The Dragons Peninsula Campaign - United States Marine Corps

104
CHAPTER b The Dragons Peninsula Campaign RICE ANCHORAGE 1 Munda’s eventual capture was a triumph over initial frustration and failure. Ad- mittedly, the campaign to take the airfield had been costly and time-consuming. But while the spotlight was focused on the Unless otherwise noted, the material in this section is derived from : CominCh Rept of SoPac Action, 21)ee42-31Jan44, hereafter CominC1/ Iixpt of SoPac Action; WesternForO No. All-43, cltd 28 Jun43; TG 3(;.1 OpO No. 1043, dtd 1Ju143; Ncw Georgiu Campaign; 2JGOF Account; 1st RdrRegt SAR 4Jul-29Aug43, dtd 60et43, here- after Ist RdrIi!c~t NAR; Ist RdrRegt R-2 Est of Situation, c. 23 Jun4J ; 1st RdrRegt R-2 Jnl 2J~n–31Aug43, hereafter IstRdrRcflt .Jnl; 1st RdrRegt MsgFile 5Jul-28Aug43, hereafter 1st RdrRegt NsgFile ; 1st RdrRegt PtlRepts 23Jul–14Aug43 ; 1st RdrBn WarD, 20Jun-29Aug 43, dtd 14 Sept43, hereafter I.st Rdrl;ti WcwD; 3/148 URepts, 7Jul–3Ang43, hereafter 3/1.j8 Rcpt; SW .-trra NavOp8-I; SE Area NavOpts- 11; Maj Clay A. Boyd interview by HistIJiv, HQMG, dtd 16 Feb51 ; Col Si~muel B. Griffith, II, itr to DirPubInfo, HQMC, dtd 121’eb51 ; CO1 Samuel B. Griffith, H, ltr to Maj John N. Rcntz, dtd 12Feb51, hereafter (7ri~tfi 2tr; Col Samuel B. Griffith, II, ltr to Col Enstace R. Snmak, dtd 3MnrY2 ; Lt Robert Il. Pape ltr to CMC, dtd 2M?eb51 ; LtCol William D. Stevenson ltr to CMC, dtd 22 Feb51, hereafter 6Vwrnson ltr; Adm Rich- mond K. Turner ltr to CMC, dtd 22Feb51 ; Col Samuel B. Griffith, II, “Corry’s Boys,” Marine f’orps Gazette, v. 36, no. 3 ( Mar52) and “.4ction at Enogai, ” Marine Corp8 Ga.zettc, v. 38, no. 3 ( Mar54 ) ; TSgts Frank J. McDevitt and Mnrrey J[ill”(lf2r, ‘“Capture of Enogai,” Marine t70rp.s {; fixtte, v. 27, no. 9 ( Sep43 ) ; Morison, Iircakin(l the Bii+m(frclis 17arricr; ONI, Com bat Narrativc8 .Y; Rentz, Marines in tl!e Central Solomon8, 691–360 0—63—9 New Georgia Occupation Force as it fought its way from Zamma Beach to the airstrip, mot her tense struggle was waged simultaneously in the northern part of the island in which the jungle combat was as bitter and as deadly. The results were much less conclusive. From the initial ship-to-shore movement of the Northern Landing Group through the following six weeks of fighting, this phase of the New Georgia campaign contributed as much to the feeling of dis:lppointment and futility as the first Munda attacks. Early plans of the NGOF called for Colonel Harry B. Liversedge’s 1st Marine Raider Regiment (less two battalions) to be a ready reserve. When intelligence re- ported a garrison of some 500 enemy troops with coast defense guns at Bairoko Harbor, the Kula Gulf landing was written into the att~ack order. Of prime concern to the Allied planners was the road con- necting Bairoko with Munda airfield. Scarcely more than an improved jungle trail, the road was nevertheless a vital link between Munda and Vila, the main source of Japanese reinforcements and supplies. Bairoko Harbor was the knot which tied the overland route to the Kula Gulf barge system. An Allied ground force between Munda and Bairoko Harbor would have the double-barreled effect of cutting OR the flow of memy supplies and reinforcements to Munda as well as keeping the airfield forces and the Bairoko garrison from rein- forcing one another. 119

Transcript of The Dragons Peninsula Campaign - United States Marine Corps

CHAPTER b

The Dragons Peninsula Campaign

RICE ANCHORAGE 1

Munda’s eventual capture was a triumphover initial frustration and failure. Ad-mittedly, the campaign to take the airfieldhad been costly and time-consuming. Butwhile the spotlight was focused on the

‘ Unless otherwise noted, the material in thissection is derived from : CominCh Rept of SoPacAction, 21)ee42-31Jan44, hereafter CominC1/Iixpt of SoPac Action; WesternForO No. All-43,cltd 28Jun43; TG 3(;.1 OpO No. 1043, dtd 1Ju143;Ncw Georgiu Campaign; 2JGOF Account; 1st

RdrRegt SAR 4Jul-29Aug43, dtd 60et43, here-after Ist RdrIi!c~t NAR; Ist RdrRegt R-2 Est ofSituation, c. 23Jun4J ; 1st RdrRegt R-2 Jnl2J~n–31Aug43, hereafter IstRdrRcflt .Jnl; 1stRdrRegt MsgFile 5Jul-28Aug43, hereafter 1stRdrRegt NsgFile ; 1st RdrRegt PtlRepts23Jul–14Aug43 ; 1st RdrBn WarD, 20Jun-29Aug43, dtd 14 Sept43, hereafter I.stRdrl;ti WcwD;

3/148 URepts, 7Jul–3Ang43, hereafter 3/1.j8Rcpt; SW .-trra NavOp8-I; SE Area NavOpts-

11; Maj Clay A. Boyd interview by HistIJiv,HQMG, dtd 16Feb51 ; Col Si~muel B. Griffith, II,itr to DirPubInfo, HQMC, dtd 121’eb51 ; CO1Samuel B. Griffith, H, ltr to Maj John N. Rcntz,dtd 12Feb51, hereafter (7ri~tfi2tr;Col SamuelB. Griffith, II, ltr to Col Enstace R. Snmak, dtd3MnrY2 ; Lt Robert Il. Pape ltr to CMC, dtd2M?eb51 ; LtCol William D. Stevenson ltr to CMC,dtd 22Feb51, hereafter 6Vwrnson ltr; Adm Rich-mond K. Turner ltr to CMC, dtd 22Feb51 ; Col

Samuel B. Griffith, II, “Corry’s Boys,” Marinef’orps Gazette, v. 36, no. 3 ( Mar52) and “.4ction

at Enogai, ” Marine Corp8 Ga.zettc, v. 38, no. 3( Mar54 ) ; TSgts Frank J. McDevitt and MnrreyJ[ill”(lf2r, ‘“Capture of Enogai,” Marine t70rp.s

{; fixtte, v. 27, no. 9 ( Sep43 ) ; Morison, Iircakin(l

the Bii+m(frclis 17arricr; ONI, Com bat Narrativc8

.Y; Rentz, Marines in tl!e Central Solomon8,

691–360 0—63—9

New Georgia Occupation Force as itfought its way from Zamma Beach to theairstrip, mot her tense struggle was wagedsimultaneously in the northern part of theisland in which the jungle combat was asbitter and as deadly. The results weremuch less conclusive. From the initialship-to-shore movement of the NorthernLanding Group through the following sixweeks of fighting, this phase of the NewGeorgia campaign contributed as much tothe feeling of dis:lppointment and futilityas the first Munda attacks.

Early plans of the NGOF called forColonel Harry B. Liversedge’s 1st MarineRaider Regiment (less two battalions) tobe a ready reserve. When intelligence re-ported a garrison of some 500 enemytroops with coast defense guns at BairokoHarbor, the Kula Gulf landing was writteninto the att~ack order. Of prime concernto the Allied planners was the road con-necting Bairoko with Munda airfield.Scarcely more than an improved jungletrail, the road was nevertheless a vital linkbetween Munda and Vila, the main sourceof Japanese reinforcements and supplies.Bairoko Harbor was the knot which tiedthe overland route to the Kula Gulf bargesystem. An Allied ground force betweenMunda and Bairoko Harbor would havethe double-barreled effect of cutting OR theflow of memy supplies and reinforcementsto Munda as well as keeping the airfieldforces and the Bairoko garrison from rein-forcing one another.

119

1!20 ISOLATION OF RABAUL

Factors involved in risking a secondaryattlrk north of the airfield had been care-fully considered before a decision to landat Rice Anchorage at tl~e mouth of theP(lntlakona (Wharton) River in KulaGldf lmd been reacllwl. Two area-theIhlndakona and the Piraka Rivrr in Rovi-ana Ln~ooll—were scout cd before the for-mer was selected. Admiral Tlu-ner’s staffreasoned tl~at a 1anding from Roviana La-goon would be unopposed but that the re-sultant, over]al]d trek would be excessivelyslow, fatiguing, and difficlllt to resupply.Further, this landing wol~ld not bring theenemy under immedinte attack. Despitethe native trails crossing the island, ~ largeforce could not travel fast enough throughthe jungle to give assistance to the ex-pectecl rnpid seizure of Munda.

On the other hand, a landing at RiceAnchorage would 1ikewise be unopposed,and the enemy could be taken under ak-tack almost immediately. This wouldforce the Japanese into one of threecourses of action: withdrawal to eitherMunrla or Vi]a, a counterattack in

stren@, or an attempt at defending the

Ilairoko Harbor area. The latter course,it was believed, woulcl be the logical enemyreaction to such a threat to the Munda-~Tila link. Defense by the enen~y atBxi roko wolll d keep that garrison f rom re-

inforcing Mundn. Though the disadvan-ta~es of making a landing on a narrow,.—confined beach on the Puudakona Riverl~early outweighed the advantages, theRice Anchorage nttack held the most, hopefor success in dividing the Muncla-13airoko

forces. ( See Map 8.)Liversedge’s gro[~p, nugmented by the

3d Battalion, l-18tl~ Infantry, was given

a multiple mission in ~~T(_)F orders.

After lmding at Rice, the Northern Land-

ing CIroup was to move overland to thesouthwest ~capturing or killing any enemyforces encountered in the Rairoko andEnogai Inlet a~ea. After establishingroad blocks across all road.< leading fromIla iroko to M{lnda, the NLG was to ad-van(w along the Mund:~-T3airoko trail asfar as possible to prevent any enemy sup-plies or reinforcements reaching Muncla,and also to block ~ny withdrawal fromthat area. Contact with the right (north)flank of tl~e lG9th Infantry was to be n~ain-tained by Liversed~e7s command.

The Marine-Army force had only a lim-ited knowledge of the terrain between Riceand Enogai Inlet and practically no in-formation on Dragons Peninsula, the areabetween Ilairoko and Enogai. For onething. no oblique angle aerial photographsof the area were available. This type ofaerial intelligence was particularly desir-able, since jungle terrain photographedfrom high altitudes directly overheadrarely revealed anything of tactical value.In addition, tl~e peninsula had not beenscoutecl. The New Georgia guides hadbeen relnctant to enter this area, fearingtreachery because of v:qg~e. rumors thattile n~tives of this area were hostile to men

fronl Roviana Lagoon.Most of the SoPac reconnaissance pa-

trols bad been more concerned with Mundawhere the main effort of the NGOF wasto be made. Those few patrols which ven-turecl into the vicinity of Enogai Inletwere forced to turn back by close brusheswith ,Japmnese patrols. Only the long,narrow I,eland Lagoon which borders the

north shore of Dragons Peninsula hadbeen pxtrolled, ancl this had been done incanoes. As a result, the dark stretches ofjungle between 13airoko and Enogai werestill an unknown area.

THE DRAGONS PEAUNSULA CAilIPAIGN 121

With the chte of the lmrding set for 4.JIIly, a one-day postponement was grantedto allow anot]ler 37th Division unit, the3d Battllioll of the 145th Infantry, to joinLiverserlge’s force. ITnexpectedly, the 4thRai(ler Battalion was still ellga:ed in the

TTiru Harbor attack ancl could not be witl~-(lrawJl in time to join the NLG.

It was a li~l~tly armed force. The onlyweapons c:nrried, other tl~an i])d ivi dIIalarms and light machine. guns, were the60n~m mm-tars of the raiders and the81mnl mortars and heavy machine guns ofthe Army battalions. Sotice:Lbly lackingin artillery support, the. NLG expected tonave air power available upon request.

Sl~ortly mfter midnight, 5 July, a cover-ing bomb:llc{lnellt of Ii(Jl~>rrll):lIIX:lraand13airoko by a cruisel-clestroyer force l}egal~on schedule. Pron]pt, retaliatory tire fromenemy shore bat tcries at Enoga i surprisedthe task force, however, because the pres-ence of large g(ms at F.nogai as well asBa iroko hacl not been reported. In a mat-ter of moments, part of the covering tireswas shifted to these. new targets and thebombardment cent inuecl. The destroyert$’trong was the only task force casualty;it was hit at 0046, not by she]lfire but bya torl)edo Iired by a ,Japznese destroyerrunl~irtg along Kolombang:lra’s northeastShore.z The ship sank fast, but n~ost ofthe crew was sa\7ed.

The actual landirrg of the Live.rseclgef~r{)ul)st:lrted about 0130 in the midst of nmtorrential downpouv and spora(lic shell-fire. For a snort time the success of theamphibious vent ure seemed in serious

‘ 310st s{)urres credit this sllccessful torpe(loin~t{) enemy destroyers }vhich were fleeing tl)e taskforce : however, one official po~t~~ilr source as-si~ns credit to a submarine. S:lvI1istDiJ-, Off ofthe CXO, NI), Ulliicll ~tat~.s A’f17,f17C1/lYMt01001/

i)l. lrorkl li’frr II ( lvtishingtou, 1955) , p. 53.

doubt. Rice Anchorage could not beIocat ecl iu tl~e darkness and rain. Thetrallslmrt group slowed and waited-un-comfortably remembering warnings of alllrking enemy submnrine force—while onedestroyer with a s}~eep radar probed ahead,seeking tllc allcllortige. After a short de-1i~~, the l’lm(lakona River mouth WaS 10-

catwl ancl the transport group moved intodebarkat ioI~. positions. ils Marines andsoldiers clambered int,o landing craftalon~side the APDs, enemy star shellsgl immerecl t llrol[gl~ the rainy darkness andsllel IS splnshed among the transports.This fire the, raiders shrugged off with thecomnlent, “erratic and inaccurate,)’ 3 butit was disconcerting, too.

A sh:lllow bar blocking the entrance toRice JInchorage further delayed the land-ing operation. The landing craft, eachtowing 10 men on a rubber raft, wereforced to return to the transports tolighte,rl loads before crossing the reef.Some of the rations were unloaded beforetl~eboxts returned for a second try. Scout-ing reports had termed the beach as“small.” The raiders found this almost anunderstatement to describe the narrowstretch of landing area hacked out of thej(lngle on the south side of the river, about300 yards Ilpstream from the anchorage.While four boats at a.time beached to un-load troops and supplies, the other boatsjammed the river mouth or idled in Rice.fuchorage waiting for a turn to unload.The black night obligingly curtained thenlilling confusion.

.ishore, drenched Marines and soldiersstumbled about the confined beach, slip-ping in the mud and tripping over hiddenbanyan roots. Since enemy shellfireral)ged overhead to hit about 2,000 yards

“1st12dv13n W7arD, P. 1.

122 ISOLATION OF RABAITL

w

+8

T

Ii ....

2

\\\I

IJ m

naz

THE DRAGONS PEA’INSLT..L CA31P-LIGN 123

farther up the coast, Liversedge’s officersdecided that the Japanese at, Enog,ai werenot aware of the exact location of the land-ing ancl ~.isked the Ilse of ]Iooded fl:lsh-

Iights. Thereafter, the unloading and re.-organizzt ion l~roceeded more smooth] y.Near dawn, with almost all suppliesashore, Colonel Liversedge broke. radio si-lence with one uncodecl word, “Scranl.”The anxious .IPI)s and destroyers, unhap-pily expecting enemy retaliation at anymoment, quickly turl~ed and headed backfor the southern Solomons. The IInding,although delayed, hail been accomplishedwithout, serious mishap. One unit, an Ar-my company, w-m taken to the wrong kmd-ing area ; it went, ashore farther llort]lalong the coast. The company rejoinedthe main body later in the day.

The NLG had been welcomed ashore bya ]nixecl greeting committee. Heading alarge group of native guides and carriers-who were obviously frightened and be-wildered by the sudden inilux of so many

white men to their island-were an ikLls-

tralian coastwatcher, Flight Lieutenant J..4. Corrigan, and a Marine raider patrolleader, Captain Clay A. Boyd of Li\--ersedge’s regiment. Corrigan had been onthe island for some time, radioing reportsof enemy activity in Kula Gulf and re-cruit ing a labor force of nearly 200 natives.Small, wiry ll~enwith powerful arm, back,and leg muscles, the native carriers wereto receive one Australian shilling, a stickof trade tobacco, and two bowls of riceand tea each day for carrying ammunitionand rations for the &-LG. A few spokepidgin English, a jargon of simple wordswhich bridged the language barrier.Colorful in cotton “lap laps” wrappedaround their waists, they were intenselyloyal to the coastwatchers.

Boyd had made several scouting trips toNew Georgia. The last time, in mid-June,he ancl his men had remained with coast-watcher Hnrry l~~ickham to direct the1audings at Onaiavisi Entrance and Za-nana Beach before cutting across the islandto 1ink up with Liverseclge. After the ar-rival of the hTLG at Rice. Anchorage; he re-sumecl commanc{ of his company in the 1stRaider Battalion.

On one of his earlier trips, Boyd hadscouted a trail leading from Rice to Eno-gia, and Corrigan’s natives had thenchopped a parallel trail on each side ofthis track. Afi er the N’LCTstacked all ex-cess ammunition, equipment, rations, andblanket rolls in assembly areas preparedby the nmtives, the march to Enogaistarted over these three trails. CompaniesA ancl B of the 1st Raider Battalion (Lieu-tenant Colonel Samuel B. Griffith, II)were assigned to move along the left trail(southern ) with the demolitions plwtoonof the raiders hemding along the right(northern) path. Thus protected on eachflank, the main elements of the NLGstarted along the center trail with the re-maining Marine companies leading andthe two Army battalions following. Twocompanies, M and 1, of Lieutenant ColonelGeorge G. Freer’s 3/145 with a medicaldetachment, communications unit, and theantitank platoon remained behind to guardthe supply dump.

Scouting reports had termed the Rice-Enog~i area an open jungle with small,broken hills and few swamps. Rain-soaked Marines and soldiers, strugglingover the sharp, irregular slopes madetreacherous by the mud and hidden roots,could not agree. The meager trails, hard-ly more than narrow defiles gashed throughthe thick, soclden jungle, were trappedwith sprzwling banyan roots slick with

124 ISOLATION OF RABALTIJ

green moss, coral outcropping, vines, andunderbrush. The rain col~tinned unabtited.

The Army battalions, carrying heavierweapons and n)ore amnl unit ion and gearthan the lightly equipped rwiders, wereforced either to tr:~ve] :~t :1slower pace orto stop to estzbl ish anot]ler sllpp]y dtunp.Tile soldiers, choosing to keep going,pushecl on but dropped steadily behind.Tl~e leading NTLG ~ll~its. heatling deeperinto the hTew Georgia jungle on a courserong-h]y south from Rice, reached the GizaGiza River late in the :~ftcrl~oon nnd setup a perillleter (lefense. Sllort]y :tfterdark, all units were in bivolvw on bothsides of the Giz~ Giza. The SIG esti-mated thxt it had trave]e(l eight l)~ilesduring the first (lay. Actllally, progresshmd been only about five miles, bul umdoubtedly the hardship of j Ilngle travelhad helped give every indication of greaterdistlnce. (See Map 8.)

