IN THE PACIFIC prior to 1944, the attention of U.S. strategic planners was, to a great extent, directed toward the reduction of Truk and establishment of our forces in the Carolines and Marianas. In common with enemy positions in the Marshalls and Gilberts, Rabaul, in eastern New Britain, constituted a principal outguard of Truk. Until Rabaul, in turn, could be brought under our control or suitably neutralized, Truk and the Carolines could not be approached from the south.
Rabaul, however, not only furnished security for the Northern Solomons and Carolines, but also provided a base for possible Japanese offensive operations against New Guinea, Australia, and New Zealand. As the immediate key to the Japanese defense of the area, Rabaul constituted the major obstacle to Allied advance up the Solomons chain, and therefore became the primary objective.
Thus the very name--Rabaul--haunted the thinking of Allied leaders in the South and Southwest Pacific. It was the main remaining threat to the American-Australian life-line. No operation in the Solomons or New Guinea, no matter how successful could be considered complete as long as Rabaul remained strong, since it was always possible for resurgent Japanese troops to move from Rabaul and attempt the reconquest even of Guadalcanal or eastern New Guinea.
Neutralization of Rabaul, in consequence, was an indispensible condition of the South Pacific planning of the joint Chiefs of Staff. It was essential that the place either be seized or contained before projected operations in the Philippines or the South Pacific could be undertaken. Thus the Northern Solomons campaign, which consisted of one large operation (Bougainville ) and numerous smaller ones,1 became in effect the campaign to contain and neutralize Rabaul.
To bring about this objective, the Allied high command, as early as July, 1942, decreed that the Rabaul area be brought under air attack, to destroy its base facilities and cut its supply lines. Such an effort necessitated seizure, establishment and protection of airfields and intermediate airstrips where bombers could be based and from which covering fighters could be flown.
Having decided to bomb Rabaul to ineffectiveness the Joint Chiefs of Staff further elected to choose sites for establishment of the proposed airdromes. Both for the immediate purpose of providing fighter-cover for the bombers, as well as by reason of its broader importance to the strategy of Northern Solomons, it was obvious from the beginning that a location on Bougainville would be more effective than any other area under consideration.
Rabaul had been first specified as an objective of Allied effort by a Joint Chiefs of Staff directive of 2 July 1942. This directive assigned three tasks to South and Southwest Pacific forces; Task One, seizure of the Santa Cruz Islands, Tulagi, and "adjacent positions"; Task Two, seizure of the remainder of the Solomons, Lae, Salamaua, and
THE COASTWATCHERS became legendary figures in the South Pacific war, living for months in enemy-surrounded jungle, subsisting on what they could find, and operating on their nerve. Here is a typical group of Australian "Cobbers" posed with their ever-faithful native scouts.
the northeast coast of New Guinea; Task Three, seizure of Rabaul and "adjacent positions". Although this directive was later modified, and the concept of seizure of Rabaul was changed to one of neutralization, it fixed the pattern for South and Southwest Pacific operations from August 1942 to the early part of 1944.
Even before the war, Japanese Imperial Headquarters had likewise recognized the significance of Rabaul and the necessity of establishing a southern outguard for that stronghold. Consequently, as part of their strategic scheme, Japanese war planners had incorporated into the design for the Greater East Asia War a concept for seizure of the Solomons and eastern New Guinea area before November, 1942.2 In line with this project, almost immediately after 7 December, 1941, Japanese troops occupied Rabaul (23 January 1942).
As dictated in the enemy plan, occupation of Boungainville was to follow immediately.
By 21 January, 1942, the Japanese began to launch air attacks against the 25 Australian soldiers (commanded by Lieutenant J. H. Mackie), who were stationed in the vicinity of Buka Passage, just to the north of Bougainville
Island.3 By mid-March the Japanese had landed raiding parties at various points along the coasts of Bougainville, and finally, on 30 March, Australian coastwatchers, with sinking hearts, watched gray Japanese warships come sliding out of the cold morning mists, bearing troops whose numbers were to be augmented again and again, and who were destined, dead or alive, to remain in the Northern Solomons for the duration of the war.4
Moving out over Bougainville, the Japanese quickly swallowed vantage points on the island. Subsequently, the entire island was in their grasp. Meanwhile the Japanese propaganda machine made headway with its "Asia for the Asiatics doctrine," and many natives of the island not only sympathized with the Japanese cause, but actively abetted it.5
By spring of 1943, Japanese troops occupied positions on an arc through the Aleutians, the Marshalls and Gilberts, the Solomons and New Guinea.6
As a result of the battle of Midway (4-6 June, 1942), however, the enemy had lost four large aircraft carriers,7 together with the flower of Japanese naval aviation's strength in pilots and planes. Not only did this loss prevent an assault against Midway, but it also later denied to Japan much needed carrier air support in the great sea-battles soon to be fought in defense of the Solomons area.8
By February 1943, Buna and Gona in New Guinea, and Guadalcanal in the Solomons were again under Allied control. To the west of Guadalcanal, the Russell Islands were occupied by the Allies on 21 February, 1943 and with complete Allied domination of New Georgia and the Ellice Islands toward the close of August, 1943, the southward drive of the Japanese was completely stalled.9 Meanwhile, Allied forces moved steadily along the coast of New Guinea, by April pushing 75 miles northwest of Buna.
Strategically, Allied plans called for establishment of air and naval bases to implement the northward advance. It was necessary to impede the heavy though sporadic Japanese air attacks on Allied bases and front lines, and to cut off the system of air supply that Japan had built up for the New Guinea and Solomons area.10 To attack in New Guinea without control of the air meant continued operation in face of the great Japanese base at Rabaul on the open flank of such a venture.11 With necessarily extended supply lines thus laid open to air or sea attack, it was essential that Rabaul be taken, reduced, neutralized, or contained.12
As early as May, 1942, General Douglas MacArthur, U.S.A., suggested that the Allies proceed along the southern route of approach toward Japan via the northern shores of New Guinea and the Solomons. The Joint Chiefs of Staff directive of 2 July, 1942, even then recognized that positions on Bougainville would be necessary in order to bring Rabaul under attack.13
On 28 February, 1943, the Joint Chiefs of Staff had accepted a plan of operations entitled "ELKTON Plan," formulated some six months before by MacArthur. This concept envisaged seizure of operating airdromes in New Georgia to provide land-based air support for subsequent operations against the Northern Solomons. The Joint Chiefs decided that:
Airdromes in southeastern Bougainville are required by South Pacific Force for operations against Rabaul or Kavieng and to support Naval Striking Force. Such bases exist in the Buin-Faisi area. Enemy airdromes in New Georgia are interspersed between the Guadalcanal bases and the bases in the Buin-Faisi area. These must be captured or neutralized prior to the assault on the southern Bougainville bases. With the enemy in possession of a line of supporting airdromes at Kavieng, Rabaul, Buka, and Buin-Faisi, it is improbable that amphibious forces can operate successfully in the New Georgia area prior to neutralization. Therefore, the operations against New
Georgia are visualized as following the operations to secure airdromes in the Vitiaz Strait area.
The Joint Chiefs went on to declare that--
this operation will be conducted by South Pacific Forces in accordance with plans developed by them. Strong air support of the operations by Southwest Pacific air forces from bases in New Guinea, may be anticipated. The time of initiation of the operations will be controlled by the Commander-in-Chief, Southwest Pacific Area.
Thus, some eight months before the operation was undertaken, the Joint Chiefs of Staff had decided to attack the Bougainville area.
By March 1943, therefore, the Joint Chiefs of Staff, in revising its orders of 2 July 1942, directed that all operations against Rabaul by both the South and Southwest Pacific forces would be conducted under the supervision of General MacArthur. All operations of South Pacific forces in the Solomons were placed under the direct command of the Commander, South Pacific, who would confer with MacArthur and follow his general directives.
Admiral William F. Halsey, Jr., USN, Commander, South Pacific Area, had to decide, therefore, how to make Rabaul useless to the Japanese. As a related task, he realized that he would have to neutralize the airfields at the northern and southern tips of Bougainville in order that heavy bombers could get through to attack Rabaul. Finally, he hoped to cut the flow of supplies to Japanese troops located in bypassed areas of the Solomons.
