Subsidiary Operations

Treasury Islands Operation

IN HIS operation plan issued 12 October 1943, Admiral Halsey directed Admiral Wilkinson, in addition to other operations, to "seize and hold Treasury Islands on Dog minus 5 days, capture and destroy enemy forces, and establish thereat radars and minimum facilities for small craft as necessary."1 On 15 October Admiral Wilkinson designated Rear Admiral George H. Fort, USN, as Commander Southern Force, and placed him in over-all command of the Treasury Islands operation.2 IMAC had previously been assigned responsibility for land operations in conjunction with the Fleet against the Northern Solomons, and the 8th New Zealand Brigade Group of the 3d New Zealand Division had been attached to IMAC for the purpose of executing the Treasury Islands landings.

Admiral Wilkinson ordered motor torpedo boats (MTB's or PT's) based at Lambu Lambu, Vella Lavella and at Lever Harbor on northern New Georgia, to screen the approach of the convoy from its staging area to its target, by patrolling a picket line from the Shortland Islands to the island of Choiseul.3

In the meantime Admiral Halsey directed the task force under the command of Admiral Merrill to cover the operation against surface attack. He also directed shore-based aircraft to provide air cover and support, and neutralize Japanese airfields in the Bougainville area.4

The 8th New Zealand Brigade Group arrived at Guadalcanal from New Caledonia on 14 September 1943, having conducted, enroute, landing exercises at Efate, New Hebrides. These exercises were carried out from APA's (much larger ships than those actually used for the final operation),5 which provided valuable lessons eventually put to good use, for the units had never experienced combat action, to say nothing of a landing operation.6

Headquarters of the 3d New Zealand Division, parent organization of the Brigade, was already established at Guadalcanal when the Brigade arrived. When the Brigade commander, Brigadier R. A. Row, NZEF, reported to his commanding general, he discovered that his unit had been assigned the task of seizing and occupying the Treasury Islands on or about 1 November. Shortly after receiving this information, Brigadier Row was informed that his Brigade would come under the command of IMAC, and that the overall operation would fall under the direction of Commander, Task Force 31 (Admiral Wilkinson).7

On 28 September 1943, General Barrett, then commanding IMAC, informed Brigadier


Row of the general nature of the proposed operation, and made a tentative assignment of troops to be attached to the brigade. At this stage of planning it was proposed that the Corps should effect simultaneous landings on Treasury Islands and at Empress Augusta Bay, Bougainville, but, as has been seen, this plan was later altered by advancing the date for the Treasury Islands operation to 26 October, and ultimately to 27 October.8

Brigadier Row received General Barrett's letter of instruction concerning the operation on 29 September, and immediately began his detailed planning. At this stage, all intelligence regarding the Treasury Islands had not yet been collected, but sufficient was available, however, for the formulation of a plan in broad outline.9

Since only eight APD's. (destroyer transports) , two LST's, three LCT's, and eight LCI's were available for transportation of the first echelon of the Brigade (which numbered approximately 7,700 officers and men), it was determined that supplies and equipment would be cut to the barest minimum, and that only about 50 to 60 percent of the troops could be taken in the first echelon.10

At the time of the first meeting of General Barrett and Brigadier Row, intelligence concerning the Treasury Islands had been gleaned mainly from the report of the Navy and Marine patrol which had landed on Mono Island on 22 August 1943, and had reconnoitered until withdrawn 28 August.11 Further information was obtained from the reports of American aviators who had been shot down in that vicinity, and had made their way to the Treasury Islands, whence they later escaped. Latest information available at this time was dated 6 September, but aerial photographs were being taken daily, while IMAC worked hurriedly to prepare hasty terrain maps.12

Evaluation of intelligence indicated that the only beaches suitable for any type of landing craft were located on Mono Island between the Saveke River and Falamai Point. There also were beaches at Maisi and Soanotalu and one 1200 yards west of Toalomo Point, but none of these seemed capable of accommodating LST's. There appeared to be two adequate shallow-draft beaches on the northern shore of Stirling Island.13

While the topographical intelligence was fairly extensive, information concerning the enemy was meager. There were said to be about 135 Japanese, armed with rifles and light machine guns, quartered at Falamai. These were presumed to be maintaining observation posts at various points, mainly on the south side of Mono.14

To obtain up-to-the-minute information, therefore, Sergeant W. A. Cowan, of the Brigade Intelligence Section, Corporal Nash, of the Australian Naval Intelligence Service, and two native members of the British Solomon Islands Defense Force, landed on Mono Island from a PT boat during the night of 21-22 October, and were evacuated by the same means the following night. This patrol reported that the Japanese had recently landed reinforcements and that their total strength was now about 225 men; that the Japanese Headquarters area was situated just west of the Saveke River; that there was an observation post on Laifa Point; that guns--probably 40mm dual purpose guns--had been emplaced on both sides of the main beach at Falamai; and that there were no enemy on Stirling Island. When it withdrew, the patrol brought with it several friendly Treasury Island natives, to serve as guides for the landing force. Immediately utilizing information gathered by this patrol, adjustments were made in landing plans.15

During the night of 25-26 October, Sergeant Cowan and three enlisted men, accompanied by two natives, again landed on Mono Island from a PT boat, this time with the triple mission of


cutting lines of communication between the Japanese Headquarters and Laifa Point observation post, of establishing at 0400 on 27 October an observation post between Falamai and Malsi in order to observe enemy movements, and finally of organizing the natives as guides and make them available to combat units. Sergeant Cowan successfully carried out all assigned tasks.16

The Brigade received IMAC Operation Order No. 1 on 16 October and immediately began preparations for the operation. It was clear at the outset that the main assault should, if possible, be made in the vicinity of Falamai, because that was the only beach suitable for LST's, possessing sufficient dispersal areas for the large quantities of supplies and equipment that had to be carried. Provided the Japanese could be subdued by supporting fires during approach of landing craft, an assault at Falamai was also the best way of engaging the enemy quickly. It was, however, necessary to land also on Stirling because that island offered the only positions from which field artillery could effectively support infantry on Mono Island. Furthermore, areas on Stirling could be utilized as positions for antiaircraft guns covering Blanche Harbor. Although LST's could not beach on Stirling Island, it appeared that with a little engineering work, a suitable beach for those ships could be created. Subsequent events proved this estimate to be correct, for by the time the Brigade had been withdrawn from the Treasury Islands, two good landings had been built, while several less satisfactory ones were available.17

Initial plans did not provide for any landings other than those at Falamai and Stirling. However, the detachment of COMAIRNORSOLS, responsible for establishment of long range radar in the Treasury Islands, reported on 14 October that radar, to be of benefit to the Bougainville operation, would have to be established on the north or northeast side of Mono Island. At about the same time Brigade was informed by CTF-31 through IMAC that establishment of this radar was to be given a high priority in planning the Treasury operation.18 It was consequently necessary to plan a subsidiary landing to provide a covering force for the group which was to set up the radar. From reconnaissance reports and aerial photographs it appeared that a small beach at Soanotalu would be sufficient for beaching an LCT. The commander of Argus Unit No. 6 (the unit responsible for erection and operation of the radar) was of the opinion that the radar could be put ashore and sited effectively in that vicinity. CTF-31 arranged to divert an APD (the ill-fated McKean) to Soanotalu, and the plan was modified. One company of infantry, reinforced with one section of machine-guns, the radar personnel, and a detachment of Company A, 87th NCB would now land as a separate unit known as "Loganforce," under the command of Major G. W. Logan, NZEF. Radar and engineering equipment were to be brought to Mono on an LCT in the main convoy, and be sent to Soanotalu during the night of 27-28 October, if the situation there should be reported favorably.19

As finally determined, the order of battle for the operation included about 7,700 officers and men.20

Final allocation of shipping consisted of eight APD's; eight LCI's; two LST's; and three LCT's. Cargo was to be loaded on eight LCM's and two APc's. This number of ships required that the Brigade be transported in echelons. The first echelon was made up of approximately 3,700 troops and about 1700 tons of cargo.21 Nearly all embarked units were well below normal strength; infantry battalions, for example, were reduced to about 600 men each.22 There was room for but one battery of field artillery, one battery of anti-aircraft artillery, and a nucleus of the command groups. Each infantry battalion was allowed only two trucks and four jeeps, a paucity of transportation which later caused much delay and confusion during unloading phases of the operation. Succeeding


ADMIRAL HALSEY AND BRIGADIER ROW, NZEF, confer in a tent on Guadalcanal during the planning phases of the Treasury Islands operation.

echelons were scheduled to come forward to the Treasury Islands at intervals of five days.23

Final orders for the operation were issued by Brigade on 21 October 1943. These directed an assault landing on Beaches ORANGE 1 and 2, located between the Saveke River and Falamai Point on Mono Island, by the 29th and 36th Battalions, with the 29th Battalion landing on the right at Beach ORANGE 1. The LST's of the second wave of the attack force were to land on these beaches. The 34th Battalion was to land on Stirling Island on Beaches PURPLE 2 and PURPLE 3, with the Brigade Headquarters landing in the Second Wave on PURPLE 3. A simultaneous landing was to be effected on Beach EMERALD at Soanotalu by Loganforce, while field and antiaircraft artillery, loaded on LST's, was to be trans-shipped to Stirling Island as speedily as possible. Beach PURPLE 1, Stirling Island, would be developed later.24 The administrative order, issued in conjunction with the Brigade operation order, set up the beach organization, to include working parties, loading of bulldozers, and pooling of transportation in order to facilitate unloading operations.25

During the planning stage considerable difficulties were encountered because of the great distance and lack of suitable communication between the headquarters of IMAC and the Brigade. In addition, roads on Guadalcanal were most difficult, while distance caused weak and inefficient telephonic communication.26 There


was no centrally located joint staff or planning group, although the presence of a liaison officer (Lieutenant Colonel J. Brooke-White, NZEF) at Corps Headquarters helped the Brigade considerably. Furthermore, some supporting units were not assigned to Brigade until a very late date.27 In most cases these units were still in the process of organization. The high degree of cooperation demonstrated by attached units went far, however, toward surmounting most of these difficulties.28

From 14 to 17 October, the Brigade practiced final landings on Florida Island. Four APD's and eight LCI's were used. The value of this training was amply demonstrated a few days later when loading of supplies and embarkation of troops for the operation proceeded without incident.29

Transport units under Admiral Fort were divided into five groups, each with its own tactical commander. Having loaded, these departed independently, timed to arrive at Blanche Harbor on 27 October from 0520 (time of arrival for the first group), to 0830 (time for arrival of the fifth group).30

The several groups made independent passages without incident except that a flare was dropped near the LCI gunboats between Simbo and Treasury Islands,31 which area the groups passed before dawn as scheduled. The element of surprise was very important, since about 25,000 Japanese were reported in the Buin-Shortland area with 83 barges at their disposal.32 These forces could reinforce the Treasury Islands in a few hours. Complete radio silence was therefore maintained except for three orders by very high frequency voice radio, which included an order delaying H-hour by 20 minutes.33

Practically every unit venturing west or north of Vella Lavella at night previous to this operation had been detected by Japanese "snoopers" (reconnaissance float-planes). It was therefore considered almost certain that our approach would be discovered during the night; but despite the flare dropped near the LCI gunboats, it seems probable that the Japanese made no contact, and that surprise was complete. The covering cruiser group under Admiral Merrill, to the westward, was spotted by "snoopers", however, and many flares and float lights were dropped near that formation.34

At 054035 seven APD's of the 1st Transport Group lay to 1,300 yards off Cummings Point, Stirling Island, bearing 060° T., just south of the entrance to Blanche Harbor, and commenced debarkation of troops.36 Upon arrival of this group, the weather was overcast, heavy rain was falling, and visibility was poor. Sunrise was approaching, however, and there was a moderate northeast wind. Within the next two hours the weather cleared and remained fine for most of the day, but in the late afternoon, and again during the night, heavy rain fell.37

The Eaton, which had been designated as fighter director ship, proceeded to her station, while the Pringle and Philip began their scheduled shore bombardment within five minutes after the transport group arrived.38 The air spotter was already over the targets, and at 0600 a fighter cover of 32 planes arrived on station.39

Since it was impracticable to operate destroyers in the narrow waters of Blanche Harbor, the fire support area was located west of the entrance.