That ni~ht, men of the h“orthern kld-

ing Group listened to the distant soundsof a naval battle in Kula Gulf. A U.S.cruiser-destroyer force Illd intercepted agroup of 10 ,Tapmnesedestroyers, 7 of themtrmrsport ing reinforcements. In a shortbut, violent action, the 1;.S. force lost thelight cruiser He?ena. The ,Japar,ese losttwo destroyers but managed to ]ancl 850troops at Vila.

.~t daybreak on 6 July, the NLG stirredfrom its wet bivouac znd resllmed windingits way through the dripping jungle. to-warcl Enoxai. The trails chopped by Cor-rigan>s n:ltil~esended abruptly :1(tile river,:~nd the Mtrines were forced to slash theirway thmllgh the mxngrow swamp lyingbetweel~ the Giza Giza and the Tan~akanRivers. Rain cent inue(l to drizzle throughtile jllngle canopy. The battalions becameone thin, stra~~l in,g line sn:lking its waytllrou~h tl}e swal]lp on an indistind trail.

The rains drowned the radio equipment,:~rlcl communication wire laid along thetrail was grounded as the protective cox--eri n~ peeled off in the hands of the in-fantrymen who used the wires as guide-lil~es. Runners (’arrj-il~g fie]d messageskept Liverse(lge in contact, with his base:~tRice.

Tile N’LG lmd divided into two segnlelltsearly that morning. Lieutenant ColonelD[~lbert E. Scl~llltz lmd been directed tot:dm l~is M I;attalion, lMtll Infantry,along allotller trail to tile southwest to clltjtl~e lfllllcl:l-l~:lirol<o road and establisl~ ablock there. Tllc remail~der of tl~e NIJ3cent innwl towar(l t l~e Tama kau River.Captain lhyd, leading Marine (’ompilll~

1), re}tclle(l tile river shortly before noon.JIucI1 io his dism:ay, the smzll stream hehad scol~te(l early in June n-as now a rag-il~g, floo(led tor-rent. The raicle.rs andsol d iers paused while equipment was c al--ried across or ferried on rafts made frombranches and poncl~m. Then the infantry-men be~an crossing the ri~er, single-file,orer z fallen tree which bridged theswollen stream. :~ rope stretched fromb:mk to bank provided an llrrsteady gy~ide-line, and stron~ swimmers fished from theriver those individuals who were unfor-tunate enough to slip and plunge into thewater.

The crossing delayed the lTL(+ until late

in the afternoon. While intermittent, rainscontinued, the LiverSedge force biv-

ouacked for the night of 6–7 July in themidst of a sw:unp. Mudcly and tired, the

raiders and soldiers swallowed canned ra-tions and hndclled ill wet ponchos underbanyan roots, wait ing for dawn.

Late tl~at nigl~t, in znsw-er to a plea fromATGOF headquarters, Li~e.rsedge broke

r:ulio silence to give a progress report. A

THE DRAGONS PENINSULA CAMP.41GN 125

listening watch had been set l~pat i~ll l~alts,but the NLG commander llwl not nswl theradio in the liope that his cross-collntry]narch was still a secret. .-llthollgl~ Liver-scdge carried mwlimmpowered radios( T13X), contact with Hester sonle 20 milesaway was nmde with diffkulty. Such conl-municltion problems were to seriouslyhandicap NLG operations. A lligll-pow-ered r~clio, (leek-loaded ml one of theiiPDs, had not been unloaded dllring tileanxious landing operations and was no]vback at GIladalcnnal — a logistics over-sight, which was to be regret te(l n)any

times.First contact with the enemy came

shortly before noon on the 7th of ,July.Liverse.dge’s wet and hungry men strug-~yledo~lt of t]lc swamp earl~~in the lnort-ing nnd moved along x tortllolls ridge par-alleling the west bank of l?nogai Inlet.The delay calwed by the Tmnalmu and theswamps was emphasized when tl]e soundsof an air strike at Enogai were heard. Thishacl been the designated day for Liver-seclge:s assault of that stjrongpoint. .1 ftermoving through the native village of 3fa -ranusa I without incident, the point pla-toon of the NLG suddenly encounteredseren ,Japanese riflemen. Surprise toboth forces was zpparent, b[lt the Marinesrecovered first. In a brief tight, two ofthe enemy were killed before the rest fled.Uniforms identified the dead as membersof the SNLF, probably from the Kure fit}lat l?nogai.

Realizing that the fight hzd warned theEnogmi gwrison of an attack through theback door, Liversedge increased the speedof the adv:lnce. Griffith was directed totake his raider battalion forward asquickly as possible to take advantage ofan-y remaining element of surprise and to

screen the advance of the rest of the force.The next brush with the enemy came assuddenly m the first. The demolitionsplatoon, meeting a strong enemy patrol,withdrew slightly to high ground and en-g:aged the Japanese in a hot fire fight.Boyd’s Company D then flanked the en-emy and killed 10 before the Japanese fled.Tbo brief fight cost the raiders three killedand fOLlr wounded. By nightfall, Grif-fith’s Ilfxrines hxd occupied the native vil-lage of Triri on Enogai Inlet. Liver-sedge’s (31>w7as set up at Maranusa I withthe hTLG reserve units, Companies K andL of the 145th Infantry. Hasty perime-ters were placed around each village.

The absence of defensive works at Tririfurther convinced the Marines that the,Japanese at Enogai had not been expect-ing nn zttack from the direction of theinlet. The only item of value found atTriri was a detailed enemy map whichpinpointed the location of four 140mmguns at Enogai. As Griffith’s battalionprepared hasty defensive positions, thedocument, was rushed to Liversedge atMaranusa 1. The NLG commander im-mediately radioed for an air strike toknock out tl~ese weapons, but his message

failed to raise a response from eitherh“GOF headquarters or the 43d Division.An Army radio station at Viru, hearingthe request for a relay, accepted the mes-sage for transmission to CornAir NewGeorgia.

Early on the morning of 8 July, Griffithhurried two platoons down divergentpat,hs north and west from Triri to am-bush any enemy probing attacks. TheMarines on the west trail scored first. AJapanese p a t r 01 of near-companystren$th, blundering along the trail with-out advance security, walked into the trap.

126 ISOLATION OF RABAUL

Premature firing, however, spoiled anysurprise effect and the enemy withdrewwithout difficulty. Within a few minutes,a full-scale counterattack had been di-rected at the Marine amblwhing party, andGriffith rushed Boyd’s Company D for-ward to help hold the trail. In the mewn-time, Colonel I,iversedge picked up hiscommand post and the two Army compa-nies and rushed to Triri to be closer to theconflict.

The fight continued for three hours, theclose jungle terrain handicapping the ob-servation and maneuvering of both forces.Company C (Captain John P. Salmon),moving forward to relieve Company T)under fire, broke the deadlock with a60mm mortar barrage and continuous ma-chine gun fire. As the raiders moved for-ward, the enemy disengaged and fled downthe trail. Fifty dead Japanese were leftlittered about the scene of the fight.

The Marines did not pursue. Enogaiwas the first objective. While the Armycompanies took over the defense of Triri,the raider batt alien hastily reorganizedand resumed the march toward Enogai—nlong the north trail where the second am-bush force had set up. The trail, how-ever, ended abruptly in an impassableswamp. Reluctantly, after considerabletime had been spent in trying to find anacceptable trail to Enogai, the battalioncommander ordered the return to Triri foranother start the following day.

Meanwhile, the Japanese force had rein-forced and reorganized for another attackon Triri. Late on the afternoon of the !%h,an estimated 400 Japanese struck quicklyat the left flank of the thin perimeter es-tablished by the two Army companies.The lines of Company K of the 145th slow-ly began to give way under the continuing

pressure of the enemy assaults. CompanyL, on the right, received only scatteredsniper fire. The demolitions platoon ofthe raider battalion, which had remainedbehind with Liversedge’s CP, rushed toassist Company K in its defense just asGriffith’s battalion returned. On ordersfrom Liversedge for a quick counterat-tack, Griffith directed First LieutenantRobert Kennedy’s platoon from CompanyB to circle back and hit the left flank andrear of the Japanese. Kennedy’s counter-move completely surprised and crushedthe enemy’s left flank. The Japanese fledonce more. Another 20 enemy dead wereleft behind. CompanY K, which had threesoldiers wounded, estimated that 75 addi-tional Japanese had been killed in the at-tempted breakthrough. Kennedy’s pla-toon suffered no casualties.

CAPTURE OF EiVOGAI 4

After a quiet night at Triri, the Marinesagain started toward Enogai the follow-ing morning. A radio team with a TBXand headquarters personnel of the raiderregiment rema.ined behind with the Armycompanies, but Liversedge moved out withGriffith’s battalion. The raiders had moreluck this day. A good trail, apparentlyunknown to the Enogai garrison, was dis-covered and rapid progress was made by

the Marines. Sounds of an air strike atEnogai indicated that the request for thedestruction of the enemy guns there was

4Unless otherwise noted, the material in thissection is derived from: i?GOF .Accomzt; NewGeorgia Campaign; I.st RdrRe@ MsgFiJe;1st RdrRcgt SAR; Ist RdrRegt Jnl; IstRdrlln WarD; Griflth ltr; Steuenson ltr;

Pape ltr, OP. cit.; Boyd interview, OP. tit.;Rentz, Marines in tltc Central Solomon*;

Griffith, “Action at Enogai,” op. cit. ; McDevittand Marder, “Capture of Enogai,” op. a’t.

128 ISOLATION OF RABAUL

being executed. Shortly before noon, Le-1and Lagoon was sighted and the Marinebattzlion turned east toward the enemydefenses at Enogai Point. After severalhours of cautions approach, the raiderswere halted by the stutter of two light ma-chine, guns. The Marines pallsed for bat-tle orders. As they waited, the, volume ofenemy fire picked up, The Enogai de-fense line was being reinforced.

The attack WNSmade without mortarpreps.ratioll—Conlpany A (CaptainThomas .i. Mullahey) with its left flankrest ing on thq 1agoon, Salmon’s CompanyC in the, center, and Company B (CaptainEdwin B. Wheeler) on the right flank.Boyd’s Complny D was held in reserve.The frontal assault, made with grenadesan(l machine g{lns, was beaten back. Withthe jungle daylight fast closing into deeptwilight, I.iverwdge called off the, assault.Griffith was told to holcl in place and re-sume the xttack the follon’ing morning.

Tl~e Marines’ defensive positions, facingcommanding grOL~nd)were not to Liver-

sedge’s liking, but the NLG commanderwanted to keep the pressure on the enemyduring the night and decided to risk a.Japanese connterattmk. The gamble paidoff. The night passed without incident ex-cept for the s:ldden crash of a huge, bomb-weakeneci banyan tree in the commandpost area which crushed one raider, in-jured three others, and completelysrnashecl the command’s TBX.

Breakfast on the morning of the 10thwas not a problem for the raiders who hadnot, eaten since the morning of the 9th.There was no food. After a few quiet or-ders from Griffith, the 1st Raider Batt-alion renewed the attack. Wheeler’sCompany B on tile right front reported11oopposition and movecl forward rapidly.Companies .i and C, as expected, however,

were hit by intense fire from rifles andautomatic weapons. The two companiespaused for a 60mm mortar barrage to soft-en the enemy line before plunging on.Company B, ilt last meeting strong defens-ive fire, raced through a small native vil-lage on the inlet’s shore south of 13nogpi.Dead enemy were sprawled throughout thevillage. A number of machine guns weretaken and turned about to put more fireon the fleeing Japanese. The break-through put raiders almost in the rear ofthe enemy lines. Opposition facing Com-pany C in the center abruptly faltered,then scattered.

As enemy resistance began to crumble,the raider zttack gained momentum. Be-Ilind a withering fire of automatic weap-ons ancl machine guns, the raiders movedthrough Enogai. Mortarmen, in positionson the high ground overlooking the village,dropped 60mm mortar shells along theshoreline of ~~LdaGulf, trapping the vil-]:~ge defenders be,tween t,w~ofires. Strag-

glers, attempting to swim across LelandLngoon, were machine gunned by the raid-ers. By earl-y afternoon, the coast defensepositions were in raider hands, and onlytwo small pockets of enemy resistance re-mained. These the Marines contained,postponing mopping-up operations untilthe next dny. Late that afternoon, Com-pmny L of the 145th strl~ggled into Enogai.each soldier carrying rations, ‘bandoleers ofammunition, and three extra canteens ofwater. Without food for more than 30hours, the rniders had been reduced tocat ch ing drinking water in ponchos duringtl~e intermittent rains.

The food was part of an air drop whichthe rear headqllzrters at Triri had receivedearly on the morning of the 10th. Liver-s~dge had requested the drop the previousday. The original three-day sLlpply of ra-

THE I) RAGOX’S PEA’INSTTLA CA?tfPAIGh”

tions carried ashore at Rice Anchoragehad been stretched over five clnys, and

fresh water w:(S n] so scarce. W’olmdecl

were fed worlny rice wllic]~ ]Ild ken

follncl xt Triri. The situntion l~ad become

tense-so serions, in fact, that tile Marineswere fzr more concerned with tile proslwctof cent inllec[ dinlinishec{ rations than theywere with the threat of having anot]lerenenly garrison in their rezr at 13airoko.

Anxiety increased when tl~c planes ap-peared over ‘rriri on schedulr but couldnot locate the purple smoke grenadesmarkinx tile NLG positions. An airliaison officer finally macle contact, with theplal~es ancl (Iirected the air drop. Par~-chutes drifted down, and soldiers al~d Ma-rines dodgecl tl]e welcome ‘~bonlbi ng” toco] lect the blllky packages. The, first con-tainers opened held only mort m shells,and the troops howled their disappoint-ment. K-rations and chocolate bzrs soonfollowed, however. An immediate reliefparty wzs organized to cmry supplies andwater to (lrit%th’s battalion, then hotlyengaged at Enogai.

That, night the Marines clined on K-rations ant{ tJapnnese canned fish? rice, zndsake. ‘1’he capture(l enemy r:~tions l~ere1iherally seasoned with soy sa~~ce foundin several large b:~rrels. Art i~’lesof Japan-ese unifornls were usrcl to rep] nce themuddy and tattered Marine uniforms.The even ing passed without, further activ-ity, the Marines resting easily behind zperimeter defense an~’horecl on Leland La-goon on the right flank and Eno~~i Inleton the left. The tlefeuses faced to~vardIlairoko. During the lligllt, .Japanesebarges were hearcl in I{ul:l Gulf and t,he

raiders scrambled for positions fromwhich to repel an enemy counterlandin.g.The <Japanese barges, however, were only

seck~ng to evacuate stragglers from the~andspit bet~yeen Leland Lagoon and KulaGlllf.

Tile f 01lowing nlorning, mop-up oper-:~tions begxl} with rompanies A and Dmoving ~lli~lil~ tllrollgh the two remaim

il~y points of ol)lmsition, :~ltholl:ll com-

pal]y I) was lmr{l-hit initially. Only x

few ,Ia]mnese were flllslled by the other

patrols, and these the Marines killed quick-

ly. Tl~e 1st R:\iders now owned all of

E]]og:l i Point between Lelnnci I.agoon nndthe inlet. ,JapaJI(vw c:wllct]ties were est,-

nlnted at 350. Tile rziders, in movingfrom ‘Ikiri, h:~d lost 47 killed in action and74 wounded. I?ollr ~t]~ers were n~is~ing

i~nd presllmed dead. The wolmded wereplace(l in :ti(l stations housed in thetllatcllecl llllts at F.nowai.

The fol~r 140mm naval guns, three .50caliber ant iaircraf t guns, and numerous)n:~cl~i~~egllns. rifles, and small mortarswere raptured, in addition to large stocksof ammnnit ion, food, clothing, two trac-tors, and a smrclll igllt. Alliecl bonlbnrd -lnents and bombings l)ad not materiallydxmnged an v of the F.nogai inst :lllxt ions.

Tile ,T:Ipa~~eseretaliated quickly on thenlorniug of tl~e llth with a bombing at-t:wk wllicll lasted for lllore tl~:m an hour:11](I left the Marines with 3 more menkilled a]ld 15 wounded. Three AmericanPBYs }~ere called in tl~at afternoon to

evammte tl~e more serimlsly injurd, andtafter landing at, Rice .fncllorage, the big

flying boats taxied along the shoreline to

Enog:li where (he woundecl were loadedaboard from rllbber rafts. Shortly beforetakeoff, the 1’11Ys were bombed nnd

strafed by t~~o etlenly floatp] znes. The

l~arilles on snore iired everything theyluld at ihe att:~ckers, in(luding smnll arms

and c:~ptmwl weapons, but the .Japanese

130 ISOLATION OF RABAUL

went unscathed and the P13Ys hastily cle-parted for Guadalcanal. On the same af -ternocm, headquart e.rs personnel of I.iver-sedge’s CT arrived at F,nogai and, at 2100,seven kmcling craft from Rice made theinitial sLlpply run into the inlet.

TRAIL L’LQPI{ A (’TZON 5

After split t ing with Liversedge’s mainforce ear]y on the morning of 6 ,J~Lly,I,ieLL-tenaut Colonel Sclll~ltz st:arted his 3d 13at-talion, l-18tlI Infantry clowl~ ii trail whichhis sketch mzl) snowed wol~ld put, him inposition to intercept .Japanese traffic. overthe Mun&-Bairoko trail and establ isk theroad block which Livers@-e had directed.The .Irmy battalion was harclly on its waydown the new trail when one of Corrigzm>snative guides-l ookin y at Schultz’ map—insisted tlmt the map was wrong. TheArmy commander, relaying this informmtion to Liverseclge by field message, re-ported tlmt he was going to press on inthe hope that the trail would cross theMunda-13airoko trail at some. point.

The soldiers moved down the inlandtrail without undue difficulty; the groundwas more rolling and less swampy than inthe coastzl area. Crossing the Tamakauprovecl no problem farther upstream, ancl,late on the afternoon of 7 July, Schultz imformed I,iversedge that he hacl reached atrail jlmction which he believed to be themain Muncla-Bairoko t.rfiil and that, ablock wOL1lC] be establ idled the followingmorning. Footprints on the trail. evi-dence of recent use, convinced Schultzthat he had incleecl reached his objective.

‘ Unless otherwise noted, the material in thissection is derived frmm : NGOF Account: NewG’covqia C’(rmpuiq?! ; Isi Rdrli’cf]t S.IR ; 18t

RdrR@ Jnl: Ist Rdrl?rvt Ms.qFilr; 3/148 Rrpt;

Rentz, lfarinc.~ in the Cvt?trol Solomms.

He ~lso requested that rations be carriedto him, and reported that the native car-riers had become apprehensive and hadreturned to Rice.

The. next morning, 8 ,July, Schultz setILp his road block. Company I defended

the approach from the north and Com-pany K the approach from the south.Company L filled in a thin perimeter be-tween I and K.

The first enemy contact was made short-ly after 1300 when a squad of Japanese,sighted coming down the. trail from thenorth, was taken under fire. The fightwas brief, the enemy quickly fleeing backtoward Bairoko. Two hours later a full-scale attack by 40 to 50 Japanese hit Com-pany I’s outposts, driving them back intothe perimeter, but the enemy did not pene-trate the battalion’s defenses. The after-noon’s engagement cost the Japanese about7 killed and 15 to 20 wounded. One Amer-ican was kill ed and three others woundecl.

