Section II: Target Rabaul
* * * * * * * * * * * * * * * * * * * * * * * * * * * * * *

Chapter 7: The Central Solomons

By the second week of February 1943 the enemy had been driven from Guadalcanal, bequeathing to its defenders the tedious task of rounding up sick and wounded Japanese scattered about the island. For Admiral Halsey and General Harmon, having surmounted the great crises of October and November, this was a period of preparation for the next step up the long chain of the Solomons, and for Twining's Thirteenth Air Force it was a period of adjustment within the confines of the peculiar naval command unit known as COMAIRSOLS.* Operational autonomy for Army air remained unknown in the South Pacific. Harmon and Twining had achieved administrative autonomy but there they halted; operational control rested in the first instance with the current incumbent of COMAIRSOLS.

It was only natural that this initial step toward a single unified air force would raise many problems, involving as it did the planes, pilots, and ground crews from widely diverse services and training schools. Their doctrines were dissimilar, their equipment varied, and even their nationality differed--the force would include able New Zealanders with their P-40 squadrons and Lockheed Hudson search planes.** Over a period of many months the successive staffs of COMAIRSOLS would face, and to a large extent overcome, problems of supply, administration, and differences in combat techniques which were inherent in this situation. The net result was an air organization with an amazing variety

* See above p. 88-89.

** The RNZAF No. 15 Squadron was the first New Zealand fighter unit to reach the combat zone, moving up to Guadalcanal at the end of April 1943. No. 3 Squadron, a bomber-reconnaissance unit, had opened operations on Guadalcanal with twelve Hudsons on 24 November 1942.


of aircraft whose effectiveness would surpass anything the enemy could put into the air against it.

Union of effort between Corsairs, Wildcats, P-40's, P-39's, P-38's, F6F's, B-17's, B-24's, SBD's, TBF's, and B-2 5's, plus several other types, was not achieved easily. But it was done, and a large share of the credit must go to Adm. William F. Halsey. From the outset Halsey had placed heavy emphasis upon the principle of unity of command, insisting that each commander of a task force have full authority over all components of his force, regardless of service or nationality. He sought to create what he termed a "South Pacific fighting team," and he succeeded1 although air force men believed that this principle occasionally did violence to proper employment of AAF units. It was improbable that all echelons and all personnel of any service could share in the enthusiasm of Halsey and Harmon for this procedure. It was not ideal and service loyalties loomed large in the minds of many lesser men, but a healthy spirit of cooperation prevailed with many others, making it possible to brigade the four flying services together. The narrative of the air campaign up the Solomon Islands thus is a skein of many threads, all leading into COMAIRSOLS. Operations of the Thirteenth Air Force were often indistinguishable from the general pattern of COMAIRSOLS--its missions were flown with the other components; heavy bombers many times struck their targets almost simultaneously with the dive and torpedo bombers of the Marines, protected in the air by fighters of all services; and in nearly every major interception of Japanese air strikes COMAIRSOLS committed fighters of every available type. This was the apparatus with which Halsey and Harmon were prepared to carve out new platforms for their air weapons farther up the Solomons.

It was immediately apparent to the theater commanders that a long and painful road lay ahead before the enemy could be evicted from all the Solomons and from Rabaul, which was the logical objective toward which the South Pacific forces were moving early in 1943. The next major goal on the road to Rabaul was New Georgia, even then under almost incessant air attack, and it was obvious that additional airfield sites were required.2

The Russell Islands provided the necessary intermediate step, and on 21 February the first waves of the 43d Division landed on the Russells to begin the long advance toward Rabaul. Southwest Pacific forces co-operated by striking nightly at Buin, Kahili, and Ballale over the period


Central Solomons


of the invasion; Guadalcanal fighters provided the air cover, which was not opposed by the Japanese, who seemed to possess no advance knowledge of the invasion.3 Admiral Fitch (COMAIRSOPAC) had warned Admiral Mason (COMAIRSOLS) that any daylight strikes against Buin from Guadalcanal must have sufficient fighter escort to prevent disproportionate losses; he had not forgotten the two daylight strikes by the new B-24's of the 307th Bombardment Group on 13 and 14 February, efforts which had taken a heavy toll of the over-all bomber strength. No less than five B-24's of a total of fifteen in the two missions had gone down, an experience which brought a halt to all day-light bombing attacks upon this area until more adequate fighter cover could be provided.4

But now the Japanese did not make any real attempt to halt the Russells construction project. By wasting eight planes they destroyed some equipment but that was all; for some unknown reason the enemy was not yet ready to move down in force, confining himself to a series of ineffective night raids in March.5 March was relatively a quiet month in the Solomons. The enemy faced a formidable task in his attempt to maintain his position in the island chain; he must take immediate steps to strengthen what he yet held and to regain what he had lost.6 Allied air operations worked in such a way that each fresh base became an additional cancer in the structure of the enemy's defense line, sending out its tentacles to destroy relentlessly the equipment and personnel opposing it. Between the two forces no equilibrium could be reached, let alone maintained, for one would consume the other. There was every evidence, despite the apparent lull, that Japanese commanders intended to make a real effort to cling to their positions. No longer were they obliged to hold their Bettys and Sallys back in Rabaul and Kavieng. Intense efforts by their construction crews had prepared shelter for an estimated 461 aircraft by mid-February, and of this number 72 were for medium bombers.7

Strategy and Command

Amid the speculation over Japanese intentions, the commanders of all services both in the field and in Washington took stock of their resources and the methods of applying them. For the first six months of the Solomons campaign the task had been to press the enemy back from a position where he might severely damage or at least threaten the line


of communications to Australia. Now it was time to look ahead to more distant goals, of which the first major one was Rabaul. For its reduction General MacArthur had prepared an over-all strategic plan based upon utilization of the forces in both the South and Southwest Pacific. This ELKTON plan,* as it was known, regarded Rabaul as the apex of a triangle, one leg of which extended southeastward through the Solomons, the other westward along New Britain. The Allied advance would move in short steps along each leg of the triangle.

In this plan there was no interest in capture of land areas as such. The goal in each case was to establish one more rung in the ladder of air bases necessary for land-based aircraft in providing direct cover and support for the next step. For Halsey's forces this meant that New Georgia's airstrip in Allied hands would provide air cover for subsequent landings upon Bougainville, while the installation of Allied land-based aircraft on New Britain and Bougainville would permit the aerial assault upon Rabaul and Kavieng. Throughout the parallel campaigns, employment of the aviation of both forces would be coordinated by MacArthur, who could shift his air efforts in support of either force requiring the maximum assistance at the moment.8

MacArthur had asked for air strength greatly in excess of the forces that could be provided, calling for no less than ten groups of aircraft for Halsey's contemplated New Georgia operation.9 Army planners, unchanged in their original conviction that attainment of the maximum possible bomber offensive against Germany must not be jeopardized by diversions of combat aircraft to theaters of lesser strategic importance,10 agreed to no more than a modest increase in Pacific air strength. On 23 March the Joint Chiefs informed Halsey and MacArthur of changes which would increase each of the two South Pacific heavy bomber groups to four squadrons, twelve planes per squadron, for a total of ninety-six. The medium group would be augmented from forty-six to fifty-seven planes, and troop carrier strength would be raised to two full squadrons of thirteen planes each.11 After examining the revised schedule of commitments, representatives of the Pacific commanders agreed that with fresh forces they could extend their advance along New Guinea to Madang and they could gain a foothold on southeast Bougainville.12

There remained numerous unsettled details. Ever since the summer of 1942 the South Pacific Area had functioned as a test tube of Army-

* See above, pp. 129-30.


Navy cooperation, and now that a fresh series of campaigns was contemplated, it was necessary to clarify the position each service should occupy in the general scheme. Equally important was the fact that the South Pacific was a crossroads and a point of contact between the great Southwest Pacific Area under Army direction and the vast Pacific Ocean Area under Admiral Nimitz. For more than six months it had served as the theater to which the U.S. Navy had committed the bulk of its combat strength; now it was to become one-half of a joint operation with the MacArthur-controlled theater. Admiral King was disturbed over the situation created by the proposed execution of CARTWHEEL, as the ELKTON operations were known. There was no disagreement over the decision to place all the Rabaul operations under MacArthur; COMSOPAC would control directly all the Solomons operations according to general directives prepared by the Southwest Pacific Commander.13 But the Navy spokesmen were meticulously careful to see that the over-all directive to MacArthur was phrased so as to retain complete control over all naval forces assigned to the South Pacific. Admiral King recognized the advantage of unified control, but he regarded as impossible any attempt to view operations in the Solomons carried out by forces of the Pacific Ocean Area as separate from those under way elsewhere in the Pacific.14

Because MacArthur believed that amphibious forces would be unable to operate successfully in the New Georgia area under the threat of the enemy's air bases on Bougainville, on New Britain, and at Kavieng, he maintained that these first must be neutralized. And since their neutralization could be effected only from the Vitiaz Strait area, he recommended that the assaults on the enemy bases in the Huon Gulf precede those directed against New Georgia.* MacArthur feared that large-scale operations in the South Pacific during his own period of intense activity in New Guinea would necessitate sending air forces to support Halsey just at the time when they could not well be spared. On the other hand, as Marshall pointed out, the field commanders must of necessity take full advantage of every opportunity to push ahead in areas where resistance seemed weak--and there was reason to believe the Japanese could not remain strong in both areas at once. In any case, Halsey had no intention of remaining idle. He would continue to exert pressure upon the Japanese with his land-based aviation and he would be ready to move into New Georgia and southern Bougainville if this

* See above, p. 130.


could be accomplished without bringing on a major operation.15 The final solution to the question of procedure was suggested by Admiral King: the supreme commander of the Southwest Pacific Area would submit for approval to the Joint Chiefs of Staff his general plans, including the composition of task forces and the sequence and timing of major offensive operations.16

By the end of March the debate had ended. The tasks were clearly defined; ahead lay the objectives. Whether in New Guinea, in Bougainville, or in New Britain they were much the same. Each assault would involve the isolation, neutralization, and final capture or by-passing of a small island of Japanese air strength clustered around an airstrip which had been carved out of the jungle. Experience warned that the defense would be fanatical, and if the assault were made frontally without adequate air preparation, it could prove costly. But whereas the enemy's ground forces might absorb losses, his merchant marine and naval service could not do so without seriously weakening the entire Japanese strategic plan. MacArthur's staff reasoned that as soon as the enemy reached a point where exploitation of the conquered territories had become seriously affected by shipping shortages and maintenance of outposts had become excessively costly, he would abandon the Bismarck-Solomons line and fall back upon the Japanese mandates as the main line of resistance.17 Thus if the attacks were to be executed at minimum cost, the first step must involve a continuation of the effort to isolate each successive objective, and for this task the greater burden would fall upon the Fifth Air Force in the New Guinea area and upon Halsey's land-based aviation, including the Thirteenth Air Force, in the South Pacific.

How best to employ their own units in the approaching campaigns was the problem which plagued the senior AAF commanders in the South Pacific. General Harmon had successfully activated an air force, but this had not altered the fact that operational control of all land-based aircraft in the South Pacific remained with COMAIRSOPAC. Harmon did what he could to maintain a reasonable amount of autonomy for his young air force. From the outset he had insisted to Admiral Fitch that the AAF retain direct responsibility for and control of all matters affecting administration, supply, movement, and training, and he maintained the right to insist on the observance of sound air doctrine and technique.18 But as the spring of 1943 approached, it became apparent to Harmon that observance of these precepts was less than


satisfactory, due, he thought, to lack of understanding upon the part of subordinate commanders and staff officers. His appeal to Halsey resulted in a direct conference with Admiral Fitch, where for the better part of two days Fitch and Harmon personally ironed out the difficulties, emerging on 4 March with a statement of policy which went far toward meeting Harmon's objections.19

The key to Harmon's recommendations lay in the statements that the highest degree of effectiveness would be achieved by vesting combat command of the various air forces in their respective services; secondly, that any necessary disruption of normal command channels should be held to a minimum; and thirdly, that the air forces should be employed in the function for which they had been organized, trained, and equipped. Short of over-all operational control, Harmon desired the closest possible participation by his own air commanders in planning the details of missions, and he sought a clear recognition of the right of AAF leaders to advise the local air commander (in this case, primarily COMAIRSOLS) as to proper formations, bomb loadings, escort and combat techniques. As a final request he asked for a restatement of the policy he had advocated since the earliest days of the Solomons campaign: that AAF bombardment aircraft concentrate on bombardment missions and forego search and patrol functions. Admiral Fitch received the recommendations and offered his full concurrence,* advising Harmon that in the future his own direction of activities affecting the Thirteenth Air Force would be in accordance with the principles suggested by the Army commander.20

Harmon's concern was for the Thirteenth Air Force, but this unit was only one component of the South Pacific air organization. In order to clarify the function of all the air services in the area, COMAIRSOPAC on 11 March issued a new air operation plan for the Solomons air units.21 This plan assumed that from his defensive positions the enemy would continue his air raids upon Tulagi, Guadalcanal, and upon Allied surface forces; he would persist in his night bombardment of Allied airfields, and he might even attempt "commando" raids against exposed

* Harmon requested that the position of commanding general of the Thirteenth Air Force be reaffirmed, and that Fitch issue to Twining all directives pertaining to training and administration of AAF units. It seems probable that there was a tendency on the part of Fitch's staff occasionally to deal directly with subordinate AAF units rather than through the AAF commander. Harmon states that throughout his tenure of COMAIRSOPAC, Admiral Fitch adhered strictly to the agreement "and could not, while prosecuting the war effort, have treated the Army component of his force with greater consideration and justice . . . than he did."


positions. To check him, three tasks were assigned to the air forces. First, it was assumed that the destruction of enemy shipping constituted the most effective employment of aircraft, since this could deprive the enemy of his logistical support, thereby causing his Solomons bases to shrivel and starve. Next, the Thirteenth Air Force, the Marines, naval air units, and the New Zealanders should strike at enemy shore installations within range to reduce Japanese air strength. Finally, these units should support the operations of Allied surface forces in the Solomons.

To Bomber Command of COMAIRSOLS went the responsibility for conducting long-range day and night attacks against surface forces, airdromes, and ground installations as ordered, and also for executing necessary search and patrol missions. This involved primarily AAF heavy and medium bombers. Few AAF planes were directly concerned with instructions to Strike Command, except as fighter escorts. The function of this command was to carry out repeated dive, glide, or low-level bombing attacks upon enemy surface units and airfields; for such work the Marines' equipment was most useful. COMAIRSOLS Fighter Command was the most diverse of all the Solomons air organizations and its duties were manifold, including assault, defense, and escort, both for surface craft and bombers, as well as operation of all air warning service units, fighter directors, and all equipment concerned with the interception of enemy aircraft. COMAIRSOPAC would control by general directives the tempo of all operations, and all local unit commanders were reminded that their aircraft would operate under the operational control of COMAIRSOLS, who himself was a task group commander under COMAIRSOPAC. These constituted the rule book for AAF planes in the campaigns for the Solomons.