Map 23
Treasury Islands Assault
8th NZ Brig Grp
27 Oct 1943

All firing at this stage of the landing was on prearranged targets. Preparatory fire of the Pringle was delivered approximately according to schedule, with good battery performance, but no assistance, unfortunately, was obtained from the air spotter, who subsequently reported a radio failure.40 While it later developed that much of the Pringle's fire was too far back from the beach to be effective for the combined effort, she covered the remainder of the assigned area reasonably well, nevertheless. Preparatory fire of the Philip was disappointing in accuracy, timing, and quantity. At 0623, three minutes before assault waves hit the beach, the destroyers ceased fire. The Pringle maneuvered clear of the fire support area to patrol west of Blanche Harbor, while the Philip proceeded independently to patrol on east-west courses approximately 6,000 yards south of Stirling Island.41

In the meantime, loaded boats from the APD's rounded Cummings Point under arched tracers of 5-inch shells, sight of which coupled with the sound of explosions on the beach, instilled a feeling of confidence in personnel embarked in the small craft.42 Two newly-converted LCI gunboats left the 2d Transport Group during the night and joined the 1st Transport Group in time to escort landing boats of the APD's into the harbor.43 In order to reach Falamai it was necessary to proceed two miles up the harbor, which averaged only about 1,000 yards in width. No enemy resistance was expected on Stirling Island, and Japanese machine-gun positions were reported along the southern coast of Mono Island as well as on both sides of Falamai Peninsula. Assault waves therefore were routed close to Stirling to a point just beyond Watson Island, where boats bound for ORANGE beaches turned left.44 These waves received unexpected machine-gun fire from Cummings Point, Stirling Island.45 Before Watson Island was reached, craft from the Stringham and Talbot turned to starboard and proceeded independently to Beaches PURPLE 2 and 3. Despite air bombings the previous day, and bombardment by the fire support group, landing craft approaching Falamai were fired on by a number of machine-guns, some mortars, and a twin-mount 40mm automatic gun located at Falamai. Fortunately only one boat was disabled by gunfire,46 although several were perforated by bullets.47 Five Navy men and eight New Zealanders were wounded on the way in.48

At this point the newly converted LCI(L) gunboats, which had arrived from Noumea just in time to participate in this operation, saw their first action. Equipped with 3-inch guns, two of these craft accompanied the assault boats, one preceding and one on the flank. Just prior to the landing one of the gunboats rounded Watson Island and knocked out the 40mm twin-mount, thus saving many lives. The gunboats also returned the fire of several Japanese machine-guns.49 On one occasion, however, fire from one LCI(L) held up the advance of the 36th Battalion's left flank.50 Their performance during this operation, for which they had no time to train, was nevertheless described by Admiral Fort as "especially creditable".51 They continued to operate in the harbor throughout the morning.52

Brigadier Row likewise paid a compliment to the LCI gunboats when he said,

The close support rendered by the LCI gunboats undoubtedly kept down casualties during the assault. These boats protected the left flank of the first wave, and in addition shot at opportunity targets. One of them proceeded past Falamai peninsula and raked that area with fire just prior to the landing [of the] 29th Battalion.53

In the meantime, the McKean, which had left the formation at 0430 and proceeded independently to Beach EMERALD at Soanotalu on the north coast of Mono to land troops to secure the


radar site, returned to the transport area at 0708 and reported that the landing had been unopposed.54

The 2d Transport Group, completed its approach by 0630, and its LCI(L)'s, under Commander J. McDonald Smith, USN, rounded the west end of Stirling Island in two columns, preceded by the minesweepers, Adroit, Conflict, and Daring. These columns on their way up the harbor passed Higgins boats of the APD's returning from the beach. Shore fire was observed converging on gunboats accompanying the landing craft, all of which were vigorously returning the fire. The LCI(L)'s in the right column swung to starboard and grounded on the PURPLE beaches of Stirling Island, while those of the left column swung to port and grounded on the Mono Island's ORANGE beaches at about 0647. These LCI(L)'s were under rifle and mortar fire throughout the unloading period, but despite this opposition 1,600 troops and 150 tons of cargo were debarked in less than 35 minutes. By 0730 the LCI(L)'s had gotten underway in company with the AM's and were headed for Guadalcanal.55

About 0715 the LST's of the 3d Transport Group stood into Blanche Harbor preceded by two auxiliary minesweepers. Since the LST's could not sight an unloading party on shore, the ships' commanding officers each selected his own beach, and beached at 0735, according to schedule. Five minutes later both ships were subjected to heavy fire from mountain guns, mortars, and small arms, which caused some damage and wounded eight men.56

LST 399 found herself in especially difficult straits. She was bracketed by a score of shells and received two direct hits from a mortar. Furthermore a large, well-covered, active pillbox, previously considered knocked out,57 was located only eight yards from her port bow door, and Japanese riflemen were active. During a lull in the firing, the ramp was lowered, but the first man off the ship was shot down, and the two who followed him dropped in their tracks mortally wounded. It became necessary to close the ramp for protection. Several New Zealanders standing on shore beyond the arc of enemy fire, ineffectually fired their Bren guns at the pillbox. None of the LST's forward automatic weapons could be brought to bear. Personnel aboard ship opened fire on the enemy but at that point they received orders from the beach to cease firing. It was apparent that the men on the beach were not going to silence the pillbox, which had become so active that all hands aboard ship had to take cover, and delay unloading until the pillbox was reduced.58

At 0815 the LST requested permission to leave the beach, but received the reply, "Not granted". Whereupon, a resourceful New Zealander mounted a D-8 bulldozer, and with the blade raised high to protect him from fire, he rolled heavily down the ramp, several New Zealanders covering him with Bren guns as he came out. Working to the blind side of the pillbox, he lowered the blade, and plowed the pillbox and its occupants under the earth, "tamping it down well all around, effectively silencing its fire".59

During the initial shelling of the LST's, the Eaton, with Admiral Fort aboard, entered the harbor in an attempt to silence the mortar fire. Unfortunately, the destroyers could not locate the battery, which ceased to fire when the Eaton approached. The Philip, on station south of Stirling Island, was assigned a target by the shore fire control party, and fired five salvos. After this bombardment there was no more mortar fire for almost three hours.60

Both LST's began unloading operations at approximately 0830, having been on the beach for almost an hour, and under fire for a portion of that time.61 Although Brigade had provided for working parties and had organized a beach


party,62 a lack of proper understanding of respective functions was apparent.63 In addition, LST 485, for lack of a bulldozer, road, or unloading area, had to resort to unloading by hand.64 Also, not to be discounted, was the fact that the initial shelling of the ships by the Japanese had demobilized the unloading details. These factors, coupled with fascinating aerial dog-fights going on overhead during unloading operations, caused discharge of cargo to proceed inexcusably slowly. LST 399 was especially adversely affected. Here the unloading progressed rather well at first, but after the mobile equipment was ashore unloading came to "an almost abrupt halt",65 for thereafter LST 399 never had more than three trucks and few jeeps.66

At 1120 Japanese mortar and mountain gun fire was reopened from new positions on the high ground west of the Saveke, about 500 yards off the bow of LST 399, which was bracketed and then hit on the port side. Shells continued whizzing over the forecastle and striking the beach. At 1123 another shell struck the ship on the port side, this time in the capstan control room, wrecking the capstan machinery.67 At approximately 1149, on about the 20th salvo, the mortar registered a direct hit on a large ammunition dump at Falamai. A violent explosion knocked men off their feet on Beach ORANGE 2, fired the native village, and set off some nearby small arms dumps.68 Burning debris, fragments from exploding 90mm shells, and exploding pyrotechnics covered the forward part of the ship, so that the whole forecastle seemed to be on fire. Heat grew so intense that the forward guns had to be abandoned, and when it was noticed that the heat was blistering paint on the starboard bow, the forward magazine was flooded. Hoses were manned, and several clusters of fire on deck were extinguished.69 The Commanding Officer of LST Group 15, Commander Vilhelm K. Busck, USN, then signalled LST 399 to retract and rebeach 50 yards to the west, which she promptly did.70 On the beach Japanese fire destroyed one 90mm antiaircraft gun, a Bofors gun, some ammunition, medical supplies, and other equipment. One 25-pounder gun was badly damaged.71

At 1155 the Philip and the two LCI gunboats entered the channel in an attempt to silence the mortar fire, but by the time they had established contact with the shore fire control party, the mortar had ceased firing. Mortar fire was never resumed.72

After getting unloading started again on LST 485, Busck proceeded to LST 399 where he found unloading details demoralized by the previous shelling. With the aid of a New Zealand officer he finally got some work under way even though slowly. Upon rebeaching LST 399 found that her forward electrical circuits had shorted out, and that the ramp had to be lowered by hand. At 1515 Commander Busck was again called back to the ship by an appeal for him to use his authority to speed work of discharging cargo.73 A few minutes after he arrived bogies were reported; a Zeke came over pursued by two P-38's, and DD's off shore opened a gun-plane duel. For a time it was again impossible to control the unloading parties.74

By 1600 the Japanese planes were driven off, and unloading continued until after dark. At 1841, LST 485, having completely unloaded and retracted, lay to off the stern of LST 399 to protect her sister ship. LST 399 reported at 1902 that unloading was at a standstill, although she still had about 20 tons of cargo aboard,75


and was under orders to unload completely. Finding it impossible to resume unloading, she retracted at 1943 and both LST's stood out into the channel, led by the Philip and accompanied by the two YMS's. When the group was about 1,000 yards from the beach it received reports of bogies and heard planes overhead. Soon thereafter the beach from which LST 399 had just retracted went up in a mass of flames, eight or ten bombs having fallen in that vicinity.76

The 5th Transport Group, consisting of one APC, six LCM's, and an aircraft rescue boat, arrived off the west entrance to the harbor about 1830; while the 4th Transport Group, composed of one APc, three LCT's, and a screen of two PT's, arrived about 1850. Both groups were ordered to report to the Commander of the Naval Base, Treasury Islands, for beaching and unloading assignments.77

The major assault was made on the ORANGE Beaches near Falamai, on the southern coast of Mono Island,78 with the 29th Battalion on the right and the 36th Battalion on the left, just as had been planned. Assault waves hit the beach exactly at H-hour (0626) amid mortar and small arms fire.79 This wave consisted of 16 LCP(R)'s from the APD's, carrying a total of 640 men to the ORANGE Beaches. The APD's remained in the transport area for about two hours, during which time they disembarked approximately 1,600 men and at least 80 tons of supplies and equipment.80 At 0800 the APD's departed under escort of the Conway and Renshaw.81

All battalions quickly cleared their respective beaches, although, as previously noted, the left flank of the 36th Battalion was held up for a short time by fire from one of the LCI gunboats; otherwise the advance continued according to plan.82

Opposition was not strong at this stage. There were pillboxes which had to be reduced, and a fair amount of rifle and machine-gun fire was being received from the left flank; but generally the Japanese had withdrawn from their beach positions, probably as a result of the shore bombardment.83

The 24th Battalion landed at H-hour on the PURPLE Beaches of Stirling Island with no opposition. Although the beaches were bad, no difficulty was encountered. However, it had been observed during the approach of the landing boats that a machine-gun on Cummings Point was contesting the landing. Quickly securing its beaches, the 34th Battalion advanced to its beachhead line, meanwhile sending out a patrol to deal with the machine-gun post on Cummings Point. This unit did not contact the Japanese, although it did discover their positions. It was later presumed that the enemy withdrew when they saw the New Zealanders approaching, hid until nightfall, and then crossed over to Mono Island.84 Further extensive patrols were carried out during the day by the 34th Battalion, and an observation post was established on Wilson Point, but no Japanese were observed.85 One group was sent to Watson Island, and when it reported that island clear, two sections of 3-inch mortars were sent there to support operations of the 29th and 36th Battalions at Falamai. Late that afternoon, a platoon of the 34th Battalion was dispatched to Mono Island and attached to the 36th Battalion to plug a gap in the lines of the latter. By the end of D-day, artillery was supporting troops on Mono island and furnishing defensive antiaircraft fire from positions on Stirling Island.86

As the McKean had reported, our landing at Soanotalu was not opposed. Loganforce established a perimeter covering EMERALD Beach, and immediately instituted a thorough reconnaissance of the area to determine the eventual location of the radar, the mission of this force. The beach was suitable for the landing of an LCT, and arrangements were therefore made to


send in the radar and engineering equipment from Blanche Harbor during the night.87

We return now to the scene at Falamai after the initial landing.