After a quiet night, Schultz sent patrolsforward on each trail in an attempt to lo-cate the enemy. No contact was made ineither direct ion, although an abandonedenemy bivouac area was discovered abouttwo miles down the. Munda fork of the

trail. Schultz also tried to contact the169th Regiment, by this time supposed tobe well on its way to Munda field. LTn-known to Schultz, neither force was in

position to make contact. That night,after listening to reports from his patrols,

Schultz reported to Liversedge that hebelieved himself to be about six milesnorth of Munda.

Early on the morning of 10 July, thebattalion was hit on the right flank byabout 50 Japanese and then on the leftflank by a larger force of about 80 men.Both probing attacks were repulsed, the

THE DRAGONS PENINSULA CAMPAIGN 131

Japanese losing 14 killed in the two skir-mishes. After a number of similar semch-ing attacks, the Japanese suddenly un-leashed a strong attack on the right flankat the junction of Companies I and L.The Army positions were quickly overrun,but an orderly withdrawal was made.The Japznese force, estimated at morethan two companies, quickly occupied asmall ridge and set up a number of auto-matic weapons and heavy machine guns.

Company K hurriedly orgmized astrong counterattack with the battalion’sreserves, but the enemy’s hold on tbe riseremained intact. An 81mm mortar bal’-rage, which Scbult z directed to be 1>1acedalong the ridge. kept the enemy from con-tinuing the attack further. The followingday, 11 July, Company K attacked againtoward the ridge, but was driven back. Alater attempt by the same company wasalso repulsed. Casualties, however, inboth attacks were few. That night, Com-pany K was hit in return by a bayonetcharge. The banzai attack was beatenback with only three soldiers beingwounded.

Schwltz’ force, by now just as ill-fed andunkempt as Griffith’s battalion, was on1A rations. The food problem had becomemore acute on the afternoon of the llthwhen Company I of the 145th Infantry

arrived from Triri to reinforce Schultz’battalion. A food drop that same after-noon had been greatly disappointing. AsSchultz had predicted in an early reportto Liversedge, the jungle prevented air-craft, from spotting either flares or colored

panels. Consequently, the air drop waswide of the mark. Schultz’ men, engngedclosely with the enemy, could recover onlya few of the packages, and these contwinedmostly mortar shells. Most of the ammu-

nition was found to be outdated and of thewrong caliber, and nearly all the rationswere spoiled. Little of either could beused.

The next morning, Company I of the145th moved up to the rear of the 148th’spositions and then lungecl forward towardthe ridgeline, following a heavy machinegun and mortar preparation. The positionhad been abandoned. The absence of anydead or wounded enemy indicated that, thewithdrawal had been effected during thenigl~t. The positions around the trailblock were restored, and, with the arrivalof some natives with rations from Triri,the situation began to look brighter. De-fense of this area had cost Schultz 11 killedzud 31 wouncled. ,Japznese caslmltieswere estimated at 150.6

I!?NOGAl : 1%19 JULY 7

Another attempt by the Japanese to re-inforce Vila and Munda through KulaGulf was partially blocked shortly aftermidnight on 12–13 July. An Allied forceof 3 cruisers and 10 destroyers ambushed4 enemy transports escorted by several de-stroyers and a light cruiser. Enemy tor-pedos damaged two LT.S. cruisers, theHonolulu and iSt.Lou~g, and the New Zea-

0Trail block casualties reported by Schnltz in148th Infantry’s .kfter Action Report, quoted inNew Georgia Campaign p. }’–21, totaled 11killed and 29 wounded with 250 estimated Japa-nese casualties. The figures given here are fromthe original day-to-day reports from the trailblock.

‘ l:nless otherwise noted, the material in thissection is deri~ed from : NGOF Account; .New

Georgia Campaign; Ist RdrRegt f3AR; Ist Rdr-Rrgt ,Jnl; 1st RdrRc@ Ms.qFite; Ist RdrBn

[V([rI); ~th l?drl~n Jul&5 Tf’arl); SE Area Nav-() /).s—1 ; SE .lrca XavOp-II; ONI, C’ombaf

Narra t iws X; M{jrison, Brcalcing tlrc BismarchxBarrier; Rentz, ,! fari//cs in the Central S’070mo??s.

132

land cruiser .knder. The U.S. destroyerGwin was sunk, and two others were &un-aged slightly in a collision. Tho Japaneselost only one cruiser in the engagement,and managed to land 1,200 troops on Ko-lornbangara. The battle, ho]vever, per-suaded the Japanese to abandon furtherattempts to run the gantlet of Kula Gulf.Thereafter, the enemy resorted to attemptsto sneak barges through the waters westof Kolombangara. The battle also less-ened the threat of a counte.rlandirrgagainst Liversedge:s force.

At Enogni, the possibility of such anenemy attempt had been considered andthe defenses of the captured village.strengthened and extended. Marinesstrung capturecl barbed wire from EnogaiInlet across tl~e point to I,eland lAgobnaud constrllcted <lefensive positions behindthis line: b~lt.the ,Japanese (li(l l]ot attemptto re~ain tile area. l?nemy bonll)ing at-tacks, too, becanle less frequent.

13nogni bec:lllle tIIe new NT14(icomlnan(lpost. Liversedge clirected that suppliesat Rice Anchorage be moved to the nev-C’P, and, w-ith the exception of a small de-tail, 3/145 moved to Triri. Rice then be-came a rel:ly point wlmre .fPDs anchoredto unload supplies into landing craf t. Thebusy small boats then skirted the shoreline.to Enogai, carrying supplies to the NLGand evacuating wounded on the returntrip.

For some til~le,Liversedge had been con-cerned almut his tactical situation. His

originml orders llacl given him tile dualmission of capt ~lril~g or (Destroying Jap-anese in the Bail. oko-I?nog:li area whilebloc.kil)~ the Mnl~tl:t-13airoko trail, but theclistance between his comrnancl post at En-o~li an(l Schultz’ trail block was too great

for eilective cent rol. A new landing area

ISOLATION OF RABAUL

on the upper reaclms of Errogai Inlet maderesupply and evacuation of the trail blockeasier by eliminating much of the overlandhike, but the combined boat trip and marchstill took considerable tinle. Moreover,General Hester on 9 JLI1~ had insisted that

Liversecl~e keep his battalions within sup-porting distance of each other. So, as soonas Enogai had been captured and a defen-sive perin~eter established at Triri, theilf~rine colonel tllrned his attention to thetrail block where Schllltz’ battalion hadsuddenly found itself facing first a deter-mined enelny of considerable strength andthen no enenly at all.

Following the withdrawal of enemyforces f lom the trw il block area on 12 JLLIY,

no fllrtller Japanese troops had been en-countered. Combat patrols, hitting alol~gthe 111~1~(1:~-Il:~iroliotrail in both direc-tions, failed to make contact. With Mundaun(ler heavy attack, this seemed surpris-ing since it appeared logicnl that the Jap-anese WOUIC1make some attempt to rein-force the airfield. Disturbed by the re-ports from the trail block, Liversedge senthis operations otliccr Lielltel~allt ColonelJoseph J. McCaff cry, to check Schultz’position. Mc(’affery left Tri ri on themorning of the 13th accompanied by partof the reginlel}ta] statf :11)(I the 145t11’sCompany K.

IIc later radioed I.iversedge that the sit-uation at, the trail block was “okay,” andthnt the clefense of the trail m-astight andnot split M had been reported. Rationswere neec{ed baclly, since natives co~~ldnot

carry enolqgh supplies to support the aug-mented trail block force and the frontlines could not be ~veakcned to supply cal--riers. An air drop was requested.

By this time, however, Liversedge was

already en route to the trail block for a

THE DRAGONS PENINSLILA CAMPAIGN 133

perso])al reconnaissance. The NLG com-mander left, Enogai with a small patrol onthe, 15tl~of .July and, after bivouac on theTriri trail, joine(l McCaffery and Schultzearly cm tl~en]orning of the l[; th. (hle dayat tlus defensivo position \vas enough toconvilme Livcrsedge that, the trail blockSllOllld be abandolleci. ~clllllt z’ biLt t:lli Oil,

Ilnal)le to contact the 169th and at a con-

si(lerable distance by boat an(l foot frolntile Sllpply })ase at E])o~ai, was (letinite]yollt 011a sllak-y tactical limb. Jfomover,3/148 W:W i]] a we:dwnd ml)dition, an(lmany sol(liers were ill from eat ill~r con-taminated food. Their :lhility to ward offil sustained :ltt~~li \T:LSqllt?StiOllilble. P\e-sllpply was z problem, too : very little ofthe rat ions droppe(l ~vere recovered. Tlwinten(led purpose of the, trail block seemedto hnve been served:

The presence of our force at the road blocksince 8 .J(lly 1]:](] materi:~lly :lssis ted in tl)e

capture of Fluogai I)y holding enemy formsat 13airoko in position and prewuting themfrom reinfor(>ing their Enogai gnrrison. Itfurther established the fact that the enemywas not using the Bairoko-Munda trail as:L supply route.’

on the morning of the 17th, executingColonel Liversedge’s orders, Schultz di-rected his battalion to abandon the trailblock, ancl the two companies of the 145thRegiment and 3/148 ,retraccd tile Imth toTriri. There the soldiers changed clothes,bathed, and ate a good meal after nearlytwo weeks in the jungle. Their rest was

to be short-lived, though..~t Enogai, the Marines, nom rested and

well-supplied, had been actively patrolling

‘ Ist RdrR@ S’AR, p, 4. The Army’s officialhistory of this campaign notes in rexard to thiscontemporary ju(lgment that “kn(J\vledge ~ainedafter the event indicates that none of these be-liefs was warranted.” Millerj Reduction of Ra-haul, p. 104.

the trails toward Bairoko. Enemy con-tacts after the capture of Enogzi had been1imited to an occasional brush between op-posing patrols, which resulted in brieftire tights with few casualties to eitherside. The rai(lers lost one killed find onewounded cluring the period 13–17 July.tJapznese planes, l~owever, continued tomake Rice iknchorage and Enogai a fa-vored target. Each night enemy float-planes droned over the NLG positions todrop bombs from altitudes of about 500feet. No damage was inflicted, and no cas-ualties resulted.

Patrol reports definitely established thefact that the Japanese intended to defend13airoko Harbor. Several patrols re-l)orted glimpses of ,Japanese working par-ties construct ing emplacements and dig-ging trenches east of the harbor. Two-man scouting teams, attempting, to get asclose to Bairoko as possible, returned withthe il] formation thxt the high ground eastof the ,Japanese positions had not been oc-cllpied by the enemy and that two goodtr~ils leading to this area had been found.The scouts reported that a battalion cou]dreach this position in two and one-halfl~o(lrs. There was still no reliable estimateof the size of the defending force at Bairo-ko, however.

Upon his return from the trail block onthe 17th, Liversedge was greeted with thenews that Lieutenant, Colonel Currin’s 4thRaider Battalion would arrive the next{lay to augment, the NLG. The NLG com-mander had requested this reinforcementshortly after the capture of Enogai. Ma-jor William I). Stevenson, the regiment’scommunication officer, had hitchhiked aride OILone of the PBYs carrying casual-ties out of Enogai on 11 July and had goneto Guaclalcanal to relay Liversedge’s re-quest personally to Admiral Turner. Ten-

134 ISOLATION OF RABALU,

tative approval for the reinforcement wasgiven. After conferring with Currin,Stevenson returned to Enogai with sup-plies and mail?

Early on the morninx of the 18tb, fourAPDs anchored off Euogai Point and the4th Marine Rniders debarked, bringing ad-ditional supplies and ammunition withthem. I.iversedge, who had expected afull bnttxlion, was taken aback when C!ur-rin reportecl his bmttaliou nearly 200 menunde.rstrength. The captures of ViruHarbor and lrangunu, as well as recurringmalaria, had taken their toll. Liversedgeput the NI.G sick and wounded aboard theAPDs to return to Guadalcanal andturned his attention toward the seizure ofBairoko Harbor. The orders were issuedlate that afternoon, after a conferencewith his battalion commanders at theEnogai CP.

Approach to Bairoko was to be madeby two columns. Two full-strength com-panies (B find D) of the 1st Raider Bat-talion and the four companies of the 4thRaider Battalion were to make the maineffort, advancing along the south shore ofI,eland Lagoon straight toward Bairokoand the north flank of the ,Japanese posi-tions. Schultz> bzttalion was to movefrom Triri toward Flairoko to hit the southflank of the Japanese positions. Freer’s3/145 was to remain in ;eserve at Triri andEnogai. The departure time was set for0730, with an air strike scheduled for 0900to precede the actual attack on the harbordefenses.

As soon as LiverSedge’s orders had beengiven, Schultz and Freer returned to Triri,and Currin and Griffith began a last re-connaissance. A reinforced ‘platoon from

Wheeler’s Company B under the com-

‘ Stevenson ltr.

mand of Second Lieutenant William J.Christie moved down the sandspit betweenLeland Lagoon and Kula Gulf to get intoposition for the morning’s attack and toprotect the seaward flnnk. At 1600, an airstrike by 18 scout bombers and 19 torpedobmmbers pounded the east side of BairokoHarbor while 8 mediums strafed Japa-nese supply dumps and bivouac area. Thestrike marked the fourth time since 15,Jllly tlmt 13airoko had been worked overl~y ComAirSols planes.

That night 13nogai was rocked in re-turn by enemy bombing and strafing at-tarks that lasted nearly seven hours. TenMarines were wounded. The NLG won-dered: Had the enemy accurately guessedtl~e date. for the NLG attack or were the,Japanese just giving as good as they hadreceived in the air attacks of the previousdays ? If the former, enemy intelligencework had been much better than theNLG’s.

Alt bough the Liversedge force knew thegeneral location and nature of the Japa-nese defenses at Bairoko, there was a dis-turbing lack of intelligence about the sizeof the ,Japanese garrison. The pre-land-

ing estimate had been about 500 enemyat the harbor. The 350 Japanese en-countered and killed at Triri and Enogaiwere identified as members of the Kure6th 17ATLF. Schultz’ attackers at the road

block had not been identified, but werebelieved to have been from the Bai-roko garrison. The NLG concluded——wrongly—that only about two reinforcedcompanies held Bairoko.

At the time of the Rice Anchorage land-ing, Enogai was lightly defended by a

detachment from Commander Okumura’sK~~re 6th SNLF. When Liversedge’sforce split on the second day, the Japa-

THE DRAGONS PENINSULA CAMPAIGN

nese believed that two regiments were at-

tacking Dragons Peninsula and ordered

half of the %2 E/ntfa?;on. I,?t]t I?pg;menf

from Vi]a to Okumura’s assistance. Thereinforcements included a machine guncompany. The new troops were to havebeen rushed to Enogai to defend the coastdefense guns but the. move was made toolate. By the time the i?d Battalion unitsreached Bziroko, Enogai had been cap-tured. When LiverSedge’s intentions tocontinue the attack toward Bairoko be-came more evident, more reinforcementswere rushed to the harbor. These includedseveral companies of the $2d Battalionl

.&5thl?eg~ment and the 8th Battery of the6th Field Artillery Regiment which hadrecently arrived from Bougainvillea.

Since contact with the enemy had beennegligible after the capture of Enogai, theNLG had no basis for comparison ofstrength and were not aware of the addedenemy capability to defend Bairoko. Pa-trols did not, aggressively test the Japa-nese defenses; in fact? no probing attacksagainst the outposts guarding Bairokowere attempted. The only enemy prisonertaken during this period was a badlyburned pilot, rescued from an offshore is-land and immediately evacuated. In ef-fect, the NLG was facing an unknownquantity in its attack against the harbor.

“Z HAVE COMMITTED

THE WOliK&” 10

The approach to Bairoko by the raidersbegan over trails and terrain now familiarthrough much patrolling. Whee]er’s Com-

‘0 Unless otherwise noted, the material in thissection is derire(l frolll : 3’r/r Cror, qioCampa ifpl;NGOF .4 mount; (k)nlAir%ls Jf-arD, ,Ju143:CornAir New Georgia S.4R, 29Jun–13Aug43;IstRdrltcgt SAR; Ist Rc7rRegt Jnl; Ist RdrRegt

135

pany B led the approach march with Com-pany D ( I1OWcommanded by First Lieu-tenant Frank A. Kemp, Captain Boyd]laving been evzcuate.d with malaria), thedemolitions platoon, Currin’s 4th RaiderBatta]ion, and the regimental commandpost following in column. The two com-panies in the 1st Battalion had beenbrol~ght np to nenr-full strength for theattack by taking men from Companies Aand C. These nnderstrength companiesmnlained behind with the 145>s CompanyL at Eno~ni. (See lfap 9.)

AS the NI.G file moved throngh thedripping jnn~le , scrambling over sharpcoral rocks and climbing low but, steephills and ridges, the Marines waited tohear the first sounds of bombing and straf-

ing which wo~lld indicate that the 0900 airstrike on 13airoko’s defenses was being ex-ecuted. The raiders waited in vain-therewould be no strike.

Unknown to Liversedge, his request, wasapparently made too late. The support

strikes by C’on~.iirSols for 20 ,JLdy werealready scheduled and the planes allotted.The NLG commander, howewr, did not,kno~v this. Considerable difficulty wasenco~ll~tered in transmitting the messmgeon the afternoon of the If)th, but the mes-sage was finally cle:ued. Scheduling ofthe strike was not confirmed, Liversedge.’scommunication officer recalls:

Acknowledgment was requested, as I re-member, but this acknowledgment did notcome until night. It was actually nothingmore than an acknowledgment of the re-

.Vs~Filc; 1st RdrRegt PtlRepts, op. cit.; .jthRdr131& Ju143 TTJarI); Griflth Ur; LtCol AnthonyWalker ltr to CMC, dtd 23Feb51; LtCol EdwinR Wheeler ltr to CMC, dtd 20Mar52; ONI, Com-bat A-orratiues X; Rentz, Marines in the Central

6’olomons.

691–360 O—63—lO

136 ISOLATION OF RABAUL

ceipt of the message by the staff officer onduty at the headquarters addressed.11

Without air support, the odds for suc-cess in capturing 13airoko Iengt hened con-siderably. Disappointed but determined,the two Marine batt al ions kept, movingforward.

The first shots crone shortly after 1015.

A ,Japanese o{~tpost opened fire on the

hTLG column, nnd Wheeler and Kempquickly deployed tl~eir companies into at-tack formation. The outpost was over-

run. Without pause, the raiders contin-ued forwnrd. feeling their wzy through

the tangled jungle. .It 1040, Griffith in-

forn~ed Liversedge by messzge tlmt he wasdeployed ancl pushing forward ag~instseveral machine gulls.