The month of March provided a breathing spell for both sides. The photo planes maintained their close daily check on air and shipping movements in and out of the Buin fields and anchorages; heavy bombers in light strength continued to heckle the installations at Ballale, Kahili, and Shortland Island, and Strike Command hammered away at Munda.22 Whenever profitable targets appeared, fighters moved out on sweeps, as they did on 28 March. Photos brought home by the 17th Reconnaissance Squadron indicated that the Shortland-Poporang area off southern Bougainville was the principal Japanese seaplane base in the Solomons, and here the prints revealed twenty-seven planes at rest. Fighter Command immediately ordered out a dawn strike of eight


P-38's and a like number of Corsairs. Engine trouble and weather forced out seven of the F4U's and three of the P-38's, but with his six remaining planes, Capt. Thomas G. Lanphier, Jr., of the 70th Fighter Squadron, led the flight over the Shortland and Poporang targets at 50-foot altitude, leaving eight float planes burning on the water in the two anchorages. Intense AA fire gained the Japanese gunners nothing and Lanphier's fighters turned for home. Six miles east of Shortland the flight spotted an enemy destroyer, which immediately began desperate evasive maneuvers. Again attacking at 50 feet, two P-38's silenced most of the AA fire, after which the rest of the flight followed through with four passes by each man. Ten minutes later the vessel lay dead in the water, aflame from stem to stern, and with a 15o list. The sole partial casualty was the P-38 of Lt. Rex T. Barber of the 339th Fighter Squadron, who made one pass a trifle too low, shearing off the top of the warship's foremast with his wing tip; but the resultant loss of three feet of wing caused no trouble and the P-38 made a normal landing.23 When the month of March had ended, the score of enemy aircraft destroyed had dropped to an all-time low of approximately sixteen; not until May of 1944 was this figure again to sink to a comparable position.24 COMAIRSOLS Fighter Command could report that its P-39's had shot down eight of the total, and not a loss was incurred for the entire month by any of its planes.25

April provided a different story. Admiral Halsey was losing no time in preparing for a fresh offensive thrust and the Japanese knew it. Down on Guadalcanal massive stockpiles of fuel, ammunition, supplies, and equipment accumulated in dumps, despite serious unloading difficulties.26 Lying off the Lunga shore were the cargo vessels and transports; across the channel at Tulagi were the warships of the task forces which moved in and out of that expanding naval base. All these offered fat targets for the Japanese Vals and Bettys if only they could break through the fighter defenses; and even if they could not reach this richer harvest, they might strike at the advanced air installations rapidly nearing completion on the Russells.

And strike they did. They tried it on the first day of the month, when Fighter Command sent forty-two fighters, six of them P-38's, up over the Russells to fend off the Zekes and Hamps coming over in two waves. For nearly three hours there was continuous air action, action which cost the enemy heavily; the final score showed twenty of his Zeros down, for a loss to Fighter Command of six planes in combat, but


three pilots were rescued immediately.27 The Japanese would have to try harder to push through, and they were having trouble enough in supplying their own forward bases on New Georgia and at Vila on Kolombangara, where their small cargo vessels were frequent targets for the fighter pilots up from CACTUS. On 7 April, after a week of sporadic night bombing, they were ready for another daylight assault on Guadalcanal. This was a big one. Coast watchers on New Georgia gave first warning of the approaching enemy, reporting no less than 160 dive bombers and fighters moving down the Slot. Ahead of them lay rich targets: a convoy near Rua Suru off the east coast of Guadalcanal, shipping at Koli Point, and a task force lying at Tulagi. Ahead, too, lay the defenders. Fighter Command sent up every available fighter on Guadalcanal, seventy-six of them, and all bombers were moved to the southeast tip of the island to avoid damage.28

As the enemy approached, his planes separated into smaller flights, so that fighting occurred off the Russells, near Tulagi, and over the convoy east of Guadalcanal. Up on top were the P-38's led by Captain Lanphier; spaced out below were the Corsairs, Wildcats, and P-39's. When the wild melee ended, thirty-nine enemy planes had fallen, thirteen of them to the AAF planes, and total Allied loss was seven aircraft and one pilot, Maj. Walden Williams of the 70th Fighter Squadron.29

Two weeks after this aerial defeat of the enemy there occurred one of the most extraordinary interceptions of the entire Pacific campaign. Intelligence sources had discovered that none other that Adm. Isoroltu Yamamoto himself would be flying down to the Kahili area on an inspection trip of the South Pacific. The decoders in Washington had precise information on Yamamoto's itinerary and they knew he was due at Ballale at 0945 on the 18th. On the 17th the message reached Halsey, who immediately informed COMAIRSOLS-the order was to get Yarnamoto.30

It was at once apparent that only the P-38's could reach far enough up the coast of Bougainville to intercept the admiral, but any attack upon his plane in the Buin area would prove extremely hazardous; here Kahili Field swarmed with fighters ready to cover Yamamoto's movements. Eighteen P-38 pilots-eight each from the 12th and 339th Squadrons and two from the 70th-were chosen. Four were designated the attack section under Captain Lanphier of the 70th, the remaining


fourteen were to provide cover under the command of Maj. John W. Mitchell, who was also over-all flight leader.31

Briefing was meticulously done, and every detail was reviewed, for the slightest error in timing would result in failure. The plan called for an overwater wave-hugging flight of 435 miles by a circuitous route which would avoid all danger of detection by land-based enemy coast watchers. If Yamamoto followed his schedule punctually--and he was known to have a passion for punctuality--then at 0935 he should be over a point some thirty-five miles up the coast from Kahili.32

Two hours and nine minutes after take-off at 0725, as sixteen P-38's flew in toward the coast of Bougainville barely clearing the water, there ahead appeared the enemy almost as if the entire affair had been prearranged by mutual consent. Two Bettys turned to escape, while their six Zero escorts tried in vain to cut off Lanphier's attack section. Lanphier exploded one fighter, then dived on one bomber, sending it flaming into the jungle, while Barber disintegrated the other Betty. Escape of the P-38's was doubtful, since now the advantage of altitude lay with the Zeros, but by hedgehopping, skidding, and sideslipping, the attackers pulled away under a heavy counterattack. Only Lt. Raymond K. Hine failed to return from this flawlessly executed mission which had cost the Japanese their highest ranking naval officer, victim apparently of Captain Lanphier's guns and Major Mitchell's flawless timing of the flight.33 Rear Adm. Marc A. Mitscher, COMAIRSOLS since 4 April, was delighted with the results of the mission, and his efforts gained the Navy Cross for Mitchell and the members of the attack section.34

By May there was evidence that the Thirteenth Air Force was beginning to operate somewhat as its commanders desired; fresh units had reached it and there was a clearer understanding all around of the command relationships and employment of air power in the area.35 General Harmon's report of 1 May to General Arnold displayed a note of optimism, and Twining, too, could share in the happier view of his affairs, notably in the field of supply and maintenance, in large part because of the authorization of the XIII Air Force Service Command, which had been activated on 14 April.36 He could report substantial improvement in living conditions at all stations, and he doubted that there would be any repetition of the serious sickness rate which had laid low so many men of the 11th Bombardment Group. Nevertheless, the


malaria rate was persistently and extremely high, and would remain so until the month of June.37

Air Force headquarters were at Espiritu Santo, where Twining could remain in close proximity to COMAIRSOPAC, but the operating sections of the bomber and fighter commands had moved up to Guadalcanal. By the end of April, the bomber command was establishing itself on the new Carney Field at Koli Point, while the fighter command was located at Fighter No. 2 (Kukum) near Henderson. Twining soon could count on operating two heavy groups at full strength, plus one medium group and two reduced fighter groups, and all of these were rotated through Guadalcanal.38 This arrangement spread the squadrons thinly over a vast area and Twining knew it, but he believed that the advantages made it desirable. Rotation of the flying personnel helped to maintain the crews in better physical condition, and the units in the rear area were useful in training and indoctrinating new crews or for reorganizing units returning from combat.

An additional reason for holding AAF planes back from Guadalcanal was the limited number of facilities at the forward base.39 Lack of proper dispersal facilities long had troubled AAF commanders and inspectors sent out from Washington, but it was only after some sharp lessons at the hands of the enemy that the situation was remedied. Presumably, Halsey's Base Planning Board, which carried responsibility for all air base construction in the area, would coordinate the needs and views of all services, but Twining had to report that the Navy had never shown much interest in dispersal and continued "to line their own planes for the slaughter."40 On the night of 24 March enemy bombers had added their weight to the argument for better dispersion. From high altitude two or three bombers damaged or destroyed a number of heavy bombers as well as lighter aircraft, and the incident keenly disturbed General Arnold. Over 300 planes were on Guadalcanal at the time, a number which outdistanced the pace of airdrome construction, and Twining reported that even the captured Japanese aircrews refused to believe their eyes at the sight of the long rows of fighter planes parked on Fighter No. 2.41

Arnold's complaints passed to Halsey, who immediately arranged for the assignment of additional dispersal areas for AAF planes. By 21 April, Twining considered that dispersal was reasonably adequate, although there was still ample margin for improvement.42

Gradually during the spring months COMAIRSOLS was able to


Guadalcanal Airfields


increase the tempo of operations against the Japanese positions on Bougainville and in the central Solomons, despite the absence of sufficient numbers of fighters to provide escorts for heavy bomber operations during the daylight hours. There was much nightly harassment of the enemy by the B-17's and B-24's, and in March COMAIRSOLS staged something new--a mine-laying operation against the Buin area. On the nights of 20 and 21 March, Marine and Navy TBF's carried mines up to Bougainville, sowing them inside the 20-fathom curve of the Buin-Tonolei area, while nine B-17's of the 5th Group and nine B-24's of the 307th Group pestered with fragmentation bombs the defending searchlight and AA crews on Kahili Field. These missions were highly successful, and with slight modifications in technique and bomb load they would be repeated again.43 By mid-May air affairs occasionally moved at a fairly fast pace, with the heavies out each night and the enemy bringing his fighters down the Slot in growing force during the daylight hours. The 13th of May was a day of intense activity,* one costing the Japanese sixteen fighters of a force of some two-dozen-odd out on a fighter sweep.44 But it was an intermittent business and the enemy's air effort would promptly subside after a period of effort, causing Allied intelligence to conclude that Japanese air strength was increasing in New Guinea as it diminished in the solomons.45 Perhaps the attrition was more than the enemy could support. To accelerate it COMAIRSOLS ordered a second series of mining operations, which were executed in the Buin area on the nights of 19, 20, and 23 May, once more with considerable success and very light losses, and with the heavy bombers on the third mission placing their bombs precisely at the proper place at the proper time.46

During the latter half of May, Japanese retaliation for the aerial assaults upon their sanctuaries was sporadic and weak, but the general lull ended with the coming of June. For one thing, the new Russells strip was ready, permitting more frequent and effective fighter sweeps over the Buin area. For another, the deadly SBD's now carried 50-gallon auxiliary fuel tanks which permitted them to reach Buin from their Solomons bases.47 These developments aroused the enemy. He

* It is of some interest to note the multiplicity of types of aircraft then available to Fighter Command. For this interception of 13 May, thirty F4U's, thirty-two F4F's, eleven P-38's, twelve P-39's, eleven 8-40's (AAF), and six P-40's (RNZAF) were sent up. To fly the AAF fighters, detachments of the 67th, 68th, 44th, 6th, and 339th Fighter Squadrons and Headquarters, 347th Fighter Group were attached to the 70th Fighter Squadron on 7 April 1943. (Form 34, 5-11 Apr. 1943, 70th Fighter Sq.)


shifted more of his air strength back to the Solomons, and on 7 June he began to use it, for he could scarcely be unaware of the preparations under way on Guadalcanal. On the Rabaul airdromes air strength again rose to a high level, showing 225 aircraft. In the Rabaul harbor area lay nearly fifty ships, and the enemy's search planes were very active, increasing both in number and range.48

On 7 June, Rabaul's air commanders began a week's air assault against Guadalcanal which surpassed anything yet attempted. The pattern was unchanged: large forces of fighters (Zekes and Hamps) would escort the dive bombers against shipping off Guadalcanal. Fighter Command's Russell Patrol normally would open the action which might extend right down to Guadalcanal itself. These attempts were costly to the attackers, twenty-three Zeros going down on 7 June, four of them to RNZAF P-40's of No. 15 Squadron, now in its first major action.49 Moreover, every single pilot was recovered from the nine Allied planes lost by Fighter Command, which set the day's exchange at 23 to 9, a rather costly business to the enemy.50 On 12 June came the second heavy thrust, this one costing the Japanese thirty-one planes for a loss to Fighter Command of six planes and two pilots;51 next a lull for three days, and then by the 16th the enemy was ready. Search planes counted 245 planes at Rabaul and found the other fields jammed with aircraft.52

Coast watchers' reports indicated something more than a normal fighter sweep was under way, and they were right. Ships lay off Tulagi and Guadalcanal. To attack them the Japanese converged on their targets with an estimated 120 aircraft; and Fighter Command, with ample forewarning, had in the air a total of 104 defending fighters of all services.53 The resulting clash constituted the greatest single Allied aerial victory of the Solomons campaign, as the air over Savo Island, Tulagi, Cape Esperance, and Koli Point was filled with enemy and Allied planes whirling about in dogfights, many of them amid the flak of ground and ship gunners. By 1403 the enemy was in full retreat, leaving behind so many of his numbers that the defenders were hesitant to believe their own results. An accurate estimate was impossible; much duplication occurred, but Fighter Command listed no less than 49 Zeros and 32 dive bombers as victims of its planes, with the F4F's claiming 30 kills and the P-40's 25--all this at a cost of 6 Allied planes, AA gunners claimed seventeen more for a total of 98 out of the original estimated force of 120.54 Although the enemy escaped with no


more than a bare handful of planes, he left behind a reminder of his visit. Very real damage occurred on the ground and on three vessels off Guadalcanal, two of which had to be beached off Lunga. Altogether losses afloat and ashore reached twenty-five killed, twenty-nine wounded, and twenty-two missing, but they could have been far more disastrous.55 And now the enemy would have much handier targets. The invasion of New Georgia lay only a few days away.

In his planning General MacArthur had set 15 June as a target date for initiating the dual set of ELKTON operations, which in the Solomons phase would carry the Allies up to the southern end of Bougainville and would consume the remainder of 1943.56 But acquisition of a base on Bougainville was dependent upon the presence of adequate air cover. So long as the nearest major air base was no nearer than Guadalcanal this cover could not be provided; the jump was too long, the Allies lacked a preponderance of air superiority, and COMSOPAC did not possess a decisive superiority in naval surface forces, particularly in the all-important carrier category. The obvious intermediate step was Munda.57

Here on New Georgia Island the Japanese had expended much effort since they were first reported in the area in August 1942. Originally using it as a staging point on the supply route to Guadalcanal, after the November defeat they undertook construction of an airdrome near Munda Point on the southwest coast, a location which made the field almost immune to an invasion from the sea. Ever since discovery of the cleverly concealed strip on 3 December, Munda had been hit almost daily by all types of aircraft, supplemented with ostensibly destructive bombardments by surface craft, but neither bombs nor shells could hock out Munda for longer than a day or two at best.58 Nevertheless, despite the easy construction possible at .Munda and Allied inability to disable the strip permanently, the Japanese never had been able to conduct major operations here or from the strip laid out late in December 1942 over at Vila on near-by Kolombangara.59

 Down on Guadalcanal all services pushed ahead in their preparations for the next advance. By July, two bomber and two fighter fields were ready, and of equal importance was the further improvement of facilities for bulk gasoline storage, including the construction of a submarine fill-line which brought aviation gasoline direct from tankers moored off Koli Point to a tank farm. As advance fields, there now were the two strips on the Russells, and it was obvious that even with the deficiencies


of Carney Field on Guadalcanal and certain operational difficulties, this New Georgia operation was to be supported far better than the original landing on Guadalcanal.60 Admiral Halsey issued his basic operational plan for the New Georgia assault on 3 June 1943. Rear Adm. Richmond K. Turner commanded the amphibious forces; Maj. Gen. John H. Hester led the New Georgia Occupation Force, whose major unit was the rein- forced 43 d Infantry Division, while held back on Guadalcanal as general reserve was the 37th Infantry Division. Hester had in addition two Marine raider battalions, the 1st and 4th. It was believed that altogether these forces should be able to wrest the entire island group from the enemy within a 30-day period. In case of dire need, use could be made of the veteran 25th Division then actively engaged on Guadalcanal in a training program to rectify its errors of the first campaign. In any case, there should be enough ground force available for the task.61 To be sure, Halsey did not regard the means as sufficient to justify a frontal assault on the Vila-Munda area, an opinion which led to the decision for the Western Landing Force to capture Rendova Island first. Simultaneously the Eastern Landing Force would land at Vim Harbor, Wickham Anchorage, and at Segi Point, where a new fighter strip would be constructed.62 A subsidiary operation had as its objective the cutting of communication lines between Kolombangara and Munda to prevent the Japanese from reinforcing their Munda garrison. To effect this end a task force under the command of Col. Harry B. Liversedge, USMC, a force composed of two Army infantry battalions and one Marine raider battalion, was to land at Rice Anchorage, then move westward and secure Enogai Inlet, thus breaking enemy communications through Bairoko Harbor. Subsequent to the ejection of the enemy from New Georgia, the forces were directed by Halsey to capture at the first favorable opportunity the remaining enemy positions in the New Georgia group.63 These were the goals for the amphibious forces whose troops were to be covered by a powerful task force of surface units and by an air force whose strength now surpassed anything yet available in the Solomons campaign.