At approximately 0735, when the LST's beached, the Japanese had begun laying down mortar and mountain gun fire which damaged the LST's and disrupted unloading operations. Shortly after the Japanese fire started, Company A, 36th Battalion, launched an attack on the Japanese Headquarters area west of the Saveke River, supported by 3-inch mortars, and fire from the Philip, on station south of Stirling Island. As the attack developed, mortar fire ceased temporarily, and it was assumed that the Japanese had moved their positions to higher ground northwest of the Saveke River. A heavy machine-gun post sited on the beach was destroyed, and a considerable quantity of rations and equipment was captured as a result of this action. Survivors were observed retiring to higher ground northwest of the Saveke.88

Japanese mortar and mountain gun fire on the beach was resumed at approximately 0900 and was again accurate. Although it was extremely difficult to ascertain the origin of this fire, it appeared likely that it was coming from the area to which the enemy had retired, i. e. high ground northwest of the Saveke. Accordingly, at 1000, two platoons of the 36th Battalion were ordered to proceed to that locality to search for suspected gun positions, find them, and destroy them. At about 1100, one of these platoons, under Second Lieutenant L. T. G. Booth, NZEF, discovered a spot from which two Japanese 75mm mountain guns were firing. Mounting a determined attack, the platoon carried the position, and captured the guns; the enemy gunners fled. Leaving a detail to guard the area, Booth led the remainder of his platoon to still higher ground, and some 500 yards farther on he heard the sound of a mortar being fired. By following the sound of the firing, Booth located the mortar position and led his men in another assault, and at 1210 captured this position as well. A 90mm mortar and a considerable quantity of ammunition were captured as a result of this second attack. Nine Japanese were killed, while the New Zealanders lost seven wounded. The other platoon, which had advanced from the beachhead perimeter by a different route, did not contact the enemy, but observed the second attack of Booth's platoon, being too far away at the time, however, to render support.89

There was no material enemy resistance during the afternoon, and by 1800 both the 29th and 36th Battalions had secured their perimeter and dug in. The platoon from the 34th Battalion arrived shortly thereafter and was placed in the lines of the 36th Battalion, while mortars of the 34th Battalion were in general support of the other two Battalions from positions on Watson Island. Artillery registered at 1730, and a field hospital was set up on Stirling Island. As many casualties as possible were evacuated on the LST's, while the Brigade settled down for the night.90

On the whole it had been a successful day. The Brigade attained its first objective, yet suffered comparatively few casualties in view of the fact that the landing was opposed. Twenty-one New Zealanders and nine Americans were killed, while 70 New Zealanders and 15 Americans were wounded. No estimate of Japanese casualties could be made.91

Fighter cover throughout the day had been excellent, and the troops were not attacked from the air, although the ships were not quite so fortunate.92 Since threat of counter-landings from the Shortland Islands, only 18 miles away, had concerned Brigadier Row from the beginning, it was deemed to be essential to hold the Blanche Harbor area as strongly as possible.93 Accordingly, he placed his 6-pounder anti-tank guns in position to cover entrances to the harbor, and assigned all other guns a secondary mission of engaging any Japanese surface craft that


should appear. An additional sense of security was felt because of the presence of Lieutenant Commander R. B. Kelly's MTB94 Squadron 9, which patrolled west and northeast of Mono Island, prepared to deal with any Japanese attempt to come across from the Shortland Islands. Furthermore, this squadron could give ample warning of any such attempt.95

During the night of 27-28 October, the enemy were quite active on the perimeter's left flank and to a lesser extent on the right flank. Beach areas were taken under fire by knee mortars,96 rifles and machine-guns. However, no determined attack developed at any point; the enemy seemed rather to be endeavoring to ascertain the Allied dispositions. They had, of course, lost most of their supply of rations when they lost their Headquarters area,97 although the 36th Battalion had not had time during the day to remove them. Since this area was outside the defensive lines, the Battalion placed a number of booby-traps and covered the area with plotted mortar concentrations. When the Japanese attempted to re-enter their old Headquarters area, they suffered casualties.98

As this was the first experience that the troops had had in night fighting, there was a considerable amount of indiscriminate shooting. Our men had not as yet become accustomed to jungle sounds and, of course, could not distinguish these sounds from those made by men.99 This shooting, plus several air raids during the course of the night, resulted in two men killed and nine wounded.100

On the following day, the 29th and 36th Battalions made extensive searches forward of the lines, but none of these contacted Japanese. Patrols did find, however, evidence to indicate that enemy had recently been in the vicinity of high ground up the Kohele River and also high ground northwest of the Saveke.101 Since it was felt that the Japanese were capable of using these areas from which to infiltrate New Zealand positions, artillery and mortar concentrations were arranged to cover those localities and likely approaches nearer the perimeter. These concentrations were fired from time to time during the night. During the afternoon, Company A, 29th Battalion, was sent on a combat patrol to Malsi and its place in line was taken by a company of the 34th Battalion, which had come over to Mono from Stirling.102

Heavy rain fell during the day; the ground became muddy. That night small parties of Japanese managed to infiltrate the perimeter at a point near the boundary of the 29th and 36th Battalions, working across to the New Zealand water point on the Saveke, probably with the intention of gaining access to the old Japanese Headquarters area.103 They did not succeed in this, however, for they finally were driven out by firing that was going on behind the lines. In addition to this activity, the Japanese launched one low level strafing attack and several bombing raids, in which 30 to 40 bombs were dropped, but no damage or casualties resulted. One Japanese aircraft was shot down by the 208th Light AA Battery.104

Although the troops were extremely worried by snipers behind their lines on 29 October, it appears that there were fewer of these than the amount of excitement they caused would indicate. However, the 36th Battalion definitely accounted for several. Throughout the day patrols again operated forward of the perimeter, but again with negative results.105

Brigadier Row made a personal reconnaissance at Soanotalu on 29 October, and as a result decided to reinforce that area with a company of the 34th Battalion. Accordingly, during the night of 29-30 October, a Company Headquarters and two platoons were sent to


Loganforce via boat, where the radar was almost ready for operation.106

There was little Japanese activity in the Falamai area during the night of 29-30 October, and once again it occurred principlly on the left in the vicinity of the old Japanese Headquarters area.107 Although the New Zealanders claimed to have inflicted casualties on the enemy, no bodies were found the following day; it was therefore presumed that the Japanese had removed their dead and wounded. Air activity was less than it had been the previous evening, only three bombs being dropped; these fell in the 29th Battalion area without effect.108

On 30 October there was little or no activity near Falamai. Opportunity was therefore taken to adjust the perimeter by shortening the front of the 36th Battalion and drawing it nearer to the Saveke River. This was a good natural line of defense, and the alteration was a considerable improvement to the left flank, which had been active the past three nights.109

Company A, 29th Battalion, reached Malsi at 1100, having made no contact with the enemy during the preceding two days. During the day a 36th Battalion patrol, proceeding towards Laifa Point, found a Japanese 37mm antitank gun and 500 rounds of ammunition in a beach position west of the Japanese Headquarters area.110 This weapon was in perfect order, and apparently had been abandoned hurriedly upon the patrol's approach. That night a single aircraft bombed and strafed the Falamai area three times, without causing damage or casualties; but there were also some other enemy planes in the vicinity for a lengthy period, a factor which prevented artillery from firing its prepared concentrations, for fear of betraying their positions.111

It was now 31 October, and the situation appeared to be well in hand. The perimeter at Falamai was secure, Malsi had been occupied, and at Soanotalu good defense positions covering both radar and beach had been prepared. Brigade then ordered a series of patrols in strength determined to seize high ground at the head of the Saveke and Kohele Rivers, and to sweep every piece of country into which the Japanese may have gone.112

The experience of Loganforce was in sharp contrast to that of the troops at Falamai. The landing at Beach EMERALD was unopposed and only minor contacts were made with the enemy during the first few days; but, when these occurred, they were with a substantial opposing force.113

At day-break, 28 October, the LCT carrying the radar and engineering equipment arrived safely at Beach EMERALD. An immediate start was made on construction of a roadway up the steep incline between beach and radar site. The LCT was unloaded and departed without incident. Apparently the Japanese had never discovered the exact nature of New Zealand activities at Soanotalu, particularly the fact that a radar was being established there. There followed several days of hard work, and by 31 October the radar was fully operating with an adequate radius.

Meanwhile, the infantry company set up a perimeter defense covering the beach and was now sending out patrols to the east and west. Early on 30 October the 34th Battalion reinforcements arrived, whereupon Major Logan established a separate perimeter covering the radar site. On 28 and 29 October, there had been several minor contacts with small enemy parties toward the west. Mortar and artillery fire effectively prevented infiltration at night.114

During the night of 1-2 November, a Japanese force estimated to number between 80 and 90 men, made a determined effort to break through New Zealand positions and seize landing craft on the beach. A captured diary indicated that this force had retired from the Falamai area to high ground northwest of the Saveke, thence to the Soanotalu River Valley, on 28 October. The enemy had carried with them all their mortars, knee mortars, grenades, and


ammunition. While traveling down the Soanotalu River Valley, they found the Soanotalu beaches occupied by our forces. Stopping, they then sent out patrols in an effort to determine Allied strength and dispositions. This resulted in the previously mentioned minor clashes.115

When Major Logan became aware that the Japanese knew of his presence at Soanotalu, he had already been apprised of the experiences of the Falamai forces during the first few nights. Accordingly he issued instructions that his men were to fire only in cases of emergency. New Zealand 3-inch mortars were registered close in to the perimeter on likely avenues of approach. Grenades were issued. New Zealanders were alerted for an attack. This was the situation in which Loganforce found itself on the night of 1-2 November, when the enemy attacked.116

The Japanese assaulted our perimeter in two groups, one numbering about 60, the other numbering about 30. New Zealanders tossed grenade after grenade at the enemy in an effort to prevent a breakthrough, but part of the first group succeeded in penetrating the lines and reaching the beach. On the beach, however, the Japanese were engaged by a New Zealand detachment under Captain D. J. Kirk, and were wiped out before they could do any damage. The remainder were unable to penetrate the Allied lines, but many died in the attempt. In this encounter, which lasted for some five hours, the Japanese lost a minimum of 50 killed while our losses were one killed and nine wounded.117

On the following night (2-3 November), the Japanese again attacked the perimeter, although not with the same intensity that they had attacked the preceding night.118 Once again they were beaten off by the use of grenades, and once again they suffered severe casualties. This was the last organized attack on Soanotalu. Succeeding days were spent by New Zealand patrols in mopping up the remnants of the Japanese forces.119

By 12 November, when civil administration was reestablished, the Treasury Islands were fully secured, and as we have seen, on 1 November, the day of the Cape Torokina landing, the flank of the forces advancing toward Bougainville was secured by the radar on the Treasury Islands, and the PT's based there.

The mission upon which the Brigade had set out was accomplished.120

Choiseul Diversion

The 2d Parachute Battalion raised a great deal of smoke with but very little fire by the Choiseul diversionary raid which was mentioned in Chapter I. The primary mission of this operation, it will be recalled, was to create a screen behind which IMAC could execute its principal assignment--an assault on Bougainville.

A Choiseul landing had been suggested in the early stages of planning of the Northern Solomons campaign as a main assault--an intermediate step before further landings on the east coast of Bougainville.121 Subsequent patrol reports on the area indicated few suitable sites for airfields and motor torpedo bases.122 Furthermore the tremendous attrition to Japanese air forces had become increasingly apparent, and Japanese shipping therefore came to be considered more vulnerable than ground forces. Finally, in September, after Admiral Halsey learned of General MacArthur's objections to a landing on Choiseul, he ordered that planning


Map 24
Northwest Choiseul

be directed toward a landing on Bougainville proper without going through any intermediate steps.

The concept of a landing on Choiseul was therefore abandoned in principle, but certain elements of the scheme were retained at the suggestion of Major James C. Murray (then Staff Secretary, IMAC), who maintained that it was necessary in order to throw the Japanese off the scent of our intended landing on Bougainville. Since U.S. intentions in that general direction were obvious, it became essential to conceal the particular point of attack as long as possible, thus preventing the Japanese from taking anticipatory countermeasures.