From the enemy point of view, continued occupation of the Northern Solomons by the Japanese assured them of being able to supply bypassed positions in the Southern Solomons and
DIAGRAM, TASK ORGANIZATION, SOUTH PACIFIC FORCE, NOVEMBER 1943
The Solomon Islands
New Guinea, and of being able to evacuate surviving garrisons at their leisure.14
As Admiral Halsey put it:
Rabaul, a Japanese naval and air stronghold, appeared at this time to be the logical objective towards whose seizure or neutralization all efforts of both the South and Southwest Pacific Forces would be directed. Until Rabaul was seized, or at least naval and air control of the New Britain area was established, the planned advance of the Southwest Pacific Forces along the New Guinea coast was impracticable.15
Seizure and Consolidation of the Central Solomons
In the meantime, events were occurring with great rapidity on both the Solomons and New Guinea. On 9 February 1943, Japanese Imperial Headquarters announced that Guadalcanal had fallen. Successful landings had been made on the Woodlark-Kiriwina Islands and Nassau Bay, on New Georgia and Rendova, in the Russells, and Vella Lavella. Lae and Salamaua were under ground and air attack, and the great Japanese base at Wewak was under almost constant aerial bombardment.16
It was time for action; time for a move which would make Rabaul virtually ineffective.17
Describing the strategic situation which existed at this time, Fleet Admiral Ernest J. King said:
The evacuation of Guadalcanal on 8 February 1943 was by no means an indication that the Japanese were retiring from the Solomons. On the contrary there was ample evidence that they would make every effort to retain their positions in the Solomons and in New Guinea. Conversely, having pushed them out of the Southern Solomon area, our next undertaking was to push them out of the Northern Solomons.18
The Russell Islands were seized on 21 February by the 3d Raider Battalion and Army units. Allied forces undertook a continued land, air and sea effort in the Solomons, beginning with landings in New Georgia by the 1st and 4th Raider Battalions, the 9th Defense Battalion, and the Tank Platoon, 10th Defense Battalion, MAG-22, and Army units. As a preliminary to the landings in New Georgia, strong intermediate points before Rabaul, more particularly bases in the Bougainville area, were subjected to a heavy naval bombardment on the night of 29-30 June. Farther south, the Woodlark and Trobriand Island Groups were also occupied by the 12th Defense Battalion and Army units. Landings were effected in New Guinea at Nassau Bay on 29 June.
Because of severe losses suffered by Japanese Army Air units during this period, and because of the fact that the Japanese Navy was responsible for the Solomons defense, the Japanese Army withdrew its air units from the Rabaul-Bougainville area in August.19
The fall of Munda airfield on 5 August, almost exactly one year after initial landings on Guadalcanal, and six weeks after inception of the New Georgia campaign, climaxed the campaign in the Central Solomons.20 Although Japanese in force occupied the neighboring island, Kolombangara, and had an airfield there, their effectiveness was nullified not only by the continual bombing and shelling they received, but also by the fact that Vella Lavella, an island about 14 miles to the northwest, was occupied by Marine (4th Defense Battalion), Army, Navy and New Zealand units on 14 August, despite Japanese resistance in the air. On 25 August, the New Georgia campaign ended when American forces occupied Bairoko Harbor. By 27 August Arundel Island was occupied.21
As soon as practicable after occupation of Munda, air operations from that base were inaugurated in support of amphibious operations elsewhere in the Solomons and to neutralize enemy supply dumps and airfields on Bougainville. On 3 September 1942, Vella Lavella was declared secured, thus by-passing approximately 10,000 Japanese troops on Kolombangara.22
Since this turn of events made Kolombangara untenable for the enemy, the latter attempted to evacuate troops by nightly barge runs, but this proved costly, particularly after American destroyers began an anti-barge campaign in September.23
The Central Solomons campaign ended on 6 October when the Japanese had completed the evacuation of troops from Kolombangara and Vella Lavella. Since most of the troops evacuated from Guadalcanal, Kolombangara, Vella Lavella, and other islands in the Southern and Central Solomons ultimately reached Bougainville,24 and since Japanese activity increased greatly in that vicinity, attacks in the Bougainville area were indicated. To that end air raids and naval surface bombardments were launched.25 In an effort to oppose these operations, the Japanese poured approximately 700 planes down the funnel of Solomons attrition between June and October, 1943.26
With Allied attacks in the Pacific increasing in severity, the Japanese, about May, 1943, formulated on operation plan called the "Z" plan. This plan established a front line of defense stretching from the Aleutians, Wake, the Marshalls, the Gilberts, Nauru and Ocean, to the Bismarcks, the area behind this line being considered vital to defense of the Empire. The Combined Fleet27 was to serve as the mobile reserve of the defense forces and was to sortie without delay upon contact with Allied forces anywhere on or within the line just described. In the Bismarck-Solomons Area, the Japanese had disposed units of the Eighth Fleet and most of the Eleventh Air Fleet.28
Modification of the "Z" plan resulted from U.S. landings at Munda and Kolombangara and at Lae and Salamaua, for it was apparent to the enemy that Rabaul's effectiveness as a major base was declining. This modification reduced the so-called "vital area" by withdrawal of the Japanese front line of defense to the line Kuriles-Marianas-Carolines. Since Bougainville no longer lay within the so-called "vital area", the Combined Fleet was no longer to be committed to its defense.29
As the Allied advance progressed steadily toward the "vital area", by-passing areas of resistance before the front line of defense, Japanese Imperial Headquarters found it necessary on 30 September 1943 to issue an order which reads in part as follows:
Make every effort to hold the important southeastern area extending eastward from the eastern part of New Guinea to the Solomon Islands by repulsing all enemy attacks in the area. To accomplish this purpose: (a) Consider Rabaul as the center and make every effort for a protracted defense of important positions in the Bismarck Archipelago and Bougainville areas. (b) Endeavor to hold out in the northern New Guinea Area by reinforcing important positions in this area. (c) Endeavor to destroy the attacking enemy before landings are made by using air and surface forces. (d) In case the enemy succeeds in landing operations endeavor to destroy him before he consolidates his position thereby disrupting his plan for counterattack. (e) Endeavor to concentrate military supplies for high speed transport to the above-mentioned positions, particularly those in New Guinea.30
In light of this enemy situation, the structure of the Northern Solomons campaign, therefore included one principal landing--Bougainville--and several peripheral operations--Treasury Islands, Choiseul, and Green. Bougainville was the keystone of this structure. Except for general remarks, therefore, the initial discussion will treat primarily the Bougainville phase of the campaign, while peripheral operations are described separately in a later chapter.
Bougainville and Its Protectors
The Japanese could have had no illusions about Allied intentions, for preparations to wrest the final Solomons' stronghold from their grasp were obvious. Arrows indicating the direction of Allied advance pointed toward Rabaul, and Bougainville lay directly in the path of that advance. It was only natural that, following moves from Guadalcanal to New Georgia to Vella Lavella,
the Bougainville area should be next attacked.
For over a year and a half the Japanese had been industriously organizing positions on Bougainville. The enemy appeared prepared to launch an asault against Australia from these positions, and also from these same positions he was prepared to oppose attacks on the nerve center of his southern perimeter, Rabaul. On the other hand, the problems to be encountered by any attacker against Bougainville were more complex, and the situations more involved than those encountered previously.
Bougainville is the largest island in the Solomons. In 1943 it had adequate facilities for basing fleet units and aircraft. Close at hand to the main base at Rabaul and other subsidiary bases around that area, it lay within easy reach of the Japanese fleet at Truk. As visualized by American planners, it was probably well defended in accessible sectors, while a mobile reserve was probably readily available to counterattack landings elsewhere. Further, during the approach of any attack force enemy air and surface units could sally forth and engage convoys in battle.31
On the other hand, the Allies could take advantage of the fact that (although this was not entirely evident at the time) Japanese sea and air power were on the wane; that successful landings had been made in the Central Solomons and in New Guinea; that the enemy was suffering from attrition to his air forces; that Japanese lines of communication and supply could be brought under attack at any time; that even though the coming strike at Bougainville was obvious, the exact point of landing on a large island might be concealed until the landing was an accomplished fact; and that interior communications on Bougainville were entirely inadequate for quick reinforcement by Japanese reserves. The most troublesome factor confronting the Allies, in fact, was the scarcity of shipping available.
The risks involved in a landing on Bougainville were calculated, but, with approval of General MacArthur, who had requested that Admiral Halsey investigate the feasibility of attacking Bougainville proper rather than landing on the strongly defended Shortland Islands, it was decided to land at Empress Augusta Bay on the premise that this location fitted exactly into the ultimate air plan and, at the same time, was so inaccessible from established Japanese positions on Bougainville that an estimated three months would elapse before the Japanese could organize an effective counter-offensive on the ground.32
Plans for landing in the Northern Solomons underwent many changes before they emerged in final form. First conceived as an attack designed to seize a heavily defended area and destroy opposing forces located there, the plans gradually were reduced to more conservative aspirations as time went on, but finally were developed into a bold, extensive operation.33
During preliminary study of the proposed operation, a new concept emerged; it was believed that the Japanese were more vulnerable in the air and on the sea than on the ground. Study also indicated that the forces originally allocated for the operation were insufficient for the assault as envisaged.34
With these ideas in mind, first on 26 July and then again on 5 August 1943, the objective of the operation was changed:
to furnish a base for further operations and to deny to enemy air and naval forces the use of South Bougainville, South Pacific Forces will, on or about 15 October, seize, occupy and defend Shortland, Faisi and Ballale Islands, and such other small adjacent islands as may be necessary.