On 18 June, COMAIRSOPAC assigned to all his air units their tasks for the New Georgia operation. As usual the fighters would be hard pressed to carry out their assignment; they would provide cover for all the forces in the Guadalcanal and New Georgia areas and all shipping between these two points.64 Admiral Fitch warned them to guard


particularly the attack transports and the supply ships moving in the Slot, and the fighters were requested as well to maintain an air cover over those surface combatant groups involved in the New Georgia operation. Furthermore, COMAIRSOLS was advised to be ready on six hours' notice to provide aircraft for dropping supplies and equipment into the New Georgia area.65

In this advance beyond the immediate vicinity of Guadalcanal, it was necessary to devise a fresh method for controlling aircraft in the forward area. Fitch was moving forward to Guadalcanal, but he could not handle New Georgia operations from Henderson. Accordingly, there was established a new unit known as Headquarters, New Georgia Air Force (COMAIR New Georgia), composed of personnel drawn from the Forward Echelon, 2d Marine Aircraft Wing, and this was attached to the New Georgia Occupation Force. COMAIRSOPAC directed that operational control of all aircraft assigned to tasks in the immediate vicinity of New Georgia would pass upon take-off to this new organization, which would also control all direct air support of the ground operations in the central Solomons.66 There were additional preliminary arrangements, such as establishment of four air liaison parties to advise ground commanders as to suitable targets and forces for air attacks; and there was provision for fighter control in the forward area.67 The whole undertaking was coordinated with the Southwest Pacific, where MacArthur's forces were to seize Woodlark and Kiriwina Islands on 30 June--that is, D-day for Rendova--and General Kenney's Fifth Air Force was to strike at targets in the northern Solomons, New Ireland, and eastern New Britain areas.68

The air and ground forces facing Halsey's task forces were considerable. It was estimated that of the 40,000 Japanese in the Solomons some 8,000 to 10,000 were on New Georgia, and of these probably one-third guarded Munda.69 Moreover, there were advantages of distance favoring the enemy. His powerful Bougainville bases were closer, his troops had been granted time to construct coconut-log pillboxes, blockhouses of coral rock, and strong defensive positions, and once again he had filled up his air bases with replacement aircraft.70 To overcome the air defense, on the morning of 30 June COMAIRSOPAC had ready to fly in the forward area a total of 455 combat aircraft out of an assigned total of 533.71 As D-day approached, B-17's and B-24's concentrated upon Poporang, Ballale, Kahili, and the Buin area, Marine dive and torpedo bombers struck at Munda, Vila, and Rekata Bay, while B-25's


and TBF's carried out daily low-altitude armed shipping searches up and down the Slot. Japanese aircraft and submarines did what they could to interfere with Halsey's flow of supplies into Guadalcanal, but they fell far behind in the race.72


At 0642 on the morning of 30 June, six transports lay off Rendova Harbor on the north side of Rendova Island. As the landing forces moved ashore, quickly overcoming light resistance, Fighter Command assumed its burden of defending the amphibious forces from inevitable Japanese reaction, and it entered the most active four-day period in the history of the Solomons campaign.

Thirty-two Allied fighters covered the operation from altitudes ranging from 5,000 to 20,000 feet, but they had little work until 1100 when the enemy sent down his first fighter sweep of the new offensive. Sixteen of his thirty-odd Zeros were shot down, leaving Rendova relatively quiet for four hours.73 Then at 1500 came another attempt, this one by a mixed force of fifty dive bombers, torpedo-carrying Bettys, and Zero fighters. The Bettys circled the task group, concealing them- selves against the island background, then swept in at minimum altitude and high speed to drop their torpedoes at 500-yard range. Corsairs and Wildcats tore into the bombers, aided by the AA gunners, and one by one every plane went down, but not before they had scored a hit. McCawley, the flagship, took a torpedo amidships, sinking at 2023 as the result of nine torpedoes from friendly PT boats.74 Early in the evening the enemy was back again, this time with a motley collection of some thirty Vals, Rufes, and Daves, plane types which indicated that the enemy air commanders were dipping low in their bins to continue the assault. Eighteen more went down before the Allied fighters, raising the enemy's loss for the month to a total of 254. This was a heavy price; Allied losses for the month had run to thirty-six fighter planes and thirteen pilots, whereas the enemy had sacrificed many multiplace planes, but he seemed willing and able to pay it.75 Next day his dive bombers were back again for another try at the shipping, but with very little success, and it was apparent that he had underestimated Allied ability to maintain adequate and continuous fighter cover 170 miles from Guadalcanal. Except for a sneak raid on the third day of the landings, one which caught the troops unloading LST's on the beaches, the Japanese had little to show for their effort.76


The daily patrol of thirty-two fighters, supplemented by additional planes when necessary, turned aside the best the enemy could send down, but it was no easy task for Fighter Command to maintain this patrol, which called for a commitment of 96 aircraft, leaving between 80 and 100 to meet all other requirements. Harmon doubted that it could have been achieved without the aid of the two Russells strips.77 Certainly there was no help from the weatherman during this period; in fact it had been necessary to abandon the heavy bomber neutralization of the Bougainville bases because of the wretched weather. Over Rabaul conditions were not much better; air photo coverage during the first week of the invasion was impossible. From New Guinea, General Kenney was able to complete a few strikes against the air installations around Rabaul town, but up to 4 July it had been impossible to send so much as a single heavy bomber strike from the Solomons bases to Bougainville, and even when the weather cleared no fighters were available for escort.78 The weather, with its low ceilings, favored the enemy pilots, permitting them to slip in close before detection, and it prevented interference at the home bases. Had the Bougainville strips been heavily damaged, it is highly probable that Japanese reaction to the landings could have been greatly diminished. Even under the prevailing handicaps General Twining found the results very encouraging, citing a loss by D plus 9 of 190 enemy aircraft to an Allied loss of 32, but by then his fighters were in a state of semi-exhaustion.79

Twining saw one answer to his problem-more P-38's to form a full P-38 group-and Harmon passed along his request for a complete group of seventy-five aircraft plus a reserve of 50 per cent by 1 September. He stressed the nature of amphibious landings, their requirement for adequate fighter cover, and the necessity for counter-air force missions by escorted bombers which were lacking in the Rendova landings; hence, the critical need to augment the twenty-nine P-38's on hand as of 9 July.80 There was no denying Twining's predicament. As matters then stood the fighter shortage contributed heavily to Fighter Command's inability to isolate the battlefield, but as in all similar circumstances, General Arnold was forced to weigh the request against the needs of the several theaters of war. In the current situation the sole source of P-38's lay in the commitment to North Africa, and this Arnold did not wish to weaken.81 Once more Harmon and Twining would carry on with less than they believed necessary.

Under the effective cover provided by the fighters, the Seabees of


Acorn 7 rushed construction of the new strip on Segi Point, so that by D plus 10 it was possible to report that the 3,300-foot strip was available for limited operations, Harmon was delighted. Now, he felt, there was an appreciation of air operations radically different from that encountered upon his initial arrival in the theater in July 1942.82 Segi had its troubles; it was an unsafe field with a high accident rate, but it is probable that this strip lying only forty miles from Munda saved many a damaged plane and worried pilot, who might not have been able to hold out for another eighty miles down to the Russells.83 Despite all its shortcomings, this field offered a refreshing contrast to the original pace of construction at Henderson Field, and there were other signs of progress as well. Harmon took increasing satisfaction from the manner in which the Thirteenth Air Force was being integrated and employed as a unit in the scheme of operational control exercised by Admiral Fitch. Very soon in the future General Twining would replace Admiral Mitscher as COMAIRSOLS. It would be a trying assignment, but one which for the first time would place an AAF commander in direct operational control of all the air units in the Solomons. The shift in command was scheduled for 25 July, and to meet it Harmon attempted to bolster Twining's staff so that the rear echelon operating from Espiritu Santo would be able to carry on without distracting the Thirteenth Air Force commander from his new responsibilities as COMAIRSOLS.84

All this represented a decisive advance over the critical months of 1942. Whereas in the early days of the struggle for Guadalcanal the enemy had commanded the sea approaches to the island and all of the air except a small area around Henderson Field, now he was unable to gain mastery of either element. Thus far he had failed to overwhelm the fighter defenses. When he tried to send down the Tokyo Express on the night of 5/6 July, he met a task force under Rear Adm. Walden L. Ainsworth, and there in the Kula Gulf occurred another of those wild night actions so typical of the Solomons naval battles. Enemy losses were judged high but later were scaled down to two destroyers, at a cost to Halsey of the Helena and the Strong. One week later came a second attempt and again Ainsworth drove off the Express, this time paying for it with his light cruiser Honolulu and the destroyer Gwinn and suffering heavy damage to several other units of the task force.85

These clashes represented the surface counterpart of the air cover provided by Fighter Command and, although costly to Halsey, they


achieved their purpose. They removed the threat to the Allied landings on the north coast of New Georgia, and by preventing the enemy from continuing his use of the Kula Gulf route in supplying the garrisons at Vila and Munda, they forced him to send ships and barges all around Vella Lavella to the west side of Kolombangara, concealing them in anchorages along the south coast of the latter island. En route they became fair prey for ranging fighters and bombers.86

For a few days in the second week of July the Japanese reduced the scale of the daily air assaults, and then having gathered their strength they came down on 15 July with the largest single formation since the epic battle of 16 June. Twenty-seven bombers, mostly Bettys, were covered by forty or fifty Zeros, who desperately tried to clear the air of the forty-four Allied defenders, but again with no luck. No less than fifteen of the Bettys went down in the combat which raced northwest over Vella Lavella, and they were accompanied down by thirty Zeros, or more than half their escort, at a total cost of three COMAIRSOLS fighters and pilots.87 This was an astonishing ratio of loss: 15 to 1 in terms of aircraft and nearly 40 to 1 for personnel. Could the Japanese long sustain such attrition? If their air commanders seemed unable to learn not to bunch their aircraft in combat, they indicated no signs of weariness in filling up the fighter pens at Kahili, Buka, and Rabaul, for despite grievous enemy losses, Allied reconnaissance planes reported heavy fighter stocks available.88

Nevertheless, there was some evidence that the Japanese were feeling the pressure of their lavish expenditure. After the 15th they virtually abandoned their attempts to attack the New Georgia positions in frontal air assault. To be sure, the beaches and ships were not left un- molested, but henceforth the enemy would use his fighters with less abandon, avoiding reckless destruction, and he would inflict all possible damage during the hours of darkness.89 After 15 July there would be no lack of red alerts on Rendova and New Georgia as the ever reliable coast watcher upon Vella Lavella monotonously passed along his laconic report: "Hear plane bearing NE, course SE."90

The weather, which had insulated the enemy's airfields against the heavy strikes planned against them and saved his shipping from the prowling B-25's and TBF's, could not hold forever. Counter-air force operations went on against Rekata Bay and Munda during the earliest stages of the operation, but no heavy bombers could reach the Buin area until 5 July, and even then the nine B-24's found no shipping targets


dropping on Ballale and Munda instead. But by the following day the airfields on Bougainville began to feel the weight of the heavies. Henceforth, B-24's and 8- 17's were over the Bougainville installations with increasing strength and frequency, both by day and by night. Furthermore, Japanese engineers had to contend with delayed-action bombs imbedded in their runways, as well as with the fragmentation clusters which shattered Zeros and Bettys parked in unprotected areas.91

B-25's of the 42d Group now entered the battle and it was immediately apparent that the long hours of training in minimum-altitude tactics conducted by Col. Harry E. Wilson down in Fiji were yielding substantial dividends. No Japanese vessel was safe from the Mitchells, and day after day as they ranged up and down the Slot they caught small cargo vessels, barges, and even an occasional destroyer, strafing and bombing from 75- to 100-foot altitude, nearly always with fatal effect.92 Twining was highly pleased with the B-25's, indicating only two worries: interference with their operations by the Navy, and lack of adequate parts to keep the medium bombers in the air, but despite these handicaps they continued to perform well.93 They accomplished much more than the execution of shipping strikes; in fact, the majority of their missions were bombing and strafing attacks upon ground targets on New Georgia or Kolombangara, varied by an occasional mission up to Ballale or over to Rekata Bay, but when surface targets appeared, B-15's of the 42d Group were ready not only to hit them but to strike with such accuracy that the vessels either sank immediately or were left badly damaged and afire.

While medium and light bombers continued to threaten the enemy's shipping and the Tokyo Express during his attempts to keep alive the besieged garrisons, Strike Command now was able to throw far heavier forces against the harbor areas at the south end of Bougainville. Two major efforts were made during mid-July, one on the 17th and another on the 18th, of which the first was the more successful. For several days prior to the I 7th the heavies had been hitting the Kahili, Ballale, and Buka areas-thirty-six were up on the night of the 13th and three nights later thirty-one more dropped forty tons of fragmentation clusters on Kahili, where the enemy's night fighters were growing troublesome.94 Then on the morning of the 17th, Strike Command sent a carefully planned and coordinated strike against the shipping lying off Buin. It was a powerful force for the Solomons air units: 36 SBD's,


35 TBF's, and 7 B-24's, all escorted by 114 fighters. B-24's struck first, claiming hits on two cargo vessels, and their crews could see the Zeros struggle to take off from Kahili to protect the ships in the harbor from the assaults of the dive and torpedo bombers. The defenders were too late. At altitudes ranging from 300 to 7,000 feet Corsairs picked off the hapless Zeros, while over the anchorages the SBD and Avenger pilots picked their targets and made their runs. It was a fine day's work. In the air the ten-minute engagement had cost the enemy at least forty-four Zeros and four float planes; in the harbor the attackers claimed serious damage or sinking of probably three destroyers, a PC, two cargo vessels, and an oiler, and one AK had been run ashore. The total cost was five Allied planes.95

Next day the bombers were back again, this time fewer light ones, more B-24's, and no less than 134 fighters, the most powerful escort yet sent north. Employing similar tactics, except that fifteen B-24's were detailed to hit the Kahili runway and revetment areas, the low-level bombers again inflicted heavy damage on ships in the harbor, while twenty-one Zeros fell to the escorting fighters. This time the cost was higher; nine fighters and one TBF failed to return, but since these engagements occurred over the enemy's home fields nearly 340 miles from Guadalcanal, they were the more remarkable.96

Buin was not yet completely untenable for the Japanese surface units, but it had become highly dangerous, and it was the task of COMAIRSOLS to keep it so. Again on 22 July a strong force hit the area, damaging a number of ships. B-24's missed the warships off Buin that afternoon in an independent attack, but the implication was plain.97 Buin was no longer a safe haven. Three times in one week striking forces involving 150 planes each had been sent against the shipping at Buin, and the lesson could hardly be overlooked by the Japanese commanders.98 By the end of July the amount of shipping there was substantially lower than the average of previous months. Nevertheless, the enemy continued to send down surface craft, smaller ones now, and COMAIRSOLS planes continued to sink them. Unfortunately the weather interfered from time to time, as it did on the 29th, when it caused cancellation of a coordinated strike of 220 planes, but from mid-July forward the local commanders at Buin could expect a heavy assault at any time favorable weather conditions and the presence of surface targets happened to coincide.

