Miti, but the enemy pilots appeared to be experienced and all of them escaped. Over Lolobata, 80th Squadron pilots, counting fifteen to twenty Oscars and a Lily, in twenty-five minutes of combat destroyed three Oscars and the Lily. The 431st, also covering Lolobata, jumped five or six fighters, which seemed to have just taken off, and shot down three in a low-altitude dogfight. Meanwhile, the 35th and 433d Squadrons had been covering Galela in support of the mediums. The 35th Squadron surprised three Tonys and an Oscar approaching for a landing; and, as a Japanese observer recorded, flying "very low and at great speed and as though they owned the place," the P-38's shot down the four planes without difficulty. Pilots of the 433d Squadron reported in some disgust that they had not seen a single airborne enemy plane. Only one of the Lightnings did not return safely to base, and the pilot of this 431st Squadron plane, who ditched because of mechanical failure, was rescued by a Catalina.116 Prior to this mission the Fifth Air Force had been apprehensive that the P-38's could not make the 1,280-mile round trip to the Halmaheras without difficulty, but Charles A. Lindbergh had been working with the 8th and 475th Groups and had shown pilots that they could extend their range by use of economical speeds for cruising (cruise control). Following his directions, all but two of the fighter squadrons were able to return to Biak-Owi without stopping to refuel at Noemfoor.117

Dispersal of Japanese planes in wooded areas made assessment of destruction of grounded planes difficult, but tabulation of pilot observations and analysis of raid photos finally set the number of such planes destroyed or badly damaged at forty-five. Counting fifteen other planes shot down by fighters, the raid had eliminated sixty Japanese aircraft. Whitehead, although pleased with the manner in which the strikes had been executed, was somewhat disappointed that the Japanese before the attack had removed some of the planes noted on 22 July.118

Now convinced that the enemy dispositions in the Halmaheras were purely defensive, Whitehead turned the attention of the Fifth Air Force to the Amboina-Ceram-Boeroe area where, according to Allied estimates on 28 July, the Japanese had approximately 151 planes at Haroekoe on the like-named island, at Liang and Laha dromes on Amboina, at Kairatoe, Boela, and Amahai dromes on Ceram, and at


Namlea Township and Old Namlea dromes on Boeroe. Additionally, the Boela oil fields, located on the east end of Ceram, comprised a strategic target which, if not one of the more important sources of Japanese fuel, was the first to come within effective bombing range in the SWPA. On 9 July, Whitehead had relieved the 380th Group from its commitment against Babo, and during the month this group had executed four attacks against Old Namlea, four against Namlea Township, three against Amahai, one against Liang, and two against Laha. These strikes had been hampered by slow-moving tropical fronts, but the group claimed seventeen planes destroyed or rendered unserviceable on the ground. On 14 July, seventy-four A-20's of the 3d and 312th Groups, staging through Mokmer, had executed a deck-level attack on the Boela airfield, oil wells, and oil storage tanks. AA had shot down one of the light bombers, but the strike had been successful in firing Japanese oil stores.119

On 28 July, Whitehead sent the 38th and 345th Groups to Haroekoe, covered by 9th and 80th Squadron P-38's which were "to stick around and shoot down Nips ... flushed ... off Liang and Laha." Weather turned the Mitchells back, but the two fighter squadrons flew over Amahai, Haroekoe, Liang, and Laha, circling each airdrome at 8,000 feet and keeping just out of effective AA range. The pilots saw very few Japanese airplanes on the dromes, and only two airborne planes were sighted, one of which was shot down by the 9th Squadron. On 28, 29, and 30 July, a part of the Fifth Air Force's Liberator force was dispatched to bomb Boela targets, but these missions encountered no Japanese fighters. A 36th Squadron fighter sweep over Amboina found no aircraft on Liang. Puzzled by this negative information, Whitehead intended to keep sending strikes into the area to determine whether the Japanese air units had withdrawn temporarily or permanent1y.120

During the week prior to the Sansapor landing, Fifth Air Force planes of every type were employed in attacks against shipping. Beginning on 27 July and continuing to the end of the month, the 498th Bombardment Squadron sent two B-25 missions each day to search for shipping around the Halmaheras. The 63d Bombardment Squadron, which had been sending armed reconnaissance planes almost nightly to the Palaus, began night missions to the Halmaheras and Ceram on 25 July. The 36th Squadron sweep to Amboina on 29 July fired a 5,000-ton cargo vessel and another small steamer. Other units attacked


barges from Cape Sansapor to Sorong. The net results claimed in the week's shipping attacks were: one 5,000-ton freighter set afire; one 500- to 1,000-ton freighter left burning; one 2,000- to 4,500-ton freighter sunk; one 4,000- to 5,800-ton freighter strafed; two smaller freighters damaged by strafing or near misses; nine luggers definitely sunk and six others probably sunk; six barges destroyed and seven damaged. Innumerable sailboats, canoes, fishing boats, and the like were strafed, results being usually unobserved.121