Thus, the target was limited to the Shortland area.35
Preliminary Marine plans for the operation went ahead with formation of the provisional units needed to execute the plan, e. g., the 2d Marine Raider Regiment and the 2d 155mm
Artillery Battalion. Logistic plans were made to form Branch No. 3 of the Fourth Base Depot, and station that unit, along with the Corps Evacuation Hospital, on still unsecured Vella Lavella. Marine units for the operation were assembled on Guadalcanal.36
On 6 September 1943, Admiral Halsey directed Vice Admiral Aubrey W. Fitch, USN, Lieutenant General Millard F. Harmon, USA, Major General Charles D. Barrett, and Rear Admiral Theodore S. Wilkinson, USN, to make a survey of the strategic situation and to submit recommendations for future planning. The following day these officers recommended that "the concept of the attack on Shortland should be abandoned" and that "the following alternate plan should be placed in effect:
Step 1--Complete projected airfields in New Georgia and Vella Lavella.
Step 2.--Continue and increase air effort to neutralize by air enemy airfields in the South Bougainville Area, and to put heavy pressure on airfields in the Buka Area.
Step 3.--On D-day simultaneously seize and occupy Treasury Islands and Choiseul Bay Area; install long range radar at both of these positions, construct airfields at one or both positions as found feasible, establish motor torpedo boat advance bases and staging points for landing craft at both positions; all in order to contain and strangle Southern Bougainville.
THE HIGH COMMAND. Assembled on Guadalcanal in 1943, during the planning for the Northern Solomons campaign, this group includes many officers who played important roles in the operations to come. In the front row, left to right, are: Brigadier General A. F. Howard, Rear Admiral Theodore Wilkinson, USN, Major General Charles D..Barrett, and Major General Robert S. Beightler, USA.
Step 4--By air action, neutralize airfields in the Buka area."
The officers further recommended that after accomplishing step four, the Allies advance up the axis of either Choiseul-Kieta or that of Treasury Island-Empress Augusta Bay, determination of which was to be in accordance with the reaction of the Japanese to the original plan of attack.37
This new concept immediately underwent study by planners. However, the board of planners, when it made its recommendations on 7 September, had not reckoned with the rapid success of the Vella Lavella operation, which had forced the Japanese to evacuate Kolombangara without a struggle. This favorable eventuality necessitated further re-examination of the entire strategy.38 The changed situation thus dictated that any operation in northern Choiseul, the Shortland Islands, or the Treasury Islands, could depend upon reserves of troops and supplies in the Vella Lavella area. Finally, occupation of both the Treasury Islands and Choiseul would make desirable the establishment of IMAC headquarters at Vella Lavella. Lack of adequate shipping, however, later altered this plan.39
On the other hand, it was felt that occupation of Choiseul coupled with a subsequent attack on the Tenekau Bay area (located between Kieta and Numa Numa on the east coast of Bougainville) would unduly delay completion of the Solomons campaign. Moreover, fighter planes, operating from airfields located on Choiseul, would be unable to cover bomber strikes on Rabaul. Since the overall view of the operation looked toward denial of Rabaul to the Japanese, planners turned their thoughts toward establishment of a beachhead on Bougainville Island proper, by-passing Japanese strongholds in the Shortland-South Bougainville area.40
Upon completing his analysis of "ELKTON III", Admiral Halsey decided that the immediate mission assigned to his South Pacific Force was neutralization of Buka, at the northern end of Bougainville. According to his interpretation, such neutralization could be accomplished if airfields were constructed on the northern tip of Choiseul Island or in the Treasury Islands or both. As a result of this interpretation, a plan for the seizure of these areas was prepared and presented to the Commander, Southwest Pacific Area, in lieu of the Shortlands project, which earlier had been abandoned. However, when plans for attacking Choiseul and the Treasury Islands were submitted to General MacArthur, he objected to their intent, because he felt that the main objective, Rabaul, could not be brought under effective air attack from the proposed bases.41
On 17 September, members of Admiral Halsey's staff, including the Admiral's Marine War Plans Officer, Colonel William E. Riley, met with MacArthur and were told:
Please inform Admiral Halsey that I feel the intent of the J. C. S. directive would not be satisfied by seizing and operating from Treasury and Choiseul Bay; that I believe myself the occupation of a suitable base for air operations on Bougainville proper and essential to further operations against the Bismarcks; that I feel it essential, in order to comply with the intent of the J. C. S. directive, that such a base be seized on Bougainville and that the operation for its seizure be initiated during the present calendar year.
MacArthur felt the Bougainville operation was so essential that he released Admiral Halsey from any real or implied obligation to support the proposed attack of the Southwest Pacific Force at Cape Gloucester, in order that the South Pacific Force could employ all its means in prosecution of the campaign. When asked if there were any consideration that influenced him to prefer an attack on either the east or the west coast of Bougainville, General MacArthur replied, "No, that is entirely as Halsey decides. To me it makes no difference."42
Thus, Admiral Halsey's representatives found that General MacArthur shared their distaste for a frontal assault on the heavily defended areas to the south, and favored instead an attack on a more lightly defended beach.
IMPENETRABLE JUNGLE RISING FROM SWAMP confronted Marines who landed at Empress Augusta Bay. The Northern Solomons campaign was a two-fold struggle in which the terrain presented opposition as formidable as that of the Japanese.
One more short-lived step ensued before the final plan was adopted. Admiral Halsey, on 22 September 1943, cancelled all previously-adopted plans and substituted the following planning alternatives for Phase I: (1) "Seize and hold Treasury Islands and northern Empress Augusta Bay area, Bougainville Island, and construct airfields in the vicinity of Empress Augusta Bay;" or (2) "Seize and hold Treasury Island and Choiseul Bay Area, install radars, PT bases and staging points thereat, and construct airfields in the vicinity of Choiseul Bay Area in preparation for Phase 2." A second phase to either alternative was designed to secure a hostile airfield in the vicinity of Tenekau, Bougainville Island, during the latter part of December, 1943.43
Alternative two had to be disregarded. Early reconnaissance revealed that Buka was too distant from Munda and Barakoma to be considered as an objective, for fighter cover could not be provided over that target; Kahili was also rejected as being too strongly defended for the size of the force available for an attack in that area; the Shortlands were rejected not only because of estimated strength of probable Japanese opposition to an attack in that area, but also because of insufficient beach-area for a landing. Kieta and Numa Numa, on the flanks of the proposed operation for carrying out phase two of the 22 September plan, had been commercially developed, and the Japanese had installed comparatively strong garrisons and improved airfields there. In addition, the longer approach by sea, coupled with an annoying shortage of shipping, would render execution of phase two impracticable. On the other hand, a landing on Choiseul, although tactically sound, would not in itself constitute a direct threat to Rabaul but would be merely a preliminary step towards
additional landings to the northwest. Thus, a period of time would be consumed and excess material wasted without furthering the strategic intent of all the planning--denial of Rabaul to the Japanese.44
Tactical and logistic considerations fixed Empress Augusta Bay on the west coast of Bougainville as the final objective. This area presented, however, disadvantages not previously encountered; but to offset this, intelligence had revealed that the area was not likely to offer strong opposition to a landing; defenses were of a negligible character; beach trails and inland foot-trails were only irregularly patrolled; and the Puruata-Cape Torokina area appeared to fall into a natural defensive region approximately eight miles by six miles in dimension.45 Furthermore, direct information concerning the locality was available, for coastwatching had been inaugurated in May, and this continued to provide some intelligence on terrain and enemy activities through the planning stages.46 In addition to aerial reconnaissance which furnished useful information throughout the operation, a ground patrol was landed by submarine (U.S.S. Guardfish) in the area above the Laruma River; aided by native guides, this patrol scouted the area and tentatively selected an airfield site. Reports by this patrol further confirmed selection of Cape Torokina as the landing beach.47
Major General Charles D. Barrett, who had succeeded Lieutenant General A. A. Vandegrift as Commanding General, IMAC, on 15 September 1943, completed his concept of the operation plan by 27 September, and issued it in a letter of instruction to General Turnage, Commanding General, 3d Marine Division, on the same day.48
The Commander of Task Force 31 (3d Amphibious Force), Rear Admiral Theodore S. Wilkinson, USN, was placed in overall command of the operation as directed by Admiral Halsey's order of 22 September, and, as instructed, he began to coordinate all detailed planning.49 In an additional operation plan dated 12 October 1943,50 Admiral Halsey set 1 November as D-day, and ordered Admiral Wilkinson to seize and hold the Treasury Islands
DIAGRAM, TASK ORGANIZATION, TASK FORCE 31
on 27 October, in order that protection would be afforded to convoys proceeding to Bougainville. In the same plan Major General Ralph J. Mitchell, commander of shore-based air in the South Pacific area, was directed to support the operation with land planes by flying reconnaissance missions and providing air cover and support for land and surface forces engaged. Carrier aircraft under Rear Admiral Frederick C. Sherman, USN, were to support the operations of General Mitchell's planes by an air-strike against the Buka and Bonis airfields, and basing planes ashore when the Cape Torokina airstrips were prepared to receive them.51 A surface task force commanded by Rear Admiral Aaron S. Merrill, USN,52 was to furnish cover for the amphibious landings, and to render support by bombardment of Japanese installations in the Buka-Bonis and Shortland areas during the night of 31 October-1 November. Prior to and during this operation, submarines of the South Pacific Force (Task Force 72, borrowed from Admiral Barby's Seventh Fleet), under Captain James Fife, Jr., USN, were to attack Japanese shipping and surface units and to furnish reconnaissance information. It was also planned that aircraft of the Southwest Pacific Area would bomb Rabaul in the meantime.53
Plans of 12 October directed Admiral Wilkinson to be prepared to establish a patrol torpedo boat base on northwestern Choiseul on five days' notice.54 It was feared that the landing on the Treasury Islands, located on a direct line toward Bougainville, would serve to point out to the Japanese the main objective in the Bougainville area and cause them to send reinforcements there.