It was apparent that the scale had tipped sharply against the enemy,


whose daylight retaliation against Rendova was now almost negligible. By skilful timing the Japanese managed to break through on 2 I July, coming in with sixty-odd planes only a few minutes after the Rendova fighter patrol had gone off station at 1700, but by placing P-39's and P-40's on Segi it was possible to put an end to these dusk attacks which the enemy had attempted on several occasions.99 On 25 July the Japanese tried one more daylight strike, sending down thirty to forty dive bombers with thirty Zeros to strafe the beaches, but the Rendova patrol intercepted, knocking down eight of the escort and forcing the Vals to shed their bombs aimlessly.100 Thereafter they subsided except for night harassments.

In pursuance of Halsey's plan, command of COMAIRSOLS passed from Adm. Marc Mitscher to General Twining on 25 July, a day which marked a milestone in the record of the Solomons campaign.101 COMAIRSOLS had come a long way since the early days in the Solomons when each of the services was anxious to maintain its own prerogatives, when the AAF commanders struggled to attain some modest measure of operational control over their own aircraft, and when the file of cable requests for additional planes grew fat by the end of each month. The Thirteenth Air Force even now continued as a training and administrative air force compelled by circumstances to forego all operational control over its own aircraft in the forward area, where the cutting edge of the South Pacific air units was in daily contact with the enemy. But Air Command Solomons, into which the Thirteenth had fed its men and its planes, had done remarkably well. Since 2 April the air strength at the disposal of COMAIRSOLS had more than doubled, moving up from 235 to 539 planes, and the daily average of fighter aircraft assigned had risen from 108 during February to 281 during July. In the 26-day period opening with the landing on Rendova on 30 June, Fighter Command alone claimed 316 enemy planes at a cost of 71 aircraft and 40 Allied pilots, of whom 11 were AAF men.102

For the first time in the Solomons operations the air units were to be commanded by an AAF general. Twining brought with him his own staff, 70 officers and 200 men from all services, leaving behind him at Espiritu Santo Brig. Gen. Ray L. Owens as deputy air force commander of the Thirteenth. His chief of staff was Capt. Charles F. Coe, USN, and his Strike Command was led by Col. David F. O'Neill, USMC; but both Fighter and Bomber Commands of COMAIRSOLS


went to AAF officers. Brig. Gen. Dean C. Strother now commanded the former, Col. William A. Matheny the latter, and both retained their respective commands within the Thirteenth Air Force.103

Twining's new command, while not exhausted, was rather well used up, especially its fighters, which had carried a very large share of operations during the first twenty-six days. Immediately he began to move the P-38's to a rear training area, retaining only a few as night fighters; the fresh P-38 unit would return by 1 September. Meanwhile, he found the performance of the mediums and heavies gratifying, although there remained the unsolved problem of congestion in the forward area. Recommended improvements had not yet occurred, the airfields on Guadalcanal left much to be desired, and the second bomber strip at Carney continued under construction and would not be ready for several weeks.104 But the general outlook was bright as COMAIRSOLS faced the enemy. His pilots had inflicted appalling losses upon the Japanese, whose gunnery remained poor and whose unarmored planes, lacking self-sealing fuel tanks, blew up with pleasing regularity. No less than 267 had been claimed in July alone.105 For the remainder of the year COMAIRSOLS would exceed the planned monthly average of 150 planes destroyed, but Twining would need all his strength. The enemy offered slight hope that he would abandon the field without a bitter struggle.

The Assault on Munda

The early phases of the New Georgia operation had proceeded auspiciously. Shielded by the superb fighter cover, the amphibious forces had made their landings on Rendova, then at Oleana Bay in the Wickham Anchorage area, at Viru, and finally, after some delay, at Rice Anchorage. By 5 July this last force, under Colonel Liversedge, was ready to advance south on Bairoko, whose seizure would prevent the enemy garrison from receiving reinforcements from near-by Kolombangara.106

These forces represented flanking movements. Halsey and his planners had determined that should the Japanese not react too strongly to these minor landings, it would be possible to push ahead with the assault upon Munda. Since the reaction had not come, Admiral Turner and General Hester were authorized to proceed against the major objective. Accordingly, early in the dark hours of 3 July, leading elements of the 43d Infantry Division moved across the Blanche Channel to Zarena


Enemy Defense: Munda Field


Beach, approximately six miles east of Munda on New Georgia. By 8 July all units lay behind the Barike River, in position for the attack upon the estimated 4,000 enemy troops, who also had to face Liversedge's three battalions coming down from the north.107 Thus far it was reasonable to assume that the schedule could be maintained and that within a 30-day period the entire New Georgia group could be wrested from the enemy.108

Unfortunately, neither the Japanese nor the terrain permitted realization of these sanguine hopes. Here on Munda the enemy enjoyed shorter lines of supply than at Guadalcanal; his strong Bougainville bases were closer, his naval forces were at hand, and he had organized an ingenious and thorough defense system based upon numerous mutually supporting pillboxes and log dugouts, all well concealed and powerfully constructed. Even his military tactics had improved. His troops, physically fit and well fed and equipped, seemingly had access to an inexhaustible supply of ammunition, and they were prepared to die rather than yield.109

New Georgia's terrain provided the enemy with a powerful ally. It is characterized by dense jungle with thick, almost impenetrable undergrowth and low ridge lines, possessing no well-defined spurs or landmarks, nor any open country such as exists on the north coast of Guadalcanal. For the ground forces all this meant jungle warfare of a type even more vicious than that encountered in the first campaign; here there were severe restrictions upon freedom of movement and visibility often was limited to a few feet.110 Such factors as these did not help the air forces in their attempts to operate closely with the painful advance of the ground units.

As the battle progressed it quickly became apparent that consistent close support of ground troops from the air, as originally planned, was impracticable.111 The jungle simply did not permit it. Denseness of growth made the detection of suitable objectives impossible until friendly troops were too near the prospective target for safety; this was not the open warfare of the flat or rolling lands of western Europe, where in good weather pilots of the tactical air forces could view miles of enemy armor winding along below them. This was New Georgia, where the battalions painfully inched their way forward through the jungle, frequently under extremely heavy mortar and machine-gun fire coming from positions they could not locate, and where pilots overhead could see neither friend nor enemy. Neither the air liaison


parties on the ground nor the air observers in the spotter planes could identify enemy positions; in fact, there was not a single instance during the entire operation when aerial observers were able to repor movement of enemy troops.112 A part of the difficulty arose from the scarcity of reliable ground maps. The standard map for the operation was a gridded aerial mosaic which included no detail other than the coast line, and in reading it troops frequently were even unable to locate their own position on the map-much less that of the enemy. Some minor success derived from the use of smoke shells as target designators, but this method depended upon close timing and upon the maintenance of reliable radio communications between air and ground units. Both were lacking. Hence it was necessary to employ the bombers primarily against those points which were too conspicuous to permit concealment, such as bivouac areas and supply dumps, or against known artillery positions; these at least were well clear of friendly troops, and whenever possible were selected a day in advance to allow time for proper briefing of the air units involved.113 In these attacks upon the New Georgia targets, Strike Command furnished TBF's and SBD's, the former carrying 2,000-pound "daisy-cutters" and the latter a similar type of 1,000-pound bomb, but frequently both heavy and medium AAF bombers joined the Marine planes.114

As the twin drives pushed toward the Munda strip, it was apparent that progress was slower and resistance more stubborn than had been anticipated. On the north, Colonel Liversedge's battalions had taken Enogai on the 9th with the help of the light bombers, but thereafter this force was unable to interrupt completely the enemy's line of communications, neither denying him reinforcements nor preventing his withdrawal. Similarly, the forces advancing westward were making slow progress in the face of heavy resistance, constant nightly infiltration, and harassment amid the dense jungle; by 12 July the 169th Infantry Regiment was unable to advance farther and the 172d, which had disengaged from the main attack and turned south through the mangroves to establish another beachhead at Laiana, was without food or water or means of evacuation.115

Progress was slow, much too slow to please Halsey. On 13 July he orally directed General Harmon to assume full command of and responsibility for ground operations and to take whatever steps he deemed necessary to facilitate capture of the airfield. Harmon at once proceeded to New Georgia, where he relieved General Hester and


gave to Maj. Gen. Oscar W. Griswold, commanding general of XIV Corps, command of all forces on New Georgia. Henceforth, the assault progressed more favorably even though it continued at a slow pacee.116

With the help of the light and medium bombers the ground units painfully edged toward Munda strip, which the 169th could see by 16 July. Many of the air missions were summoned by Colonel Liversedge in the Bairoko area, but the immediate Munda area absorbed increasingly heavy bomb loads as the attack troop units slowly closed in on the field.117 On 25 July came the final push on the Munda strip, an attack carefully coordinated with a naval bombardment by seven destroyers and with the heaviest air bombardment yet seen in the South Pacific.118 Against three target areas covering less than one-half square mile, a force of 171 light, medium, and heavy bombers dropped more than 145 tons of bombs within a period of slightly more than a half hour. Later in the afternoon, ten more B-24's hit Bibolo Hill, just north of the strip, to add to the morning's destruction.119 But even under this weight of metal the Japanese resisted, apparently determined to exact the highest possible price as they faced the 43d Division approaching from the east and the 37th driving in toward the coast north of the field. Even when the 43d had reached the outer taxiways of Munda by 1 August, cautious Harmon was reluctant to predict the date of Munda's fall or the time necessary to evict the enemy from all New Georgia.120 But now his pessimism was overdone. On the afternoon of 5 August, less than nine months after the Japanese had landed on 21 November 1942, all organized resistance ceased around Munda field. There remained a pocket of enemy resistance at Bairoko Harbor, but when two battalions of the 25th Division occupied the east bank of the harbor on 25 August, they met no opposition. The enemy had evacuated his forces to Arundel and to Kolombangara, where he had begun work on an airstrip the previous January.121

Now it was over. The new task was to rush the Munda strip into a condition suitable for operation by the planes and crews which had aided in its capture, for possession of Munda brought all the Solomons bases within range of the light bombers and fighters. The New Georgia operation was a sharp portent of the direction in which the tide of air strength had turned; the enemy had lost his grip only 120 nautical miles from his strong bases at Kahili and Ballale, whereas Allied aircraft had worked from their major bases on Guadalcanal, 180 miles distant, supported


by the forward strips on the Russells, 120 miles down the Slot from Munda.122

The air phase of the operation had been of greater magnitude than ever was possible in the Guadalcanal campaign. During the 37-day struggle Allied air units reported the destruction of 358 enemy aircraft, incurring a loss to themselves of 71 fighters and 22 bombers of all types.123 Fighter cover had been superior. Both COMAIR New Georgia and the air liaison parties had called for strikes, whose execution required 1,833 bomber sorties by planes of all services, although the greatest number of these (1,649) were flown by the light bombers of the Marines; and this figure excluded the missions ordered directly by COMAIRSOLS against similar targets.124 Throughout the New Georgia campaign the heavies continued to hit at Kahili, sometimes in coordination with the light bombers, and Allied fighters were forced constantly to remain on the alert against Japanese fighter and dive-bomber sweeps, which became troublesome in the final week of the offensive. Allied ships off Rendova provided tempting targets but all efforts to hurt them were ineffectual; on 4 August a series of running fights with Zekes cost the enemy three fighters lost to Corsairs, and eight more shot down by ten AAF P-40's, with Lt. L. B. Shuler personally accounting for four.125

There was divided opinion over the effectiveness of the air effort. Even prior to capture of Munda, General Harmon had praised the contribution of the air forces, particularly the fighter cover. But the terrain itself had worked against close employment of the air weapon, and the Japanese once more demonstrated ability to absorb tremendous punishment. Moreover, the practical aids to useful employment of the air weapon were absent. Adequate aerial photos had not been made available, several days elapsed before any aerial coverage could be obtained, and in many instances when the negatives arrived they had not covered the desired areas.126 Those responsible for the provision of close support for the ground troops were not entirely content with the results achieved, despite the handicap of terrain factors. The operations officer of Strike Command believed that the commander of ground troops failed to appreciate the utility of the air weapon, and because missions could not be dispatched unless specifically called for, on many days air power sat idle on the strips. Even when requests finally came through, the targets were from 1,000 to 2,000 yards in front of the lines.127


It was apparent also that the heavy air attack upon the enemy positions at Bairoko prior to the assault by the Liversedge force had failed either to dislodge or disrupt the Japanese sufficiently to make the attack successful. Despite the very high count of enemy planes destroyed in the air and the ships and barges sunk by the B-25's and light bombers at Webster Cove, at Hamberi, up at Kahili, or out in the Kula Gulf, it proved impossible to cut off New Georgia completely from the enemy's adjacent bases. He continued to evacuate a number of his troops from Bairoko. Nor were the invaders ever able to stop the nightly harassment by light bombers, an activity which constituted probably the most effective Japanese air operation. It was particularly troublesome in the Enogai area, where almost nightly the float planes from Faisi or Vila dropped light antipersonnel bombs, inflicting a number of casualties and much loss of sleep.128

For the first time in the Solomons campaign, excluding the improvised B-17 assistance at Mt. Austen, the air transports participated actively in the battle. Unlike the Guadalcanal experience, this time air commanders had foreseen the necessity for providing air drops, although ground commanders were informed that air transport should be used sparingly and only in the absence of all other means of transportation.129 The emergency quickly arose when the 169th and 172d Infantry Regiments drew up along the Barike River area, where deep swamps and very thick jungle reduced progress to a snail's pace; men were exhausted and units were weakened as the result of hand-carrying supplies for distances of 2,000 to 2,500 yards. Then the 172d ran into further trouble as it turned to the beach and had to rely in part upon air drops.130 The major problem for the C-47 pilots was to secure an adequate supply of parachutes and containers and to drop them within reach of friendly troops. Some of the materials went to the Japanese and some hung beyond recovery in the tall trees, soon making it necessary to notify the ground units to recover all possible parachutes and containers, so scarce was the supply of these items. But by 16 July more than sixty tons of supplies, including ammunition and water, had been dropped to the northern force and to the battalions east of Munda. These drops were made by C-47's of the South Pacific Combat Air Transport (SCAT) organization, to which planes of the 13th Troop Carrier Squadron were assigned as needed.131