The original plan for a landing in force on Choiseul was therefore modified to that of a landing in simulated force by a small but energetic unit, which would conduct raids to mislead the Japanese as to our real intentions. Further, if necessary, the tiny beachhead on Choiseul could be exploited, and patrol torpedo boat bases and fighter fields could be established there. This landing, it was hoped, would lead the enemy to believe either that it was our intention to attack the Choiseul area only, or to attack the east coast of Bougainville. In light of all these considerations, the proposed plan was submitted to Admiral Wilkinson who approved the idea. Accordingly, IMAC issued Operation Order No. 2 on 22 October 1943, providing for a landing on Choiseul by the 2d Parachute Battalion (Reinforced) on 27 October 1943.123


Two amphibious patrols had been sent to Choiseul while that island was being considered as a main objective for our attack. One group had scouted the central portions of the southwestern and northeastern coasts, while the other had scouted the north, northwestern and northeastern coasts of the northern portions. In addition to information which these units had gathered, coastwatchers on Choiseul were able to provide a great deal of valuable information and advice.124

The commanding officer, 2d Parachute Battalion, 1st Marine Parachute Regiment, Lieutenant Colonel Victor H. Krulak, was summoned from Vella Lavella to IMAC Headquarters on Guadalcanal several days before the operation was to take place.125 There he conferred with General Vandegrift, Admiral Wilkinson, Colonels Wade, Snedeker, Linscott, and Coleman.

For a period of several weeks prior to this meeting, a Choiseul-based coastwatcher, Sub-Lieutenant C. W. Seton, of the Royal Australian Navy,126 had been gathering intelligence information and forwarding it to IMAC headquarters.

As a result of the conference, it was decided that the actual landing should be made on beaches in the vicinity of Voza Village, on the northern portion of the island's southwest coast. Seton had reported a good beach there; natives were friendly and could be organized to help invading forces, and nearest Japanese positions were lightly held. Furthermore, approaches to the beach were easily accessible, and in the vicinity were inlets that might be suitable for establishment of a patrol torpedo boat base.

Upon taking leave of General Vandegrift, Krulak was told, "I desire an immediate and credible appearance of a large force. Take immediate action. Get ashore where there are no Japs."127

On his return to Vella Lavella, Krulak summoned his company commanders and staff, showed them the orders, and gave detailed instructions for the operation. For the next four days officers and men labored to assemble equipment, make final plans, and digest all available information. Study of the area was facilitated by the arrival of Seton, who had been taken out of Choiseul by PBY on 24 October. To facilitate loading, arms, ammunition, and supplies were placed on the beach in four individual stacks. Eight LCM's were borrowed from the Vella Lavella boat pool, and arrangements were made to have these proceed to the battalion's beach in separate pairs.128 Since Japanese planes were active daily at this time, embarkation was set for dusk.

Late in the afternoon of 27 October, Marines of the Parachute Battalion loaded one-half of their equipment and supplies into their borrowed LCM's and at 1080, when the APD's Kilty, Ward, Crosby, and McKean129 lay to off the mouth of the Juno River, the LCM's were promptly brought alongside. As a result of detailed planning, and by dint of hard work on the part of all hands, the 725 men of the battalion, with all equipment and supplies, embarked in 45 minutes.130

With the destroyer Conway acting as escort, the convoy sailed at 1921, and proceeded in


column through dark night and over calm sea.131 At 2310 a single enemy plane dropped a bomb near the rearmost APD in the column. The force of the explosion caused the ship to shudder and shake. A young Marine standing near the ship's rail, thinking as did many other Marines aboard that the ship's guns were firing, asked, "What are they shooting at?" "My boy," answered one of the battalion's officers, "you have just been bombed."132

At a point about 2,000 yards off the southwest coast of Choiseul, near the little village of Voza which lay behind an islet known as Zinoa, the convoy hove to, while the Conway took up a patrol station about 3,000 yards to seaward. In accordance with the battalion operation plan, a boat with a reconnaissance party under First Lieutenant Rea E. Duncan went ashore at 2352, followed at 0019 by the first wave of the landing force (Company F).133

A night fighter had been assigned to the force, but remained at so high an altitude that it was unable to intercept either the first air attack or the Japanese twin-float plane (a Jake) which crossed over the Conway at an altitude of about 200 feet, dropping two bombs which exploded off the port quarter at about 0145. The enemy plane did not renew its attack, and the destroyer did not take it under fire because, as was duly reported--

(1) it was thought at first to be our night fighter until bombs were dropped and it was very close aboard, (2) the proximity of TransDiv 22 which was dead in the water and evidently undiscovered and (3) another close-in attack was not made on this vessel.134

At 0120 the battalion, with the bulk of its supplies and equipment, was ashore. By 0209 the convoy was ready to retire, leaving behind four LCP(R)'s with Navy crews for the Battalion's use.135

With the aid of some 80 native bearers under Seton's direction,136 the Marines moved off the beach into the jungle to the rear thereof. At 0600 a base of operations was established on a high jungle plateau about a mile to the northwest of Voza, and outposts were set up on the beach north and south of the village.137

During initial establishment of the beachhead, enemy reconnaissance planes discovered the landing and bombed the area without effect. This reconnaissance was repeated after daylight--apparently without success due to a combination of factors. In the first place, troops and supplies were well dispersed and camouflage discipline was exercised; second, a dummy beach was set up quite some distance away from the actual location of our force; third, after all supplies and equipment were ashore and the transports departed, natives carefully obliterated every sign of a landing by brushing out the very footprints in the sand.138

During the night of 28 October, Japanese planes strafed the Marine boat-hideout at Zinoa Island on two occasions, but were unsuccessful in causing damage.139

The natives immediately confirmed IMAC intelligence reports that the nearest enemy installations were a barge staging and replenishing base at Sangigai, about eight miles to the south, and a defensive outpost for the Choiseul Bay area at the Warrior River, about 12 miles to the


north. Nevertheless, confirmatory reconnaissance patrols were dispatched to the north and south at 0630 on the morning of 29 October and again the following day. Reports of these patrols verified the information on hand. In the meantime, however, Seton's scouts reported that the Japanese to the north were withdrawing behind the Warrior River.140

In order to locate the PT boat base-site as ordered, a Navy officer (Lt. (jg) Keresey) who had been assigned to the battalion to make tentative selection, accompanied an overland patrol to Moli Point. No suitable area was located.141

On the same day a reconnaissance patrol moved southward with the dual intention of further apprising the Japanese of our presence on Choiseul, and of gaining information for use in planning an attack on Sangigai. This patrol, led by Lieutenant Colonel Krulak, consisted of Major Tolson A. Smoak (Operations Officer of the battalion), 17 enlisted men, and several native guides. Just short of Vagara River, Krulak directed four Marines, led by Platoon Sergeant Frank J. Muller, and four natives, to move inland, take up positions in the hills behind Sangigai, and there make sketches of the area. He further directed Muller to calculate time and space factors involved in a move over the reconnoitered route. The remainder of the patrol continued onward along the beach. As the latter group approached the mouth of the Vagara River, the party discovered a detachment of about ten Japanese unloading a landing barge. This appeared to be an excellent opportunity of apprising the enemy of the presence of Marines on the island. Accordingly, the Marines, unobserved by the enemy, took up firing positions and selected targets. At Krulak's signal, each man opened fire. In less than a minute no enemy were left on the scene, and the barge was sinking. Before the Marines returned to their base camp, they counted seven Japanese on the beach.142

In the meantime, Sergeant Muller's detachment, reconnoitering Sangigai, had worked its way into position. Complete and accurate field sketches were made of the area, and a detailed calculation of time and space factors were submitted upon return to the base.143

At dusk the same day, a squad of Marines under Sergeant Norman F. Law, Company F, proceeded southward to locate a position in the vicinity of Vagara River from which to signal ashore the LCP(R)'s bearing the force which was to strike Sangigai the next day. This patrol was also to serve as an advance guard, and to set up an ambush for any Japanese patrols which might have been dispatched as a result of the action at Vagara River on the preceding day. When Law's group had cleared our lines by some 2000 yards, it ran head-on into just such a patrol proceeding toward the Marine positions. Firing broke out immediately and the Japanese retired. When the fight was over, Law attempted to locate his men, but could find only two. These three then returned to our lines and reported to Krulak.144

An attack on Sangigai had been under consideration as early as 22 October when first plans for the Choiseul diversion were being made. Krulak laid final plans for this attack following his patrol of 29 October, and requested an airstrike for dawn the next day on located enemy positions about one-quarter mile northwest of Sangigai. At 0610, 30 October, 12 TBF's and 26 fighters dropped about two tons of bombs.145

The primary purpose of the Marine attack on Sangigai was to make the Japanese believe that Americans were present on the island in force, while the secondary mission was to destroy the local base facilities and disrupt the regular barge route built up along the coast of Choiseul by the enemy, and which depended in considerable measure on Sangigai as a replenishing point.146

Although it was estimated that 150 Japanese


Map 25
The Sangigai Raid
30 Oct 1943

were based at Sangigai, and the coastwatchers had warned that the enemy could--and probably would--reinforce the area shortly from the southeast, Krulak decided to launch the attack nevertheless, for he realized that he had "to make credible appearance of a large force." He therefore issued the attack order.

About 0400, the battalion (less Company G and small headquarters detachments) left camp. As it entered Voza contact was made with the remainder of Sergeant Law's squad--which had spent the night in Voza village outpost. Since this group had suffered no casualties, the men rejoined the battalion. Moving to the beach, the force waited for landing boats to move it to the vicinity of the Vagara River. Before the craft could leave the boat pool on Zinoa Island and join the Marines, several Allied planes strafed them, damaging one boat. One of the Marines, however, had an American flag which he promptly displayed. Upon recognizing their error, the fliers "buzzed" our force and executed a "thumbs-up". Fortunately no Marines were hurt in this encounter.147

Due to disablement of the landing boat, however, it became necessary to abandon the plan to move afloat to the Vagara River. An order of march was therefore made on the spot. Company F (Captain Spencer H. Pratt), with a section of regimental machine-guns and a section of an Experimental Rocket Platoon148 attached, led the advance while Company E (Captain Robert R. Manchester), and remaining attached units followed. The entire force, commanded by Lieutenant Colonel Krulak, was guided by Seton and two native scouts.149 At about 1100, Company F hit Japanese outposts in the vicinity of the Vagara River. After a short fire-fight, the enemy withdrew into their main base at Sangigai.

At this point Krulak put into effect his attack plan. In essence this was a small-scale double envelopment. Company E would proceed down the coast and launch an attack on the village from the north, while the remainder of the force would circle inland around Sangigai and attack from the east.150 As ordered, Company F moved to the left front (inland) and headed for high ground immediately behind the village, while Company E continued down the coast. Movement was difficult for the enveloping column. Its route, passing through the mountainous interior of the island, encountered numerous fast flowing streams and heavy rain forest. Weighed down by heavy rocket ammunition and impeded by trailing lines and slippery footing, the Marines inched their way forward. H-hour (1400) was rapidly approaching, and the column was still a considerable distance from its designated assault position. At the appointed time there was no sound of firing from the beach column. This disturbed Krulak, who therefore moved up to the point; there he found Pratt, evidently also worried. Shortly thereafter, however, the sound of Company E's rockets reached the group, and within a few moments the two native guides who were with the point signaled the presence of enemy ahead, although Company F was almost an hour away from its objective.

Company E had launched an attack as planned, initially firing its rockets and mortars into Japanese positions on high ground about 500 yards northwest of Sangigai. A few minutes later the company entered the enemy base without opposition and found that the Japanese had hastily withdrawn toward their prepared defenses in the mountains to the northeast of the base, either in an attempt to flank the beach column and attack its open (inland) flank, or to occupy previously prepared positions sited to defend Sangigai's northern and eastern approaches. In either event it later developed that this movement brought the enemy face to face with our enveloping force, and it was at this moment that the natives sighted enemy.151

At about 1430, therefore, a vigorous fire-fight broke out and continued almost an hour. The


enemy had reached the vicinity of their positions, and immediately began to occupy them. Some Japanese riflemen climbed trees; others fired from rifle pits and bunkers built among the roots of banyan trees. As soon as the extent of the enemy position was determined Krulak ordered Pratt to attempt an envelopment of the enemy left. For this purpose Pratt dispatched the 2d Platoon (Lt. W. E. King). Its movement was supported by the attached regimental machine-gun platoon which had worked two guns forward, and placed them in immediate operation.152 The enveloping attack met with some success, and the Japs were forced either to abandon their forward positions or be trapped in them. During this phase of the fighting both Colonel Krulak and Captain Pratt were wounded. After having been driven back about 150 yards, the Japanese undertook two "Banzai" rushes, which, though they penetrated the Marine positions, resulted in great loss. Immediately thereafter the 3d Platoon Co F (Lt. W. F. Naylor) was directed to move around the enemy right flank and prevent any further withdrawal. This was generally successful, although about 40 survivors were observed running southward in most non-Samurai fashion.153 A final count showed 72 enemy bodies after the fight.