Five minutes later, the raiders were ina violent, all-out battle. A sudden erup-tion of intense and accllrate fire from closerange raged at them. The Marine attack-ers were pinned down, closely pressedagainst banyan roots, logs, and coral out-c.roppings, unable to move against the

“ Stevensonttr. All NLG records, both Marineand Army, indicate that such a strike was ex-pected. ComAirSols and ComAir New Georgiarecords, however, do not reveal any notation ofthe request. This particular incident, which il-lustrates IW.G liaison difficulties with higherechelons, remains nnrf?solved. Evidently a staffofficer at ComAirSols, adhering to a policy thatair support requests had to be received before1600 on the day prior to the date of execution,took no action on the request. The XIV CormG–3 Journal of 19 JUIy cent sins a message fromLiversedge, sent at 2235, 18 July, requesting a12-plane strike on the 19th and a “large striketo stand by for July 20 A M and SBD’S to standby for immediate call remainder of day. ” Corpsheadquarters replied that a “large strike stand-by” for the 20th was “impracticable.” Quotedin .Miller, Reduction of Rabaul, p. 1.30n.

withering fire from automatic weaponsand machine guns which raked the jungle.Recovering qnickly, the Marines returnedthe fire, the battle racket becoming louder.As the intensity of the firing increased, thedin was punctured by hoarse shouts andcurses as the Marines tried t.o maneuveragainst the murderous fire pouring fromthe jungle facing them.

Confronting the raiders was a series oflog and coral bunkers dug into the risingground under banyan roots, and well cam-ouflaged with palm fronds and branches.The ridge ahead blazed with fire fromthese low fortifications. Similar to thoseencountered by the NC~OF in its approachto Munda, the emplacements supportedexch other with lanes of interlocking fire.Further protection was furnished by Jap-anese soldiers in trees overhead who snipedat the Marines with l-Tambu (.25 caliber)light machine guns. Okumura had pre-pared his defenses well.

The pitched battle went on, both sidesfiring at a rapid rate. Wheeler’s company,with its right flank near the end of thelagoon, was unable to move forward andCOU1d not make contact with Christie’splatoon on the sandspit. Heavy firingacross the lagoon indicated that Christie,too, was engaged. Kemp’s company, on

the left, finally regained fire superiority,however, and begwn to inch forward in anattempt to take high ground to the front.

As casualties began to mount in both com-panies, Griffith moved his sole reserve unit—the demolitions platoon under MarineGunner Angus R. Goss—to the left flank

for protection from attacks from that di-rection. At 1105, after 20 minutes of furi-ous combat, Griffith reported to Llvers-e.dge: “Harry: I have committed the

THE DRAGONS 1’ENIAWI;I,A CAMPAIGN137

138 ISOLATION OF RABAUL

works. . . . Movement forward cent inues.Sam.” ‘2

By noon, tile first line of Pllemy resist-ance crumbled, broken under the relel~tlesspressure of the raider units. ITnable touse the 60nml mortars because of the jun-gle canopy, lacking the ne~vflan}e-thro]yerweapons, tnd without air or artillery sup-port., t]le Mari IIW brencheci ~klmlur:~’s cle-fensive ]inc by lmocking out first one pill-box and tl)en anotl~er by demolitions nndover~vlle]ming small-arms fire. BUt losseswere heavy and progress was slow.

Shortly after noon, with the 1st Bat-talion clearly needing quick assistance,Liversedgc con~mitted Currin’s battalionto the fight. Company P (CaptzinWalker) was in close sl~pport behind Grif-fith’s battalion. and thus able to movequickly into the line. Kemp’s CompanyD, which had moved stemdily ahead de-spite numerolls casualties, was receivingheavy fire on its left flank and Walker nowxttacked toward this opposition. Goss’demolition platoon, in turn, circledthrough the rear of the 1st Battalion totake up a new position on Kemp’s rightflank to bridge the gap between CompanyD and Company B.

Walker’s fresh company, uncier ordersto attack southwest to the shores of theinlet before turning north to hit the ene-my;s right flank, was barely able to moveforward before criss-crossing fire fromboth right, and left flanks held it back.Whi]e Walker scouted his front lines to

determine the location of the machineguns facing him, Captain Snell moved hisCompany N into position behind Walker’sunit to refuse the left flank and supportW’alker;s attack. The battle continued in

“ IstRdrRcflt,lJsyFi/e, Griffith to Liversedge,{ltd 20Ju143.

full fury all along the line, the raidergains measured a yarcl at a time. Con-tact, with the platoon on the sandspit stillha(l not been made. Christie’s unit, facinga marshy swanlp backed by a strong lineof Japanese fortifications, COUIC1not ad-vance.. Seven enemy machine guns, pour-ing t deadly fusilade, over the swamp andalong the shores of the sandspit, resistedevery attempt at forward movement.

In the next two hours, the raider attackslowly pllncbed through two different de-fensive lines, uncovering a number ofbunkers on the reverse slopes. CompanyD, riddled with casualties by the heavyand continuous enemy fire, scrambled tothe top of a ridge line which overlookedthe hzrbor at Bairoko, about 500 yardsaway. But between the raiders and theirobjective lay another series of formidablefortifications. Hoping to cement Kemp’sposition on the commanding terrain, Liv-ersedge directed First Lieutenant Ray-mond L. Luckel’s Company O into the gapbetween Company D on the ridge andCompany P. Both companies had beenhit hard by several machine guns in thisarea, and I~uckel’s company was orderedto silence these weapons. k Company O

lunged forward, the maneuver reduced fireon Company P and Company N. Walkerand Snell then moved their companies for-ward to take a small ridgeline to the leftfront.

At this time, the NLG front lines archedin a wide U pointed towards the harborwith Company D as the leading unit. Onthe left flank, Currin had three companies,bent around to the southwest. Griffith’stwo companies and the demolitions pla-toon, on the right, had managecl to movenearly to the end of the lagoon, but aslight gap still existed between the bat-

THE DRAGONS I’ENINSULA CAMPAIGN 139

talion and the lagoon’s shoreline. Livers-edge, in all attempt to plug this gap nnd

try once Illore to contact (?hristie, n~ovedFirst Lieut en:~nt Leonard W. Al ford witha reinforced platoon flom Coll]pa]]y O tothis flank. Al ford’s platoon ll]ade a sl)ir-itecl attack, but the volume of el)emy fireprevented Illoven)ent beyond that ofWheeler% con~pany. The move, howwer,tied Christ ie’s platoon closer to the mainNLG line.

Atj 1445, spora(lic but accurate mortarfire from enemy positio]is on the il]let sLld-

denly changed il)to an intense barrage tl]atshook the attacking lines. ‘rhe Marines,without wmpons for coll]]terl)ntte.l”j7 fire,could only press closer into their shallowpositions hehin(l scant cover on the ridgelines and try to weatl~er the pol]nding.Estimated to be 90mm rou]~clsj the sl~ellsinflicted further casualties, mainly fromtree bursts overhead. The barrage JVaS in~-

mediately followed by a screnming coun-terattack. Kemp’s comp~ny, bearing thebrunt of the enen]y chi~]”~~,was pinnecl be-

tween searing fire fron~ the front and the

mortar shelling. \Yithdra~v ing to the firstridge taken, Kemp orgal~ized a counterat -tack of his own, ancl with a badly deplete(lcompany stormed lx~ck to l]is old positionin a sudden rush. The quick conquest wasthe first visible crzck in the ,Japanese de-fenses. Mnrines reported tile enemy flee-ing, ma~y of them without weapons.Griffith sent x hasty note to I.i~-ersedge,advising the iW,G comrna,n(ler that theaddition of just one company (I. of 3/145)WOU1d take Pmiroko I)y night. Tl~e ,Japa-nese, C~riffltllbelieved, wele on the rim, butcasllalt ies were heavy among the raidersancl rein forcen~ents Ivould be neede(l.

Unfortunately, there were no rezdy re-serve units. Nothing had been l~eard fromthe Army battalion which was supposed to

hit the south flank of the enemy, but soundsof firing from that direction indicated thatSchultz was engaged. I’reer’s battalion,scattered between Rice, Triri, and 13nogziwas not in position to help, even if thosebases coldd have been left unguariie(l.Company L at Enogai had been orderedto the front lines with ammunition, ra-tions, and blood plasma at 1400, but hadnot yet arrived. Liversedge would have totake the Japanese position with the. troopsalready at hand.

Following Company D’s return to itsformer position, the 4th Batti~lion foundmovement easier, and Companies N andP managed to move forward in the face ofstiffening fire to extend the NLG linesmore to the southwest. Bllt the move wascostly: both companies received heavy cas-ualties. Company Q (Captain Lincoln N.Holdzkom), the sole remaining companyas yet lmcommitted, moved up to the rearof the other three 4th Battalion units tobe in position for an attack when directed.

By 1600, the ,Japanese had been p~lslled,still defimt. and dangerous, into m m-eson the l~airoko Harbor headlands mbout300 yarcls wide and 800 yards long. Theirback to the sez, the enemy defel~ciers keptup x sustained and nlurdelolls machinegun and mortar fire that showec] few signsof slackenil~g. In an effort to strike one.last, conclusive blow, Liverseclge orderedCompany Q into the lines. Holclzkom’scompany moved arouncl the left flank ofCompany N in an attack straight into theteeth of hewvy enemy fire. Action alongthe rest of the front line dwindled as thefury of the attack on tile sollth flank in-creased. Now all combnt units had beencommitted; only the demolitions platoonof Currin’s battalion remained as securityfor tile command posts in cwse of an enemybreakthrol@l. l~lleeler’s Conlpany R, re-

140 ISOLATION OF RAFlA12L

questing reinforcements for a last attackjwas told that no help was available.

The outcome of Liversedge’s last at-tempt to take his objective was not long indoubt. Despite the vigor of Company Q’sattack, the overwhelming fire of the enemywon. Badly depleted in a matter of mo-ments, Company Q was forced to retire.Repulsed, the company reeled back, vir-tually noneffective through its losses. Thetactical situal ion had been opportune forone last heavy punch to knock out the en-emy defenders, but without artillery, airw~pport, or other heavy weapons, theraider battalions could not deliver it.

During the, early part of the Marines’attack, (?olone] T.iverswlge he:ircl nothingfrom Schultz, w-11owas supposed to havebit the enemy’s other flank. From hiscommand post just behind the raiders>front, lines, the NIX+ commxnder tried tocontact Schultz by telephone to order reg-istration of the battalion’s 81mm mortarson the harbor’s defenses. The wires, l~o\\--ever, were dead, apparently groundedsomewhere in the relay linking Liversedgeto Enogq~i and then to Schultz. And, inthis crucial moment, tl~e TRXS carried bythe raider regiment failed to rezch eventhe short distance back to Enogai. Cha-grined by the absence of contact with3/148, and desperately needing assistancein his bid to capture 13airoko, Liw+rsedge.at 1345, directed McCxffery to take a smallpatrol and try to contact Schllltz as soonIs possible. For tile operations officer, thisentziled a rugged trip to F.nogai, then wboat ride to Triri, and a subsequent, marchto Schnltz’ position,

The first word Li versedge had fromSchult.zl a field message fronl F.nogai atabout, 1500, was not encouraging:

Harry: Steve [Stevenson] has contactwith Dutch. Dutrh has been hit 3,400 yards

from Triri. Steve told Dutch to keep push-ing and try to connect with our outfit. Ar-tillery fire is falling betweenRice and Triri.LaHue.”

Schultzl battalion had departid Triri onschedule that morning, Company K lead-ing the colurnl) down tile Triri-Bairokotrail. Progress was slow, the heavy ma-chine guns and nlortars carried by the sol-{liers adding to the difficulty of movementover the slippery jungle terra in. By noon,the battalion had rezched a point on thetrail where el~emy posit ions had been re-ported, bl~t tile .Japanese had apparentlyevacuated the area. The only enemy con-tact Jvas a :1 impse of a ,Japanese pat,rolof about, 15 men moving hurriedly downthe trail al~ead of the colt~mn, but no shots\vere tired. ,Short Iy afterwards, however,the chatter of an enemy light machine gun

sent the colnm]) oti the tr<lil. fjevev<tl prol)-ing attacks were made to determine theenemy position, and at 1515 Schultz senta messa~ye to 12nogai for relay to Liver-sedge:

Light Horse Harry : Have met Nips about31~ miles down trail. Have not yet hitMunda-Bairoko trail. Strength of enemyundetermined, but know they have four au-

tomatic weapons. We are attacking. Willkeep you informed as situation develops.They hold high groumi to our front. DutchI)el.”

Schultz then moved his companies intoattack formation al]d ordered a mortarbarrage on the Japanese positions. The

“ Ibid., Capt Foster C. I.aHue to Liversedge,dtd X).Tui43.

“ IIJi/1.. Srbnltz to Liverse{ige, (itd 20.Jni43.I.iverse(lge, ;I fornlw Olympic athlete, was~~idely kn(nyn I)y the nic,knsme used in the mes-WWW. I.ieutenant Colonel Srhultz signed mostof his nless:lges with the name used here, evi-(Ie]]tly a conllling of a nickname and the contrac-tion {jf his .ziven name, Delbert.

THE DRAGONS PENINSULA CAMPAIGN 141

pre-atta.ck bombardment was to start at1600. Enemy strength, Schultz decided,was about one company. Shortly after themortar barrage began, LiverSedge wasable to contact Schultz directly by tele-phone and advise him of the situation thatfaced the Marines on the right flank.Schultz must establish contact with themain positions at Bairoko--and soon—Liversedge told him, or the attack on Bai-roko would fail.

The Army commander, not knowingwhether his present attack would succeed,reported that. he did not think it possiblethat contact with the Marine units couldbe made before nightfall. McCaffery, whohad reached Schultz after the attack hadbeen started, could only urge that Schultzpush forward as rapidly as possible. Thebattalion’s attack carried forward only afew hundred yards before stiffening enemyresistance stopped the advance. Schultzthen ordered his men to dig in and hold the~~ound taken. He had, he fibmred, reacheda position from which he could launch anattack the following morning.

For LiverSedge, Schultz’ failure to at-tack aggressively on the left flank was thefinal blow in a series of sharp disappoint-ments. To his front, the battle din hadsubsided into an uneasy calm broken oc-casionally by the stutter of a machine gunor the sharp report of a rifle. While bothforces—the Japanese compressed into acorner and the Marines clinging tenacious-ly and tiredly to shell-pocked ridges wonthrough sacrifice and courag~waited forthe next move, Liversedge asked Griffithto reconnoiter the front lines and reportwhat action could be taken. Griffith’s rec-ommendation: withdraw.

By this time the Raiders (1st and 4th)had nearly 250 casualties, or about 30 per-cent of the force. We had another 150 men

tied up getting them evacuated to aid sta-

tions and to Enogai. There was nothing todo but pull back to reorganize, re-equip, getsome rest, try to get something to cope withthe Jap !30mm mortars, and get the woundedout.

The decision to pull back was made byHarry the Horse on recommendation fromme after I had talked to Currin and his andmy company commanders and had made apersonal reconnaissance of the front. Harryhad a mission and was understandably loathto abandon it. The final determining factorwas the Japanese capability to reinforcefrom Vila Stanmore during the night bybarge. We were already up against a stonewall, low on ammunition and out of water,and had a responsibility to 200 wounded men.In any case, reorganization was a paramountrequirement. I feel that the decision to with-draw was entirely sound and the only sensi-ble one to have made.n

Victory had been close. At 1630 Grif-fith had joined Kemp on his hard-wonridgeline overlooking Bairoko. The har-bor was about 300 yards away—but stillunattainable. For more than seven hours,the raiders had been in continuous attack,trading punch for punch with the enemyand had almost won. Exhausted andnearly out of ammunition, with almost asmany men wounded as were still fighting,the raiders could only retire, carryingtheir dead and wounded. The positionswon through courage and indomitable willcould not be held during the night be-cause them were no other troops ready topick up the fight. Regretfully, Liversedgeordered the withdrawal of his forces.

The retirement began shortly after 1700.First to leave were the litter cases, about90 in number. Marines from the battalionand regiment al headquarters companiescarried the wounded off the ridgeline incrude stretchers made from folded pon-

142 ISOLATION OF RABAUL

chos and tree branches. The walkingwounded followed, a thin stre~m of lurch-ing, bloody men who had remained in thefight despite injuries. While CompaniesN and P held the main positions. CompanyQ pulled back. Companies O and D dis-engaged next. Despite a continued spatterof enemy mortar and machine gun fire, theretirement was orderly, Marines assistingthe wounded and each other whenever nec-essary. As they moved back, the men sal-vaged weapons and ammunition whichhad been dropped in the fight.

As the abrupt jungle darkness closed in,the rest of the raider companies disen-gaged to retire to the high ground east ofthe end of the lagoon. A rough defensiveperimeter was set up both flanks restingon the lagoon. Company L of 3/145,which arrived at 1800 with badly neededmedical supplies and water, also movedinto the defensive line. Christie’s platoon,pulled back a short distance on the sand-spit, blocked a possible enemy counter-attack from that direction.

After seeing 80 walking wounded startthe long and tortuous night march back toEnogai, the Marines settled down into anuneasy rest in their shallow foxholes.That night Liversedge made another re-quest for air support. To forestall anyswift counteract ack by the Bairoko de-fenders, the NLG commander asked thatthe area between the NLG perimeter andthe harbor be worked over by a bombingand strafing attack the next morning.Liversedge then concluded his request

with: “You are covering our with-drawal.” “

The night of 2&21 July passed withonly one enemy attack to test the hasty

‘= Wevenson ltr.

perimeter. A light Japanese force at-tempted to penetrate the defenses on thewest flank, but was repulsed by Com-panies B and D in a sharp fight thatwounded nine more Marines and killed an-other. Four dead Japanese were foundthe next morning.

At dawn on the 21st, another group ofwalking wounded started toward Enogaiwhere three PBYs waited. The main bodyof the NLG followed, the Marines carry-ing the more seriously wounded men onstretchers. Shortly after the gruelingmarch began, a group of Corrigan’s na-tives appeared to take over the stretcherbearing. Progres was slow and exhaust-ing as the natives and Marines, burdenedwith extra weapons and packs, laboredover the rough terrain. A stopWa,smadeevery 200 yards to rest the wounded andthe carriers. The main body of troopshad gone about halfway to Enogai whenthe Marines were met by Company I,3/145, which had hurried from Triri totake over the rear guard. The roughmarch was further eased when a numberof the wounded were transferred to land-ing craft about halfway down Leland La-goon. After that, the march speeded andby 1400 all troops were within the defen-

sive perimeter at Enogai. Christie’s pla-toon, which retired down the spit, also ar-rived without incident.

Schultz, who had been surprised at theabrupt change of events, had kept his

soldiers on the alert for a morning attack

if a switch in orders came. When theorder for withdrawal was repeated,Schultz turned his battalion around andwithin several hours was back at Triri.

During the march toward Enogai, theMarines had been heartened by the soundsof continuous bombing and strafing attacks

THE DRAGONS PENINSULA CAMPAIGN

at 13airoko. Although Liversedge’s re-quest for air support the night previoushzd been received at 2244, well past therequired deadline for such requests, theCornAir New Georgia headquarters ap-parently read the appeal in the NLG mes-sage and the request was passed to Con~-AirSols. Every available plane, includingsome outmoded scout planes, was divertedto attack the enemy positions at Bairoko.The strikes began at 0950 on the 21st, and1asted until 1710, long after the raidershnd remched the base It F.nogai. In all, 90scout bombers, 84 torpedo bombers, 22medium bombers, and 54 fighter planestook part in the continuous air attack. Atotal of 135 tons of bombs were droppedon enemy positions, and strafing attacksby the mediums started a number of firesin Suppl<ydumps and bivouac areas. Theonly resistance by the ,Japanese was aflight of 17 fighters which attempted to in-tercept the last flight of medium bombers,but was driven off by the Allied fightercover.