To throw the Japanese off the scent, therefore, Major James C. Murray, Staff Secretary of IMAC, suggested that elements of the Choiseul landing be retained in the final plan, even though it was not intended to seize the island, or any position thereon. Thus, the diversionary raid on Choiseul was conceived.55 Choiseul's size and position would suggest its use as a base against Shortland or Bougainville, and if Bougainville, U.S. seizure of a base on Choiseul would indicate a landing on the east rather than the west coast.
Operation orders for the Choiseul diversion, therefore, were given by IMAC on 22 October,56 directing the 2d Parachute Battalion, 1st Marine Parachute Regiment, FMF, to land at Voza, Choiseul, during the night of 27 October, not only to conduct a raid but also to be prepared to establish a permanent base if the situation should so warrant.57
In order to facilitate and expedite planning, all Task Group commanders were authorized to work directly with the various staff sections of the units they were supporting.
Final concepts of the campaign were dictated by the decision that the operation should look more definitely toward reduction of Rabaul. It was decided to establish a beachhead on Bougainville proper; as a protective feature for Bougainville-bound convoys and as a boat refuge, the Treasury Islands were to be taken; the landing on Choiseul was modified to a diversionary movement "a series of short right jabs designed to throw the enemy off balance and conceal the power of the left hook to his mid-riff at Empress Augusta Bay."58 Japanese strength, availability of airfield terrain, the position astride the enemy system of communications between Rabaul and the Solomons, airline distances which placed it at the geographical center of the Bougainville arc of Japanese air and naval installations, availability of advance naval base facilities, and the comparatively favorable convoy routes, plus the opportunity of maintaining U.S. shore-based aircraft over the target for a maximum time, all dictated selection of Cape Torokina as the projected beachhead.59
In speaking of this final concept, Admiral Halsey said,
Enthusiasm for the plan was far from unanimous, even in the South Pacific, but, the decision having been made, all hands were told to get going.60
The final concept of the plan had now emerged. On 15 October 1943, IMAC issued the operation order for the landing on Bougainville.61
This order directed a landing by the 8th Brigade Group, 3d (NZ) Division, on the Treasury Islands on 27 October to capture or destroy Japanese forces located in the area; to establish long range radars and an advance naval base in order that a staging refuge for landing craft and operating facilities for motor torpedo boats could be put into operation. The 3d Marine Division was to land on 1 November on beaches between Cape Torokina and the Laruma River, and on Puruata and Torokina Islands, to capture or destroy Japanese forces located there and to build and protect airfields, radar sites, and an advance naval base in that area. This operation was undertaken to obtain positions for further offensive operations. Prior to and during the operation, aircraft of the Southwest Pacific Force were to fly missions against Rabaul, and units of Task Force 33 were to carry out reconnaissance, search, and striking missions on the surface. Task Forces 37, 38, and 39 were named the support of Task Force 31, while submarines of Task Force 72 were directed to execute attacks against enemy shipping and surface units in order to give additional support to the operation.
Initially the order specified the coconut plantation east of Cape Torokina as the site for the airfield, but a Naval Construction Battalion (71st NCB) was instructed to reconnoiter, as soon as safety permitted, for the purpose of finding another landing strip.62
If a weak link existed in the chain of offensive preparations which IMAC began forging in the fall of 1943, it was the bare minimum of available
GENERAL A. A. VANDEGRIFT, Commander of the I Marine Amphibious Corps, who shaped the Landing Force plans for Bougainville, Treasury Islands, and Choiseul. After the sudden death of Major General Barrett, General Vandegrift found himself unexpectedly charged with executing his earlier plans.
shipping which might, in the event of any early losses, jeopardize adequate support of the landing force. It was therefore decided that in addition to the staging-storage-bivouac area west of the Poha River on Guadalcanal and the Fourth Base Depot in the Russells, an advanced supply base would be necessary.63
Supply items peculiar to the Marine Corps were stocked at the Fourth Base Depot. The principal source of general supplies, however, from which probably a tenth of the supply tonpage would be drawn was Guadalcanal.
Therefore, to utilize available shipping during the preparatory phase and to reduce the amount of shipping which might be needed after inception of the Treasury-Bougainville operations, it was decided to establish a depot forward of Guadalcanal, at a sufficient distance to permit movement of supplies to the target area by LCT's (landing craft, tank), if a shortage of LST's (landing ship, tank) made it necessary. It was
also considered desirable to have a minimum of ammunition, rations, and petroleum as far forward as possible in the event an emergency made it necessary to move them to the fighting fronts hurriedly.
Although the Japanese still had a sizable force on Vella Lavella, that island afforded the best location for such a depot. An engineer reconnaissance party was dispatched to reconnoiter the east and west coasts of Vella Lavella, and it subsequently reported that the east coast from Barakoma to Orete Cove furnished landing, storage, and bivouac areas suitable for the project.64
On the basis of this finding a Task Organization was formed to establish a hospital and Base Depot. This unit was moved to Vella Lavella in echelons. Major James C. Murray accompanied the first of these as staff representative for CG, IMAC, and reported upon arrival to CG, 3d NZ Division, then Island Commander, arranging with him allocation of areas from the Juno River to Bipora Village.65 Within these areas Major Murray carried out his assigned mission.
Upon arrival of the second echelon, Colonel Alton A. Gladden, who had been designated commanding officer of Corps Troops, Vella Lavella, assumed responsibility for further direction of the project.66
While planning was underway, meanwhile, combat intelligence preliminary to the operation was being gathered on a scale not heretofore attempted. Many patrols were landed on proposed objectives, while map and aerial photograph studies were made of the areas. Interviews were held with individuals who were acquainted with the localities which were under study. Slowly the early trickle of information grew into a flooding tide.