In the meantime, Company E had occupied Sangigai. Utilizing demolitions, Manchester's men completely destroyed the base, with all supplies stored there, sank a new barge, and captured a number of important Japanese documents in the local headquarters, among them highly important charts of the minefield and barge routes in the Bougainville-Choiseul area.154

Upon completing Sangigai's destruction, Company E retired to the Vagara River where it boarded our landing boats (now again all operational) and returned to Voza.

After the enemy had broken off the fighting north of Sangigai late that afternoon, Company F buried its dead, and excess rocket ammunition, and retired along its route of approach to the Vagara River mouth, where Krulak deduced the homeward movement of Company E.

Krulak then decided to remain where he was for the night, because it was already dark. The Vagara area was defensible, and the troops, having moved some 12 miles through the jungle and fought a sharp battle, were fatigued. Furthermore, with him were 12 men wounded in the fighting--two seriously--who needed immediate attention. Therefore the troops dug in, while base camp was contacted by radio and arrangements made for the boats to return the next morning to pick up the remainder of the Sangigai force.155 The night passed without event. Shortly after dawn the boats arrived as requested and by 0730 31 October the entire battalion was again reassembled at the Voza base. The six dead Marines who had been buried west of Sangigai were later exhumed by the natives and reburied on the beach near Vagara.

At 1500, 1 November, a Dumbo plane landed near the beachhead at Voza and evacuated the wounded men. The pilot of this plane was made an officer-messenger and the charts, captured at Sangigai, were entrusted to him for delivery to Admiral Halsey.156

Immediately following the return of the Sangigai force on 31 October, patrols were dispatched and ambushes erected to thwart the almost certain Japanese counteraction. Meanwhile a reconnaissance patrol from Company G, led by Major Warner T. Bigger (Battalion Executive Officer), moved via boat as far north as Nukiki, but encountered no enemy in that direction. Natives reported, however, that not only were there installations to the north, but also that immediately after the fight to the south--at Sangigai--the Japanese had begun to re-occupy the village. This latter report was apparently confirmed by virtue of several brisk patrol actions fought on


1 November and the fact that Marine ambushes were successful in trapping and killing a number of enemy, particularly in the vicinity of Vagara.157

Since rice for the natives was running low, and the supply of hand grenades and demolitions was below the minimum required for combat, the battalion now requested resupply by air drop. This was accomplished on 1 November with the dropping of 250 hand grenades, 500 pounds of TNT, and 1,000 pounds of rice.158

Following the negative report of Major Bigger's patrol the day before, a force commanded by Bigger and consisting of 87 Marines--mainly from Company G plus several from Headquarters Company--and two natives, was dispatched on 1 November to Nukiki in two boats. This force was instructed to proceed from Nukiki to the northwest side of the Warrior River, destroying enemy troops and installations along the way and, if possible, to come within bombardment range of the Japanese base in the vicinity of Choiseul Bay in order that this locality could be shelled with 60mm mortars. If the base at Choiseul Bay could not be shelled, then Guppy Island, south of the Bay, was to be used as an alternate target.159

The attack proceeded generally as planned.

The Company G patrol moved into the mouth of the Warrior River, which according to available hydrographic information was deep enough for small schooners. To his surprise, however, Bigger found that the boats continually grounded and the coxswains were forced to gun the motors. Because it was impossible to hide the craft in the river mouth as had been planned, troops were disembarked and the boats were sent back to a cove near Nukiki to be hidden. The patrol now felt sure that the Japanese were aware of its presence, because of the noise made by the motors when the landing craft had grounded.

Landing on the east bank of the Warrior, Bigger ordered his radio set up and communications established with base camp. All excess gear, including demolitions, was cached near the radio, and a detachment of four men guarded it. After mortar ammunition was distributed among all the troops, Company G set out along the river toward its objective. By mid-afternoon Bigger realized that the natives guiding the patrol were hopelessly lost, so he decided to bivouac in the middle of a swamp for the night.160 Having made this decision, he sent First Lieutenant Rea E. Duncan and a squad of men back along the route of approach to the base on the point at the Warrior River mouth, in order to radio information concerning the day's activities to Krulak, and to direct the boats to return to Voza for the night. In the meantime, a native, acquainted with the Choiseul Bay region, reported to Seton. Seton, of course, realized that Bigger's native guides were unfamiliar with the area, so he in turn sent the new man to contact and assist the patrol.

Duncan made his way back to the mouth of the Warrior, over tortuous trails, where, upon arrival, he set up a bivouac for the night.

Upon waking in the morning, Duncan discovered, to his consternation, that the enemy had moved in behind the main body of his patrol, and that he, the cache and the TBX radio had been surrounded. Duncan and his squad slipped through the enemy and made their way back to Nukiki, where they embarked in the boats. Upon arrival at Voza, he informed battalion of Major Bigger's plight. Upon receiving this information, Lieutenant Colonel Krulak requested that fighter support and PT boat cover be given to the boats engaged in the withdrawal of the force from the Nukiki area.

During all this, Major Bigger was unaware


that Japanese were behind him. On the next morning (2 November), Major Bigger--oblivious of the situation in his rear--set out again for Choiseul Bay. The native sent by Seton joined shortly thereafter. Upon reaching the coast, Bigger sent Sergeant Rahland Wilson, Corporal Winston Gallaher, and several other men back to the Warrior River mouth by way of the coast to have the TBX radio instruct the boats to proceed westward from the river mouth to meet and pick up the patrol that afternoon. This group also discovered the presence of enemy and had in turn to fight its way through to Nukiki (losing one man, Corporal Gallaher) where, late that afternoon, it met Duncan's group, returning in boats from Voza--en route to the Warrior River area.

The remainder of Bigger's patrol followed the new guide to the coast, which, they discovered, was only a short distance away, and proceeded up the coast toward Choiseul Bay. Just as the southern tip of Redman Island (the southernmost island in the chain stretching across the front of Choiseul Bay) came in sight, an outpost of four Japanese was sighted. A short burst of fire killed three immediately, while the fourth fled into the jungle. The patrol now pushed on hurriedly, and it soon became evident that the jungle was thinning and that detection was therefore probable.

Major Bigger accordingly decided to execute his alternate mission and shell Guppy Island, the barge replenishing-center and fuel base for the Choiseul Bay area. Overhanging vegetation, however, extended to the water's edge. It was necessary, therefore, to set up the mortars in the sea with only their muzzles protruding from the water's surface. The shelling was accordingly accomplished as ordered, with 120 rounds of high explosive shell,161 and large fires were observed to have been started, one of which was obviously a fuel dump.

Return to the mouth of the Warrior River was uneventful. Upon arriving on the west bank, and after establishing local security, Major Bigger gave permission to several men to go in the water to bathe. These had entered the water and had proceeded no farther than 15 to 20 yards from shore, when they were fired on by Japanese who were then upstream on the same bank of the river. This was the first indication to Bigger of enemy in the vicinity.

The Japanese closed on the Marines. A firefight broke out, in which no Marine casualties resulted, while the enemy withdrew leaving behind a number of dead. Now believing that it would not be feasible to land boats in the river itself, Major Bigger sent First Lieutenant Samuel M. Johnston, Platoon Sergeant Frank J. Muller, and Private Paul Pare across the river to the east bank (where Duncan's patrol was believed to be located), with the mission of bringing the boats to the east bank when they arrived. Just as the three men reached the east bank of the river they were fired on by Japanese on that bank, one of the men (Platoon Sergeant Muller) being killed instantly, and Johnston being wounded. Pare, who fortunately was not quite so strong a swimmer and had not reached the east bank was able to escape and return to the west bank, but before he made his way back, he observed two Japanese running out in the water and capturing the injured Johnston. Muller's body sank beneath the water; Johnston's body was found later and buried by the Coastwatchers.162


Map 26
Bigger's Patrol
1-2 Nov 1943

Believing that the fire on the Johnston detail was from the Duncan group, who had not identified the swimmers, an American flag was displayed by Marines on the west bank. Thereupon vigorous fire came from the east bank wounding one man. This was the first definite indication that the patrol was cut off from Voza.

During the above fire fight, the boats were sighted south of the east bank of the Warrior River.

A storm was gathering and a heavy sea was rising, as the boats, guided by Lieutenant Duncan, approached the mouth of the Warrior. These received the fire that had been directed at Bigger's patrol. The Navy officer in charge of the boats wanted to turn back when fired on from the beach, but Duncan prevailed on him to continue his approach. Marines ashore attracted the attention of personnel in the boats, which then moved toward the west bank, with all available weapons firing on the east bank causing the enemy to cease fire. Surf was so rough when the boats beached that one was seriously damaged. With its motor flooded out, it was being blown in towards the enemy shore. When all hope seemed gone, a PT boat arrived. Marines scrambled aboard, the LCP(R) being scuttled on a coral head. The PT's continued to furnish cover for the surviving boat for the remainder of the trip to Voza, where they arrived at 2130. During this journey the one wounded man died. The Japanese had lost a total of 43 counted dead as a result of all this.163

On 1 November, the day Bigger left for Warrior River, the Marines launched an attack on Vagara Village, now held by an aggressive enemy unit which seemingly had become aware of the small size of the Marine force on Choiseul, and appeared determined to cut it off. The Marines (a platoon under Lt. John J. Richards) succeeded in driving this unit back towards Sangigai. Eight dead Japanese were counted on the field, while one Marine was lost.164

Strong combat patrols were therefore sent out the following day, to the northwest and the southeast in an effort to strike enemy columns now apparently advancing in order to launch an attack against our forces. Japanese patrols became increasingly aggressive. Ambushes were prepared, and one succeeded on trapping and killing three Japanese.

Intelligence reports from coastwatchers and natives indicated that there were between 800 and 1000 Japanese at Sangigai and some at Moli with more moving into that locality. By late afternoon, 3 November, the enemy had advanced to the general vicinity of Kuluni in the north and to within 500 yards southeast of Voza in the south. After analysing the accumulated information and realizing that mission had been accomplished, General Vandegrift ordered the 2d Parachute Battalion to withdraw during the night of 3-4 November. In order to defend the beachhead selected for the scheduled withdrawal, the raiding forces began to install mine fields and booby-traps. Two fields, including 1100 pounds of TNT and 243 booby-traps, were laid. During hours of darkness prior to the time scheduled for arrival of the evacuating LCI's these protective measures became audibly apparent to the raiding force. It was presumed that customary enemy infiltration was taking place, somewhat hampered in this instance by the placed charges.165 Three LCI's under Commander James McD. Smith, USN, beached at about midnight of 3 November; reembarkation of troops and supplies, which began about 0138, was uneventful.166 All supplies and equipment were brought back by the raiding force excepting rations, which were given to the faithful native carriers, who had done a magnificent job


for the force. Embarkation was completed in 12 minutes, and the battalion was returned to Vella Lavella by 0730, 4 November 1943.167

The Battalion accomplished its mission in a highly satisfactory manner.168 Since operations had been conducted on a wide front (25 miles), and since COMSOPAC had issued a communique which indicated the landing of a larger force, the Japanese command, it appears, thought that a strong American move was underway in the Choiseul area.169 Further it was definitely ascertained that about 180 tons of supplies had been destroyed at Sangigai, to say nothing of the undetermined destruction wrought by the shelling of Guppy Island. Two barges were sunk and the permanent buildings at Sangigai were burned. Immensely valuable minefield charts of the Shortland area were captured at Sangigai. In addition, the battalion disrupted the barge withdrawal of Kolombangara refugees along the coast of Choiseul, by cutting the overland routes from the southeast to the Choiseul Bay embarkation point.170

Our losses in this campaign were nine killed, 14 wounded, and two missing (later declared dead).171 A minimum of 143 Japanese dead were counted, while others probably died unobserved.172

Most important of all, however, Japanese forces had gathered to counter a supposed Allied effort on Choiseul, while in fact the Northern Attack Force had landed at Cape Torokina to strike an enemy now off-balance.

A Step Upward--Green Island

After establishing the Bougainville beachhead, Admiral Halsey took under consideration the additional steps designed to complete the conquest of the Northern Solomons for containment of Rabaul. At that time the joint Chiefs of Staff plans indicated that the next move for South Pacific forces would be the neutralization of Kavieng, a Japanese stronghold on New Ireland in the Bismarcks.173

Prospective operations in the Central Pacific, namely, those in the Marshalls, Carolines, and Marianas, however, appeared to preclude the availability of necessary fleet support for any such proposed landing until approximately 1 May 1944.