Evacuation of the wounded from Eno-gai continued despite attempts by Japa-nese planes to strafe the big, lumberingPBYs which landecl in 13nogai Inlet. Theinterruptions delayed, but did not halt,the removal of wounded for hospitaliza-tion at Guadalcanal. M’ith all the troopsin bivouac at Triri or Enogai, a soberingcount of wounded and dead was made.The lst, Battalion with two companies inthe attack had lost 17 killed and 63

wounded. Currin’s battalion counted 29dead and 137 wounded. In the actionalong the trail south of Bairoko, Schultz

lost 3 killed and 10 wounded.”

“ 1st RdrRegt AS.IR ; Ist RdrBn WD ; ~thRdrBn Ju14.7 lVD; $/148 Rept.

143

The raiders had faced an estimated 30machine guns in coral and log emplace-ments, c]everly camouflaged with narrow,hard-to-detect firing slits. Only 33 enemydead had been counted during the day-longattack, but the evidence of much blood inthe bunkers which had been reduced in-dicated that the Japanese casualties hadbeen considerably higher.

The following day, 22 July, Liversedgereceived orders from Griswold to remainat Enogai and Rice Anchorage. Activepatron ing was to be continued, and theNTGOF was to be apprised of any hostiletroop movement from Bairoko to Munda.Evidently, no further attempt to take thewell-fortified harbor would be made fora while. With these orders, the conflicton Dragons Peninsula settled down to astate of cautious but active watchfulness.

Occasional fire fights flared as oppos-ing patrols blmlped into each other, butclose contact between the two forces wasinfrequent. The .Japanwe reclaimed thehigh ground overlooking Bairoko and re-constructed their fortifications. Evident-ly hoping to keep the NLG off balance,the enemy harassed the Enogai positionsnight ly with bombing attacks by one ormore planes. Some nights the number ofsuch attacks or alerts reached as high asseven. The Allies, meanwhile, poundedBairoko with short-range shelling from

three destroyers on 24 July and bombedthe harbor defenses on 23 and 29 July and2 August. For the most part, however, theoperation reverted to a routine of enervat-ing patrolling and air raid alerts. Of par-ticular benefit was a rest camp establishedby Corrigan’s natives near Rice .~nchor-age where Marines were able to relax for

three days away from the weary monotonyof patrols and air raids.

ISOLATION OF RABAUL144

ElVI) OF A CAiJf PA Z(7LV ‘8

The virtlml stnlemate on Ihwgons Pe-ninsulz ended on 2 .%llgust. The XIV

Corps, poised for x last headlong break-throllgh to Mundfl fiel cl, directed the NLGto rush another blocking force betweenMun& and 13airoko to trap any retreating

enemy. After a hurried night conferencewith his battalion commanders, Liver-se.clgeordered Schu]tz; battalion on a quicknmrch down the Munda-13niroko trailfrom Triri. The 4th Raider Battalion, atRice, ret urned to reserve positions fit

Enogzi and Triri. Schultz’ battalion,leaving Triri on the 3d, moved quick]y pastits old positions abandoned on 17 July toanother trail junction farther southwest.Here he established a road block. On 5.~ugust, as Munda fell, I.iversedge joinedhim with wreinf orcin.g ~roup (CompaniesI and K) from the 145th Infantry and areinforced platoon from each of the two

raider battalions. The first enemy contactcame on 7 August when a patrol fromSchultz> battalion encountered Japanesebuilding a defensive position and killedseven of them.

Contact between the forces capturingMunda and Liversedge’s command wasmade on 9 Auagust when a patrol from the1st Battalion, 27th Infantry, commandedby Lieutenant Colonel Joseph F. Ryneska,appeared at Schultz’ road block.

“ ITnlessotherwise noted, the material in thissection is (lerived from : .Vrw Georgia Campaign ;.VGOI,’ .$cmmnt; 1st RdrR(>qt N.1R; Ist [email protected])![ : Ist I:{?rRrgt :~1.s,qFil~ : Ist li’drI{~t WarI) ;

-4th R(lrI\n Jy:l rI). lug-l:] : f Ith I)rflln TVarD ;

llth DefBn QImrterly A.\ ReI)t, dtd 13Se@3 ; Maj

~I:lr~in D. Girardeau, USA, ltr to CMC, dtdflFeb5i ; Gvifitll ltr ; Rentz, Marines in the Cen-tral Solomons.

The following day, 10 August, on Gris-wold’s orders, operational control of theNLG passed to the 25th Division. Con-trol of Schultz’ battalion passed to the27th Infantry, and Ryneska’s battalionjoined the NLG in Schultz’ place. Leav-ing the road block position to be defendedby Ryneska’s outfit, Liversedge and hisMarine-Army force returned to Triri andEnogai. There the Marines had been ac-tively patrolling to determine if the enemywas preparing to make another deter-mined stand at Bairoko. Heavy bargetraffic, however, and lack of aggressive re-sistance indicated that Bairoko was beingevacuated. Meanwhile, the nightly enemyair raids continued with practically thesaule results as before: “No casualties, nodamage, no sleep.” “

on 9 Augustl a light antiaircraft batteryfrom the 1lth Defense Bat.talien arrivedat Enogai. The 50 Marines with 40mmantiaircraft guns and .50 caliber machineguns were a welcome addition to the base’sdefense. The first night that the batterywas in action: the 40mm guns scored a hiton a surprised Japanese plane which hur-ried away trailing smoke. The gleefulMarines scored the hit as a “probable.”Thereafter, the nightly enemy raidersclimbed considerably higher; and as thealtitude increased, the accuracy of thebombing decreased.

The final assault on Bairoko was madeon 24 August after two regiments of the25th Division (161st and 27th) hadpushed inexorably toward the last Japa-nese stronghold. In the late afternoon of24 August, R-yneska-whose battalion hadadvanced steadily toward the objective onthe Munda-Bairoko trail—sent a messageto Liversedge that he was one hour’s

THE DRAG(3XS PEA’13WULA CAMPAIGN

mnrrh fronl the smlthern end of the harhor

and that llQ was ~oiug into Elairoko on the

followil~,c n~ol.nin~ “come hell or high wr-

iter.” 3(’ Ryneska’s message was followed

by another nlessa~e from the 3cl Battalion,

l$5th (llo~lc ollllll;lllclecll ly~~ajot.~farlin

D. ~rirardeatl) which l]md advanced fromEnogai over tile raiders’ routeof 20,Tuly.ii comp:lny from that bnttalion reportedthat. it ll:lclelltered 13airol<o lTitllollt oppo-sition. The harbor had been evacuated.(’omposite raider companies, forn~ed fronlthe effective members of each battalion,were in reserve at Triri ancl Enogai butm-em not Ilcedwl for tl~e tin:~]phases. Thelon~ fight for 13airoko was over.

On the 28th, General Collins, command-ing the 25th Division, arrived at Enogainnd after an appraisal of the situation or-dered the Marines withdrawn. Thatnight and em]v on the 29th of Augnst, theraiders went ~board APDs. By 1130 on

the 30th, the raiclers were back at Guzdal-

cnllal. Tl~e last entry in the 1st RaiderRegiment ,Journal, at midnight of 31 .lu-gust 1943, is significantly eloquent: ‘{lstMarine Rai cler Re~iment relaxes (bunks,movies, beer, C]1OW) .>! 21

The Marine raicler battalions which re-turnecl to Guadalcaual w-ere a pale shadowof the two units which had originally beenassigned to the NI~G. illa]nutrition, UU-

avoidab]y poor sanitary conclitions, expo-sure, fatigue, and continued loss of sleepancl malaria had taken their toll. Battlecasualties had been unexpectedly high-25 percent of the total command of the 1stBattalion, 27 percent of tile 4th. Griffith’sbattalion had lost 5 officers killed and 9woundecl, with 69 enlisted n~en killed and

““ l.st RdrIfc~l .Vs~FiTr, Ryneska to Liversedge,dt(l 24.\ug-13.

“ 1stRdrl?cgt J)/l.

145

130 wounded. Cnrrin’s battalion, in threeoperations ( ~Tiru,Wickham, and Ilairoko)ha~l 2 officers killed and 8 wounded, 52enlisted nleu killed and another 160~vollllded. (>f tile 591 men remaining intl~e 1st Batt alien, only 245 were judgede (~ective by batt aliou mediczl officers.Ollly 154 M:~rines out of the 412 officersand men in the 4th Battalion could beCltl~S~d as efiwtive. The cloctors con-cluded that furi l]er commitment to combatat this time N-xsimpossible:

Not more than fifty percent of the prf.s~ntpersonnel \vould& able to move out on an):lrch Ivitl]out extreme exh:lustion and ofthese. the llndernliuin~of ljh~sicaland nerV-ous stamina has ljeen so ~reat aS to rendernone of them (,:1lu~l)leof exerting sixty per-wut of tkeir usual offensive effectiveness.n

COA’(’LIz7islolvhs

Th~ contributions of the NLG to theeventlull snccess of the hTew Georgia cam-lJaign fippe~r slight< in a post-operationalreview, The trail block, as originally sit-uated, lost all surprise value and useful-ness after one engagement. The Japanesedid not, contest its presence further, andsimply moved reinforcements to Mundaover another route. As later reconnais-sance proved, the actual location of thetrail block should have been another 1,200

yards farther southwest at the junction ofthe main Munda-Bairoko trail.

Liversedge’s force, in attacks on Enogai

and Bairoko, inflicted a large number ofcasualties on the enemy and forced the

Japanese to commit additional troops tothe Dragons Peninsula area — troopswhich tl~e enemy could have used to ad-vantage in the defense of Munda. This,

2’InformalRept, Bn Surgeons to CO, 1st RdrRegt, dtd 8.kug43.

146 ISOLATION OF RABAUL

perhaps, was the principal lwnefit derivedfrom the NI,G’s operations at Enogai mndBairoko.

The failure of the attack on 13airokocan be ascribed to the burden of handicapsunder which the hTLG labored-lack of in-telligence, poor communications, the vitalneed for supporting air and artillery, andinsufficient support from higher echelons.Each handicap, in its turn, contributed tothe eventual failure.

operational planning was lmndicappedby the failure of the NGOF in makingmaps, mosaics, and aerial photogmphsavailable to the NLG prior to the landing.Other than the operational mosaic, theLiverseclge force received only one. high-Ievel stereographic set of prints of Bai-roko~ which revealed nothing. And, asI,iversedge later pointed out, no provisionwas nuade for the NLG to receive furtherintelligence.z3

Realistic estimates as to enemy strengthancl reinforcement capabilities were hLck-ing. On a par with the assumption thatMunda w-ould be captured in a matter ofdays was the equally poor reasoning thatthe Japanese would not stoutly defendagainst an attack on their major port ofentry into N“ew Georgia. Pre-attack pa-trolling by the Marine xnil Army battal-ions was extensive but, M Liversedge ad-mitted, not aggressive enough to force theenemy to reveal the addecl strength of theBairoko defenses.

The serious disadvantage imposed bycommunication failures in the drippingjungle ba] keel the operat ion constantly.Contact with NGOF headquarters at Ren-dova was difficult, and S1.G messages usu-ally had to be relayed by a variety ofstations, including those at Segi and

“ Ist RdrRty)t SAR, p, 17,

~LUldCil C:lll~]. Not zI1 the communicationswoes were equipment failures, however. Insome instances, transmission of messageswas refused. After Enogai was captured,I,iversedge reported, permission to trans-111it three n rgent messages to the NGOFw-as not grznted, and the NL~X was di-rected to Clelr the message with znotherstation, unknown to the NI,CX. The urgentmessages to the NGOF were fin.dly clearednfter 15 hours of waiting.zl

The zttack on Bziroko, started and con-tinlled withmlt, air bombardment, the onlysupport in~ weapon available to the NLG,raises questions which existing records donot answer. Since his request for air prep-arat ion on the objective had apparentlylwen rejected and there was no assurancethat another req(~est would be honored,Liversed~e undoubtedly believed that a

higher echelon had deemed air supportunnecessary for the zttack. As the next,day, 21 JLlly, was to prove, however, airsupport — and lots of it — was available.The only restriction, apparently, was thatrequests hfid to reach the headquarters ofComAirSols on Guadalcanal before theend of the working day.

Another question unanswered was thecomplete absence of any supporting ar-tillery. .&]though it, would have been im-possible to pull artillery pieces through the

jungle from Rice to Enogai, there. seemsto be no reason why artillery could nothave been unloaded. at Enogai after thatvillage wm captured. It is believed thatone battalion of lt)5mm howitzers couldhave been spfired from the many battalionsthen at ~~unda. Based at Enogai, theseguns would have made a vast difference inthe attack on Bairoko.

‘4Ibid., pp. 1>13.

THE DRAGONS PENINSUI,A CAMPAIGN 147

hTaval glmfire support, m known lwter inthe war, was at this point in mid-19+3 stillin tile exploratory st:l<ges: “reliable, fool-proof communications and the clevelop-rnent of glmnery techniques for the cle-1ivery of accurate, indirect tire from afloatonto llnseen targets>>‘s :Isllore had not beenfully worked ollt yet. .\s }wfore, recordsdo not indicate the IPasoIls why Alliedplznners waitecl llntil after the repnlse atB;li roko to plaster thilt enenly point withair and naval hombnr(lrnents.

Although the Marine battalions wereforced to admit failure in taking the as-signed objective of Bairoko, the seven-holm attack by men armed with only gre-nades, rifles, znd light, machine guns ‘cagainst all ellellly of near eql!a] nllrnerical

streltgth barricaded ill heavily fortified

bunkers st an(ls as one of the finest exanl-

ples of personal courage in Blnrine annals.

It is to the laiders’ credit that victory over

these overwhelming odds was at one point

very nearly in their grasp. ~rhetller the

“ Col Robert. D. Heinl. Jr., ltr to Head, HistBr,G–3, HQMC, dtd 3M:uW?.

‘c 31:1jor Gf.ller:~l Rolx?rt S. 13eightler, who com-m:lnde(l the 37th Infnntry I)iyis ion at hTew(lef)r~i:l, nf)tecl th:tt he h:ld ‘{perso]l:illy urged”

(“( IIow1 Liversdxe “to a(lequately eqnip the Ma-riue Imttalions \vith heavy nutom atie weapons”before they left Gl]n<lalcannl. 31aj Gen Robert S.l~ei~htler. I“S.i, ltr to .\(:ofS, G–3. IIQMG, dtfl1511ec60. hereafter Bciflh tlcr [tr.

harbor could have been taken by more ag-gressive action by tile 3d Battalion, 148th1nfantry is pure conjecture. The recordsindicate that fiction of the left, fknk wasnot coordirtatecl with the raider attack, andthat apparently tile urgency of the situa-tion was not realized by Schultz. Why the3d I+~ttalion, 145th 1nfantry was neverused except, as a support force and notcommitted to combat is al~other questionwhich was unanswered in reports of theaction. Equally ptLzzling is the fact thatthe Army battnlion’s 81mnl mortars werenot employed to support the raiders’attack.

In nny event, an evaluation of the Drag-ons Peninsl~la campaign does not discreditthe troops and their leaders who foughttl~ere. Rather than being remembered forfailllre, the- Dragons Peninsula operationand the attack on Bairoko in particularare a testimonial to the personal courageof the Northern Landing Group, whichachieved at leaSt P:lt’ti:k] SLICWSS, a]thOLl@

almost l~opelessly lmndicapped by innu-merable shortcomings in the initial phrn-ning and in tile support subsequently re-ceived. Falllty intelligence which under-estimated the enemy, fanlty task organizat-ion wl~ich n~glected the incll~sion of re-qui red fight illg elements. and somethinglCSS than full support, by higher headqnar-tm-s are the m:lin shortcomings whichann]ysis reveals.

CHAPTER 5

End of a Campaign

In a matter of days after its seizure bythe hTGOF, the airfield at Mnnda-shell-craterecl, with stripped and fire-blackenedpahn stllmps ollt ]in ing the ru]lways-wasconverted to :~n illlied b:~se for furtheroperations ill the Central Solon )ons. *il-most as soon as enemy resist allce aroundthe airfield was ended, tl]e busy bulldozersof the hTxvy>sconstruction battalions weresmoothing the coral landing strips ancl re-pai]ing revetments for use by ConlAirSolsp]tnes. As the 25th Division turned northto follow the enemy’s wit hdrawa] towardBairoko, the 43d Division took over de-fense of the airfield and began mop-up op-erations on the offshore islands.

Separated from New Georgia by onlya few yords of shallow water> Ilaanga Is-land north of Muncht Point was a. ready-

‘ ~nl~ss otherwise noted, the material in thiswction is doriwd from : CinC’Pac–CinCPO.\Rept of 01}s, PO.L, for Sep43, dtd 17Dw43 :PO]*[ii){’11 l:r~,t of 8oPac .1 ctif,$l; (70msopac

.4?1g–fJcp43 WnrI)s; NGOF Account; ~3d In f-Dil. Hi.st; 9th DefILn WarD, .\ug-Se1A3 : Iltl/I)efBn lt’arl); 9th I)rfBn Tank OTIS; loth I)efBnRept on Action of the TkPlnt in the N’ew GeorgiaCampaign, dtrf 3Se1A3 ; -L3d InfI)i~Rept to (2G,XIV Corps, Employment of Tanks in ArnndelOps, dtd 23 Srp43 ; SE .-1rm 37(/uOps-11; S’mxN-

teenth .1 rm?l OIM—Z; ,JICPOA Item No. 1973,Translation of Captured J~pan(?Se Document, dtd2Z>-OT-4El:CIIC SoPacFor Item Nos. 799 and 814,

dtd 260ct43, Translation of Captured ,JapantweDocuments; New Gcorgic Contpaig}$; Karoleyitz,2~flI InfDirIIi.~t; Rentz, .l[ari?/cs in the Crn-tral solonlons; Zimmer, ~37’,s Historq.

made sanctuary for Japanese fleeing thebigger island. As such, the densely -wooded appendage was a stepping stonealong the ,Japfinese route of retreat. Theoriginal island garrison h~d been small—about 100 ~irmy xnd Navy troops—but thegeneral exodus from Munda swelled thepopulation. Tag ends and remnants ofMunda’s defenders fled to the islancl eitherto go olrer]allc~ to\\-ard Arundel or await

evacuation by barge. (See Map II, Map

,Sectio l].)On 11 .iugust, as the 43d Division

widened its cleanup efforts around the air-field, a p:~trol confirmed reports of Japa-nese activity on 13aanga. The followingday, a company-sized ll~lit moved by lanCl-ing craft to the island. As the soldiersdisembarked, n withering fire from thejungle fe.]led abollt half of the force andforced its withdrawal. Two days later,while an artillery barrage from 155mmguns hastily emplaced at Munda paved theway, two battalions of the 169th made anunopposed dawn landing on the shore

opposite the site of the ill-fnted assault ofthe 12th. As the infantrymen moved in-

land, crossing the island from east to west,resistance stiff enecl. An estimated 400Japanese manned a strong line of hastily-bl~ilt fortific:tt ions blocking the advance.

On 16 August, two battalions of the 172dRe~iment wel~t to 13aanga to reinforce theattack. ~b n]ore artillery units ( illClLld-

ing tile 155mIli gun batteries of the 9th

Marine l)efense Ratta]ion) moved into po-

148

END OF A CAMPAIGN 14!)

sition at Munda and on the offshore. is-Iands, and syst emat ica}ly knocked outevery known enemy gun emplacement, re-sist~nce dwindled. Il~creascd barge trafficon the night of 19 A~qy~st indicated that,the ,Japauese were withdrzwinx. The fol-lowing day, the southern part of the islandwas quickly occi~pied, and two battalionsthen moved north along opposite coast-lines. Only scattered stragglers were enc-ountered; the enemy had abandoned13aanga. The 43d Division lost 52 nwnkilled ancl 110 wounded in the week-longbattle?