Accumulation of Intelligence
Various means were employed to collect combat intelligence for the Bougainville operation. Amphibious reconnaissance patrols were landed by submarine, seaplane, and motor torpedo boat to penetrate defended beaches; these were the first to be undertaken on so wide a scale in the South Pacific and they key-holed the Shortland area, the Treasury Islands, Choiseul, Santa Isabel, and the Kieta and Numa Numa areas.67 Extensive aerial reconnaissance provided much valuable information in regard to southern and northern Bougainville. In order to obtain specific information relating to terrain, hydrographic conditions, enemy dispositions and defenses, and any other information that might be of use, three patrols were landed from a submarine in rubber boats on Shortland and on the Treasury Islands on the night of 22-23 August. These patrols were withdrawn by the same means on the night of 27-28 August.68
On 3 September 1943, at 2230, a patrol composed of Captain J. R. Harper, Lieutenant R. C. Parlier, USN, Lieutenant Morse Holladay, Platoon Sergeant W. A. Dunsworth, Corporal D. L. Kettler, and Private First Class T. G. Simpson debarked from a patrol torpedo boat into native canoes at a point southwest of Rhodes Island, and landed on Santa Isabel just opposite Bero Island, crossed the main island and reconnoitered the Rekata Bay area as far south as Suavanu Point, and Papatura Fa Island. Having completed reconnaissance of the Suavanu Point area. Harper's men moved via native canoe to the Kia Bay area where they reconnoitered but found only negative information. Upon returning to its starting point, the group left the Santa Isabel area on 9 September at 2230.69
On 6 September 1943, at 2300, a patrol composed of Captain Gordon Warner, Captain William R. Helmer, Lt (jg) K. P. Hill, Corporal Edgar C. Terrell and Corporal William Weber, was discharged from a PT boat into a native canoe about one-half mile southeast of the Ruravai Island mission, off the southwest coast of Choiseul. Landing at the mission wharf on the southeast side of the island about 0115, the patrol spent the remainder of the night in the wharf boat shed. Next morning the group proceeded via canoe in a northwesterly direction along the shore toward Posarai Village, to observe shore line, beaches and jungle. On reaching Posarai Village at 0830, Warner's men proceeded to move along inland waterways to Todoro Lagoon, where they disembarked and continued northwest by native trail to a village on the Doroko River, just north of Mount Sambe. At this point the patrol made camp for the night.70
On 8 September 1943, the party left the Doroko River and proceeded south to the high cliffs overlooking Zoga where an observation post was established in order to observe Japanese positions, the beach, harbor channels, and terrain. After spending the day thus occupied, the patrol left its observation point and traveled in a northwesterly direction to the village of Gorabara, where it arrived at 1800 and set up camp for the night.
The following day the five men traveled for a period of two hours via canoe to Panai, where they landed and followed a native trail to a point of observation about one mile north of Kakasa. Information was gathered in regard to the Kakasa area, before they retired northwest at 1530 in order to make camp on the bank of the Verulatu river. Ascent of Mount Maitabe to the north consumed the entire working day on 10 September, and it was not until 12 September that the group reached the coastwatcher's station just north-northwest of Kanaga village.71 From Kanaga village, the party proceeded via canoe downstream on the Kanaga river to the north coast of the island, along which it proceeded along the inner reef to Nanano Island, where it arrived at approximately 0900, and finally made the contact at 1215 with a Navy patrol bomber, which had been directed to rendezvous with it.
During their travels the men observed some Japanese activity, and made a study of the beaches and terrain which indicated that there were few good airfield sites or landing beaches on the parts of the island observed. Upon returning to Guadalcanal, the patrol reported its findings to the planners.72
Two patrols, number one being composed of Commander W. Painter, (CEC) USNR, Captain J. J. Delahanty, Lt. (j. g.) F. B. Rauch,
USN, Pilot Officer Leatham (of the Royal New Zealand Air Force), and Private A. N. James; number two consisting of Captain R. Enich, Lieutenant W. T. Manley, Lt. (j. g.) E. Greathead, USN, and Sergeant J. Mandel, were landed from a Navy patrol bomber at Nanano Island off the northeast coast of Choiseul on 22 September 1943, at 1500, where the first patrol was joined by Sub-Lieutenant C. W. Seton of the Royal Australian Navy, the coastwatcher in this area.
From this point the groups proceeded by native canoe to the village of Mamarana, near the northwestern tip of the main island, where they arrived at 0600 the next day. Upon landing, the parties moved overland toward Mount Gaili in the west, and, on arriving at that position at 1100 on 24 September 1943, separated, the first reconnoitering to the west, the second to the south. General mission of patrol number one was to reconnoiter in the Choiseul Bay area and then work northwest along the coast away from Japanese-occupied areas in an effort to study the terrain with a view to establishment of radar stations, airfields, bivouac areas, dump areas, and landing beaches. Patrol number two had the same general mission as patrol number one, except that the area in which it was to conduct its reconnaissance lay between the mouth of the Warrior River Point and Moli Island, thence overland to Bubukana.
Patrol number one bivouacked east of Terikukure Plantation at 1700 on the evening of 24 September 1943, shortly after hearing Japanese activity. The following day it reconnoitered the Choiseul Bay area and in the evening observed a B-24 attacking Japanese positions there. On 25, 26, and 27 September, the area along the beach from Poroporo to Maramana was studied. On 28 September the party proceeded to the rendezvous at Salamanda to meet patrol number two.
In the meantime, patrol number two had passed 24, 25, and 26 September reconnoitering along the beach southwest to the vicinity of Moli Island. The following two days were spent traveling overland from Vudertaru to a point on the beach near Pola, where it boarded a canoe and proceeded to the rendezvous at Salamanda, and rejoined patrol number one. Patrol number two did not observe any Japanese land or sea activity but did spot Japanese planes on 25, 26, and 28 September. Accompanying natives, however, reported Japanese activity in the vicinity of Moli Island, and at mouths of various rivers in the area.
After joining forces, the groups returned to Nanano via Susuka and boarded the Navy patrol bomber about 1500. They had observed several fair airfield sites in the vicinity of Choiseul Harbor, and had definite knowledge of enemy barge activity in that area. They had also located several good landing beaches and, from conversations with natives and the coastwatcher, were able to estimate that there were about 300 Japanese in the Choiseul Bay region and about 1,000 in the Kakasa Region.
Along the southwest coast of the island, the patrols reported, there was no suitable site for a landing strip; however, there were some level areas on the island, but these were swampy, and a great deal of work would be necessary in order to drain them. On the other hand there was a suitable area slightly northeast of Poroporo, although it would probably require 30 days' work on the part of an entire construction battalion or its equivalent to grade and surface a strip of 150 by 4,000 feet, and a suitable beach road had to be constructed before work on the airfield could begin. It was also noted that a strip could be built on Kondakanimboko Island if the landing beach could be improved, the channel between island and mainland were filled in, and a road were built from the beach to Sipasai and Taro Islands.
Reported further were some locations where radar could probably be placed, but these would have to be spotted after troops had come ashore. The nature of the terrain had prevented direct observation to locate those positions accurately.
These findings were immediately conveyed to planners upon the patrol's return to Guadalcanal, and the planners incorporated the findings into operation plans.73
Two patrols composed of Navy and Marine
Corps personnel were sent to Bougainville on 23 September 1943, one going to the Kieta area on the northeast coast and the other going to the Cape Torokina area on the southwest coast. On the northeast coast the party was composed of Captain Bertram S. Behrens, First Lieutenant Oscar Salgo, Lieutenant Stuart (RANVR), Lt. (j. g.) Woods, USN, Lt. (j. g.) Reed, USN, Platoon Sergeant Gwost, Sergeants Hagler, Schramm, Bisol, Ober, Clark, and LaMonte, Corporals Aaron and Splain, Private Roberts, and four native guides, embarked in the Submarine U.S.S. Gato. This group had as its mission reconnaissance of probable landing beaches in the vicinity of Cape Mabiri, from Asitavi to Metong Plantation. To accomplish this mission, the party divided into two sections each night and reconnoitered specific sections of beach area. In the meantime the submarine took soundings off shore for the purpose of charting the coast. As the men moved from submarine to beach, they took soundings on the way in. When ashore, patrol members would examine the beach area and the land directly behind it.
A considerable amount of Japanese activity in the area was noted, but despite this, a reasonably accurate and complete survey of landing beaches and approaches thereto in the area was made. The men returned to the submarine each night, and at 0145 on the morning of 28 September, returned for the last time.74
The party on the southwest coast consisted of Captain Harry B. Barker, First Lieutenant James C. Clemmons, USA, Lts. (j.g.) Kenneth P. Hill and Edward C. McBurney, Ensign Morris N. Palmer (all three USN), Second Lieutenant William C. Langdale, Ensign Antone Jossten (RANVR), 30 enlisted men, and four natives. This group was to reconnoiter beaches and terrain between the Torokina River and Atsinima Bay, five miles inland and the islands offshore. Embarked in the submarine U.S.S. Guardfish, the men landed in rubber boats and operated ashore from 2000, 23 September 1943, to 2000, 26 September 1943.