With Munda strip in Allied hands, the immediate tasks were to defend it and to reconstruct it. Early examination of the field indicated


that damage had been lighter than originally anticipated and that a minimum expenditure of labor would be required to place the field in operation. Heavy interference by the enemy might be expected. But despite the short distance to Kahili, and despite the fact that the Japanese had nearly 5 00 aircraft in the Rabaul-Kavieng-Bougainville area, they were unable to execute a single effective attack against the men and equipment on and around the field-in fact they scarcely even made the attempt. One effort occurred on 10 August when an estimated fifty enemy fighters came down to make a low-level attack upon the bull- dozers, but eleven P-40's and three P-39's turned them aside. Three days later they tried again but failed, and thus Munda's reconstruction moved rapidly ahead.132

The 73d Seabee Battalion had inherited the task of widening, resurfacing, and regrading the original 4,000-foot runway and the two taxiways which the Japanese had added. On 9 August, Acorn 8 joined the 73d, and now as equipment arrived it was put to work. Power shovels bit into Kokengolo Hill to make the dugouts, bulldozers pushed out the roads, and rollers and dozers spread out the coral on the strip which was dry but soft. Construction went ahead with extreme rapidity, and on the afternoon of 14 August two AAF P-40's of the 44th Fighter Squadron landed at Munda to remain overnight, thus becoming the first operational aircraft to use the newly captured field.133 When the Army's 828th Engineer Aviation Battalion joined the naval construction force early in September, the project forged ahead even more rapidly, and by 5 October planes were able to use the entire 6,000 feet of runway. The end of each day's labor brought rain, and each night brought down the enemy's raiders who did their utmost to disrupt the work, but it moved rapidly ahead regardless, although it was necessary to abandon night work.134 As rapidly as possible refinements were added to Munda, new parking aprons and new wide taxiways appeared, and to all of this the 828th made a major contribution, aided on some projects by the 131st Engineer Regiment and cooperating with the 73d Seabees on othes.135

Very quickly Munda became a key installation in the Solomons chain. Its excellent coral, easily available in quantity, permitted construction of a runway capable of withstanding very heavy operational loads, and Twining used it to its fullest capacity. Before many weeks had passed, traffic at Munda exceeded that of any other field in the South Pacific; during the month of October the average daily departures


and arrivals approximated 400, and on one day the peak reached 564 aircraft of every type. Its major handicap was the hill which interfered with take-off to the northeast, but despite this hazard Munda remained the best of all the fields.136

Seizure of Vella Lavella

While the engineers tackled the coral of Munda with their earth movers, the troops of the XIV Corps faced the task of evicting the enemy from the Bairoko area on New Georgia and from the other islands of the central Solomons. Halsey and Harmon were forced to decide immediately whether to send the corps directly against the Japanese base at Vila on Kolombangara, or to move around it. Vila was too swampy for development by COMAIRSOLS, hence very quickly the decision was reached to send the assault troops entirely around Vila to seize Vella Lavella instead.137 Accordingly, an amphibious force under Rear Adm. Theodore S. Wilkinson, consisting primarily of the 35th Infantry Combat Team, by-passed Vila, landing at Barakoma on the southeast coast of Vella Lavella early on the morning of 15 August.138 Other units landed on Baanga, where the enemy apparently had located some artillery with which he had been shelling Munda field. From 25 August to 23 September units of the XIV Corps were engaged in clearing Arundel and the smaller islands lying near by. No immediate occupation of Kolonibangara was contemplated; instead, a plan was developed whereby air and artillery bombardment alone would reduce the Vila area, and this scheme was followed until 5 October.139

On New Georgia all organized resistance had ceased by 25 August; Baanga, Arundel (after stiff fighting), and Wana Wana all were secured by 21 September; and by 6 October all patrols sent over to Kolombangara had returned negative reports. Next day a battalion of the 27th Infantry Regiment landed here and immediately occupied Vila.140 The policy of by-passing had paid handsome dividends. The enemy had lost heavily in barges and personnel as he clashed with Allied surface craft patrolling north of Vella Lavella, but the greatest gain was in the saving of Allied lives. The enemy could have made an Allied frontal assault on Vila a most costly business, for the place was studded with pillboxes, trenches, machine-gun emplacements, and coastal guns. Instead of facing the triple line of pillboxes on Kolombangara, the troops went ashore on Vella Lavella against minor opposition, gaining the site at Barakoma for an additional fighter strip and


forcing the enemy to abandon all his central Solomons positions except that over on Rekata Bay.141

All this ground activity was accompanied as always by a heavy drain on the air forces. The fighters had pretty well insulated the Allied bases against enemy daylight attacks, but they still could achieve very little at night, when the Japanese demonstrated increasing skill.142 Their float planes regularly heckled the PT boats operating against barges in the central Solomons, and they bombed the Russells and Guadalcanal at every opportunity and not always without damage.143 But none of these raids could equal the effectiveness of the attacks which General Twining now could throw against the Japanese from the bases on Guadalcanal, the Russells, and latterly Munda.

The second phase of air operations in the New Georgia campaign began as soon as it was possible to operate planes from the newly captured Munda strip. On 14 August the COMAIR New Georgia command post was officially opened on Munda Point by Brig. Gen. Francis P. Mulcahy, USMC; next day, from a tunnel driven into Kokengolo Hill, Mulcahy conducted the first full day of operations, when fighters were rushed up to cover the landings on Vella Lavella.144 The new beachhead lay dangerously close to Kahili; there would be no problem for the enemy to strike it. Mulcahy planned to hold a 32-plane cover over the landing operations, but he found it impossible to do better than a 12-plane patrol on the second day. On the first day his fighters had knocked down approximately twenty-five aircraft in two interceptions over Vella Lavella, and eight Corsairs had destroyed ten more at Kahili, but the enemy came right back.145 Operations from Munda continued hazardous, with enemy raiders overhead nightly, and beginning on the 16th the entire area, including the 73d Seabee camp, was subjected to intermittent artillery shelling during the daylight hours. Fortunately, this last hazard disappeared on the 19th with the capture of Baanga Island.146 But the night raids persisted and general operating conditions at Munda were far from satisfactory, notably communications and transportation. Plane-servicing and maintenance crews were inexperienced, spare parts were extremely scarce, and the incomplete taxiways were cramped and in poor condition, but the rapidity with which all these were overcome contrasted strongly with the slow progress of the Guadalcanal effort.147 Even more critical was the difficulty encountered in securing coordination and proper operation of the fighter direction centers during the


early phases of the Vella Lavella landing. The Barakoma air warning unit had been flung together with great haste on a 36-hour notice; properly trained personnel were not available, nor were spare parts, and when one of the most essential sets suffered a direct bomb hit on 17 August, radar detection became even more awkward.148 The Japanese took full advantage of the difficulties and of the weather as they continued to strike hard at the LST's drawn up along the Barakoma beach. From their Kahili base they could easily hit the area, and hit it they did-108 times in the first month, using an estimated total of 319 planes. Much of the air raid warning had to come from observation posts; warnings varied from none at all up to fifteen minutes, but with the usual four to eight minutes most men were able to reach their foxholes.149

In these Barakoma raids the enemy never consistently inflicted serious damage or casualties, and he paid heavily. In one week ending on 26 August he lost forty-three fighters, five dive bombers, and one float plane, at a cost to COMAIRSOLS of one F4U in combat.150 It proved impossible to prevent all attacks from reaching Barakoma-it was too close to Kahili. Radar warning at ninety-five miles was reasonably good, scrambles were prompt, and planes were off the ground in excellent time; but in the absence of continuous air cover and adequate radar facilities at Vella Lavella to furnish Fighter Command with an earlier warning, Allied planes could not prevent the attacks. The best that could be achieved by existing facilities and by maintaining ground alert at Munda was an interception after the enemy had reached his objective, and on occasion he struck a telling blow.151 The heaviest attacks regularly coincided with the arrival of additional assault echelons, of which the ninth lost one LST on 10 October, with fifty-two men killed or missing. But despite all attacks on the Barakoma site, construction moved ahead, and even before the strip was ready Navy fighter pilots returning from Bougainville with crippled planes had landed on the field or in the sea near by.152

While the fighters protected the newly won strip, medium and light bombers with their escorts were out stalking barges and having a good hunt. By August it was increasingly apparent that widespread employment of barges had become vital to maintenance of the enemy's outposts. He could move his heavier vessels from Rabaul over to Buka, thence down the east or west coast of Bougainville to Buin, or even directly to Buin from Simpson Harbor at Rabaul, although this last


route was used primarily by warships. But it was risky business to send heavy vessels down the Slot. The attempts to do so during July had resulted in disaster in the Kula Gulf, and thereafter the enemy relied very heavily upon barges of varying sizes, of shallow draft, and a fair turn of speed. They moved in groups down from Buin southeast to Timbala Bay on the north coast of Vella Lavella carrying men and supplies; then skirting the west coast of the island, they passed Wilson and Gizo straits to Kolombangara.153 When attacked, the groups scattered, hiding in small bays and creeks where their crews covered them with branches, making them extremely difficult to spot from the air. Allied pilots derived particular pleasure from tracking them down at every plantation on the west coast of Vella Lavella, and at Kakasa and Sagigai over on Choiseul.154 For the B-25's they provided a primary target. Day after day Colonel Wilson's B-2 5's ranged up and down the barge lanes, usually in pairs, though occasionally in larger flights with fighter escort, seeking out the hidden craft and often guided by the advice of local coast watchers who would report by radio the shifting location of the barges. At night PT's challenged them, sometimes in minor pitched battles. Fighters received instructions to strafe them as they returned to their bases upon completion of a patrol, although when heavy fighter cover was essential it became necessary to allow good targets to pass untouched. The B-25's enjoyed fewer diversions from their primary task, and by the end of August the two forward squadrons of the 42d Group had sunk or fatally damaged seventeen barges.155

The antibarge campaign went well. Pilots of P-39's found their heavy nose cannon particularly useful against the light surface craft, and B-24's prowled against them at night. Halsey was impressed and pleased with this phase of the Solomons campaign, and well he might be, for on 4 October four P-39's and four Corsairs destroyed sixteen barges on a single mission along western Choiseul.156

Air operations during August had cost the enemy an estimated 235 planes--55 of them went to the B-24's alone- -and in the last three months no less than 781 planes had been claimed by all forces in the Solomons.157 In September, Twining trebled the pace against the four Bougainville fields, including the new one spotted by Corsair pilots on 30 August near Kara, about seven miles northwest of Kahili. He sent up twenty-four bomber strikes, of which fourteen were by B-24's,158 and the accelerated offensive must have stung the enemy, whose offensive and defensive operations reached a new high level by the third week


of the month. Japanese night flyers now were highly skilled and active; on the single night of 14 September, Guadalcanal, Barakoma, and Munda all were attacked, COMAIR New Georgia reporting seventy-nine alerts, all raiders coming over in flights of two or three. Searchlights probed for them, AA batteries fired 2,900 rounds, but no planes fell. The Japanese used all their tricks.159 Tuned dipoles suspended from parachutes were dropped to baffle the radar operators; feints were made at Munda from one direction, then the attack delivered from another at a lower altitude; or the enemy would use Allied radio frequencies to confuse the defending fighters.160

Normally, it proved impossible to prevent the Japanese from inflicting some damage on the planes and personnel at Barakoma, Munda, or Guadalcanal. P-70's had promptly proved useless for high-altitude interception and to make some use of them they were put to work with the PT's, long heckled by the night raiders. The surface craft would serve as "bogey-bait" to attract float planes, which then would be attacked by the P-70, but the results were disappointing.161 P-38's and F4U's enjoyed a few successes; Corsairs shot down a bomber over Munda on the night of 19 September, and the following night Lt. Henry Meigs II, flying a P-38, broke all local records by shooting down within sixty seconds two Bettys over Fighter No. 1 on Guadalcanal.162 But these exploits merely emphasized the absence of any really adequate night fighter defense. None of the enemy efforts would equal the damage caused by Allied night raiders over the Japanese bases or the success of the Black Cats, the Navy PB4Y's, or the newly arrived SB-24's in their searches up the Slot and around Bougainville. The SB-24's represented the culmination of a low-altitude radar bombardment project undertaken in 1942 by the Radiation Laboratory of the National Defense Research Council, in close collaboration with Col. Stuart P. Wright, AAF liaison officer at the 1aboratory.163 These planes were fitted with special radar-sighting devices which permitted operation of the bomb-release mechanism irrespective of visual sighting of the target below. General Arnold assigned ten of them to the South Pacific under the command of Colonel Wright, Harmon welcomed them, and on 22 August they landed at Carney at a time when it was convenient to place them as a complete unit in the 394th Bombardment Squadron (H) of the 5th Group.164 Within five days pilots and radar men were ready to test their plane loads of electronic machinery against the enemy's barges and surface craft moving in the


blackness of the night. On 27 August they opened their own private campaign, and henceforth these planes were out over the shipping lanes almost nightly, bombing from 1,200 to 1,500 feet, often with surprising accuracy. Taking off from Carney Field in the evening, their flights averaged eleven hours each as they patrolled up the Slot to cover the shipping lanes leading into the Buin-Faisi area and to harass the convoys or single vessels moving down the east and west coasts of Bougainvi1le.165

Each succeeding week's operations served further to substantiate the claims made by Colonel Wright for his planes, which rapidly acquired an impressive reputation throughout the South Pacific. Perhaps their most distinguished performance came on the night of 28 September when five of the eight SB-24's aloft that night attacked a 10-or 11-ship convoy just north of Cape Alexander, sinking one destroyer and damaging other ships sufficiently to cause the convoy to reverse its course.166 Both Harmon and his bomber commander, Colonel Matheny, were highly pleased with the unit, whose advance claims had now been proved and whose equipment thus far had stood up remarkably well under the rigors of nightly operations.167 But the radar planes had created special problems of supply, of maintenance, and of employment, since for reasons of secrecy and because of their weak defensive armament, they could not be brigaded alongside normal B-24D's in daylight missions.168 Neither could regular maintenance men be allowed to service their equipment, and to the XIII Air Force Service Command the unit represented a most awkward problem.169 Harmon's solution lay in the creation of a special squadron to handle this type of equipment, and although his initial requests were turned down, his persistence--and doubtless the splendid record of the aircrews--convinced General Arnold. On 1 January 1944 the new 868th Bombardment Squadron (H) was activated, carrying on henceforth as a separate squadron directly under the XIII Bomber Command.170

Meanwhile, the hybrid unit continued to operate with success. The SB-24's sank or damaged numerous surface craft, they forced the enemy to place a night fighter cover over his shipping, and they provided an almost infallible means of locating night convoys and task forces.171 The project proved the faith of its developers; plane for plane, its SB-24's sank or damaged more surface craft at night from low altitude than did the heavy bombers attacking in daylight. By March 1944 the rapid retreat of the Japanese and their unwillingness


to risk shipping in the Solomons area would result in technological unemployment for the "Snooper Squadron," and at this time the radar planes would be authorized to serve as pathfinders for the high-altitude bombers.172

By the first week of October the enemy could no longer sustain his barge losses. Attempting to retrieve his surviving troops from isolated Kolombangara on the nights of 6, 7, and 8 October, he suffered severely, losing thirty-one craft to destroyers and twenty-five to the air forces,173 with fresh disasters yet to come at the hands of Allied destroyers. On 6 October, twenty-four B-2 5's had introduced the Kahili defenders to tree-level attacks with fragmentation clusters, and the week's score showed twenty-six enemy planes e1iminated.174 These events set a hot pace, one that seemed momentarily to retard Japanese air operations; by the 9th only a handful of planes rested on the Bougainville fields, and there was a sharp drop at Rabaul as well.175 Could the enemy maintain a pace in which he paid with an average of five planes for every Allied loss? He could and did. By 11 October Rabaul carried 294 planes; next day the Bougainville fields showed seventy-one, and the engineers were hard at work on Kara and the Buka and Bonis fields.176 A long fight lay ahead.