Obviously, provision for shore-based fighter cover over the Kavieng area and for shore-based air strikes preliminary to an amphibious assault was highly desirable. To this end, intermediate targets were sought which would offer suitable sites for the required landing-strips. By the same token, the objective finally selected would have to fall within the range of planes flying from our most advanced airstrip--Torokina--in order that a continuous land-based fighter cover could be maintained over the beachhead as well as the supply lines. This requirement was rendered essential because of the non-availability of carrier air-support.174

On 20 December 1943, therefore, at a conference between members of Admiral Halsey's staff175 and General MacArthur, it was suggested that the Green Islands be seized first, and, later on, islands of the St. Matthias Group. At


this time, MacArthur made it clear that he did not intend to take Manus Island (one of his next primary objectives), until such time as the South Pacific was prepared to undertake operations against Kavieng or the St. Matthias Group.176 In view of the fact that a delay in launching either of these was imminent, MacArthur suggested that Admiral Halsey land in the Green Islands or engage in any offensive move that he saw fit.

When informed of MacArthur's proposal, Admiral Halsey sent the following letter to Admiral Fitch, Admiral Wilkinson, and General Geiger:

As you already know, present estimates indicate the FOREARM [Kavieng] operations cannot start much before 1 May 1944 due to inability of Cincpac to make available to us the necessary Fleet support. This delay in the execution of FOREARM will work more to the advantage of the Jap bastards than to us as it gives them an additional two months in which they can strengthen their defenses.

As an alternate plan, I am entertaining the thought that we can accomplish our mission of denying Kavieng as well as other Jap installations in the South thereof by occupying one or more of the islands in the St. Matthias Group. Of course, it is readily apparent that such occupation cannot be initiated until such time as Cincpac can make certain required Fleet support available to us. Simultaneously with our occupation of these islands, the Southwest Pacific Force should occupy MERCANTILE [Manus].

In the meantime, I want to occupy one of the islands to the north of Buka, from which PT boats can operate against the Bougainville supply line and from which air can operate against Kavieng. Green Island appears to provide these requirements. This operation should not require much more than a reinforced regiment and possibly two CB Battalions plus 1 Acorn. It might be advisable to employ a portion of the New Zealand Division, staging either from Treasury or possibly from Vella Lavella.177

To some officers, substitution of the St. Matthias Group as an alternate objective for the principal effort appeared of little value with respect to the isolation of Kavieng and Rabaul. Furthermore, they believed it doubtful that fighters stationed at a base in the Green Islands would be able to cover Kavieng. Admiral Wilkinson, therefore, with the concurrence of Admiral Fitch and General Geiger, suggested that the proposal be abandoned, and that plans be laid for a landing either at Borpop or Boang Island.178

On 28 December, however, after analysing the situation, Colonel William E. Riley, Admiral Halsey's War Plans officer, made the following observation:

There is no doubt but that the occupation of this area [Borpop] and the utilization of air facilities thereat would assist materially in the neutralization of both Rabaul and Kavieng. However, the grave possibility that such an operation would result in the employment of approximately two divisions and additional naval support, not now available and will--in all probability--not be made available in time for this type intermediate operation, makes it necessary to eliminate Borpop as a possible intermediate operation.179

Colonel Riley went on to point out that a fighter umbrella could be thrown over Green during a landing, an accomplishment impossible in regard to Boang. Furthermore, he said:

Though Green Island is not as far advanced as Boang Island, yet it is quite possible to provide fighter cover for strikes on Kavieng. In addition, dive bombers, based on Green Island can operate most effectively against shipping at Kavieng.180

After weighing the advantages against the disadvantages of each proposed operation, and after taking into consideration his concept of a future landing in the St. Matthias Islands, Admiral Halsey decided to follow Colonel Riley's recommendation, and directed that plans be made for an attack on Green Island.

Situated northwest of Buka Island and only 117 miles directly east of Rabaul, the Green Islands constitute a circular atoll about four miles in diameter. In 1943, the Japanese were known to have been using the area as a barge replenishing and staging point. Little else was known of the topography or hydrography of the islands; less of the enemy based there.

To assure a thorough survey despite the presence of the enemy, a reconnaissance in force was ordered. To avoid betraying our intentions, the date of the reconnaissance was set as 31 January


1944, the last moment when resultant information would assist in preparation or final attack plans.181

Even before the 31 January raid, however, it was necessary to know whether or not landing craft could effect passage of channels entering the lagoon. Accordingly, during the night of 10 January two PT boats conducted a hydrographic reconnaissance of the islands and discovered that the southernmost of the channels was suitable.182

On 31 January, therefore, about 300 troops of the 30th New Zealand Battalion,183 embarked aboard three destroyer-transports, landed and conducted operations ashore for 24 hours. During this period excellent landing beaches and a promising airfield site were discovered.184

Just one week previously (24 January), Admiral Halsey had issued his operation order, which set D-day as 15 February and designated the 3d New Zealand Division (less the 8th Brigade), which had until 4 January been under operational control of IMAC,185 as the attack and garrison force: Admiral Wilkinson, as CTF-31, was placed in overall command,186 while General Barrowclough was in tactical command of the landing forces.

Provision was made for operation details by subordinate commanders, and by 12 February all was in readiness and ships began loading.187 Although successive echelons of the attack group were heckled without result by enemy planes during the approach, the New Zealanders arrived safely at their destination and by 0641, 15 February 1944, landing boats of the APD's crossed the line of departure and headed for the beach.188

While the landing craft were making their initial run toward shore, Marine pilots of VMF-212 and VMF-216 patrolled the air overhead. At 0645 about 15 Japanese dive bombers189 were sighted approaching to attack the shipping lying just offshore. Without hesitation, the VMF-212 flyers dived through the ships' protective antiaircraft fire to intercept the enemy. In the ensuing melée First Lieutenant Phillip C. DeLong destroyed three Vals in quick succession, while Captain William C. Carlton, First Lieutenant Theodore J. Horner, and First Lieutenant Thaddeus J. Trojnar each accounted for another. The remaining attackers with bombs still in their racks were driven off by VMF-216. This success was achieved by the Marines without sustaining damage or loss to themselves. For the remainder of the day, 38 F4U fighters from VMF-217, VMF-218, and VMF-222 continued to furnish cover for the landing forces, but no further enemy air opposition ensued.190

Meanwhile, assault units landed as planned on beaches inside the lagoon at Pokonian and Tangalan Plantations. Within two hours, 5800 men, with all their supplies and equipment, were ashore. Combat and reconnaissance patrols quickly pushed out, but no resistance, other than slight fire from two enemy barges, was encountered. By nightfall a secure perimeter was established beyond the southern boundaries of the plantations.

The following day, the New Zealanders moved southward from their positions and discovered a small Japanese garrison of about 70 troops near the Catholic mission on the south end of Nissan Island. For the next three days operations continued in the vicinity of the mission, culminating in elimination of the enemy resistance on 19 February. In this area, the New Zealanders captured several 20mm guns, machine-guns, mortars, and large quantities of small arms. Sixty-two dead Japanese were counted, while New Zealand forces sustained a loss of three killed and 11 wounded.191

In the meantime other operations continued elsewhere. On 17 February, a small landing party was placed on Sirot Island, from which LCI gunboats had received small arms fire on D-day. A


Map 27
Green Island
15 Feb 1943

small enemy group was contacted and in the resulting action 29 Japanese were killed. New Zealand forces lost five killed and four wounded. The same day a boat patrol, sent around the neighboring small island of Pinipel found and destroyed two unmanned Japanese barges. Two days later a patrol discovered 14 enemy on the island of Sau, adjacent to Pinipel, and killed them all. This completed mopping up operations, although for several days afterwards stray Japanese were found.192

As combat troops fought the enemy, the 33d, 37th, and 93d Seabees labored to construct an airstrip, tank farms, roads, and other base installation. Work on these projects proceeded so rapidly that by 4 March, only 18 days after the initial assault, an Allied plane was able to make an emergency landing. On 7 March, the strip was ready to stage aircraft for strikes against Rabaul. Concurrently, a bomber strip was being laid and naval base facilities were rapidly taking form. A PT squadron arrived on 17 February and commenced nightly patrols.193

To counter our activity, the Japanese attempted to launch air attacks against our beachhead, but entirely without success. During the night of 17-18 February, for example, enemy planes sortied over the Green Islands. Marine planes of VMF(N)-531, however, proved to be an effective cover, destroying at least two of the attacking planes that night.194

By 15 March the Green Islands position was completely secure and was operating as a major base in the neutralization of Rabaul. Aircraft and PT boats based at Green were continually attacking enemy shipping and barge traffic, thus adding to the slow strangulation which the joint Chiefs of Staff had planned for Rabaul. The intermediate base to be seized before the final attack of the Northern Solomons-Bismarcks campaign was firmly in our hands.

The Last Step--Emirau

As planning for the Green Islands operation proceeded, Admiral Halsey continued developing his concept of a landing on one of the islands of the St. Matthias Group. In late December, therefore, Admiral Wilkinson was directed to prepare for the seizure and occupation of Emirau Island.

While working out his solution of this problem, Admiral Wilkinson received orders to abandon his present endeavors and concentrate on framing a scheme for an amphibious attack on Kavieng. In early March, however, it was realized how strongly Kavieng was defended and held. To reduce such a stranglehold would require more troops and material than were then available to South Pacific Forces. Along with these factors came the realization that the effectiveness of Kavieng would be nullified if the Allies should occupy islands north and west of that area, since Cape Gloucester to the south had already been secured.

Another reason for this change of plans was the 1st Cavalry Division's successful reconnaissance in force on 29 February, which resulted in establishment of American positions in the Admiralty Islands. We now had the required area to the west; all that remained to seal the fate of Kavieng was to occupy a spot to the north whereby Japanese in the Solomons and Bismarcks would be completely cut off from the Empire.

On 14 March, therefore, Admiral Halsey directed Admiral Wilkinson to "seize and occupy Emirau at the earliest practicable date, not later than 20 March. . . ."195 In view of the urgency indicated by the progression of events, the 4th Marines--although already earmarked for another campaign--were selected as the landing force.196


Since it was essential to bring the staffs of various echelons of command together immediately for detailed planning, it was fortunate that Headquarters of IMAC and CTF-31 were both in close proximity to that of the 4th Marines. The staffs gathered. Emirau Island landing plans, previously discarded, were now reassembled and studied. Final plans were quickly formulated.197

Operation orders evolving from these plans envisaged a simultaneous landing of the 1st and 2d Battalions, 4th Marines, on two southern coast beaches of the island's eastern tip, while the 3d Battalion would remain afloat, serving as a reserve to support either landing.198

Although no preliminary naval gunfire bombardment or air bombing was projected, provision was made for both should either become necessary. Boat assignment tables, landing diagrams, logistics data, and other details were worked out.199

Emirau,200 the target, is an irregularly shaped island located in the southwest portion of the St. Matthias Group (1°38' S--150°0' E), about 75 miles northwest of Kavieng, and about 250 miles north of Cape Gloucester (in March, 1944, the closest Allied base). Approximately eight miles long by four miles wide, the island is relatively level, the maximum elevation being about 120 feet. Undulating and densely wooded, Emirau, nevertheless, possessed several suitable sites for airstrips. Many smaller islets and reefs surround it. On the south coast is a fringe of mangrove swamp.201 In 1941 approximately 320 natives lived on the island.

In general, embarkation plans provided for the transportation of the two assault battalions in nine destroyer-transports, and a reserve battalion (3d Battalion, 4th Marines) plus part of the service troops in an APA (transport). Remaining service troops and supporting elements would be carried to the target in LSD's.202

Cargo and equipment was distributed throughout the various ships of the force. Nine destroyers and two tugs were furnished to provide the gunfire support and act, enroute, as escort of the landing force.

Although the Marines were ready to embark on schedule, their equipment and supplies previously prepared for the proposed operation against Kavieng, had to be reorganized and segregated for loading. Despite this difficulty, loading proceeded expeditiously.

D-day was set for 20 March 1944.

Shortly after dawn on D-day, a flight of fighters from VMF-218 made a perimeter search of Mussau Island, 12½ miles to the northwest, for enemy float planes, radar or radio installations; likewise an air reconnaissance of Emirau was made. Results of both of these flights proved negative, as no Japanese activity was observed.

No definite information concerning the enemy was available. Although believing Emirau to be unoccupied, the assault battalions were prepared to deal with any eventuality, landing with rifles loaded and bayonets fixed. Upon reaching shore, the Marines established beachheads in the immediate vicinity of their separate landing points and shortly joined forces, thus securing the eastern end of the island before nightfall. Concurrently with the main landings, a small detachment went ashore on Elomusao, a small island off the eastern tip of Emirau covering the approaches to the right flank beach. During the approach of this detachment, the amphibian tractors in which the Marines were embarked, fearing opposition, opened fire--shortly augmented by fire from a destroyer. One man was


Map 28
Emirau Landings
4th Marines (Reinf)
20 Mar 1943

wounded as a result. It was later learned that this precaution on the part of the amphibians was unnecessary, since no Japanese had been on Emirau since January.