While ashore, this party observed that terrain behind the beaches was extremely swampy, and that few suitable sites for airstrips existed, neither did it find the reported road along the coast. On 11 different occasions, Japanese planes were sighted flying along the coast. Other Japanese activity noted was the presence of one sentry at the mouth of the Laruma River, and the sound of heavy guns in the vicinity of Mutupena point, and glare of antiaircraft searchlights sweeping the sky from the vicinity of Mavavia.75 In the meantime the Guardfish had made an extremely accurate surey of hydrographic conditions offshore. All information obtained was duly reported upon return to Guadalcanal.76
On 1 October 1943, a low level aerial reconnaissance mission was flown over the Cape Torokina region by a PB2Y piloted by Lieutenant (j.g.) Hager and carrying an observation party consisting of Lieutenant Colonel S. S. Wade, Lieutenants (j.g.) Kenneth P. Hill and Edward C. McBurney. This party also noted that swamps existed inland from the beaches, and that the small coconut grove just east of Cape Torokina might be a suitable site for a small landing strip.77
Lieutenant Colonel James M. Smith, 3d Division Headquarters, and Captain John I. Delehanty, 21st Marines, landed at Atsinima Bay on the night of 27 October with a coastwatcher party under direction of Lieutenant Keenan (RANVR)78 in order to determine strength, location, and disposition of Japanese forces and location of their coast defense guns in the Cape Torokina area. This information was to be communicated via radio to Commander, Task Force 31 by 1200, 30 October. If the Japanese did not have more than 200 to 300 men in the area, a large fire, visible to seaward, was to be lighted on the beach. Furthermore, this party was to locate high, dry, ground behind the beach between Cape Torokina and the Laruma River, and to be prepared to guide assault troops to it. Native scouts were to be sent to the Jaba area
to observe and report on Japanese activity noted in recent photographs of that area. This information was to be made available to the Commanding General, IMAC, and the Commander of Task Force 31, immediately after landing of assault units had been accomplished.
Existence of high ground about 1,000 yards east of the Koromokina River, suitable for radar installations was to be determined. Immediately upon landing of the first wave of invasion forces, the party was to rejoin our forces by approaching from the west displaying an American flag.79
Intelligence information regarding the strength and disposition of Japanese forces in the Cape Torokina area was meager, and the very nature of the jungle in that area prevented secrets hidden beneath dense foliage from appearing on aerial photographs. True, there were signs of activity on Cape Torokina itself, but the character of the photographs prevented an exact determination of the extent or characteristics of activity there. Equally obscure was the size and extent of swamp areas immediately inland from the beach, and to complicate matters there was practically no reliable hydrographic information available.
Charts of the northern Empress Augusta Bay area were vague, and an aerial photographic survey of the region had disclosed that the coast line was eight to ten miles removed from positions indicated on the charts. Early reconnaissance by the submarine Guardfish had indicated presence of uncharted shoals. Consequently, approach to the target had to be scheduled to occur after daybreak.
Intelligence estimated enemy strength as follows: Northern Sector (Buka area), 5,000 troops; Eastern Sector (Numa Numa-Kieta area), 5,000 troops; Ballale and adjacent islands, 3,000 troops; Shortland area, 5,000 to 6,000 troops; Southern Sector, 17,000 troops. The only enemy concentration of any strength in the Empress Augusta Bay area was estimated to be 1,000 troops at Mosigetta. These appeared to be, for the most part, engaged in cultivating extensive rice fields.
Throughout September there was ample evidence that the enemy was strengthening his Bougainville defenses. He appeared determined to continue defense of the Buin-Shortland-South Bougainville area. This was indicated by (1) rapid replacement of aircraft combat losses, (2) reinforcement and expansion of facilities of advance bases in the northern Solomons, (3) maintenance of supply, even at the cost of critically needed ships, (4) continued development of existing air bases and construction of new ones.80
At this time, however, the enemy was under severe Allied punishment from the air, and it was increasingly evident that he was gradually placing less confidence in his ability to maintain the safety of his air bases in southern Bougainville. Strength of aircraft at Buka continued high, but concentrations at Kahili, Ballale and Kara began to dwindle. While this was clearly a recognition of Allied air and naval superiority (captured documents admitting it), the Japanese conceded nothing to Allied land forces. The enemy was preparing his bases for a frontal assault, and now he committed himself to a passive rather than an active defense, digging in, taking shelter from artillery and aerial bombardment and making only minor air raids at night.
It was anticipated by now that the Japanese were, after Kolombangara, preparing to counter further bypassing or flanking movements and had adopted one of two courses: either a large increase in the number of their coastal areas provided with strong, static defense facilities, or a considerable increase in mobile troops to repulse landings on lightly defended areas; intelligence indicated that the second course was held in high favor on Bougainville.
Not until later in October, however, were there any indications that the Japanese had started to
apply this factor to the Empress Augusta Bay area. Intelligence photos for the first time indicated that there was any military activity in the Cape Torokina region. It was in no wise alarming, indicating no enemy awareness of his danger, and consisted merely of a few emplacements for antiaircraft guns, a slit trench, and some minor building construction. Further defensive installations, possible 20 gun sites, were discovered behind the point of the Cape in the closing days of October. There was nothing to indicate, however, that the enemy had fully awakened to his peril.81
Assault Unit Planning
As intelligence poured in, planning continued unabated. CTF-31 and CG, IMAC, ironed out details so essential to a successful operation. The task organization for the operation was set up by letters of instruction from General Barrett to General Turnage and Brigadier R. A. Row,82 of the New Zealand Army.
General Turnage immediately instituted the detailed planning necessary to execute his assigned mission. Preliminary planning had begun previously upon receipt of a verbal warning order, later confirmed by a letter of instruction. According to these orders, the 3d Marine Division was directed to land in the vicinity of Cape Torokina to seize a beachhead between the Torokina and Laruma Rivers extending inland approximately 2250 yards and over a front of about 7350 yards.
Troops were to proceed toward the objective in two echelons via Efate, New Hebrides, where rehearsals would be conducted. Due to the shipping shortage, only those items absolutely essential to maintenance of life and combat were to be taken to the combat zone by the first echelon. To do this, personnel landing in assault would carry on their persons only two days' rations and two units of fire.83
On 28 September, General Barrett issued a letter of instruction to Brigadier Row concerning the Treasury Islands operation.84
In addition to providing for combat details, the final order, issued 15 October 1943, assigned code names to the operation.85 One week later, on 22 October, the 8th Brigade Group was directed to land on the Treasury Islands on D-day minus five days, prepared to furnish one rifle company "as garrison for PT boat advance base in NW Choiseul" on four days notice from CTF-31.86
Since the tentative task assignment of 27 September had envisaged an early landing by the 37th Infantry Division,87 the 3d Marine Division was to prepare to carry out its assigned mission in coordination with the 37th upon the latter's arrival subsequent to D-day, in order that the perimeter could be extended and additional airfield sites, naval base facilities, and long-range radar sites could be incorporated into the defensive sector.88
Tactical planning met with several immediate limitations. Transportation allotted IMAC for movement to the combat area for initial landings consisted of only eight APA's (attack transports) and four AKA's (attack cargo vessels). It therefore became necessary to plan to transport the Division to the landing area in succeeding echelons.89
Evaluation of Japanese strength and known dispositions on Bougainville, coupled with known characteristics of the Cape Torokina area, i.e., jungle covered swamp and lack of roads, evoked the conclusion that whatever defenses the Japanese had established in the area must be localized
and restricted by terrain to small detachments. It was not expected that large bodies of reinforcing troops, such as could endanger the bulk of the Division, could be brought to the area readily and committed, except by sea. Whatever initial opposition the enemy offered to our landing would have to come from the air. Due to the proximity of Japanese air bases, however, this reaction was fully expected.
Tactical considerations dictated that the 3d Division embark three task units in the available ships. Two of these consisted of reinforced infantry regiments of four landing teams each, while the third unit consisted primarily of the 3d Defense Battalion, FMF, which was to furnish antiaircraft artillery support. With each task unit it was necessary to embark logistic personnel, such as engineers, air, naval base, signal, and service troops, whose primary mission was immediate development of air and naval base facilities. Furthermore, loading of each ship was restricted to about 500 tons of supplies and equipment, and all elements of the Division embarked thereon would have to be landed simultaneously, so that ships could be unloaded within a five to six hours' time-limit imposed by CTF-31.
Since swamps were known to exist immediately behind the beach, and space for unloaded supplies would therefore be at a premium, landing plans of the 3d Marine Division therefore envisaged employment of 12 beaches, one per ship, 11 extending west from Cape Torokina some 7350 yards, and one on the north (inner) shore of Puruata Island. (For beaches and detailed scheme of maneuver, see map 3.) Each combat team was to overrun and destroy all Japanese positions in its respective sector, seize a shallow beachhead initially and institute immediate vigorous reconnaissance patrolling to front and flanks, meanwhile unloading ships as rapidly as possible and installing beach defenses. At the same time, the units were to be prepared to deploy to right or left, or to make a lateral shift in either direction as a unit.90
Task unit A-1 (3d Marines, reinforced), organized as noted, into four landing teams, embarked on 13 October--before IMAC had issued Operation Order No. 1--from Tetere, Guadalcanal, in its attack transports with task unit Headquarters and transport division flag in the U.S.S. Jackson, famous throughout the fleet for the jingle about "relaxin' in the Jackson". Cargo accompanying each landing team was limited to 520 tons.