Observing all this from Washington, General Arnold found the situation somewhat confusing by the end of September. He knew that steadily the strength of the Thirteenth Air Force had increased and he believed that Twining now had available notably superior forces in the air. Furthermore, combat scores indicated that for some time the enemy had been losing planes at a varying rate of 2 to 1 to 5 to 1, yet now the favorable ratio seemed to be falling off, and in some engagements COMAIRSOLS was trading plane for plane. Why? he asked. What prevented more vigorous offensive operations? 177

General Twining replied, basing his report upon ten weeks' experience as COMAIRSOLS. He had the planes--this he admitted; he had available bombers in numbers and types sufficient to neutralize enemy bases if only these aircraft could be employed to their full capacity, but once again it was the shortage of fighters that prevented maximum bomber employment. He controlled 200 fighters in the forward area, a number deemed adequate to furnish escort for all his bomber strikes in addition to the sundry other air tasks were it not for the air cover necessary to protect shipping into Barakoma and to protect the destroyer task forces falling back down the Slot after their nightly operations


around Vella Lavella. These tasks simply consumed every available fighter, preventing the consistent heavy daylight strikes on enemy air bases which he believed were essential to fulfillment of the air mission. Recently the Japanese had begun to refuse combat in the air unless present in greatly superior numbers. Hence, Twining felt that enemy air power must be destroyed upon the ground. But because of the uncertainty of fighter employment and availability, it was impossible to execute any systematic plan of operations.178

The answer lay in the dispatch of more fighters. Already he had three of the five South Pacific AAF fighter squadrons in the forward combat area, and he expressed some concern over the Navy's reluctance to increase its own contingent of one squadron during August and September.179 For Twining this remained the paramount problem--not enough fighters. There was much pride in the performance of all branches. There was the great new base at Munda, and the new one at Barakoma which placed the air forces only eighty-five miles from Kahili, bringing Buka and Bonis within fighter range.180 But behind all the successes lay the feeling that there were not enough fighters to do the job. Demands upon the air forces had increased beyond any previous period because of the scattered bases and the variety of missions requested, and now Twining held a new directive ordering him to reduce the airfields in southern Bougainville by 1 November. He wanted more planes; lacking them he would "of course plug along and do it somehow."181

In Washington, General Arnold fully appreciated Twining's problems. He reassured COMAIRSOLS that despite the bombardment of messages he implied no criticism of the field commander, and he passed him a "well done." But Arnold advised him that he would have to manage with the squadrons currently available, because every fresh unit which the AAF could equip and train "must be thrown against the German until he is beaten."182 Thus, the requirements of global war bore down hard on the men in the South Pacific.


Table of Contents ** Previous Chapter (6) * Next Chapter (8)

Notes to Chapter 7:

1. Incl. (ltr,, Comdr. 3d Fleet to COMINCH, South Pacific Campaign-- Narrative Account, 3 Sept. 1944). The South Pacific Area, 20 Apr. 1942 to 15 June 1944, in USMC Hist. Div. files (hereinafter cited as Halsey Rpt.).

2. Halsey Rpt.; Harmon, The Army in the South Pacific. At Casablanca the Combined Chiefs of Staff had ordered the seizure of Rabaul. (Statement of Gen. Wedemeyer, Discussion of Availability of Forces for Implementation of ELKTON Plan, n.d.)

3. Hq. 43d Infantry Div., Report of Occupation of the Russell Islands, 8 June 1943; COMAIRSOPAC Opn. Plan 2-43, 15 Feb. 1943, in Ftr. Comd. Guadalcanal War Diary, USMC Hist. Div. files; 13th AF A-2 Periodic Rpt., 31 Jan.-28 Feb. 1943.

4. COMAIRSOPAC Opn. Plan 2-43, 17 Feb. 1943; Guadalcanal Escorted Missions during December 1942 and January 1943 through February 24, 1943, in A-2 Library.

5. Hq. 43d Inf. Div., Report of Occupation of the Russell Islands, 8 June 1943.

6. SWPA Intel. Sum. 92, 7 Apr. 1943.

7. SOPACFOR Office of Naval Air Combat Intel., Development of Jap Fields in Solomons Area, 15 Dec. 1942-15 Feb. 1943, dtd. 21-27 Feb. 1943. General Harmon reported eighty-two fighters on Buka, Kahili, Ballale, and Munda as of 21 February, although other reports placed seventy-five fighters on Kahili and Ballale alone. (USAFISPA G-2 Per. Rpt. 5, 13-20 Feb. 1943; CM-IN-12514 [2-25-43], Harmon to Marshall, 4998, 24 Feb. 1943.)

8. Plan for the Seizure and Occupation of the New Britain-New Ireland-New Guinea Areas, 28 Feb. 1943, especially Annex B. This plan represents a stage in the development of the ELKTON plans: I (12 Feb. 1943), II (11 Mar. 1943, and III (26 April 1943). ELKTON III was the plan actually followed, but the differences were in detail rather than principle.

9. ELKTON Plan; JCS 238/1,18 Mar. 1943.

10. Incl. C (JCS 238/1, 18 Mar. 1943).

11. CM-OUT-8736 (3-23-43). bk. msg., JCS to CINCSWPA, CINCPAC, and COMSOPAC, 2226, 23 Mar. 1943.

12. JCS 238/2, 20 Mar. 1943. At the Pacific theater conference held in Washington, Maj. Gen. R.K. Sutherland spoke for General MacArthur, Capt. Miles R. Browning, USN, represented Halsey, and Rear Adm. R. A. Spruance came as deputy CINCPAC to Nimitz.

13. JCS 238/5/D, 28 Mar. 1943.

14. JCS 238/4, 27 Mar. 1943; JCS 70th Mtg., 28 Mar. 1943, Extract from Minutes.

15. JCS 238/4, 27 Mar. 1943, App. C; CM-OUT-8736 (3-23-43), JCS to CINC-SWPA, 2226, 23 Mar. 1943; JCS 70th Mtg., 28 Mar. 1943, Extract from Minutes.

16. JCS 238/4, 27 Mar. 1943, App. C; JCS 238/5/D, 28 Mar 1943.

17. ELKTON Plan, Annex A, G-2 Estimate of Enemy Strength and Reinforcement Rate in New Guinea-Bismarcks Area, 26 Feb. 1943.

18. Harmon, The Army in the South Pacific.

19. Ibid., Incl. E, Principles and Policies Governing Employment of Air Force Units, 4 Mar. 1943; ltr., Gen. Harmon to Adm. Fitch, 3 Mar. 1943. This conference was held on 2-3 March 1943.

20. Harmon, The Army in the South Pacific.

21. Task Force 63 Opn. Plan 4-43, 11 Mar. 1943. Until 16 March, COMAIR-SOPAC was known as Task Force 63; on that date the new designation of Task Force 33 became effective. (COMAIR-SOPACFOR War Diary, 1 Mar.-30 Apr. 1943, Navy Register 8693.)

22. CM-IN-4883, 7130, 11160 (3-10, 14, 6044, 6314, 6849, 10, 14, and 21 Mar. 1943.

23. CM-IN-8213 (4-14-43), COMGEN-SOPAC to AGWAR, 449, 14 Apr. 1943; USAFISPA Mission Rpt., 12 Apr. 1943; Form 34, 29 Mar.-4 Apr. 1943, 70th Fighter Sq. The sole Marine to complete this mission was Lt. Benjamin E. Dale of VMF-124. SOPACFOR Office of Naval Air Combat Intel., Enemy Aircraft Destroyed by SOPAC Forces and Shipping Attacked by SOPAC Planes, May 1944, in USMC Hist. Div. files.

24. SOPACFOR Office of Naval Air Combat Intel., Enemy Aircraft Destroyed by SOPAC Forces and Shipping Attacked by SOPAC Planes, May 1944, in USMC Hist. Div. files.

25. Ftr. Comd. Guadalcanal War Diary, App 11, Enemy Planes Shot Down by Fighters from Guadalcanal Area, and App. IV, Results of Fighter Action on Guadalcanal.

26. Halsey Rpt.

27. Ftr. Comd. Guadalcanal War Diary, incl., Interception of Enemy Fighters over the Russell Islands, April 1, 1943; Hq. 43d Inf. Div., Report of Occupation of the Russell Islands 8 June 1943.

28. Ftr. Comd. Guadalcanal War Diary, incl., Interception of Enemy Dive Bombing Attack against Shipping, Tulagi Harbor and Vicinity, Apr 7, 1943; CM-IN-5022 (4-9-43), COMGENSOPAC to C/S USA, 8453, 8 Apr. 1943.

29. Tactical Intel. Questionnaire of Lt. Joseph F. Moore, 24 Sept. 1943; Ftr. Comd. Guadalcanal, War Diary, incl. of 7 Apr. 1943 and App IV; CM-IN-5022 (4-9-43), COMGENSOPAC to C/SUSA, 8453, 8 Apr. 1943. The P-38's accounted for eight planes here, and Lanphier was credited with three of this number.

30. Interview with Lt. Col. Thomas G. Lanphier, Jr., by author, 10 Sept. 1945; Ftr. Comd. Guadalcanal, War Diary; Hq. USAFISPA Ftr. Interception Rpt. 21 Apr 1943.

31. Lanphier interview; Form 34, 12-18 Apr. 1943,70th Ftr. Sq. and 339th Ftr. Sq.; Hq. USAFISPA, Ftr. Interception Rpt.

32. Hq. USAFISPA, Ftr. Interception Rpt.; Lanphier interview.

33. Lanphier interview; Enemy Aircraft Destroyed by Army Fighter Pilots in the Solomon Area Covering Period from August 22, 1942 to June 30, 1943, in A-2 Lib.; CM-IN-11334 (4-19-43), COMGEN-SOPAC to C/S USA, 1033, 19 Apr. 1943. Much secrecy surrounded this exploit, since its success rested on U.S. ability to understand the Japanese codes. Not one of the original intelligence reports examined ever gave any indication that Yamamoto was involved. Some difficulty arises in fixing responsibility for shooting down the victim's plane. Both Barber and Lanphier destroyed a bomber, and so did Holmes over Moila Point. But this last one fell in the water, Barber's disintegrated in the air at 5,000 to 6,000 feet, while Lanphier's crashed into the trees. A subsequent report by the Japanese stated that Yamamoto was found dead in his wrecked plane, which would indicate that it was Lanphier's work.

34. Ltr., Harmon to Arnold, 1 May 1943; Decorations and Awards made for Service during the Current War as Indicated by the Records of the Navy Department Board of Decorations and Medals, January 1, 1944. All other members of the flight received the Navy Air Medal.

35. Ltr., Gen. Harmon to Gen. Arnold, 1 May 1943.

36. Ibid.;CM-IN-8818 (4-15-43),COMGENSOPAC to AGWAR, 561, 15 Apr. 1943; ltr., Twining to Arnold, 27 Apr. 1943.

37. Col. George F. Baier, 111, MC, Report of Inspection of the Medical Activities of the Thirteenth Air Force, 11 Dec. 1943, Exh. 2, Days Lost per 1OO Flying Officers per Month.

38. The Operations of Aviation Engineers in the South Pacific, January 1942-August 1944, dtd. 15 July 1945 (herein-after cited as Opns. of Am. Engineers); ltr., Twining to Arnold. 27 Apr. 1943. The groups were as follows: heavy bombardment, 5th and 307th; medium bombardment, 42d; fighters, 18th and 347th; photo, 4th.

39. Ltr., Twining to Arnold, 27 Apr. 1943.

40. CM-IN-4047 (3-8-43), Lindsay and MacArthur to Marshall, Q4412; ltr., Lt. Angus Hopkins, Jr. to Brig. Gen. R.W. Douglas, CG VII Ftr. Comd., 10 Mar. 1943; ltr., Harmon to Arnold, 1 May 1941; Opns. of Avn. Engineers; ltr., Twining to Arnold, 27Apr. 1943.

41. CM-IN-902 (4-2-43), COMGEN-SOPAC to Arnold, 7518, 1 Apr. 1943; CM-OUT-9769 (3-26-43), Arnold to Harmon, Rane 3953, 25 Mar. 1943; CM-IN-15058 (3-28-43), Thompson to Arnold, 7319, 28 Mar. 1943. See Arnold's comment in Daily Log; CM-OUT-11089 (3-29-43), Arnold to Harmon, Rane 4006, 29 Mar. 1943; ltr., Twining to Arnold, 27 Apr. 1943. When asked why he had not attacked Fighter No. 2, where 130 fighters were parked on the one strip, an enemy airman replied that his people knew these planes for the most part were dummies since the "Americans would not be so dumb as to put that many planes in the open." (Ltr., Twining to Arnold, 27 Apr. 1943.)

42. CM-IN-902 (4-9-43), COMGENSOPAC to Arnold, 7518, 1 Apr. 1943; ltr., Twining to Arnold, 27 Apr. 1943; CM-IN-13079 (4-22-43), Harmon to Arnold, 1289, 21 Apr. 1943; CM-IN-14006 (4-23-43), Harmon to Arnold, 1550, 23 Apr. 1943. On 21 April aircraft on Guadalcanal totaled 286 distributed as follows: Henderson, Army--23, Navy--109; Carney Field, Army--24, Navy--0; Fighter No. 1, Army--9, Navy--54; Fighter No. 2, Army--58, Navy--18.

43. Mine Laying Operations by Aircraft in the Buin-Tonolei Area, 20-21 Mar. 1943; 31st Bomb. Sq. Strike Mission Rpt., 21-22 Mar. 1943; 307th Bomb. Gp. Consolidated Mission Rpt., 20-21 Mar. 1943, all in USMC Hist. Div. files; CM-IN-11680 (3022043), COMGENSOPAC to C/S USA, 6931, 22 Mar. 1943; CM-IN-12186 (3-23-43), COMGENSOPAC to C/S USA, 7005, 23 Mar. 1943.

44. CM-IN-9631 (5-15-43), Comdr. 3d Fleet to Comdr. 1st Fleet, 140661 GR 208, 15 May 1943, SCR NR 10; Air Combat Intel. Rpt. (ACIR), 9-15 May 1943; Ftr. Comd., MAG-12 Air Combat Intel. Log, Feb. 10-July 10, 1943 (hereinafter cited as Ftr. Comd. Log, with appropriate date).

45. ACIR, 16-22 May 1943.

46. Report: of Mine Laying in Kahili and Shortland Island Areas, May 19, 20, 23, and COMAIRSOLS Strike Comd. War Diary, 2 Apr.-25 July 1943, both in USMC Hist. Div. files; Form 34, 17-23 May 1943, 424th Bomb. Sq.; 307th Bomb. Gp. Consolidated Mission Rpt., 19 May 1943; TBF Intel., VT-11, 20 and 23 May 1943; 5th Bomb. Gp. Consolidated Mission Rpt., 23 May 1943.