Contact with the Japanese was reestab-lished on Arundel. One of the smallest ofthe major islands in the group and virtu-ally unoccupied by the Japanese except asa barge staging base, Arundel was withineasy distance of both New Georgia andKolombangarw Its eastern shore bor-dered Hathorn Sound and its northernfringe of n~rrow reef islands was just1,200 yards from Kolombangara-a stra-tegic position which became increasinglyimportant to both forces. For the Jap-xnese, the island was mnimportant outpostto Kolombangara nnd an invaluable evac-uation point. The NGOF wanted the is-land because Aruudel in Allied handswould bring Vila airfield within range ofartillery. (See Map 10.)

On 27 lfLlgUSt, troops from the 172d In-fantry crossed Diamond Narrows fromhTewGeorgia and landed unopposed on thesoutheastern tip. After securing the south-ern part of the islznd~ the landing forcesplit into two reinforced companies to be-gin extended patrol action north along theeast and west coastlines of Arundel.

As on New Georgia, thb dense jungleand large mangrove swamps made travel

difficult,. First, enemy contact was madeby the east shore patrol on 1 Septembersollth of Stima Lagoon. Pushing on, thepatrol fought its ~yit~ through brief skir-n~ishes and delaying actions without trou-ble. To Ilelp cut off the retreating enemy,the Qd Battalion of the 172d establishedx beachl~ead near the lagoon and rein-forced the eastern patrol. Meanwhile, the1st Rxttal ion moved by I.CMS through~Vana ~V:LnnI,a,goon to link up with thewestern patrol which had reached Bus-tling Point on the northwest coast w’ith-out so much as seeing an enemy soldier.The beachhead on that coast was then ex-pancled to include the extreme westernencl of Bomboe Peninsula.

When the 2d Battalion’s attack nearSt ima Lagoon on 5 September was ab-rl~ptly halt ecl by fierce enemy resistance,the 3d Batt al ion was landed to reinforcethe effort. Neither battalion, however,was able to penetrate the enemy’s strongline of defense which included mine fieldsand booby traps as well as many machineglms. Art illery fire from Kolombangarasupported the defense. The arrival of the1st Battalion from Bustling Point, where abattalion of the 169th hacl assumed re-sponsibility for the western beachhead,placed the entire 172d Infantry on the eastcoast and paved the way for the commit-ment of the 27th Regiment. (25th Divi-sion) on Bomboe Peninsula. Two bat-teries of 155mm howitzers and a 4.2-inchchemicxl mortar company also landed atBustling Point to support the 27th Regi-ment on that coast, while, XTGOF artilleryon New Georgia emplaced on the shores ofHathorn Sound delivered counterbatteryfire on Kolombangara to support the 172d’sattack on the east coast. Of the two in-fautry regiments, however, only the 27thInfantry was relatively fresh, although its

150 ISOLATION OF RABAUL

rifle companies were =riously under-strength and its men “well seeded withmalaria.’? 3 The 172d had been throughnearly two months of arduous fighting andT1-ilSl):I(llJ’ Illlderstreugth.

While troops from the. 169th held theBustling Point area, the 27th Infantry on12 September openecl a drive east alongthe length of 130mboe Peninsula. Thelezcling battalion, restricted to a narrowstrip of island only 400 yards wide andunable to make a flanking attack, couldonly grind straight ahead when it ran into

stiff opposition. Small gains with mount-ing casl~alties were the inevitable results.

As the front lines inched abreast ofSngekarasa Island, which parallels 130m-boe Peninsula, a second batt alien swamand waded across a lagoon to establish an-other front on that island. Unable to

erase the beachhead in a series of scream-ing Counterattacks that night, the Japa-

nese then hurriedly evacuated their barge

base on the extreme western tip of the is-land. Souncls of barge traffic each night,however, indicated that the enemy stillhad other bases on Stima Peninsula whichcould be used to resupply and reinforcethe Arundel defenders.

By dusk of 14 September, the two bat-talions of the 27th were in secure positions

astride Sa~ekarasa Island and Bomboe

Peninsula while the 172d Infantry pressedslowly northward along the east coast. In

the gap between, stragglers from the %i?9thand a battalion from Tomonari’s 13thRegiment fought determinedly to holdStima Peninsula and a corner of Arundel’snortheastern coast.

‘ MajGen l~illi:un W. Dick, Jr., USA, ltr to.\ClofS, G–3, HQMC, dtd 310(.t60, including com-ments by 31aj Gen David H. Buchanan, USA.

On the night of 14-15 September, theremaining battalions of the 13th Regi-ment on Kolombangara were loaded onbarges for transfer to Arundel to begin acounteroffensive which was supposed toregain the initiative in the Central SOlo-mons. Undaunted by the loss of ColonelTomonari and two battalion commanderswho were killed by American artillery fireas their barge beached on the Arundelcoast., the Japanese unleashed a near-fanatical attempt to break out of the pe-rimet er. The desperation thrust failed.The lT2d and 27th, reinforced quickly online by troops from the 169th, containedthe attack although the battle was touch-and-go for some time. As the attack sub-sided, the ,Japanese reverted once more todelay ins tactics to preserve their thin foot-hold on Arunde]. The repulse decided thetJapanese upon withdrawal from Arundeland eventual evacuation of the CentralSolomons.

The counterattack, however, resulted inMarine Corps tanks joining the 43d Divi-sion. Alerted earlier for possible commit-ment, the tank platoons of the 9th, 10th,and llth Defense Battalions moved theirremaining 13 serviceable machines byLCM from Munda to Bomboe Peninsulaon the 16th. While the tanks of the 9thand 10th went into bivouac, five tanks ofthe llth Defense Battalion moved up tohelp the 27th Regiment in the BomboePeninsula area. The armored attack on

17 September took the Japanese by com-plete surprise. The heavy jungle rains ap-

parently drowned the noise of the tanksclanking into attack position. Moving

forward in two waves with infantrymenfollowing, the Marine tanks crunchedthrough the enemy defenses before

abruptly turning to the left in a flanking

END OF A CAMPAIGN 151

0000

00.3.

c

ccc.

c*

.

c

.

.Q

CLa>

691–360 0—63—11

152 ISOLATION OF RABAUL

manellver to complete the rout of enemyin that sector. Infantry units advancedabout, 500 ynrds in the attack. The fol-lowing day, however, as four tanks andan infantry company jumped forward inanother assault, the enemy suddenlyopened poin-hlfink fire with 37mm anti-tank guns. Two of the 1lth Defense Bat-talion tanks were knocked out, of action,but quick and effective covering fire by thein fxntry allowed the tank crewmen to es-cape. The attack stalled.

On 19 Septemberj the remaining effec-tive tanks-two from the, 9th, four of the10th—joined those of the llth. Lined upin two ranks virtually trei~d to trexd they

started toward the enemy lines. The rear

rfink coverecl the front with fire. Concen-trated blasts of 37 mm cfinister roundsand bursts of machine ~wn fire from theleading txnks withered the jungle ahead,

stripping foliage from the enemy posi-tions and hewing out an avenue of attack.Behind this shield of firepower, the infan-try advanced rapidly. Afterwards de-scribed by 27th Infantry officers as one ofthe finest examples of tank-infantry co-ordination they lmd seen, the attack movedquickly and steaclily forward.

This fenrsome mass assault, coupledwith the Japanese decision to quit .4run-del, settled the fight for the island. Thatnight, despite near-continuous artilleryand mortar barrages, Japanese barges be-g-an evacuating the bulk of Arundel’s de-fenders. While enemy artillery fire fromKolornbangara kept the two Americanregiments from closing in, the remainderof the l~th Regiment was withdrawn the

next night. On 21 September, with only afew overlooked stragglers to contend with,the NGOF declared Arundel secured.

Instead of being a routine mopping-upjob, the fight for Arundel had unexpect-edly developed into a major operationwb]ch required the principal elements of1hree infant ry regiments as well as ar-mored and artillery support. <Japaneselosses in three weeks of fighting were 345counted dead, although the enemy musthave lost considerably more. Countlessshallow graves dotted Arundel’s northerncoast, and the Izgoons and Blackett Straityielded many other bodies of enemy deadwho had been killed in evacuation at-tempts or had drowned attempting toswim to Kolornbangara.

Allied losses for the island’s capturewere relatively light, 44 killed and 256wonnded. Army observers credited thetimely support of Marine Corps tanks forabruptly terminating the campaibm andpreventing the loss of additional Allied1ives.

With Munda taken and the Allied driveslowly turning toward Vila airfield, the~Japanese in mid- Au=gyst had every rightto expect that the decisive battle in theCentral Solornons would be fought on thebig, volcanic island of Kolombangara.But Admiral Halsey, a former NavalAcademy halfback, knew the value of an

end run in warfare as well as in football.Ten days after Munda was captured, theAllies skirted the strongly defended posi-tions prepared by the enemy on Kolom-bangara and hit at lightly-held VellaLavella.

4 Unless otherwise noted, the material in thissection is derived from : CinCPac Rept, Sep43,op. cit. ; CinCPac-CinCPOA Rept of Ops, 0ct43,(itd 20J:11114; ~ol)~iN(~h Rc’pt of 80~flC .iCtiO?L ;

~ootSoI’oc 4 u~-ScpJ3 WarDs; Corn III Phibi?or

END OF .4 CAMPAIGN 153

The decision to switch targets was madea month earlier. On 12 July, just six daysafter asking for Admiral Turner’s plansfor Kolombangara, Halsey changed hismind and directed that this island be side-stepped and J-ells Lave]la taken instead.By this time it, was obvious to the staff ofComSoPa{! that Munda was not, going tobe taken as quickly as estimated and thatthe island of Kolombangara, with nearly10,000 entrenched defenders, would beeven harder to take. Further, Vila air-field was reported to be poorly (lrained andpoorly situated. If a better airfield sitecould be found, the chance to land virtu-ally unopposed at Vella Lave]la would bea much sounder tactical moves

A reconnaissance tealll wllicll scoutedthe island in late July returned to reportthat the sm~thern end of the island nearBarakoma drained sufficiently well to en-able construction of an airfield there, andthat there were adequate b~iicl~es, bivolmc

AR, 10-19 Au$343, dtd 20Dec43 : COIU III PhibForRept of Occupation of Vellm I,avella, 12Aug-3Sept43, dtd 20 Sep43: CTF 31 OPO A12+3, dtd1lAug43 : 3“r?e Gcorf/ia ~(~mpai~?? ; NGOF Ac-rount ; A-IF FO h-o. 1, dtd llAu~43 ; ~WCYttC(’nth

.Irmy Ops-I: SE Arm XfJv Op,s-11: Frankel,J7’fh I)lfDirHist : 0. .k Gillespie, Th c O~riatHi8tory of New Zealand in the Second WorldU’ar—T’h c I’ac’ific ( Wellington : War HistoryBranch, Dep:nrtment of Interrml Affairs, 1957).hereafter Gillespie, Ncw Zrala Jtd History ; Hal-sey and B~~iln, Hu7sry’s Story ; Karolevitz, 25thIttfI)i!,Hist: 310rison, Brc({l;itiq tllcBix>))arcksBarrier; 0S1, L’ojjfbat A-arratities, The S’olomotl

Is7a)lds Campfl iqn: .YI—I{070t)lban~/(!ra [J}ld T-(’l[a

L/[rel[a. 6 .Illgt(s-i C)ctolwr 1943 (Washington.1944), hereafter OXI, t70ntbat 7Vf/rnztiwx XI:Rentz, .lforinm i)~ tl{c Cv~t)//t S(~70mon.s ; CSSBS,(’awpflig)t<s.

5 This is not the first instance of successful1).vlla.ss strate~~.r. The :I]ll]]hil}ious forte \\-hif.hlanded in the Alellt ians in lli~~ lfM3 took Attubefore forcing the eva{mation by the enemy ofstrongly-held Kiska Island.

:lrews, :lnd MTIl anchorakres in the area.l’ell:~ Lavella, the pi~trol reported, differed1ittle from New Georgia. A dense jlmgleof tangled creepers and huge trees coveredtlie island from comt line to the low butsharp mountain peaks in the interior. Oneof tlle most developed islands in the groupbefore the w:lr, Vella Lavella’s l+;uropean-lype buildil~gs included a hospital, severalmissions, and tt Ieprosariurn. (See Maplo.)

Coast watchers on the island added to thereport. Chly about 250 ,Japanese wereestimated to be, occl~pying the northernpart of the island where Jrella Lavella’s ir-regular coastline provided many coves forI)rotect ion for barges shut t1ing betweenKolon~ban~ara and Bougainvillea. The na-tives on the island had remained friendl~;to the Allies, and were well orgmized.They had, in fact, aidecl the man-y survi-vors of the USS FIe7ena who had managed

to swim to the island and had assisted intheir evacuation by fast APDs on 16 July.’

on 11 August, orders were issued byCon~SoPac for the seizure of Vella Lavellaby Admiral Wilkinson’s Task Force 31.The forces on New Georgia were directedto continue the cleanup operations in theMunda area and to interdict T-ila airfield

by artillery fire. The Allies had decidedtImt enemy troop concentrations on Ko-Ioml.)angara did not necessitate an attack,that neutralization of the island would beas effective :W occupation and not as costlyin terms of troop casualties or supplies.

“ 1’. S. ships ~Jicked up 93 men and 11 officersIt one point and 39 men and 2 officers at another~]oint on the island. Natives had protected andfed both groups. The rescue of these survivorsfrom an enemy-held island in enemy waters wasa distinct boost to morale to all South Pacificforces.

154

Further, soggy Vil a airfield was no longerdeemed worthy of capture.

The Xortl~er~~ Landing Force (NLF),7organized to attack and occupy Vella La-~-ells, con~priwd the .~rrny’s 35th Regi-mental Combat Team which included the64th Fiel(l Artillery Battalion, the 58thNaT7al Construction Battalion, and the Ma-rine Ah Ilefense Battalioll! as well as M-ditional ;irmy and N:Lvy support. units.13rigadier General Robert R. lf,’Clllre, the25th Divisioll”s assistant commander, wasnamed to head this organization.

Emb~rkztion of major l~nits began atChmdalcanal on 12 August. That samenight, an ldvance force landed near 13ara-koma to mark channels and landingbeaches and to select bivouac areas anddefensive positions. After being forcedto fight their way to snore, however,through fire from a motley collection ofsurvivors from sunken barges, the recon-naissance group hurriedly requested rein-forcements. The next night a infantrycompany landed to help them.

The main landing force departed Gua-dalcanal on 14 August on a split-second,staggered schedlllc. The slowest transport,~rotlp, L$+Ts, st:~rled first and WaS PaSsed-later by the faster .kPDs. In this manner,the transports which had departed Gau-

dalcanal in reverse order arrived off VellaLavel]a in the proper order and at the right,time.

Debarkation of troops began at dawn on15 .fugmt, the i\PDs nnloading quickly

in one hollr. The first snag in the inva-sion sche(ln]e occurred lv]lell it \V:lsdiscol~-ered that the beach could accommodate

‘ This was a new echelon of the A’(2OF an(l is

Ilot to be confused with the A’orthern I.andingGroup commanded by Colonel I.iversedge at Eno-gai and Bairt)ko.

ISOLATION OF RABAUL

only 8 of the following 12 LCIIs. The

LSTS, which arrived later at the correct

time, were forced to stand offshore wait-

ing to unload. Limited beach areas had

resulted in the very delay and exposurewhich it was hoped the staggered schedulewould prevent. There was, however, noenemy opposition ashore. As the beach-head widened, soldiers reported scattered,Japanese troops fleeing northward.

Shortly before 0800, jnst as the LCISwere in the unloading stage, the first offour frantic. ,Japanese air attacks struck.After making one pass at the protectivedestroyer screen standing offshore, the en-emy bombers and fighters turned their at-tack on the LCIS and I.STS, evidentlyti~lring that the smzller transports car-ried the bulk of invasion troops and SLlp-

plies. Al] four attacks were driven off byalert planes from ComAir~Sols and thefierce antiaircraft fire from the task forcedestroyers.

The fighter cover came from Munda air-field, which hnd begun operations only theday before. As a dividend for having wonan airfield closer than Segi or Guadal-canal, the All ies were able to keep an um-brella over the beachhead most of the day.Despite, the presence of this air cover, how-ever, the .Japznese persisted in sporadicattxcks, striking from different altitudesnnd directions. The results were negligi-ble. None of the ships in the convoy weredzmag-ed, and during the day more than

4,600 troops and 2,3(Io tons of equipment,and supplies were unloaded at Barakom:l.Twelve men were killed and 40 woundedin the day’s attacks. That night, as theconvoy withdrew S1OW1 y down Gizo

Straits, the shil)s fought, off repeated tor-l)e(lo attacks. Enemy float planes kept thearea lit with flares.

IIND OF A CAMPAIGN

The successful jump from Muncla toVella Lavella asserted Allied dominationin the Cemtral Solomons. Failing to re-pulse the landing, officers of the Eighth~l~ef and the (~e11e7/ tcenth AP71/y hzst ilycalled a conference to consider making acounterl anding on the island. One batt al-ien was :L1l that, could be sptredj it was

decided. This proposal was promptly

squelched by ~~qh th .1 reu :l?’?/2.Y. Such a

move would re~t~ire at ltmst two brigades,

the higher headqlmrters decided; and, in

view of the existing difficulties in reinforc-

ing ~n~ reslll)plying other central sOlO-

mons ~arrisous, the iden was better for~ot-

ten. ‘1’Ile only help the str<t~~lers on ~ella

Lavella received was reinforcement on

19 August by -..){)() .4 rlny :t]i(] I(N) NaVY

personnel.

The NL~ beachhead expanded rapidly.

Within the first 20 days of the operation,

6,505 troops, l@7 tons of rations, 843 tons

of gasoline and oil, 2,247 tons of’ ammuni-

tion, 547 vehicles, and 1,011 tons of other

classes of supplies were landed. Shipping

was to have been unloaded during nighthours, but the attacks on the convoys inthe narrow confines of Gizo Straitchanged that schedule. After 18 ikttb~~stj

the convoys arrived and cleparted 13ara-koma during daylight hours, protectedduring the unloading and passage throughGizo Strait by Allied planes from Munda.

There was little opposition to the ad-vance of the ATLF. By 18 August, thethree battalions of the 35th Infantry hadestablished a firm defensive. perimeteracross the southern end of the island. Be-hind this protective barrier, airfield con-struction began immediately. The Marine+lth Defense Battalion provided antiair-craft and seacoast defense.

155

As the fight for \’ells Lavella pro-gressed, the 35th began driving the enemybefore it. Toward the end of Au5gust, in-creased resistance was met on the eastcoast near Lambu I.ambu, and it was 15.September before the regiment’s assaultbattalions broke through the Japanesedefenses to overrun the barge base atHoraniu on the northeastern coast. Theenemy, however, escaped and fled north.

At this point., the 14th Brigade of theNew Zealand 3d Division landed at Bara-koma with two infantry battalions, the35th and 37th, as the main units. In areshuille of command, Major GeneralH. E. Barrowclongll of the 3d Divisionwas named as commanding general of allAllied forces on Vella Lavella. With thearrival of a third New Zealand battalion,the 30th, the American frontline troopswere relieved. On 25 September, the col-orful New Zealanders-the majority ofwhom disdained the use of steel helmetsto wear their distinctive fltit visored fieldhats-began their attack.