The transport division proceeded to Efate, in the New Hebrides, where, from 16 to 20 October, the force engaged in landing exercises. The loaded ships then stood by at Espiritu Santo until 29 October, when they proceeded to Koli Point, Guadalcanal, where, on 30 October, Commodore Reifsnider and General Turnage embarked on board the U.S.S. Hunter Liggett.91 The transport division made its rendezvous with the balance of the attack force 84 miles west of Guadalcanal at 0740 on 31 October.92
Task unit A-2 (9th Marines, reinforced) went through the same routine with its transport division during the period 18-25 October93 at the conclusion of which it proceeded to a point about 20 miles south of the eastern end of San Cristobal Island where it joined Task unit A-3 just before dark on 30 October.94
Task unit A-3 (3d Defense Battalion, reinforced) conducted landing exercises at Guadalcanal during the period 26-29 October and effected final rendezvous with the other task units on 31 October.
As a result of the rehearsals it was estimated that the average time for unloading the APA's and AKA's would be four and one-half hours, assuming a 2,500 yard-run from anchored transports to beach.95
During the period 14 to 17 October four APD's (destroyer-transports) and eight LCI's (landing
craft, infantry) were made available to the 8th New Zealand Brigade for training. Landing exercises were carried out on Florida Island.
Loading and embarkation of the New Zealand Brigade for the Treasury Islands operation occurred without incident during the period 23 to 26 October.96
Thus, by the end of October, IMAC troops were fully trained and thoroughly acquainted with plans which had been months in preparation. With a feeling of confidence in themselves, their comrades, and their weapons, they patiently awaited the dawning of that day toward which all their effort had been bent--1 November 1943.
Training was finally completed and preliminary maneuvers were underway. Troops had already been landed and were fighting in the Treasury Islands and on Choiseul.97 At last Admiral Halsey was prepared to launch the principal amphibious operation: the operation designed as the knock-out punch of the Northern Solomons campaign--Bougainville.
Landing operations at Empress Augusta Bay were coordinated with a series of surface bombardments and air strikes designed to neutralize Japanese airfields on the northern and southern extremities of Bougainville.
While these operations were underway, the
DIAGRAM, IMAC TASK ORGANIZATION
Transport Divisions were protected by Destroyer Divisions 90, 89, and 44, while Task Forces 37, 38, and 39 moved in to furnish cover and support against attack from Japanese surface forces. Task Force 72 furnished submarine support as planned. Minesweeping and protective minelaying operations were carried out. During the night of 31 October-1 November, overhead cover was provided by Navy Liberators, Ventura night-fighters, and "Black-cats".98
As a precaution against alerting the Japanese to our main intention, Admiral Wilkinson had postponed the rendezvous of the three transport divisions until the morning of 31 October (D minus one day). This took place at 0740, that day, and Commodore Reifsnider assumed tactical command of the Transport Group and screen. From that time until 1800 the force followed a northwesterly course, then steering toward the Shortlands Islands area in order to deceive the Japanese into thinking that the target was to be somewhere on southern Bougainville, associated with our seizure of the Treasury Islands and our landing on Choiseul. After night fell, the course was again altered, this time in the direction of Empress Augusta Bay. During the night eight separate bogies were chased by the night fighters, which were being constantly vectored99 by the Conway, but none of the Japanese planes apparently sighted the convoy.100
DIAGRAM, TRANSPORT GROUP ORGANIZATION
Naval Operations Off Bougainville
31 Oct-2 Nov 1943
As the Northern Attack Force ploughed through phosphorescent seas toward Empress Augusta Bay, its leaders felt secure in the knowledge that the starboard (north) flank of the convoy was secured by operations underway in the Treasury Islands, and that as a result of the Choiseul diversion101 the Japanese were unaware of our ultimate point of attack. Furthermore, Admiral Wilkinson and General Vandergrift realized that they would be forewarned by long-range radar in the Treasury Islands in case of any Japanese air attacks that should be forthcoming.
The night of 31 October was dark. Tiny freckles of clouds momentarily obliterated certain stars as an observer would watch. Below decks, purposeful but unhurried last-minute preparations were being made for the supreme test which was to take place shortly before dawn. The Marines were confident that their long months of training would assure them of ultimate victory. In their minds the issue was never in doubt.
COMMODORE REIFSNIDER EXPLAINS details of the approach on Bougainville to General Turnage, commander of the 3d Marine Division. This picture is believed to have been taken on the morning of D-day minus one, as Reifsnider's Transport Group moved northward past the Treasury Islands.
Table of Contents ** Previous Chapter (Foreword) * Next Chapter (2)
1. In addition to Bougainville, the Northern Solomons campaign embodied the following operations: Treasury Islands, Choiseul, Green and Emirau.
2. US Strategic Bombing Survey, Campaigns of the Pacific War (Washington, 1946), 3. Hereinafter cited as Campaigns.
3. Eric A. Feldt, The Coastwatchers (New York, 1946), 58-60.
4. Ibid., 60-74. None of the sources consulted gives the identification of this body of men. The point of landing was apparently on the east coast of the island, somewhere between Numa Numa and Kieta.
5. Ibid., 58-70.
6. F. T. Miller, History of World War II (Philadelphia and Toronto, 1945), 607.
7. Interrogations of Japanese Officials (2 vols., Washington, 1946) I, 13. Hereinafter cited as Interrogations. Cf. Campaigns, 7.
8. Miller, op. cit., 609.
9. Ibid., 617.
10. Campaigns, 175, 184, 191.
11. E. J. King, U.S. Navy at War: Official Reports to the Secretary of the Navy (Washington, 1946), 103.
12. Campaigns, 176.
13. Joint Chiefs of Staff, ELKTON I, Plan of Operations in the Pacific, 1-3. Hereinafter cited as ELKTON I.
14. The Bougainville Landing and the Battle of Empress Augusta Bay; 27 October-2 November, 1943 (Washington, 1946) 4-8. The work is cited hereinafter as Combat Narratives.
15. COMSOPAC, South Pacific Campaign--Narrative Account, 6. Hereinafter cited as Halsey, SoPac Narrative. It should be noted here that in the chain of command between Admiral Halsey and the Joint Chiefs of Staff was the Commander, Pacific Ocean Areas, Admiral Chester W. Nimitz, USN.
16. Campaigns, 142.
17. Ibid., 176-177.
18. King, op. cit., 63.
19. Campaigns, 142-143, 175.
20. King, op. cit., 65.
21. Ibid., 65-66.
22. Campaigns, 143.
23. King, op. cit., 66.
24. Interrogations, I, 81.
25. King, op. cit., 66.-67. See also IMAC Bougainville Beachhead, Phase I, 4. Hereinafter cited as IMAC Opn Rpt.
26. Campaigns, 143, 150.
27. This was the name given by the Japanese to the main body of their fleet.
28. Campaigns, 42.
29. Ibid., 7.
30. Ibid., 186.
31. IMAC Opn Rpt, I, 23-31.
32. Halsey, SoPac Narrative, 8. For a detailed analysis of factors which prompted the decision to land at Empress Augusta Bay, see page 11 following.
33. IMAC Opn Rpt, I, 5-6.
34. Ibid., 25.
35. COMSOPAC, Letter of Instruction of 26 July 1943, serial 001346 and COMSOPAC, Letter of Instruction of 5 August 1943, serial 001421. Hereinafter cited as COMSOPAC LofI, 26 Jul 43 and COMSOPAC LofI, 5 Aug 43.
36. IMAC Opn Rpt, I, 8. See also IMAC LofI, 17 Sep 43, serial 0036, entitled "Warning Order 18-43"; this document is cited hereinafter as Warning Order 18-43.
37. Memorandum for COMSOPAC from Fitch, Harmon, Barrett, and Wilkinson, dated 7 September 1943.
38. Combat Narratives, 4.
39. IMAC Opn Rpt, I, 6.
40. Joint Chiefs of Staff, ELKTON III Operation Plan, 26 April 1943, 1-3. This plan had been Phase II of the JCS directive of 2 July 1942. It had been prepared in SWPA for Kavieng and was approved by JCS.