47. ACIR, 23-29 May, 15-12 June 1943.

48. COMAIRSOLS Weekly Intel. Sum., 11 June 1943.

49. CM-IN-7447 (6-12-43), COMGEN-SOPAC G-2 to COMGENAAF, 749, 12 June 1943; ACIR, 6-12 June 1943; Hq. 18th Ftr. Gp., Ftr. Interception Rpt., 17 June 1943; RNZAF Dir. of Public Relations, RNZAF in the Pacific, Historical Summary, Aug. 1944, p. 10. P-40's were credited with twelve planes this day.

50. Ftr. Comd. Guadalcanal, War Diary, App. IV, Results of Fighter Action on Guadalcanal.

51. Ibid.; Hq. 18th Ftr. Gp., Ftr. Interception Rpt., 12 June 1943.

52. ACIR, 13-19 June 1943; CM-IN-10632 (6-17-43), COMGENSOPAC to C/S USA, 869, 17 June 1943; COMAIR-SOLS Weekly Intel. Sum., 18 June 1943; CM-IN-9837 (6-16-43), Comdr. 3d Fleet to CINCSOWESPAC, SCR 10 150600 Z GR 133, 15 June 1943.

53. ACIR, 13-19 June 1943; Ftr. Comd. War Diary, incl., Action Report, Fighting Squadron Eleven, 16 June 1943; CM-IN-11857 (6-19-43), COMSOPAC to COMINCH, 190114 SCR 9, 19 June 1943.

54. Ftr. Comd. War Diary, App. IV, Results of Fighter Action on Guadalcanal; CM-IN-11225 (6-18-43). Comdr. 3d Fleet to Comdr. 1st Fleet, 170529 NCR 8046, 18 June 1943; ACIR, 13-19 June 1943; COMAIRSOLS Weekly Intel. Sum., 18 June 1943.

55. CM-IN-11857 (6-19-43), COMSOPAC to COMINCH, 190114 SCR 9, 19 June 1943.

56. CCS 239/1, 23 May 1943.

57. Harmon, The Army in the South Pacific.

58. ONI Combat Narratives, Operations in the New Georgia Area, 21 June-5 August 1943, pp.1-2.

59. Ibid., p. 2; Harmon, The Army in the South Pacific; COMAIRSOLS Weekly Intel. Sum., 17-23 Mar. 1943.

60. ONI, Opns. in the New Georgia Area, p. 3; SOPACFOR, The Air Aspect of the Munda Campaign, 15 Aug. 1943. Carney was flooded out for three days in the period 14-20 June. (Form 34's, 424th Bomb. Sq., 9-16 June and 14-20 June 1943.)

61. Halsey Rpt.; Opns. in the New Georgia Area, p. 5; Hq. New Georgia Air Force, Special Action Rpt., 29 June-13 Aug. 1943, Annex A, in USMC Hist. Div. files; Task Force 31 Opn. Plan A8- 43, 4 June 1943; Operations of the 25th Infantry Div. in the Central Solomons,14 June 1944.

62. Task Force 33 Opn. Plan 7-43, 18 June 1943, in USMC Hist. Div. files; Halsey Rpt.; Harmon, The Army in the South Pacific; TF-31 Opn. Plan A8-43.

63. TF-31 Opn. Plan A8-43; Harmon, The Army in the South Pacific.

64. TF-33 Opn. Plan 7-43; TF-31 Opn. Plan A8-43, Annex D, Air Support Plan, Toenails Operation.

65. TF-31 Opn. Plan A8-43, Annex D.

66. TF-33 Opn. Plan 7-43; Hq. New Georgia AF, Special Action Rpt.

67. New Georgia AF, Special Action Rpt., Annex C, Plans for Air Support Missions for Ground Operations and for Fighter Direction in the New Georgia Area, 24 June 1943; TF-31 Opn. Plan A-8-43, Annex D; TF-33 Opn. Plan 7-43.

68. TF-33 Opn. Plan 7-43.

69. ONI, Opns. in the New Georgia Area, p. 4.

70. SOPACFOR, The Air Aspect of the Munda Campaign. COMSOPAC's estimate of 26 June credited the Japanese with 380 aircraft in the New Britain-Solomons area, of which the majority were at Rabaul and in the Solomons.

71. Ibid.

72. COMAIRSOLS Weekly Intel. Sums., 25 June, 9 July 1943; History, 42d Bomb. Gp., 15 Jan. 1941-30 June 1943; CM-IN-17927 (6-28-43), COMAIRSOLS to COMAIRSOPAC, SCR II 262336, 26 June 1943.

73. New Georgia AF, Special Action Rpt., Annex E, Daily Intel. Sums.; Lt. Burton Atkinson, USNR, Navy Fighter Control in the Solomons, in AAFSAT Intel. Rpts., March 1944; Ftr. Comd. Guadalcanal, War Diary; COMAIR-SOLS Weekly Intel. Sum., 2 July 1943; ACIR, 27June-3 July 1943. Fighter control was handled by Argus No. 11, first of the Navy's radar fighter-director units.

74.  ACIR, 27 June-3 July 1943; ONI, Opns. in the New Georgia Area, pp. 11-13.

75. CM-IN-314 (7-1-43), COMAIRSOLS to COMAIRSOPAC, SCR 10 301252, 30 June 1943; ACIR, 27 June-3 July 1943; Ftr. Comd. War Diary, App. IV. AAF planes were credited with seventy-one enemy kills, Corsairs with ninety, and F4F's with seventy-eight. RNZAF P-40's added fifteen more. A rough estimate of enemy personnel losses in June indicates an expenditure of more than 500 pilots and aircrew men. Fighter Command's figures exclude Allied losses in heavy and medium bombers.

76. COMAIRSOLS Weekly Intel. Sum., 9 July 1943; Ftr. Comd. War Diary, App. IV; ACIR, 27 June-3 July 1943; New Georgia AF Daily Intel. Sum.

77. The Air Aspect of the Munda Campaign; ltr., Harmon to Arnold, 13 July 1943.

78. Ltr., Twining to Brig. Gen. LaVerne Saunders, DC/AS, 8 July 1943; Strike Comd. War Diary, 2 Apr.-25 July 1943; CM-IN-315 (7-1-43), COMAIRSOLS to COMAIRSOPAC, SCR 9 301311, 30 June 1943.

79. Ltr., Twining to Saunders, 8July 1943. Fighter Command lists 169 Japanese planes down by 8 July, with an Allied combat loss of 27, exclusive of 7 operational losses.

80. Ltr., Twining to Saunders, 8 July 1943; CM-IN-7580 (7-11-43),COMGENSOPAC to COMGENAAF, 1372, 11 July 1943; ltr., Harmon to Arnold, 13 July 1943. In addition to the twenty-nine reported on hand, ten were due in by boat on 13 July, thirty more were scheduled for July shipment.

81. CM-OUT 5178 (7-13-43). Arnold to Harmon, Rane 6321, 12 July 1943; memo for Col. Berquist from Lt. Col. Roger L. Shearer, 11 July 1943.

82. Ltr., Harmon to Arnold, 13 July 1943; CM-IN-7005 (7-10-43), CTF 31.

83. 13th AF, Opns. Analysis Sec., Losses, Accidents and Injuries of Fighter Planes and Pilots in Relation to Flying Time, Solomon Islands, June through Nov. 1943, dtd. 15 Apr. 1944. It was necessary to withdraw P-39's from this field in September and to replace them with carrier-type fighters.

84. Ltr., Harmon to Arnold, 13 July 1943.

85.  ONI, Opns. in the New Georgia Area, pp. 29-31, 34-35; CM-IN-5302 (7-8-43), Comdr. 3d Fleet to Comdr. 1st Fleet, 070626 NCR 1406, 8 July 1943. The Japanese lost two destroyers in the first action, and their light cruiser Jintsu in the second. (USSBS, The Campaigns of the Pacific War, p. 142.)

86. USSBS, The Campaigns of the Pacific War. ACIR, 11-17 July 1943; Ftr. Comd. War Diary, App. IV; COMAIRSOLS Weekly Intel. Sum., 16 July 1943. It was believed that the Japanese army planes were involved here because some of the aircraft were painted a mottled green-brown. (CM-IN-11174 [7-16-43], COMAIRSOLS to COMAIRSOPAC, 151245 SCR NR 9, 15 July 1943.) The eleven AAF P-40's involved claimed 12 ‡ enemy planes destroyed in this action. (Form 34, 12-18 July 1943,44th Fighter Sq. Det.)

87. ACIR, 11-17 July 1943; Ftr. Comd. War Diary, App. IV; COMAIRSOLS Weekly Intel. Sum., 16 July 1943. It was believed that the Japanese army planes were involved here because some of the aircraft were painted a mottled green-brown. (CM-IN-11174 [7-16-43], COMAIRSOLS to COMAIRSOPAC, 151245 SCR NR 9, 15 July 1943.) The eleven AAF P-40's involved claimed 12 ‡ enemy planes destroyed in this action. (Form 34, 12-18 July 1943,44th Fighter Sq. Det.)

88. ACIR, 18-24 July 1943.

89. New Georgia AF Daily Intel. Sum.; The Air Aspect of the Munda Campaign.

90. Ibid.

91. CM-IN-973 (7-2-43). COMAIRSOLS to COMAIRSOPAC, 010951Z SCR 14, 1 July 1943; CM-IN-2478 (7-4-43). COMGENSOPAC to C/S USA, 1219, 4 July 1943; COMAIRSOLS Weekly Intel. Sum., 9 July 1943; Form 34, 5-11 July 1943, 72d Bomb.Sq. (H), 394th Bomb. Sq. (H), 370th Bomb. Sq. (H); Strike Comd. War Diary, 2 Apr.-25 July 1943.

92. Form 34's, 21-27 June, 5-11 July 1943.69th Bomb. Sq. (M); 28 June-4 July, 12-18 July 1943, 390th Bomb. Sq. (M); CM-IN-1770 (7-3-43), COMAIRSOLS to COMAIRSOPAC, 021103 SCR 8, 2 July 1943.

93. Ltr., Twining to Saunders, 8 July 1943; Form 34, 12-18 July 1943, 390th Bomb. Sq. (M). Twining pointed out at the same time that the Navy now was interfering very little with heavy bomber employment.  

94. CM-IN-12874 (7-18-43), Comdr. 3d Fleet to Comdr. 1st Fleet, 150547 SCR 13, 15 July 1943; CM-IN-12917, COMAIRSOLS to COMAIRSOPAC, 170013Z SCR NR 17, 17 July 1943; COMAIRSOLS Weekly Intel. Sum., 16 July 1943.

95. COMAIRSOLS Weekly Intel. Sum., 23 July 1943; Form 34, 12-18 July 1943, 370th Bomb. Sq. (H); Strike Comd. War Diary; CM-IN-13454 (7-19-43), Comdr. 3d Fleet to Comdr. 1st Fleet, 180540 SCR 6, 18 July 1943; ACIR, 11-17 July 1943. The fighter escort was typical of these operations: twenty-three P-40's, twelve P-38's, thirty-five F4F's, and forty-four F4U's. P-38G's of 339th Fighter Squadron escorted the B-24's, shooting down five Zeros, but they lost two of their number. (Form 34, 11-18 July 1943, 339th Fighter Sq.; Ftr. Comd. War Diary, App. IV.) A later estimate placed enemy losses at forty-six Zeros. (COMAIRSOLS Weekly Intel. Sum., 23 July 1943.)

96. COMAIRSOLS Weekly Intel. Sum., 23 July 1943; The Air Aspect of the Munda Campaign; CM-IN-13540 (7-19- 43). COMGENSOPAC to C/S USA, 1536, 19 July 1943; Strike Comd. COMAIRSOLS War Diary, 2 Apr.-25 July 1943; Form 34, 12-18 July 1943, 424th Bomb. Sq.; Ftr. Comd. War Diary, especially App. IV; CM-IN-14231 (7-20-43), Comdr. 3d Fleet to CINCSOWESPAC, SCR 12 190552 GR 171 BT, 19 July 1943; CM-IN-14454 (7-20-43), COMAIRSOLS to COMAIRSOPAC, SCR 8 181927Z, n.d.

97. CM-IN-16416 (7-23-43) COMAIRSOLS to All Comdrs. SOPAC, SCRSOPAC, SCR, NR 15 221224, 22 July 1943; COMAIRSOLS Weekly Intel. Sum., 30 July 1943; Form 34ís, 19-25 July 1943, 72d, 424th, and 370th Bomb. Sqs.

98. CM-IN-21689 (7-30-43), COMAIRSOLS to COMAIRSOPAC, SCR 8 290951, 29 July 1943. Both SBD's and TBF's were now equipped with wing tanks which increased their range. (COMAIRSOLS Weekly Intel. Sum., 23 July 1943.)

99. COMAIRSOLS Weekly Intel. Sum., 23 July 1943; CM-IN-16782 (7-23-43), CTF 31 to COMSOPAC, SCR NR 20 220316Z GR 190 BT, 22 July 1943. LST 343 lost ten dead and ten wounded in this attack. P-39's and P-40's would leave Segi late in the afternoon to cover the Rendova area at dusk. (Form 34's, July 1943, 70th, 12th, 339th Ftr. Sqs.)

100. Weekly Intel., Sum., 30 July 1943; CM-IN-18667 (7-26-43) COMAIRSOLS to COMAIRSOPAC, SCR 9 251103Z, 25 July 1943; ACIR, 25-31 July 1943. Four F4F's and two P-39's were lost in this interception but three pilots were recovered.

101. Ltr., Twining to Arnold, 29 July 1943

102. Strike Comd. War Diary, 2 Apr.-25 July 1943; Ftr. Comd. War Diary. From March to July inclusive, the Navy and Marines together had assigned to them an average of slightly less than three times the number of fighters operated by the AAF in Fighter Command. (Ftr. Comd. War Diary, App. I.) The enemy's loss record in the 26-day period was as follows: 222 Zeros, 1 twin-engine fighter, 23 float biplanes, 55 Bettys, 6 Sallys, and 9 dive bombers. In terms of personnel losses, assuming a crew of 5 for the Sallys and 6 for the larger Bettys, the above figures would indicate that it had cost the lives of 40 Allied pilots of all services to destroy approximately 648 enemy airmen. (Ibid., App. IV.)

103. Maj. Victor Dykes, Air Comd. Solomon Islands, n.d.; ltr., Twining to Arnold, 29 July 1943. Strother had commanded the XIII Fighter Command since its activation. Matheny took over the XIII Bomber Command on 1 July 1943.

104. Ltr., Twining to Saunders, 8 July 1943; ltr., Twining to Arnold, 29 July 1943.

105. ACIR, 25-31 July 1943; SOPACFOR Office of Naval Air Combat Intel., Enemy Aircraft Destroyed by SOPAC Forces and Shipping Attacked by SOPAC Planes. F4U's held the highest score in June and July, destroying 85 and 112‡ planes, respectively. AAF P-40's made the best Army record: 43 and 31‡.

106. 43d Inf. Div., Report of Operations; ONI, Opns. in the New Georgia Area, pp. 16-17, 19-20; CM-IN-1864 (7- 3-43), CTF 31 to COMSOPAC, 020556.

107. CM-IN-1864 (7-3-43), CTF 31 to COMSOPAC, 020556; Harmon, The Army in the South Pacific; 43d Inf. Div., Rpt. of Opns.; CM-IN-5366 (7-8-43), COMGENSOPAC to C/S USA, 1302, 8 July 1943; CM-IN-6041 (7-9-43) GENSOPAC to C/S USA, 1321, 9 July 1943; CM-IN-7005 (7-10-43), CTF 31 to COMSOPAC, 100123Z SCR 12, 10 July 1943; COMAIRSOLS Weekly Intel. Sum., 9 July, 1943; ONI, Opns. in the New Georgia Area, p. 15-16, 31.