While the 35th Battalion leapfroggedaround the west side of the island in aseries of landings, the 37th Battalion beganmoving by landing craft up the northeast-ern coast, making landings at variouspoints to cut off the fleeing Japanese.When cornered, the enemy soldiers foughtstubbornly and fiercely for survival, butit was apparent that they were not undera single command or organized into asingle unit.

By early October, the New Zealanderswere in position to put the squeeze on theJapanese, who had been backed into a jut-ting piece of land between Marquana. andWatambari bays. The two battalions ofthe 14th Brigade made contact and joinedfor the final push to crowd the enemy into

,.,

156 ISOLATION OF RABAUL

the SL?:l.A Japanese priscmcr reported that,tl~e tired illld hungry enemy force W:lswilling to surrender, but that, ,Japaneseofficers would not permit it.

On the night of 6-7 October, Alliedtroops heard voices and the sound ofbarges scraping coral, but the supportin~fires were ordered too 1ate. The nextmorning, only littered stocks of ,Jmpaneseequipment and supplies were scatteredover the peninsula. The enemy cori]eredon Vella I~il},ella—589 by ,Tapanese ac-counts—had been successflllly withdrawn.

The hTe\vZe~li(]](lers ha(l estimated tllztthe Tell a I~avell a campaign would end inless than tTT70weeks. The conclusion camesever:ll dttys early—one of the few success-f~d tinletables in the Central Solomonsfighting. The 14th Brigade cwsllalties to-tfiled 32 killed and 32 woul~ded. .Jap:meselosses for the defense of the entire islan(lwere about 250.

The price for an island of considerablestrategic xnd operational value was notexorbitant. Allied casualties for the sevenweeks of fighting were less than 150 killed,most of these in air attacks, The practiceof bypassing a strong point to bit at aweaker point somewhere else was now es-tablishecl. Future Pacific operations fol-lowed the strategy initiated in the A]eu-tians and used with success in the CentralSolomons.

WIRIA~Efi AT VELLA LA T7ELLA 8

Success of the Rendova beachhead hadproven the value of z Marine defense bat-talion in a landing effort. When the task

‘ C nless otherwise noted, the material in thissection is derived frmn : X(;OF ..1c(’ounf; 4th1)efB1l Warl ), .\ug-0ct43 ; 4th I)ef13n SAR, dt(l

!).J:III-M ; C(). Corps Troops :ind St aging .lrea,lMAC, N“arrat.ive Rept of 01)s in Vella Lavella,

organization to seize T7ellg Lavella was

planned, tile inclusion of a similar unit

seemed logical. Closest and most availa-

ble was the 4th Defense Rattalion, then

on C+uaclalcana]. Organized at Parris Is-

land in 1940, the battalion was stationed at

Ilfate, New Hebrides, before transfer to

>Tew Zealand, then @lada]canal. Its or-

mtnization was similar to other Marinee’defense battalions-155mm seacoast, artil-

lery group, 90tnm antiaircraft group, a

special weapons group of 40mm, 20mm,

and .50 caliber weapons, and a tank

platoon.

13y ni@~tf all of the first day ashore at

Earakorna, about two-thirds of the light

ant iaircraft weapons were in temporary

firing posit ions. Otller guns of MajorMcDonald I. Shuford’s special weaponsgroup were kept set up on two LSTS in

the l~arbor, an innovation which increased1he firepower of the beachhead. The ad-dition was effective. A total of five enemyp] anes -were claimed by the 4th the firstclay.

During the next, six days, other echelonsof the battalion arrived and moved into

positions to defend the beach. Only theair defense units of the battalion got intoaction, however. Tl~e 155mnl gun groups,which moved ashore shortly after the orig-inal landings, were in coastal defense posi-

tions ready to fire within a few days, butthe need never arose. The tank platoon

dtd 30Jan44; I?wd Ech, Corps Troops, IMAC,( )1x) Xo. 143, dtd Z3SelA3 : Col John H. Cook,.Jr., ltr to CMC, dtd 2MarW ; (Jo1 IWcDonald I.Shnford ltr to (XC, dtd 4Mar52; llCol ]Jonald11. Srhmuck ltr to (XC, dtd 28Apr32; LtColChtirles T’. ]Iodges ltr to C31C, dtd 21 Mar52 ;I.t(.’ol C:trl M. Johnson ltr to CMC, dtd 2MarW2 ;X<IIO (Jmy7ia C’ampui,fw; Rentz, ,~fori~trsjtl tl)<’(’clltl’111 solo))) on.!+.

END OF A CAMPAIGN 157

which landed on 21 .4ugust was never com-mitted to action.

After the initial landings, the .Japa.nesebombing attempts dwindled in frequencyand ferocity. Dllring the early part of theoperation, the enemy attacks were pressedhome with fanntical fury and many reck-less planes were knocked spinning intoVella Gulf. Later the ,Jnpanese becamemore cautious, and fewer mass assaultswere attempt ed. Since any activity atBarakoma was readily cliscernible fromKolomtmngara, the arrival of Allied shipswas generally followed closely by a strikeby a conglomerate force of enemy bombers,fighters, and float planes. Most of the at-tacks were less than vigorous, however, asthe ,Japanese pilots soon gained a healthyrespect for ComAirSols planes and the ac-curate shooting of the 4th Defense Bat-talion. By late August, those few enemyplanes that did attack usually did not getclose enough to bomb accurately.

During the Vella Lavella operation, 15iiug~~stto 6 October, the 4th Defense Bat-talion compiled an enviable accuracy rec-ord. During 121 different air attacksaimed at the island, the Marine antiair-craft gun crews knocked down the follow-ing: 90mnl gun group—20 planes; 40Lnm

batteries—10 planes; 20mm batteries-5planes; the .50 caliber weapons of the spe-cial weapons groups+4 planes; and the.50 caliber weapons of the seacoast artil-lery group—3 planes. ‘Fhe total: 42.

Other Marines, not part of the NLF,also took part in the Vella Lavella opera-tion. After the 35th RCT moved pastHoraniu, establishment of a Marine ad-vance staging point on the island was or-dered. PI arining for the Bougainvilleaoperation was already underway, and theI Marine Amphibious Corps wanted a

base closer than Guadalcana.1 to the North-ern Solomons. On 17 September, the newCommanding General, IMAC, MajorGeneral Charles D. Barrett, w-hohad takencommand of the corps on tile 15th, namedllajor Donzl{l 11. Schmuck to headthe, proposed Corps Forward StagingArea, Vellz Lavel]a. The task organiza-tion include(l elements of the Marine 4thBase Depot, a motor transport company,a special weapons battery, a communica-t ion tenm, pm-t of the Navy’s 77th Sea-bees, as we]l as two provisional infantrycompznies froln the 3d Marine Division.All told, the forward echelon of CorpsTroops included 28 officers and 850 men.

The task force was to lxnd at two points:,JLIno river and Ruravai beach, on Vella’seast coast. Part of the organization wasto begin the establishrne.nt of a base campwhile the combat elements provided localsecurity. Hastily organized, the forwardechelon made one practice landing at Gua-dalcanal before proceeding to Vella Lavel-la. On 25 September, troops went ashoreby landing craft at ,Juno river while LSTSbeached some three miles north at Ruravaibeach.

Unloading at both points proceededwithout incident lmt il about 1115, when 15,Japanese bombers and about 20 fig,htersswept over. After one brief sideswipe atthe destroyer screen offshore, the enemyplanes turned toward Ruravai. Some40mm and .5o caliber antiaircraft weap-ons had been hastily set up on the beach,and these opened with a steady fire thatwas accurate and effective. Three bomberswere downed and a fourth damaged. TWO

of the doomed bombers, however, man-aged to complete their bombing runs be-

fore crashing in the jung-le. The other

planes continued to bomb and strafe. One

40mm crew and gun was destroyed by a

158 ISOLATION OF RABAUL

direct hit and a second crew knocked outof action. Volunteers quickly manned theseconcl gun w~d continued the fire.

.As the bombing attack ended, Alliedfighter cover appeared to clear the skyin a series of running dogfights. Butthe landing area at Ruravai was a sham-bles. Exploding ammunition continuedto wreak havoc. Caswdties and damageto supplies \verehigh. One LST had beensunk outright, others had been damaged.A total of 32 men on the beach had beenkilled and another 58 wounded.

The Japanese did not let up. Each daybrought a number of pressing air attacks.Despite frequent interruptions, the con-struction of roads, LST beaching areas,and base installations continued. Thework was further handicapped by wander-ing bands of enemy stragglers, which ne-cessitated active combat patrols as well asincreased guards at all construction proj-ects. Pro5gress, however, was fairly rapid.

On 1 October, as the second echelon ofcorps troops (including the 2d ParachuteBattalion ), arrived, the Japanese struckanother heavy blow. Four air attacks dur-ing the day resulted in further damageand more casualties. One LST was sunkand another damaged. The Japanese lostonly one plane. Convinced at last of thefutility of trying to land men and suppliesover a beach inadequately protectedagainst air attacks, lMAC then directedall further echelons and supplies to be un-loaded at Barakoma under the protectionof the guns of the 4th Marine Defense Bat-talion, The supplies were then truckedto Ruravai, where wide dispersal and in-creased aircraft defense measures ensuredfewer losses.

After surviving a number of such severeair strikes during the next week (whileBarakoma was studiously avoided by the

Japanese), the Corps Staging .4rea wasreplaced on 8 October by the newly ar-rived Vella Lavella Advance Base Com-mand. Some troops were returned to theirparent organizations: others remained atthe base under the new command. Rura-vai was seldom used for its intended pur-pose, since most ships preferred theloading facilities at Barakoma. Later,however, the sawmills and hospital of thebase command proved valuable during theBougainvillea campaibn by providing tin~-bers for briclging and ready medical facili-ties for seriously wounded men. Itsconstruction had proved costly, however.In the two weeks at Vel]a Lavella, the l?or-ward Echelon had lost 17 men killed and132 wounded during the many air attacks.

JAPANESE M71TFIDBA WAL’

As the campaign in the Central Solo-mons drew closer to its inevitable end, theJapanese efforts during August and Sep-tember became those of near-desperation.The Allied attack on Vella Lavella, whicheffectively shunted the enemy forces atKolombangara to the sidelines of the war,had the added effect of nearly isolating,Japanese garrisons from the main sourcesof supplies and reinforcements in theNorthern Solomons. Aggressive action byAllied destroyer squadrons tighte.necl theblockade. Camouflaged enemy barges,trying to keep the supply lanes open bysneaking along the coves and breaks ofisland coastlines were hounded and ha-

g Unless otherwise noted, the material in thissection is derived from: CinCPac Repts, Sep-0ct43, op. cit.; CominCh Re~t of S’oPac Action;ComSo~ac Ott.+S ~Farll; New Georgia Cam-paign.; Screnteenth Army Ops—I; 6’E Area iVav-

01}s-11; ON1, Combat Narratives X; Rentz,Marinrs in the Pcotral Solomons; USSBS,r’(zmpaig~~s.

160

rassed by the vigilant MTIk and the black-hulled Catalina flying boats (“BlackCats”) which prowled the waters of Vellaand Kula Gulfs. Nearly stymied in theirbmrge supply attempts, the Japmnese fi-nally resorted to supplying garrisons byfloatplanes and submarines.

These inadequate measures and a care-ful second look at the strategic situationforced the enemy to make the only decisionpossible: general evacuation of all forcesfrom the Centrnl Solomons. The opera-tion began with the removal of troops fromthe seaplane base at Reketx on Santa Isa-bel Island in early September. An .illiedpatrol, landecl from an MT13 on 3 Sep-tember, verified the absence of enemytroops. Quantities of rations and am-li~unition found on shore indicaled thzt thewithdrawal lmd been hurried.

After scattered outposts on Gizo andGanongga Islands reh~rned to Kolomban-gara on the 19th and 23d of September, theonly remaining enemy troops were thesmall force defending Arundel, a sizeablebody of troops on Kolombangara, and thestragglers back-tracking along the coastof Vella Lavella-about 12,000 troops inall, by ,Japanese estimates.

Weighing two factors—the direction ofthe Allied effort and the capability of the13th Regiment on Arundel to conduct adelaying action-the Japanese scheduledthe withdrawal for late September duringa moonless quarter. The northern coastof Kolombangara was designated as the

evacuation point. Landing craft fromthe Buin area would ferry troops acrossThe Slot to Choiseul for further transferto Bougainvillea. Sick and woundedwould be evacuated by fast destroyers.

The Japanese schedule began none tooearly. By 27 September the fighter air-

ISOLATION OF RABAUL

field on Vella Lavella was operational al-tl~ough not yet completed, and enemytroops on Kolombangara were caught ina vise between ComAirSols planes atMunda and 13arxkonm. In addition, .41-lied 155mm guns and howitzers emplacedon New- Georgia’s northern coast werepounding a steady tattoo on Kolombang-ara’s defenses.

The effect of waning moonlight — plusthe increased barge activity-was not loston the Allies. By late September it be-came evident that all Japanese activitywas directed toward withdrawal. Imme-diately, all available Third Fleet destroyersquadrons rushed with protecting cruiseminto interception duty in Vella and KulaGulfs.

The planned withdrawal begin, but wasdisrupted many times by the sudden ap-pearance of Allied planes and ships. Onthe night, of 28 September, the Japanesemanaged to load 11 destroyers with 2,115sick and wounded for a qllick sprint tosafety at Bougainvil]e. Despite the Alliedinterference and considerable loss of smallcraft and men, the Jxpanese relayed an-other 5,400 men by landing barges toChoiseul during the next few clark nightsand an additional 4,000 men were pickedup by six destroyers. In the squallyweather and murky darkness of the period,the Allied destroyers were hard-pressedto keep track of al] enemy activity. In a

number of instances, the destroyers hadto choose between steaming toward tar-gets which radar contacts indicated assmall craft or heading towards reportedenemy destroyer forces. Sometimes con-tact could not be made with either target.Allied ships, however, reported a total of

15 barges sunk on the nights of 29 and 30September.

,.,

END OF .4 CAMPAIGN

During the night of 1-2 october, allzvailwble Allied destroyers steamedthrough The Slot seeking the main Jap-xnese ev~cuation attempt. Few contact,swere made in the pitch darkness. About20 of the 35 barges encountered were re-ported sunk. The following night the.Lllied ships again attempted contact withthe Japanese but could not close to firingrange. Aware that the enemy destroyerswere acting as obvious decoys to lure theattackers away from the barge routes, theAllied ships abandoned the chase and re-turned to The Slot to sink another 20

barges.Further enemy evacuation attempts

were negligible. and the Allies reasoned

that the withdrawal had been completed..i patrol landed on Kolornbangara on 4October and confirmed the belief that theJapanese had, indeed, successfully com-pleted evacuation of all troops. Jumbledpiles of supplies and ammunition attestedto the fact that the enemy had been con-tent to escape with just their lives. The-withdrawal, the Japanese reported later,was about 80 percent successful, the only

losses being 29 small craft and 66 men.The final evacuation attempt was made

on 6 October from Vella Lavella. A size-able enemy surface force was reportedleaving Rabaul in two echelons, and three

U.S. destroyers moved to intercept, theenemy. Another Allied force. also hurriedto the scene. Contact was made in highseas and a driving rain. In a fierce battle

which lasted less than 12 minutes, theUnited States lost one destroyer to enemytorpedoes and two other destroyers werebadly damaged. The ,Tapanese lost onlyone of nine destroyers, and during the bat-tle the transports managed to evacuate the

161

troops stranded almost within the graspof New Zealand forces.

The removal of troops from Vella La-vella ended the Japanese occupation of theNew Georgia Group. The lOSS of the

islands themselves was not vital, but theexpenditure of time and effort and the re-sultant loss of lives, planes, and ships wasa reverse from which the ,Japanese neverrecovered. There could only be a guess asto the number of casualties to the enemyin the various bombings, sea actions? andland battles. Postwar estimates placedthe number at around 2,733 enemy dead,l”but this does not account for the manymore who died in air attacks, barge sink-ings, and ship sinkings. In any event, theunits evacufited after the costly defense ofthe New Georgia Group were riddledshells of their former selves, and few everappeared again as complete units in theJapanese order of battle.

More than three months of combat hadbeen costly to the New Georgia Occupa-tion Force, too. Casualties to the manyunits of the NGOF totaled 972 men killedand 3,873 wounded. In addition, 122 diedof battle wounds later, and another 23were declared missing in action. MarineCorps units, other than the 1st MarineRaider Regiment, lost 55 killed and 207wound ed.1~

hTew Georgia lacked the drama of theearly months of Guadalcanal and the awe-

some scope of later battles in the Central

‘03-GOF Accow t, p. 29. The figure reportedhere includes the 230 enemy dead at VellaLavella.

“ Ibid., p. 29; Rentz, Marines in the CentralSol(mons, p. 174. These figures do not includenon-battle casualties.

ISOLATION OF RABAUL

Pacific. Instead, it was characterized bya considerable amount, of fumbling, incon-clusive combat; and the final triumph wasmm-red by the f~ct that a nllmber of com-mand changes were required to insure thevictory. There were few tact ical or stra-tegic successes and the personal lmrdshipsof a rigorolls jllngle campaign were onlyunderscored by tile planning f milures.And too, the ori~inal optimistic t.imetableof Operation T013&T.lII,S later becnmeall embnrrassirrg subject. For these ren-sonsj n postwar resum6 of tile kmttle forthe Central Solbmons pnles in comparisonwith accounts of later and greater Alliedconquests.

The primnry benefit of the occupation ofthe New (3eorgia Group was the :i(lvance-ment of Allied air power another 200 milescloser to Rabaul. The fields at Munda,Segi Point, and 13arakoma providedComAirSols 1)1anes with three additionalbases from which to stage raids on the ,Jap-anese strongholds in the Northern Solo-mons and to intercept quickly any retali-atory raids aimed by the enemy at the mainAllied dispositions on Gua(lalcana] and inthe Russe]ls. Behind this protective buf-fer, relatively free from enemy interfer-ence, the .Illies were able to mass addi-tional troops and materiel for futureoperations. This extended cover also gave:Mliecl shipping near immunity from at-tack in southern waters. <klthough mostfleet activities continued to be staged from

Guaclalcanal, the many small harbors andinlets in the hTew~~eorgiz Group provideclvaluable anchorages and refueling pointsfor smaller surface craft.

The capture of the Central Solomousalso afforded the All ies the un(lisputed ini -

ti~tive to set the location and time for theIIext attack. The simple maneuver of by-

pnssing Kolombangar-a’s defenses won fortile .111ied forces the advantage of selectingthe next vulnerable point in the enemy’ssupply, communication, and reinforce-ment 1ines. The ,Japanese, guarding anempire overextended through earlier easyconquests, could now only wait and guesswhere the next blow wol~ld fall. The NewGeorgia Campaign presented the ,Japanesein his true light-an enemy of formidablefighting tenacity, but not one of over-whelmin~ sllperiority. His skill at con-ducting night evacuation operations, dem-onstrated at (~uadalcmal and confirmed atNew Georgia, col~ld not be denied, how-ever. Both withdrawals had been madepracticfllly under the guns of the Alliedfleet.

On the Allied side, the campaign fur-thered the complete integration of effortby all arms of service—air, sea, andground. Seizure of the Central Solomonswas a victory by combined forces—andnone could smy -who played the dominantrole. Each force depended upon the next,and all knew moments of tragedy and wit-nessed acts of heroism. The New (3eorgiabattles marked a long step forward in thetechnique of employment of combinedarms.