Despite the fact that the tanks were not landed promptly,203 they arrived in time to scout the island as far west as Pakane and Homestead.

In the meantime the reserve battalion had been put ashore, and preparations were made for the occupation and defense of the island.

Occupation was completed early the next morning. General Noble organized his troops for the defense of Emirau: antiaircraft defenses were coordinated, protective trenches dug; trails cut; fields of fire prepared; and all other installations such as the creation of an artificial beach,204 were progressively developed.

A night reconnaissance patrol from Emirau discovered that the Japanese had established a garrison of approximately 46 officers and men on Mussau Island for the purpose of operating a small seaplane base. On 23 March, therefore, two destroyers and a spotting plane were dispatched to Eloaue, an islet off the southwest coast of Mussau, where the base was reported to be located. After a four-hour shelling, the destroyers retired, leaving a fuel dump in flames.205 Four days later a patrolling destroyer intercepted a native canoe 40 miles south of Mussau. Upon approaching the canoe, the destroyer received machine-gun fire, which it was forced to return. After a brief engagement, the canoe was brought along-side; in it were the newly dead bodies of one officer and 26 men of the Japanese Special Navy Landing Force. The following day (28 March), a reconnaissance patrol of Mussau was informed by natives that 46 Japanese had abandoned the island in two native canoes early in the morning of 24 March, hoping, apparently, to reach Kavieng.206

As succeeding echelons continued to arrive, responsibility for maintenance of the position was gradually shifted to the Army. On 11 April, therefore, the 147th Infantry assumed control as garrison force of the "Emirau Island Command."207

With the occupation of Emirau, the campaign in the Northern Solomons was concluded. Considered with operations in New Guinea and New Britain, the Emirau operation may be characterized as the cork placed in a bottle containing Rabaul and Kavieng.

Enemy reinforcements and matériel could no longer be poured into or out of these positions. At Emirau forces of the South and Southwest Pacific came together, there to unite for a drive through the Philippines toward Japan. After Emirau, neither Rabaul nor Kavieng could be regarded as effective barriers to our advance toward ultimate victory.


Table of Contents ** Previous Chapter (3) * Next Chapter (5)


1. COMSOPAC Opn Plan 16-43.

2. Task Force 31 Operation Plan A14-43, 1-3,

3. Loc. cit., 2.

4. COMSOPAC Opn Plan 16-43, 6.

5. CTF-31 Opn Rpt, 1-3.

6. Row Rpt, 16. The APA Action Report, 4 (see bibliography), and Combat Narratives, 12, erroneously report that these troops had had experience in Greece, Crete, and Libya.

7. Task Force 31 Operation Plan A14-43.

8. IMAC LofI, 28Sep43.

9. Row Rpt, 1-3.

10. TG 3d AmphFor AR, 2.

11. See Chapter I--"Accumulation of Intelligence."

12. IMAC D-2 Miscellaneous Reports, 1-6. Cf. Row Rpt, 4. The IMAC document cited here appears to have been mistitled. These reports are bound in a jacket with the title cited here, but there is no explanation for the designation "D-2" having been assigned to what are primarily Corps (C-2) documents.

13. IMAC Estimate of the Situation, Treasury Islands, 1-3.

14. Ibid., 3-5. Actually about 250 Japanese troops defended the Treasuries. These formed the 1st Japanese Base Force, which consisted of the 7th Combined SNLF, Kure #7 SNLF, 16th AAA Tai (Company), 17th AAA Tai, and Sasebo #6 SNLF.

15. Row Rpt, 6.

16. Ibid., 8. See Chapter I, section entitled "Accumulation of Intelligence."

17. IMAC Concept of Operations Plan I, 3. See also Row Rpt, 9.

18. TG 3d AmphFor AR, 4.

19. Row Rpt, 7.

20. Ibid., 8.

21. Task Force 31 Operation Plan A14-43, 3.

22. Row Rpt, 8.

23. Task Force 31 Operation Plan A14-43, 4-6.

24. Ibid., 3.

25. Combat Narratives, 19-20.

26. Row Rpt, 8.

27. For example, Company A, 87th NCB was not assigned to this operation until 14 October, and did not arrive at Guadalcanal from the Russell Islands until 18 October.

28. Row Rpt, 9. Because units were not as yet formed in several instances, officers in charge could not furnish the Brigade staff with requisite information concerning numbers of personnel, equipment, tonnage, or shipping space necessary for their units.

29. TG 3d AmphFor AR, 11. For information regarding organization of transport units see: COMTRANSDIV 12 AR, 1; TG 3d AmphFor AR, 4; and, Combat Narratives, 79-80.

30. CTF-31 Opn Rpt, 3.

31. Row Rpt, 7.

32. CTF-31 Opn Rpt, 4.

33. Combat Narratives, 14.

34. TG 3d AmphFor AR, 18.

35. Twenty minutes late, due either to adverse currents, inaccurate charts, or faulty navigation. The radar of Admiral Fort's flag-ship Eaton, had given a false range, probably due to a rain-squall (TG 3d AmphFor AR, 5). H-hour was to have been 0606, but due to tardiness, it was delayed 20 minutes by Admiral Fort (TG 3d AmphFor AR, 1). See preceding page and cf. Combat Narratives, 14.

36. TG 3d AmphFor AR, 8.

37. Row Rpt, 7.

38. Loc. cit., 7, indicates that this fire began at 0545.

39. CTF-31 Opn Rpt, 6; see also TG 3d AmphFor AR, 9.

40. TG 3d AmphFor AR, 7.

41. Ibid., 8.

42. ComDesDiv 90, Action Report, Treasury Islands, 1; TG 3d AmphFor AR, also is enlightening concerning this phase of the operation.

43. TG 3d AmphFor AR, 1-2.

44. Combat Narratives, 15.

45. Row Rpt, 7.

46. This boat was from the Crosby.

47. TG 3d AmphFor AR, 2.

48. Combat Narratives, 16. Additional information for this statement was obtained from Row Rpt, 7.

49. LCIFLOT 5, Action Report of LCI(L)'s at Treasury Islands, 27 October 1943, 6.

50. Row Rpt, 7.

51. TG 3d AmphFor AR, and 1st Endorsement to the AID Action Report both give much credit to the LCI(L)'s.

52. Combat Narratives, 16.

53. Row Rpt, 10.

54. U.S.S. McKean, Action Report--Treasury Islands Landings, 3.

55. LCIFLOT 5, Action Report of LCI(L)'s at Treasury Islands, 27 October 1943, 9.

56. LST Group 15, Action Report, Treasury, 3.

57. Row Rpt, 8.

58. LST 399, Narrative Account of the Action, 4.

59. Ibid., 5; see also LST Group 15, Action Report, Treasury, 7. Time (20 December 1943) gives credit for this act to F 1/c Aurelio Tassone, Company A, 87th NCB. No official record appears to confirm this, all stating that the act was performed by a New Zealander.

60. TG 3d AmphFor AR, 9.

61. LST Group 15, Action Report, Treasury, 8. It appears that the pill-box mentioned supra had been the cause of the delay.

62. Brigade Administrative Order No. 1 had provided for men and equipment for this task.

63. Principal difficulty seems to have been lack of suitable arrangements for a shore-party, and a complete lack of understanding on the part of LST personnel and New Zealand working parties. The LST crew could not convince the New Zealanders, few of whom were veterans, that unloading was a troop responsibility. Combat Narratives, 19-20.

64. LST Group 15, Action Report, Treasury, 8.

65. Ibid., 9.

66. LST 399, Narrative Account of the Action, 7.

67. LST 399, Narrative Account of the Action, 10. See also Row Rpt, 9, for an account of this action.

68. Row Rpt, 8-10.

69. LST 399, Narrative Account of the Action, 10.

70. LST Group 15, Action Report, Treasury, 7.

71. Row Rpt, 9.

72. Combat Narratives, 19.

73. LST Group 15, Action Report, Treasury, 8.

74. LST 399, Narrative Account of the Action, 11.

75. Row Rpt, 9, states that there was a total of 34 tons of rations and some equipment left aboard.

76. LST Group 15, op. cit., 9 and LST 399, op. cit., 12.

77. Combat Narratives, 18-19.

78. TG 3d AmphFor AR, 6.

79. COMTRANSDIV 12 AR, 2-4.

80. Ibid., 4.

81. Combat Narratives, 16.

82. Row Rpt, 7. See also note 55 and the text thereto supra.

83. IMAC, C-2, Periodic Reports, 26, 27, 28 October 1943.

84. IMAC, loc. cit., 27 October 1943.

85. Row Rpt, 7.

86. Ibid., 8.

87. Row, loc. cit., gives a complete description of this action.

88. Ibid., 8.

89. Ibid., 9.

90. Ibid., 8.

91. TG 3d AmphFor AR, 9.

92. The Cony was badly hit about 1534 by two bombs from dive bombers. See TG 3d AmphFor AR, 9.

93. IMAC, C-2 Periodic Reports, 30 October 1943.

94. Motor torpedo boat, sometimes called a PT or patrol torpedo boat.

95. Row Rpt, 10.

96. Grenade dischargers. The term "knee mortar" is used herein as a descriptive term which was adopted by American troops in the Pacific theater to describe this particular weapon.

97. IMAC, C-2 Periodic Report, 30 October 1943.

98. Row Rpt, 10.

99. IMAC, C-2 Periodic Reports, 29 October 1943.

100. Row Rpt, 10.

101. IMAC, C-2 Periodic Reports, 29 October 1943.

102. Row Rpt, 11.

103. IMAC, C-2 Periodic Report, 30 October 1943.

104. Row Rpt, 11.

105. IMAC, C-2 Periodic Reports, 30 October 1943, and Row Rpt, 11.

106. Row Rpt, 11.

107. IMAC, C-2 Periodic Reports, 1 November 1943.

108. Ibid, 29 October-2 November 1943.

109. Row Rpt, 11.

110. IMAC, C-2 Periodic Reports, 1 November 1943.

111. Row Rpt, 11.

112. Ibid., 12.

113. Row Rpt, 13.

114. Ibid., 13-14.

115. IMAC, C-2 Periodic Reports, 2 November 1943.

116. Row Rpt, 14.

117. Ibid., 14. The Japanese probably had more than 50 men committed in this engagement; the New Zealanders were able to count only bodies that were on or behind their lines, and it is safe to assume that some enemy died in front of the lines while others may have escaped.

Captain Kirk himself was wounded in this engagement and subsequently died of wounds.

118. IMAC, C-2 Periodic Reports, 4 November 1943.

119. TG 3d AmphFor AR, 11. Casualties for the Treasury Islands operation were as follows: Japan, 205 killed, 8 captured; New Zealand, 40 killed, 145 wounded; U.S., 2 killed, 29 wounded.

120. 3d AmphFor AR, 12. Although not originally planned, an airfield, erected on Stirling Island, was put into operation on 25 December.

121. Memorandum for COMSOPAC from Fitch, Harmon, Barrett, and Wilkinson, dated 7 September, 1943. See note 74, Chapter I, supra.

122. 3d MarDiv Patrol Reports-Choiseul. See also Chapter I, section entitled "Accumulation of Intelligence".

123. IMAC Opn Rpt, I, II. The battalion was then encamped at Vella Lavella. 1stLt. Paul Jansen, of Headquarters, XIII Fighter Command, was attached to the Battalion for the purpose of finding a suitable airfield site.

124. Feldt, loc. cit. gives the most complete information of activities of the coastwatchers.

125. Marine Corps, Division of Public Information, Historical Section, Interrogation of LtCol Warner T. Bigger, USMC, 1 May 1947. Hereinafter cited as Bigger Interrogation. See also same agency, Krulak Interrogation (22 March 1948) which states that LtCol Robert Williams, Commanding Officer, 1st Parachute Regiment, had received a dispatch on the night of 20 October 1943, directing him to bring Krulak to IMAC Headquarters. Krulak states that he and Williams flew to Guadalcanal the following day.

126. Feldt, op. cit., 107, 108, 111-114, 157-159; Loc. cit. Seton is called a sergeant of Australian Imperial Forces. When war broke out, Seton immediately enlisted in the Australian Army, was sent to New Guinea, and was promoted to Sergeant. Upon organization of the coastwatching unit by the Australian Navy, Seton was recalled to Australia and was commissioned in that organization.