108. Opns. of the 25th Infantry Div. in the Central Solomons, 14 June 1944.

109. XIV Corps Photo Intel. Rpt., New Georgia Opns., n.d.; Harmon, Summary Munda Operation, 1 Aug. 1943; 43d Div., Rpt. of Opns.

110. 43d Div., Rpt. of Opns.; Harmon, Summary Munda Opn.

111. TF-31, Opn. Plan A8-43, Annexes C and D, in USMC Hist. Div. files; Hq. New Georgia AF, Special Action Rpt.

112. Hq. New Georgia AF, Special Action Rpt.; CG XIV Corps, Informal Report on Combat Operations in the New Georgia Campaign, British Solomon Islands; Hq. New Georgia AF, Special Action Rpt.

113. Hq. New Georgia AF, Special Action Rpt.

114. The Air Aspect of the Munda Campaign.

115. CM-IN-7005 (7-10-43), CTF-31 to COMSOPAC, SCR 12 100123Z, 10 July 1943; COMAIRSOLS Weekly Intel. Sum., 9 July 1943; 43d Div., Rpt. of Opns.

116. Harmon, The Army in the South Pacific; Halsey Rpt.

117. Hq. New Georgia AF, Special Action Rpt.; CM-IN-15700 (7-22-43), COMAIRSOLS to COMAIRSOPAC, SCR NR 13 211122Z, 21 July 1943; COMAIRSOLS Weekly Intel. Sum., 23 July 1943.

118. COMAIRSOLS Weekly Intel. Sum., 30 July 1943; ONI, Opns. in the New Georgia Area, pp. 52-53.

119. COMAIRSOLS Weekly Intel. Sum., 30 July 1943; CM-IN-18696 (7-26- 43), Comdr. 3d Fleet to Comdr. 1st Fleet, SCR 13 250637, 25 July 1943; Strike Comd. War Diary. This force, which struck at 0630, was composed of fifty-three SBD's, fifty-two TBF's, thirty-two B-24's, twenty-four B-25's, and ten B-17's, escorted by seventy-nine fighters of all types. One B-24 and one SBD were lost. Harmon went along to drop a "personal bomb." (CM-IN-18671 [7-26-43] ACORN RED I to COMSOPAC, SCR 11 250158Z 25 July 1943.)

120. Harmon. Summary Munda Opn.

121. CM-IN-3986 (8-6-43), Gen. Griswold to COMSOPAC, SCR NR 11 050545, n.d.; 43d Div., Rpt. of Opns.; ONI, Opns. in the New Georgia Area, p. 59; Opns. of the 25th Inf. Div.

122. ACIR, 1-7 Aug. 1943. Distances in nautical miles from Munda to the enemy's fields were as follows: Kieta, 160; Tenekow, 180; Buka, 234; Rabaul, 394. Enemy planes had landed on Munda as early as 23 December 1942, but by 2 January 1943, thirty-seven Zeros had been shot down over the strip and by early March it was totally inoperative. Thereafter Japanese aircraft landed only rarely and Munda never was placed in active operation.

123. The Air Aspect of the Munda Campaign. Enemy losses were: 259 fighters, 60 twin-engine bombers, 23 dive bombers, and 16 float planes.

124. q. New Georgia AF, Special Action Rpt., Annex D, Air Support Requests--New Georgia Campaign, June 30-August 5, 1943. Total sorties by COMAIR New Georgia were distributed in the following manner:

125. COMAIRSOLS Weekly Intel. Sum., 6 Aug. 1943; CM-IN-1740 (8-3-43), Comdr. 3d Fleet to Comdr. 1st Fleet, SCR NR 12 020557, 2 Aug. 1943.

126. Harmon, Summary Munda Opn.; Opns. of the 25th Inf. Div.; CG XIV Corps, Informal Rpt. on Combat Opns. in New Georgia.

127. Interview with Lt. Comdr. Harold H. Larsen, USN, 18 Jan. 1944. Larsen served as operations officer for Strike Command for five months, beginning 25 July 1943. His criticism of the failure of ground commanders to request close support may be attributed to a pride often demonstrated by the TBF-SBD men in the accuracy of their bombing.

128. The Air Aspect of the Munda Campaign; COMAIRSOLS Weekly Intel. Sum., 16 July 1943. The Navy report credits COMAIRSOLS with the destruction of one seaplane carrier, one oiler, four destroyers, six cargo vessels, nine barges, plus damage to seven DD's, nine AKs, arid three PC's during the campaign.

129. TF-33 Opn. Plan 7-43; 43d Div. G-4 Journal, Munda Campaign, 22 June-22 Aug. 1943.

130. 43d Div., Rpt. of Opns.

131. 43d Div. G-4 Journal; Hq. 145th Inf., The First Battalion Journal for New Georgia and Vella Lavella Campaigns; COMAIRSOLS Weekly Intel. Sum., 16 July 1943; History, 13th Troop Carrier Sq.

132. Interview with Col. F.L. Beadle, 21 Nov. 1943; New Georgia AF Daily Intel. Sum. Thirty Japanese planes, all in poor condition, were captured on Munda. (ACIR, 8-14 Aug. 1943; COMAIRSOLS Weekly Intel. Sum., 13 Aug. 1943; CM-IN-7817 [8-1 1-43], COMAIRSOLS to COMAIRSOPAC, SCR NR 14 101000, 10 Aug. 1943.) Lt. Shuler, who had downed four planes on the 6th, shot down his fifth in this action.

133. Ibid.; History, 828th Engineer Aviation Bn., Nov. 1942-July 1944, incl. 5, Operations of Aviation Engineers; Form 34. 8-14 Aug. 1943, 44th Ftr. Sq. When the Navy moved to an advance position requiring establishment and maintenance of an airfield, it ordered an Acorn and a Combat Aircraft Service Unit (CASU). CASU personnel serviced and repaired the planes. Seabees attached to the Acorn built and maintained the airstrip, buildings, and installations. Each Acorn was so equipped that when joined with a CASU, it could service, rearm, and perform minor repairs and routine upkeep for the aircraft of a carrier group or a patrol plane squadron. (Naval Aviation News, 1 June 1945, p. 18.)

134. History, 828th Engineer Aviation Bn.

135. Ibid. The 73d Seabees maintained this field.

136. 13th AF Opns. Analysis Sec., Report on Study of Non-Combat Accidents, 15 Nov. 1943. From the center line of the strip to 100 feet on each side, Munda was surfaced with eighteen inches of rolled coral.

137. Harmon, The Army in the South Pacific; Halsey Rpt. Apparently, Halsey made his decision to by-pass Vila on 12 July. (Hq. N.Z. 3 Div., History of Occupation of Vella Lavella.)

138. Harmon, The Army in the South Pacific; Hq. New Georgia Occupation Force, Narrative Account of the Campaigns in the New Georgia Group, B.S.I.; Hq. USAFISPA G-2 Per. Rpt., 14-21 Aug. 1943. The 35th was followed by the 4th Defense Battalion, USMC, the 58th Seabees, and the reconnaissance troop of the 25th Division.

139. Hq. New Georgia Occupation Force, Narrative . . . New Georgia; Rpt. of the 43d Division During the Campaign to Secure Arundel Island and Diamond Narrows

140. Opns. of the 25th Inf. Div.; Hq. New Georgia Occupation Force, Narrative . . . New Georgia.

141. Hq. New Georgia Occupation Force, Narrative . . . New Georgia. Total Allied ground casualties in the central Solomons were 972 killed, 3,873 wounded, 23 missing, and 122 additional deaths due to wounds.

142. Larsen interview.

143. Larsen interview; COMAIRSOLS Weekly Intel. Sum., 13 Aug. 1943; CM-IN-11200 (8-15-43), COMAIRSOLS to COMAIRSOPAC, 140012 SCR NR 13, 14 Aug. 1943; Hq. 7th AF Intel. Bull. 100, 29 Aug. 1943.

144. Hq. New Georgia AF, Special Action Rpt., Second Phase, 14 Aug.-20 Oct. 1943, in USMC Hist. Div. files; COMAIRSOLS Daily Intel. Sum., 16 Aug.1943.

145. COMAIRSOLS Daily Intel. Sum., 16-17, 19 Aug. 1943; Hq. New Georgia AF, Special Action Rpt., Second Phase, especially Annex A; New Georgia Air Comd. Daily Intel. Sum., 14 Aug.-20 Oct. 1943; COMAIRSOLS Weekly Intel. Sum., 20 Aug. 1943.

146. Hq. New Georgia AF, Special Action Rpt., Second Phase; NG Air Comd. Daily Intel. Sum. On 17 August an estimated fifty shells fell around Munda Field. On the day previous, one shell had burst in the Acorn 8 dispensary.

147. Hq. New Georgia AF, Special Action Rpt., Second Phase.

148. COMAIRSOLS Daily Digest, 28 Sept. 1943.

149. Ibid.

150. COMAIRSOLS Weekly Intel. 29 Aug. 1943.

151. NG Air Comd. Daily Intel. Sum

152. Hq. N.Z. 3 Div., History of Occupation of Vella Lavella. It was estimated that prior to initiation of full operation, Barakoma had been directly or indirectly responsible for the saving of twenty-two pilots and aircrew men.

153. Hq. USAFISPA, Air Information Bull. 24, 18 Aug. 1943; Hq. 7th AF Intel. Bull. 100, 29Aug. 1943.

154. Hq. 7th AF Intel. Bull. 100, 29 Aug. 1943.

155. COMAIRSOPAC Daily Intel. Sums., 21-23, 27-27 Aug. 1943; Hq. 42d Bomb. Gp., Narrative Rpt. of August Activities, 9 Sept. 1943. Often the Japanese ran their barges up on the beach where they were burned out.

156. CM-IN-20417 (8-27-43), COMSOPAC to CINCPAC, 260601Z SCR 11; ltr., Twining to Arnold, 7 Oct. 1943; COMAIRSOLS Weekly Intel. Sum., 17 Sept. 1943; COMAIRSOLS Daily Intel. Sum., 5 Oct. 1943. On 11 September Halsey congratulated Twining on the progress of the anti-barge campaign in his usual pungent manner, remarking that "your anti-barge warfare is warming [the] seats of the Jap britches. Keep 'em blazing." (Strike Comd. War Diary, 26 July-19 Nov. 1943.)

157. SOPACFOR Office of Naval Air Combat Intel., Enemy Aircraft Destroyed by SOPAC Forces; ACIR, 5-11 Sept. 1943. This was the second highest monthly total of kills achieved by the B-24's in the entire Solomons campaign.

158. COMAIRSOLS Weekly Intel. Sum., 3 Sept. 1943; The Thirteenth Air Force, March-Oct. 1943, App. 5.

159. Hq. New Georgia AF, Special Action Rpt., Second Phase; ACIR, 19-25 Sept. 1943.

160. SOPACFOR Intel. Sec., The F4U- 2(N) in the Solomons, 1 Mar. 1944, in USMC Hist. Div. files; COMAIRSOLS Daily Intel. Sum., 18 Sept. 1943; ACIR, 19-25 Sept. 1943.

161. ACIR, 17 Sept. 1943.

162. Ibid., 21 Sept. 1943.

163. Draft AAF historical study, Development of Radio and Radar Equipment for Air Operations, 1939-1944.

164. CM-OUT-12466 (4-30-43), Arnold to CGSPA, Rane 4684, 27 Apr. 1943; CM-IN-2186 (5-4-43), COMGENSOPAC to CGAAF, 2899, 4 May 1943; ltr., Col. W.A. Matheny to CG 13th AF, 21 Sept. 1943; 13th AF Opns. Analysis Sec., Some Factors Affecting Snooper Performance, 1 May 1944.

165. 13th AF, Some Factors Affecting Snooper Performance; Form 34's, 23-29 Bomb. Sq.; ltr., Matheny to CG 13th AF, 21 Sept. 1943.

166. Form 34, 27 Sept.-3 Oct. 1943, 394th Bomb. Sq.; COMAIRSOLS Daily Intel. Sum., 29 Sept. 1943; ltr., Harmon to Arnold, 7 Oct. 1943; COMAIRSOLS Weekly Intel. Sum., 1 Oct. 1943; CM-IN-1629 (10-3-43), Comdr. 3d Fleet to Comdr. 1st Fleet, 010302Z, 1 Oct. 1943.

167. CM-IN-1629 (10-3-43), Comdr. 3d Fleet to Comdr. 1st Fleet, 010302Z, 1 Oct. 1943; interview with Lt. F. L. Sylvester, 26 July 1944; interview with Col. Stuart P. Wright by author, 3 Jan. 1946.

168. Ltr., Matheny to CG 13th AF, 21 Sept. 1943; CM-IN-18528 (9-26-43), COMGENSOPAC to CGAAF, 3030, 26 Sept. Sept. 1943.

169. XIII AFSC, War Critique Study, p. 57.

170. CM-OUT-12979 (9-28-43), Arnold sgd. Marshall to COMGENSOPAC, Rane 8179, 28 Sept. 1943; ltr., Harmon to Arnold, 7 Oct. 1943; CM-IN-17283 (12-28-43), CG 13th AF to CGAAF, F1986, 28 Dec. 1943; ltr., Matheny to CG 13th AF, 21 Sept. 1943; 868th Bomb. Sq., Periodic Activities Summary, 1-31 Jan. 1944. Colonel Wright was recalled from the South Pacific as soon as his project was well established. Harmon desired Wright's return in a command status. (CM-IN-3488 [10-6-43], COMGENSOPAC to CGAAF, 3261, 6 Oct. 1943.)

171. 13th AF, Some Factors Affecting Snooper Performance; ltr., Matheny to CG 13th AF, 21 Sept. 1943. During the 10-week period ending 5 November, 111 strike missions were flown by these planes. Results were observed on 93 of the bombing runs, yielding the claimed figure of 23.6 per cent direct hits.

172. Capt. Ernest R. Barriere and Lt. Sol L. Reiches, Preparation for the Use of SCR-717-13 as a Pathfinder in the XIII Bomber Command, 3 Sept. 1942, p. 2. The Navy was very cooperative and went to considerable effort to confirm sinkings by this unit. Direct hits were claimed on five DD's, two CL's, one submarine, and one carrier. (Some Factors Affecting Snooper Performance; Sylvester interview.)

173. COMAIRSOLS Weekly Intel. Sum., 8 Oct. 1943.

174. Ibid., 1 and 8 Oct. 1943; ACIR, 3-9 Oct. 1943; CM-IN-4900 (10-8-43), COMGENSOPAC to C/S USA, 3305, 8 Oct. 1943.

175. ACIR, 3-9 Oct. 1943.

176. COMAIRSOLS Daily Intel. Sum., 13 Oct. 1943.

177. Msg., Arnold to Harmon, Rane 8212, 29 Sept. 1943. It is of some interest to note that the commanding general of the AAF was concerned when his flyers seemed temporarily unable to continue their destruction of five enemy planes for each loss of their own.

178. Ltr., Twining to Arnold, 7 Oct. 1943.

179. Ibid.

180. ACIR, 19-25 Sept. 1943. The first aircraft to land at Barakoma came in on 23 September, but none were based there until 17 October. (Hq. New Georgia AF, Special Action Rpt., Second Phase.)

181. Ltr., Twining to Arnold, 7 Oct. 1943.

182. Ltr., Arnold to Twining, 16 Oct. 1943.

Transcribed and formatted for HTML by Anson Raymond for the HyperWar Foundation