On the nights of 26/27 and 27/28 July, the slower boats departing first, the assault convoys had left Toem. The voyage northward was generally uneventful, although an 80th Squadron P-38 flight drove off five Japanese dive bombers which seemed anxious to attack the naval convoys just north of Manokwari. By the predawn hours of 30 July, the naval forces were standing off Red Beach, a point midway between the Wewe River and Cape Opmarai. No enemy opposition being evident, Sibert canceled the preliminary naval barrages. Air support was also unneeded, although five 501st Squadron B-25's strafed targets near the beachhead shortly after the assault waves began landing at 0800. Thirty minutes later, a cavalry reconnaissance troop was launched in amphibious tanks from an LST and secured Middelburg without opposition; shortly afterward the same troop captured Amsterdam, also without opposition. The next day an infantry battalion, making a shore-to-shore landing, seized Sansapor village. Ground fighting in the Sansapor area never amounted to more than sporadic skirmishes, and at the termination of the operation on 31 August, the TYPHOON force had killed only 379 Japanese and captured 213 prisoners at a loss of 10 killed and 31 wounded.122

General Barnes landed behind the assault troops on D-day and immediately inspected sites near Mar village (just east of the Wewe River), on Amsterdam, and on Middelburg. By D plus 1, he had decided that a fighter airdrome could be built on Middelburg and that the medium bomber fields should be built at the Mar site. In spite of difficulties inherently connected with the construction of an airstrip on a small, reef-surrounded island, the airdrome program progressed most satisfactorily. The engineers, finding that the coarse coral sand of the island would be impossible to compact, borrowed coral off the shallow floor of the sea during low tide and surfaced the strip with pierced steel plank. The 5,000-foot runway received a crippled B-24 on 14 August and was ready for regulation landings three days later.


The job, complete with sixty hardstands, three alert areas, a service apron, and interconnecting taxiways, was finished on 8 September. Work at the Mar site was complicated by thick jungle growth and large trees, but the airdrome was operational on 3 September, at which time it comprised a 6,000-foot, steel-mat runway, four alert aprons, and seven hardstands. By 18 September a total of eighty-five hardstands had been completed, and by the end of the month the strip had been lengthened to 7,500 feet.123

Installation of the Thirteenth Air Task Force garrison at Sansapor required movement of 6,500 men, belonging to thirty-five combat and service units, from as far away as New Caledonia. The first units to move were the headquarters squadron of XIII Fighter Command, the 347th Fighter Group, the 6th Service Group, and a detachment of the 419th Night Fighter Squadron. Ground echelons of these units left the Solomons between 15 and 18 July, reloaded on LST's at Toem, and shuttled into Sansapor between 15 and 19 August. Aircrews of Detachment B, 418th Night Fighter Squadron flew to Middelburg on 18 August, followed by the flight echelons of the 347th Group, which moved to Middelburg by squadrons between 20 and 26 August. Local air cover having been established, the waterborne echelons of the 18th Fighter and 42d Bombardment Groups, plus their supporting service units, moved directly to Sansapor in transports and landed there on 23 August. Flight echelons of the 18th Fighter Group landed at Middelburg between 4 and 7 September, but the crews of the 42d Group flew to Hollandia between 1 and 4 September, remained there to fly missions coordinated with the invasion of the Halmaheras, and did not reach Mar airdrome until 14-18 September.124

During the period of the establishment of the Allied base at Sansapor, both Allied and Japanese air activities were nominal. No air support was required other than routine day and night fighter patrols from Noemfoor, and the Japanese night raiders did not appear for nearly a month after D-day. On 25 August, however, the Japanese, timing their effort to coincide with a full moon, began a series of five night attacks which killed one man, seriously wounded three others, and destroyed or damaged eight P-38's. Although Detachment H, Fighter Wing, the 33d Fighter Control Squadron, and reporting units of Company A, 565th Signal AW Battalion and Company A, 596th Signal AW Battalion had completed the permanent installations of the 35th Fighter Sector on 21 August, well before the first Japanese air raid, their


warning network could not overcome the permanent echoes caused by the ring of mountains circling the rear of the beachhead, and the Japanese raiders late in August were able to sneak in to attack the airdrome area without adequate warning.125

Once ashore at Sansapor, the units of the Thirteenth Air Task Force established their camps in the thick rain jungle, clearing the timber with axe and machete. Because of the construction under way on Middelburg, the 347th Group had to bivouac temporarily on the mainland and ferry its personnel to the island. Rations, especially perishable foods (not present in quantity until D plus 55, when a shipload came forward), left much to be desired. Rigidly enforced mite-control measures, including clearing of undergrowth, aerial DDT spraying, and impregnated clothing held cases of scrub typhus among Thirteenth Air Task Force troops to thirty-eight infections, none of them fatal. Swamps and slow-flowing rivers necessitated especially vigilant malaria-control measures. The flurry of Japanese night raids at the end of August proved annoying to troops from the Solomons who, more recently, had gotten out of the habit of having their sleep interrupted, but the fact that several of the groups originated betting pools as to the time of the next interruption showed that the raids were not taken too seriously.126

The closer collaboration of Thirteenth and Fifth Air Force units fittingly gave emphasis to the great achievements which had marked two years of hard fighting in the South and Southwest Pacific. The campaigns in New Guinea, like those in the Solomons, had begun in desperate and essentially defensive attempts to check the enemy's advance toward vitally important Allied positions. Allied air, ground, and sea forces--American, Australian, New Zealand, and Dutch--had wrested the initiative from the enemy, and advancing then along two converging lines, they now controlled the Solomons, the Bismarcks, and New Guinea. And the Philippines, where the earlier Far East Air Force had taken its initial defeats, would soon be brought within reach.


Table of Contents
Previous Chapter (18) * Next Chapter (20)

Transcribed and formatted for HTML by Alan Clark for the HyperWar Foundation