41. Letter, MacArthur to Halsey, dated 11 Sep 43.
42. Planners' Discussion--Bougainville Operation. Cf. Halsey, SoPac Narrative, 8.
43. COMSOPAC Warning Order of 22 September 1943, serial 001901. Hereinafter cited as COMSOPAC Warning Order 22 Sep 43. These changing plans, naturally, made IMAC planning most difficult.
44. 3d Amphibious Force, Seizure and Occupation of Northern Empress Augusta Bay. . . . Hereinafter cited as CTF-31 Opn Rpt.
45. Combat Narratives, 5.
46. Feldt, op. cit., 134-139, 145. The Japanese became aware of the presence of coastwatchers toward the end of August, and forced their withdrawal.
47. IMAC Opn Rpt., I, 4. See section entitled "Accumulation of Intelligence", this chapter, for further details.
48. IMAC LofI, 27 Sep 43, serial 001A.
49. IMAC Concept of Operations Plan I, serial 002A, hereinafter cited at IMAC Concept. See also COMSOPAC Warning Order 22 Sep 43.
50. COMSOPAC Opn Plan 16-43, dated 12 Oct 43.
51. Campaigns, 151. In Admiral Sherman's force were the carriers Saratoga and Princeton, the anti-aircraft cruiser San Juan and San Diego, and ten destroyers, all comprising Task Force 33.
52. Ibid., 153. Admiral Merrill's force (Task Force 39) consisted of the cruisers Montpelier, Cleveland, Columbia, and Denver, and eight destroyers.
53. COMSOPAC Opn Plan 16-43.
54. Loc. cit., gives a detailed analysis of the task assigned Admiral Wilkinson.
55. IMAC Opn Rpt, I, 3.
56. IMAC Operation Order No. 2, 22 Oct 43, hereinafter cited as IMAC Opn O #2.
57. Krulak letters of 5 Nov 43 and 22 Nov 43. For complete details concerning this diversion, see Chapter IV, section 2, following.
58. General Roy S. Geiger, in his report on the operation, made this statement.
59. IMAC Opn Rpt, I, 3.
60. Halsey, SoPac Narrative, 8.
61. IMAC Operation Order No. 1, 15 October 1943. Although planning and preparation of the order was done under the direction of General Barrett, it was signed by General Vandegrift who had succeeded Barrett as CG, IMAC, upon the untimely death of the latter.
62. IMAC Opn O # 1.
63. IMAC Opn Rpt, I, 7-8.
64. Ibid., 6-9.
65. IMAC Opn Rpt, I, 8. An amusing incident occurred in this connection. When moving from Guadalcanal to Vella Lavella, Murray carried a letter of introduction from General Barrett to Major General H. E. Barrowclough, CG, 3d NZ Division. In preparing this letter a stenographer had misspelled the General's name as "Bearclaw", its phonetic approximative. When Murray handed him the letter, General Barrowclough, a dignified Imperial officer, winced perceptibly. For a complete report of this incident, and the movement of the Task Organization to Vella Lavella, see Ltr LtCol J. C. Murray to Dir Div Pub Info, dtd 4 Nov 47, Serial 2185 over JCM/vca.
66. IMAC Engineer confidential Memorandum to CO, 19th Marines, 31Aug43; Corps Engineer confidential Memorandum to CO, 19th Marines, 3 Sep 43; CG, IMAC secret LofI to OinC, Vella Lavella Det, 77th NCB, 30 Sep 43; CG, IMAC LofI to CO, IMAC NCB, 14 Oct 43.
67. IMAC Opn Rpt, 1, 3.
68. Their reports indicated that the best landing beaches in the vicinity that were suitable for all types of landing craft were located between the Saveke River mouth and Falamai Point. Furthermore, these beaches had an ample water supply nearby and also had adequate dispersal and bivouac areas. The party saw signs of Japanese activity but actually sighted no Japanese.
In order to obtain last minute information before the GOODTIME (Treasury) operation, a reconnaissance party of two New Zealand noncommissioned officers (Sgt. W. A. Cowan of the 8th Brigade Intelligence Section and Corporal Nash of the Australian Naval Intelligence Service), and some natives was landed by patrol torpedo boat on Mono Island, the largest of the Treasury Group on the night of 21-22 October and was withdrawn in the same manner the following night. This party reported that according to natives, the Japanese had recently landed reinforcements in the area and that their strength was about 225 men; that medium caliber guns had recently been emplaced on both sides of Falamai Point; that machine guns were emplaced on Mono Island along the approaches to the landing beaches; that there was an observation post located at Laifa Point with direct wire communication to the radio station near the Saveke River; and that Stirling Island was unoccupied by the Japanese. When they were withdrawn, the party brought with them several Mono Island natives who were to serve as pilots for the attacking forces on the day of the landing.
Finally on the night of 25-26 October an advance party of three New Zealanders and two natives led by Sgt. Cowan was landed in order to cut the communication lines between Laifa Point and the radio station (Combat Narratives, 12-13) at 0400 on 27 October, to establish an observation post in order to observe Japanese movements, and to furnish native guides to the combat units. (Row Rpt. 3. The correct title of this document is "HQ 8 (NZ) Bde Gp, 3 (NZ) Div NZEF, In the Field, REPORT ON OPERATIONS--TREASURY IS., (OP 'GOODTIME')" 30Nov43).
69. 3d MarDiv, Santa Isabel Patrol Report. See bibliography for complete description of this document.
70. For map of Choiseul showing patrol routes, see Map 24.
71. LtCol Warner reports that his patrol was the first group of white men to scale Mount Maitabe.
72. 3d MarDiv, Patrol Reports-Choiseul, "K" Patrol. See bibliography for clarification.
73. Ibid., IMAC Patrol.
74. IMAC Patrol Report on Empress Augusta Bay Area. See bibliography.
75. 3d Amphibious Force Reconnaissance Report. See bibliography for complete description of this document.
76. CTF-31 Opn Rpt, 5.
77. IMAC Opn Rpt, I, 18.
78. Feldt, op. cit., 161-162.
79. IMAC Opn Rpt, I, 20. Because of excessive terrain difficulties encountered, and because of radio failure, none of the information that this group obtained was made available to the invading forces until they were finally ashore; furthermore, the fire was not lighted. Lieutenant Colonel Smith rejoined the American forces as planned on 1 November, and made an oral report to General Turnage. Smith displayed the first American flag to be shown ashore. The remainder of the patrol proceeded to the Numa Numa area to act as a coastwatcher party.
80. Ibid., 17.
81. Ibid., 4.
82 IMAC LofI to CO, 8th Brig Gp, 3d NZ Div, 803943; IMAC LofI to CG, 3d MarDiv, 27Sep43. The proximity of IMAC and 3d MarDiv Headquarters on Guadalcanal made planning efforts fairly easy.
83. IMAC LofI to 3d MarDiv, 27Sep43.
84. IMAC LofI to 8th Brig. For complete details concerning this operation see Chapter IV, Section 1, following.
85. Code names had been selected previously by COMSOPAC and appear in his 16-43. The order of 15 October assigned the code name "DIPPER" to the whole operation, "GOODTIME" to the Treasury Islands, "CHERRYBLOSSOM" to Empress Augusta Bay, and "HAIRBRUSH" to Puruata Island. By virtue of a change issued 22 October, the name "DIPPER" was applied to Bougainville only, while the name "BLISSFUL" was assigned to the Choiseul operation.
86. IMAC Change No. 2 to Operation Order No. 1, 22 October 1943.
87. IMAC LofI to 3d MarDiv, 27Sep43.
88. 3d MarDiv Opn Rpt, 1.
89. IMAC Tentative Shipping Assignment, 27Sep43.
90. 3d MarDiv Opn Rpt, 3. The word sector as used herein denotes the area commonly referred to as zone or zone of action in other campaigns. Operation orders and reports of the Bougainville campaign habitually use sector.
91. Admiral Wilkinson and General Vandergrift boarded the George Clymer.
92. 3d MarDiv Opn Rpt, 81.
93. 3d MarDiv Opn O 18-43, 6Oct43.
94. Combat Narratives, 39.
95. Transport Group, 3d Amphibious Force, Action Report, 1-4. Hereinafter cited as TG 3d AmphFor AR.
96. Report of Operations--OP GOODTIME, 6. Hereinafter cited as Row Rpt.
97. For discussion of these subsidiary operations, see Chapter IV.
98. IMAC Opn Rpt, I, 12.
99. To vector--to determine the relative positions of friendly and enemy aircraft and transmit this information to friendly planes.
100. TG 3d AmphFor AR, 5-6.
101. See Chapter IV ff. for discussion of operations in the Treasury Islands and on Choiseul.
Transcribed and formatted by Jerry Holden for the HyperWar Foundation