127. Krulak then went to the airstrip to wait for the plane to take him back to Vella Lavella. While waiting he wrote the Battalion Operation Order.

128. This was done so as not to alert the enemy, who conducted daily air patrols over the Vella Lavella area.

129. Transport Division 22, which, under command of Commander Robert H. Wilkinson, USN (with his flag in the Kilty), had just participated in the Treasury Islands Landings.

130. COMTRANSDIV 12 AR, 4; 2d Para Bn Opn O 1-43; Combat Narratives, 24-25. See also Krulak Interrogation.

131. U.S.S. Conway AR, 27-28 Oct 43, 1, 5-6; Combat Narratives, 24. Commander J. H. Pahl, USN, commanded the convoy; LtComdr Harold G. Bowen, Jr., USN, was captain of the Conway.

132. Bigger Interrogation. During the trip Seton was in constant radio communication with his colleague, the District Officer for Choiseul, Nick Waddell.

133. Conway AR, 6 and Krulak Interrogation. Duncan's patrol was to scout the beach and guide the battalion ashore by light signal if the beach were clear. Should the patrol encounter enemy, the battalion planned to go ashore one mile to the southeast. Companies F and G were ordered into landing boats while Duncan was on the beach, and headed shoreward as soon as the signal had been received. Since Company G was to have formed the first wave (according to the battalion operation order), this new situation constituted an on-the-spot change of plan.

134. Ibid., see especially paragraph 15 and the log enclosed.

135. CTF-31 Opn Rpt, 13.

136. Seton, accompanied by his personal servants, Petacari and Petanui, had joined the battalion at Vella Lavella on 24 October at 1200. The native bearers were not at Voza when the battalion landed. Seton, therefore, moved out in the pitch black night, found them and led them to the beach.

137. Krulak Interrogation. This area had been previously selected as the camp site. Here a stream runs around three sides of a sheer cliff--a very defensible place.

138. 2d Para Bn Unit Journal, 28 Oct 43; Bigger Interrogation; Krulak Interrogation.

139. 2d Para Bn Unit Journal, 28-29 Oct 43; Bigger Interrogation; Krulak Interrogation; 2d Para Bn War Diary, 28 Oct 43.

140. 2d Para Bn Opns Map; Krulak Interrogation; CTF-31 Opn Rpt, 14; 2d Para Bn Unit Journal 28-29 Oct 43.

141. 2d Para Bn Unit Journal, 28, 29, 30 Oct 43; Bigger Interrogation; Krulak Interrogation.

142. 2d Para Bn Unit Journal, 29 Oct 43; 2d Para Bn War Diary, 29 Oct 43; Krulak ltr 5 Nov 43.

143. Krulak Interrogation. To the work of this patrol LtCol Krulak attributes a large share of credit for the success of the Sangigai attack.

144. Krulak Ltr 5 Nov; Krulak Interrogation.

145. 2d Para Bn Unit Journal, 28 Oct 43; Combat Narratives, 78.

146. 2d Para Bn War Diary, 29, 30 Oct 43.

147. Historical Section, Division of Public Information, Interrogation of Major Spencer H. Pratt at HQMC, 13 Oct 47.

148. This platoon, a part of Corps Troops, IMAC, had been attached to the battalion for the operation.

149. Feldt, op. cit., 157-159; Krulak Interrogation; Pratt Interrogation.

150. MCS, AWS, SC-97, L, IP, 15.

151. Krulak Interrogation; Pratt Interrogation; Krulak ltr 5 Nov 43.

152. Krulak Interrogation. The man carrying one tripod was hit and the tripod lost. The gunner, however, held his weapon in his arms and fired it "movie style."

153. Feldt, op. cit., 157-159; Pratt Interrogation; Krulak Interrogation; Krulak ltr. 2 Nov 43.

154.. Krulak ltr 5 Nov; Krulak Interrogation. These charts were deemed highly important by Admiral Halsey, who directed that the minefield information contained thereon be radioed to him immediately since the Empress Augusta Bay convoy had already put to sea. Several days later, coastwatchers in the Shortlands area reported the sinking of two Japanese ships in the midst of their own channels, which Halsey had mined based on information contained in this chart.

155. 2d Para Bn War Diary, 31 Oct 43.

156. Pratt Interrogation.

157. 2d Para Bn Unit Journal, 31 Oct 43.

158. Ibid., 31 Oct-1 Nov; LtCol Krulak reports that this mission was performed so accurately that nothing was lost in the drop and that all containers and parachutes were returned when the battalion reached Vella Lavella.

159. Krulak ltr 5 Nov 43; Bigger Interrogation.

160. In all fairness to the natives, it should be said here that Choiseul is divided into compartments by spurs jutting out toward the coast from the main mountain range running along the entire length of the island. Natives seldom leave their individual compartments, so know little or nothing of the terrain in other areas of their island. Only two natives, familiar with the Choiseul Bay area, were attached to the battalion for this operation. Of these one had been injured in an accident and the other was out on another mission when Bigger's unit left camp. Consequently, the guides that Bigger had with him were unacquainted with the Choiseul Bay region.

161. Krulak letter of 5 November 1943.

162. There appears to be some element of doubt concerning this particular incident. An officer in a position to know (Bigger) says Johnston, Muller, and Pare, volunteered to swim the stream and that the action as described in the text took place. Official records indicate that Johnston swam over alone, contacted Andrew Sivokana (a native of the coastwatcher group), and reconnoitered the Japanese positions. He then proceeded toward Nukiki and en route was confronted suddenly by a party of Japanese to whom he surrendered.

Official records state that Corporal Winston C. Gallaher also was able to swim to the opposite side of the stream. In any event--Johnston and Gallaher were the only Marines who crossed the river and remained on the other side. Lt. Seton reported that two Marine bodies were found on the east side of the river and buried. The only identification available to Seton at the time of the burial was a stencilled GALLA on the jacket worn by one of the bodies. Headquarters, Marine Corps, presumes that this was Gallaher, and since Johnston was the only other Marine known to have crossed the river, the other body must have been that of Johnston, since Muller's body was observed sinking in the water. Sivokana definitely states that Johnston was captured, yet mentions nothing about wounds. At any rate it is apparently safe to assume that one of the bodies later found by Seton's group was that of Johnston. (Records at HQMC).

163. All information concerning the patrol described above, unless otherwise indicated, is taken from the Bigger Interrogation.

164. Krulak letter of 5 November 1943.

165. Krulak letter of 5 November 1943.

166. Combat Narratives, 24, 79. One of the most amusing incidents of the entire action took place at this time. The sailors, having been informed previously that the Marines were being hard pressed, assumed that the evacuation would take place under fire. Consequently the crews were armed with rifles and standing at the ready as the ships beached. When the Marines began to reembark their equipment, the sailors became quite excited, for they expected to be attacked at any moment. The Marines ignored their comrades in arms and proceeded with their task of loading the gear much to the disgust of the seamen. (Bigger Interrogation.)

167. Krulak Letter of 5 November 1943. Removal of equipment was complete except for ammunition of the rocket detachment, which was left ashore. This ammunition was found by Mr. Seton after departure of the battalion, and he dumped it into the sea so that it would not be captured by the Japanese.

168. CTF-31 Opn Rpt, 17.

169. Krulak Interrogation.

170. Combat Narratives, 24.

171. The muster rolls of the battalion indicate that nine men were killed in action, 12 were wounded in action, and five were missing. Of those missing, four were later indicated as missing, believed dead, and one (1stLt Samuel Johnston) as missing, believed a prisoner of war. Combat Narratives, 24, states our losses thus: nine killed, 15 wounded, two missing--one of whom was captured.

172. Krulak letter of 5 November 1943; Muster rolls of 2d Para Bn for October, November, and December, 1943; Muster Rolls of Prisoner of War Detachment for October, November, and December, 1943 and January and February, 1944.

173. JCS, ELKTON III, 1-5.

174. 3d AmphFor, Seizure and Occupation of Green Islands, 15 Feb to 15 Mar 44, FE25/A16-3(3) over 00176, dated 24 March 1944, p 1-2. Hereinafter cited as 3d AmphFor Green AR.

175. Rear Admiral Robert B. Carney, USN, Commander H. Douglas Moulton, USN, Colonel William E. Riley, and Captain W. F. Riggs, USN. The conference was held at Port Moresby.

176. Memorandum for Admiral Halsey from Admiral Carney, dated 21 December 1943.

177. Letter, Halsey to Wilkinson, dated 22 Dec 43.

178. Wilkinson memorandum dated 25 December 1943.

179. Riley, Memo for Admiral Halsey, 28 December 1943, 2.

180. Ibid., 4. A fact that later became obvious to all was already apparent to Riley: Rabaul-based Japanese air would but sporadically interfere with strikes not aimed at the Rabaul area; consequently, U.S. fighters equipped with wing tanks could fly from Green to Kavieng on a direct line without resorting to fuel-consuming evasive tactics.

181. CTF-31 Opn O 1-44, dated 25Jan44.

182. 3d AmphFor Green Ar, 4.

183. Of the 14th NZ Brigade, 3d NZ Division.

184. Comdesron 45, Action Report, serial 0048, dated 10 February 1944.

185. The 3d NZ Div (Maj Gen H. E. Barrowclough) was composed of the 8th and 14th Brigade Groups. See Section entitled "Treasury Islands", supra. See also note 65, Chapter I, supra.

186. COMSOPAC Opn Plan 5-44, 24 January 1944.

187. CTF-31 Opn O 2-44, dated 5 February 1944 and 3d NZ Div Opn O 101, dated 4 February 1944.

188. 3d AmphFor Green AR, 7.

189. Val type.

190. Daily Intelligence Summary, COMAIRNORSOLS Fighter Command, 15 February 1944. See also VMF-212 War Diary, 15 February 1944.

191. 3d AmphFor Green AR, 9.

192. Ibid., 10.

193. Ibid., 11.

194. VMF(N)-531 War Diary, February 1944.

195. COMSOPAC dispatch to CTF-31, dated 14 Mar 44.

196. A new, independent regiment, made up of recently disbanded Raider Battalions, had been designated "4th Marines" in honor of the famous China unit lost in the Philippines at the beginning of the war. Lieutenant General H. M. Smith earmarked this regiment as part of the landing force for the Marianas campaign. Since it was the only fully trained Regimental Combat Team immediately available for action, it was used at Emirau, despite Smith's objections. Admiral Halsey, however, promised to return the 4th Marines to Smith's control as soon as the Emirau task was completed. True to his promise, Halsey returned the regiment to Smith in time for the Guam landings. (See Halsey and Bryan, Admiral Halsey's Story, pp. 188-191.)

197. Admiral Halsey approved these plans during a visit to Guadalcanal on the afternoon of 15 March, an indication of the celerity and dispatch with which the Marine and Navy planners worked. Commodore Lawrence F. Reifsnider, USN, was designated Attack Group Commander.

198. The Emirau landing force consisted of a command group from the 3d Marine Division and IMAC, a Signal Detachment from IMAC Signal Battalion, a detachment of COMAIRNORSOLS, a Naval Advance Base Unit, and the 4th Marines (reinforced). Reinforcing elements included Company A, 3d Tank Battalion; a Composite Automatic Weapons Battery, 14th Defense Battalion; Company D, IMAC Medical Battalion; Company C, 3d Amphibian Tractor Battalion; a detachment of the 3d Motor Transport Battalion; and the Pioneer Company, 19th Marines. Brigadier General Alfred H. Noble commanded this force.

199. COMSOPAC Operation Plan 7-44, 18 March 1944; CTF-31 Operation Order 5-44, 16 March 1944; IMAC Operation Order No. 7, 15 March 1944.

200. Sometimes called Emira, or Squally.

201. Hydrographic Office 165, Pacific Islands, vol. II, (Washington, 1938), 446.

202. Landing ship, dock.

203. The LSD Lindewald failed to launch her tank bearing LCT's until 0930.

204. Initial reconnaissance suggested that no beaches suitable for landing craft existed on Emirau. Further reconnaissance, however, revealed a 75 yard wide coral shelf, two feet under water at high tide, on the southeast coast.

205. No other results could be ascertained.

206. A. H. Noble, Operations of the Emirau Landing Force, 15 March-9 April 1944, dated 20 April 1944, pp. 1-7.

207. Island Commander was Major General James T. Moore, USMC, who assumed command on 9 April.

Transcribed and formatted by Jerry Holden for the HyperWar Foundation