There were valuable lessons learned in

the campaign, too-lessons which were putto use during the many months to follow.As a result of the New Georgiz operation,future campaigns were based on a morerealistic estimate of the amount of men

and time required to wrest a heavily de-fended objective from a tenacious enemy.Another lesson well learned was that acommand staff cannot divide itself to coverboth the planning and administrative sup-

port for a campaign as well as the active

END OF A CAMPAIGN 163

direction of a division in combat. AfterNew Georgia, a top-level staff was super-imposed over the combat, echelons to planand direct operations.

On a lower level, the tactics, armament,and equipment of individual units werefound basically sound. As a result ofcampaign critiques> a number of worth-while equipment improvements were fos-tered, particularly in communicationswhere the biggest l:~ck was a light andeasily transported raclio set. From tilesuccessful operation of Marine Corps lighttanks over jungle terrain came a numberof recommendations which improved t:ic-tics, communication and fire c!oordinat ionof the big~er xnd more potent n~acllinmlwhich were included in the task organiza-tion for future jungle operations. Thebattle against the enemy’s bunker-type de-fenses on ?iTewGeorgia also pointed up thedesirability of tanls-mounted flame throw-ers. Experimental portable models usedin the fight for Munda had proved inval-uable in reducing enemy pillboxes. In-creased clependency upon this newly de-veloped weapon was one direct result of

its limited use in the Central Solomonsfighting.

Throughout the entire campaign, theimprovement in amphibious landing tech-niques and practices was rapid and dis-cernible. Despite seeming confusion, largenumbers of troops and mountains of sup-plies were quickly deposited on islandshores, and rapid buildup of men and ma-terial continued despite enemy interfer-ence. One contributing factor was the in-creased availability of the ships neededfor such island-to-island operation—LCIs,LSTS, LSDS, and the workhorse LCMS.By the end of the Central Solomons cam-paign, two years of war production wasbeginning to make itself felt. Equipmentand ships were arriving in bigger num-bers. The efficiency of these ships andcraft was? in part, a reflection on thesoundness of Marine Corps amphibiousdoctrines—vindication for the early andcent inlleci insistence by the Marine Corpson their development and improvement.12

“ I’{)r the story of the Marine Corps part inthe develf~lm]ent of l:mding craft, see Chapter 3,I’art I in Volume I of this series.

. . .

PART 111

Northern Solomons Operations

CHAPTER 1

Continuing the Pressure

STRATEGIC REQ UZ12EN ENTS ‘

There was little time for extended restfor iillied sea, air, and ground forces after

the final Central Solomons action in Oc-tober. AS early as the previous March, aclecision had been made tbzt the NorthernSolomons would be the target next afterNew Georgia; and by the time Munda fieldw%s operat ionl] under Allied control,plans for the seizure of a beachhead in theShorthands - Choiseul - Bougainvillea areawere in the final stages. In effect, ComSo-Pac operations of September and Octoberwere the last. act to the successful comple-

tion of the Central Solomons campaign as

*Unless otherwise noted, the material in thissection is derived from : CinCPac–CinCPO.4W:lrDs, Arrg through Nov43 ( COA, XIII) ) ;(Yonrgopur .JulL–Xov~.j IVarT)s; ThirdPlt .Narr-

Rept; 1111’hil~For AR, Seizure and (kcupationof Xorthern Err]press Augusta Bay, Borrgain-ville, 1–13SOV’H, dtd 31)ee43 ( BougainvilleaAreaopFile, Hist,Br, HQMC ) hereafter ZIZPhib-J’(,t+ AJi’; IM.\C AR, Phase I, Sec A, Rept onBongainville Operation, dtd 2131ar44 ( Bougain-villea .\reaOpFile. HistBr, HQMC ) hereafterIJIA C AR–I; E[aj .John X. Rentz, J;ouqwinvitlfi

and tile .~’ortl[cm t3010nLO?Z8, (Washington:

HistSec, DivInfo, HQMC, 1948 ) hereafterRentz, BougainviUe and tli e Northern Solo-tn OHS; Morison, Breul; ing theBiwnarcks Barrier;

Isely and Crowl, Marines and Ampl! ibious War;Halsey and Bryan, Hal,sc{/’,s Story; King rmdWhitehill, Ki//y’,s ,47ara7 L’ccorz; lliller, Rrdltc.

tion of Ra0au7; I{ist I)iv, HQMC, ““1’he Bougairl-ville Operation,” MS, cu. Feb-lfi ( HistBr, HQ31C ),hereafter HistL)iti Acct.

691–360 0—63—12

well as the overture to the forthcomingBougainvillea attack.

The importance of Rabaul had not di-minished durin: the long period involvedin taking New Georgia, but the events ofthe last six months had put the ,Japanesestronghold on New- Britain in a new light.In 1942 there had been no doubt that Ra-baul would have to be physically elimi-nated to insure an inviolate hold on the.$otltllITest-ASoutll Pacific area. By mid-1943, there wxs a growing realization thatthe enemy air and naval base might nothave to be erased by force, that neutraliza-tion would serve the Allied cause as well.Concurrent with this shift in thinking wasa proposal that, the main Allied effort bedirected through the scattered islands ofthe Central Pacific instead of through thelarger land masses of New Guinea andthe Philippines.

Both concepts bad many high-rankingproponents. The divergent views resultedin some open disagreement among themany strong-willecl commanders and staffofficers responsible for the Allies’ opera-tional strategy, bLLt, the eventual solutionwas born of the imagination and experi -ence of all, and there was no fl~rther dis-pute once the course of action had beenchartecl. The actual decision to strangleRabaul by air instead of capturing it wasmade by the Combined Chiefs of StaffLlpon reconlnlell(btt ion of the JCS. The

CCS conference at Quebec in August alsodirected that the advance through the

167

168 ISOLATION OF RABAUL

,Sol~tl~\~’est-So~lthPacific by (%meral Mac-.irth~lr and Admiral Halsey was to con-tinue while .fdmiral Nimitz aimed a newoffensive alon<g the Central Pacific zxis.The idea of two campai~rrs was an effectivecompromise. .ili hoi@ some real igntnent.of forces lvas n(?C13SSilr’~, all extensi l-(?shuffle of troops or shipping from either(heater would not be required: and :L co-

ordinated attack nlong two fronts wonldhave the advantage of keepinx ,Japnnmedefenses off lmlanw and comrnitt ed ovrra wide are,a.

Throngbout this evaluation period. Gen-eral MacArthur held fast to the originalEI,KT03T concept. On the. 411 of Sep-teml)~l’. the lyIIPl~ihFor ( Adnl;ral Bar-bey’s comnr:lrrd ) 1:~nrledSoWest,P:~c troopson the Hllon Peninsula of New Guinea toset the st:lge for eventual passage of Mac-Arthur’s forces tbroug]l the Vitiaz-Dampier Strlits. Success of the venture,though, depended upon insurance in theform of air bases within fi@ter plane dis-

t ante of Ilab~ul. Thus, MacArthur’s con-tinued surge toward the capture of Sal-amaua, I.ae, and Finschhafen virtuallydictated wtablislm]ent of ikllied air facili-ties in the Northern Solomons by Novem-ber or I)ccember of 1943. Plans for alanding in the Bougainvillea vicinity, tem-porarily shelvecl while other strate~ic con-cepts were being examined, were onceaglin restored to the stztus of ,a full-scaleoperation by MacArthur’s insistence.

The exact location for such an under-taking had been the callse for considerablediscussion and recon]la issance b.y Halsey>sstaff. Several islands lMd been proposedas targets but closer examination elinli-

rratecl them. ‘1’he island of Choiseul waslittle regarded as a major landing site be-cause of two factors: it was not within

fighter escortrange of New Britain; and

Mac.4rthur was of the opinion that anassault on Choiseul did not directlythreaten Rabaul. Buka, the small islandappendage to northern 130ugninville, wastoo far from Allied i~ir bases for a landingthere to be protected adequately. Initial

plans to seize the southern end of Bou-gainvillea were canceled because the air-fields at Kahili and Kara were too strong-ly protected to attack with the forces thenavailable to Halsey. With the drawn-outcampaig[l on Munda still fresh in hismind, ComSoT’zc was reluctant to mixw-it])the ,Jap;tnese in a prolonged strugglethat would trike too many lives and toomuch time without paying immediatedividends.

After some deliberation, Halsey pro-posed that the SoPac forces seize theSl~ortland Islands (Ballale and Faisi) asairfield sites and then interdict Kahili andKam with artillery fire as the Rendovaforces did at Murrda. This move wouldput Allied air support within tighter rangeof Rabaul. MacArthur, willing to settlefor any act ion which WOU1d help him real-ize his expressed ambition to return to thePhilippines, approved this concept. Butlater reconnaissance revealed that theShortlands had no beaches big enough orgood enough over which an amphibiousassault, could be staged, and that airfieldsites were limited.

Halsey’s top-echelon planners, abandon-ing the Shorthands idea,z on 6 SeptemberIdvanced another plan to seize the Treas-

‘ ‘l’he major factors governing the at)arldon-nwnt of the Shorthands m+ Imirnary targets w“ere1heir (listanrt, from Rabaul, “il little far . . . for:(UUC of ol[r short-le~ged planes,” and the strong

feelir)~ an~on: the I)lanuers that “~~e had to sl~endtoo nlu(h to *et too lit tie.” I,t(;er] Fiehl HarrisItr to ,\C(}fS, G–3, IIQM(l, dtd 2’iOct60.

CONTINUING THE PRESSURE 169

1’70 ISOLATION OF RABAUL

ury Islands and Choiseul as radar and PTbases and jumping-off points for further

operations against 130ugainville. Mac-Arthur, however, had concll~ded that theoriginal intentions of EI.KTONT called fora landing clirectly on Elougainvi]le andthat the interests of the JCS-approvedplan COUIC1best be executed by an earlyoperation (within the next few months)against 130ugainville. MacArthur indi-cated that any target Halsey selectedwould be acceptable, but, thzt a decision

must be reached soon.Kieta, on the northmstern coast, of

Bougainvillea, wzs a protected anchoragewith an airfield close by: but an assault onKieta inlolvecl long approaches by sea—flncl the .kllied sI]ipping short age was crit-ical. Kieta was too close to Choiseul,moreover> ancl that, island would have tobe attacked M a preliminary measure toprotect the ]nnding on 130ugainville. Em-press Angusta Bay, on the west coast, at,first glance had little to recommend it asa landing site. The beaebes along tile baywere exposed to wind and waves and heldno sheltered anchor:lge for the amphibiollsshipping required for SUC1lan operation.The terrain in] and was known to beswampy, heayily timl)ered, and with fewroutes of communication. On the otherhand, the enemy had apparently dismissedthis area as a probable landing spot andthe bay was only lightly defended. Aslate as 17 September, the Allies stood at,a figltr>tti~~ecrossroads, Il]ldecide.d abollt,which fork in the road to t:~ke.

Five days 1It er, on 22 September, Hal-sey announced a decision that, canceled allprevious p] am and alerted his forces forone of two alternate courses of action:seize and hold the Treasury Islands andnorthern Empress Augusta Bay area on

Bollgainville as airfield sites; or, as a ~c_oncl course, seize and hold the TreasuryIslands and the (loiseul Bay area as ad-vance r~dar points, torpedo boat anchor-ages, and a stxging base for landing craftbefore moving on to construct an airfieldon the east coast of Bougainvillea later inthe year. The final decision dependedupon last-minute reconnaissance efforts.

SPOTLIGHT ON BOUGAINVILLE 3

For the ,Japanese, the conviction that~ou~~inville was the ultimate Al]ied ob-jective in the Northern Solomons washardly a random guess. Lying in a posi-tion athwart the northern entrance to TheSlot, Bougainvilleas big bulk dominatedthe rest of the Solomons chain. By virtueof this ideal geographic location, the is-1and served the Japanese as an advancedsLl])p]y and refueling base for most of thesea and air operations against the Allies atGuadalcanal and in the Central Solomons.Here, too, were staged the infantry re-placements destined for combat or garri-son duty on other South Pacific islands.No less than six major airfields and a num-ber of naval operating bases were estab-lished by the ,Japanese on Iluka, Bougain-villea,and the Shortlands to help guard theol~ter defenses of the airfields and Simp-son Harbor at Rabaul. Easily supportedby air bases on New Britain, New Ireland,

‘ Unless otherwise noted, the material in thissection is derived from : Th irdFtt iVarrRep t ;

IIIPhibFm AR; IMAC AR-I; HistDiv Acct; 3dJlarDiv Terrain InfoRept, n.d. (BougainvilleaAreaOpFile, HistBr, HQMC) ; MilHistSec, G–2,FEC Japanese hlcmo.graph No. 100, SE AreaNavOps, Part III (OCMH), hereafter SE Area.NauOps—IZI; Rentz, Bougainvillea and theNorthern Solomon8 ; Morison, Breaking the Bis-mn.rcl<s Barrier; hIiller, Reduction of Rabaul;

Feldt, The Coa.staoatcher.s.

CONTINUING THE PRESSURE

and Trllk—points outside the range ofmost Allied land-based planes-Bougain-ville was also under the umbrelln of navalsupport from Rabaul and Truk. In short,the Allies needed Bougainvillea for furtheroperations against ~~abau]; the ,Japaneseneeded Ilougainville to defend Rabaul.Viewed from either camp, the island was apriority possession.

130ugainville was never a part of theBritish Solomon Islands protectorate.German expansion had claimed the isbmdin 1899, some 130 years after its belateddiscovery by the French explorer whosename it bears. Mandated to Australiaafter World War 1, the island still wasinfluenced b-y. German missionaries, forthe 130ugainvll]e natives in 1943 were rec-ognized as ]lost,ile to t,l~eA]lies and consid-erably more friendly to the Japanese.4Some 40,000 natives were on the island,{rathered nl:li]~ly in small villages of less~han 200. Energetic and industrious, theyprovided an adequate ]ahor force for thelarge plantations which, before the ,Japa-nese occupation, were owned by Europeancompanies.5

The island is the largest of the Solomonsgroup. Nearly 30 miles in width, Bougain-villea is divided clown the center by moun-tain ranges which extend nearly the entire125-nlile length of the island. The north-ern Emperor range is capped by the dis-tinctive 10,171-foot peak, Mt. Balbi. Thesecond range, Crown Prince, is less ruggedand settles gradually into a broad, widen-ing plain at the southern end of the island.This area has the best anchorages and thebiggest plantations. The two coastlines,

‘ IMAC’ .4 R–1, C–2 Est of Sit, p. 26.

‘ Rentz, Bougainvillea and the Northern tYolo-n~ons,Appendix V, contains an excellent descrip-tion of the island of Bougainvillea and its people,terrain, and geographic features.

171

east and west, are markedly different. Theeastern side of the mountain ranges slopesto a fairly open plain with good beachareas. The western side, however, is deep-ly etched by rushing mountain streamswhich carry silt into a swampy, alluvialplain bordering the coastline. The char-acteristic result is a series of deep valleysending in swamps and sand bars cut bymeandering waterways and sluggish riversof varying depths. This soggy fringearea, covered with tall marsh grass andbamboo-like growth, is trapped betweenmountains and the sea by a grey-blackbeach strip which seldom exceeds 15 yardsin width.

The island interior is enveloped by adense rain forest and choking junglegrowth which combine with the ruggedmountain ranges to discourage overlandexploration. A number of good trails tra-verse the more populated areas in the southand east, but only a few native tracks ven-ture across the inhospitable interior. Onepath cuts across the mountains from NumaNuma to Empress Augusta Bay where itconnects with the East-West trail. Thispat ]L joins the western villages of Maw-areka anti Mosigetta to Buin in the south.Fairly wide and cleared, the East-Westtrail skirts the coastal swamps and can betraveled most of the year.

Although the Japanese had occupied theisland since March 1942, only those facili-ties necessary to maintain the war in theSolomons had been constructed initially.In time, four airfields were in operation,two at each end of the island, and addi-tional troops were stationed on Bougain-villea. Scant attention, however, was givento the island region between the major air-fields. For more than a year, ,Japaneseactivity was restricted largely to the Buinarea in tbe south ancl the Buka-Ronis pas-

172 ISOLATION OF RABAUL

sage in the north. G There was little over-land travel. Barges moving along thecoastline served most of the transportationneeds. AS a result, few roads were im-proved and in the later defense of the is-land, this proved an important, oversight.

i4nstrzlian coastwai chers and a fewfriend] y nai ives maintained observnt ionposts on the southern p:~rt of the islanduntil the surnn~er months of 1943. Thenxggr-essive ,Japanese patrols, assisted bylmfriendly native guides, forced the iilliedscont,s to abandon their radio equipmentand withdraw to the interior. i As a re-sult, military information about the islandand its defenders was cut off abruptly just,when it became most neecled.

Intelligence estimates on the number of,Jzlmnese soldiers and sailors in the area~vwied widely. Interceptions of radio mes-s:~ges provided lnost, of the inforrnat ion

on troop dispositions, and this intelligencewas angn~ented and checked by enemy doc-uments captured in the Central Solomonsand by prisoner of war interrogations. Al-lied guesses placed the total number of de-fenders in the vicinity at 98,000—2,000 at,Choiseul, 35,000 or more at Bougainvi]leand the Short 1ands, 5,000”at New Ireland,and the reumining 56,000 at Rabaul. Theestimates on the strength of the Bougain-vil]e forces, based on order of battle infor-mation from prisoners. rnngwl between35,000 and 4+000. The biggest conce]ltra-tion of defenders was in the southern partof the island, where an estimated 17,000

‘ US.kFISPA Ohj Folder, Ilongsinville Island,dtd 1Auw13 ( Fkmg’ainville ArenolmFile, HistBr,IIQMC).

‘ ComSoPac WarD, 22Jun43. An interestingsequel to this instance is the 13 December entry :‘“The native responsible for the capture of Alliedpersonnel in Bougainvillea in ,Tnur was seized andshot.” ComSoPac nrc4S WarD.

soldiers of the Seventeenth Army wereheadquartered. .4nother 5,000 troops werebelieved to be in the Buka-Bonis area, witha similar number at Kieta. The onlyknown enemy concentration on the west<w:lst,was at Mosigetta, where about 1,()()(),Japanese-beliered to he laborers-wereengaged in cultivating the extensive ricefielcls of that coast. Less than 300 troopswere estimated to be in the Cape Torokinavicinity of Empress Augusta Bay. TheShorthands defense force was estimated at3,000 to 6,000, most of these naval persomnel from Eighth Fleet headquarters andSNLF units.

Postwar records of the ,Japanese indi-cate that the Allied estimates were close.The. Ruin area actually had about 15,000troops of the Seventeenth Army and 6,800of the Eighth Fleet headquarters and baseforce personnel, primarily for defense ofKahili and Kara airfields. About 5,000men were deployed in the Shorthands.The airfield on Ball ale was defended en-tirely by naval personnel with seacoastart illery.

The troop dispositions were in line withthe enemy conception of the plans of the.\llie.s. After Guadalcwral was evacuated,xnd tl~e ,Japanese became aware of the re-

sponsibil ity of defending what they hadso easily grabbed, the Shortland Bay areawas decided upon as the strategic key in

the defense of the Northern Solomons..fccording]y, the southern part of Bou-

gainvil]e and the Shortland Islands re-ceived first priority in troop allotments.At that early date, the enemy believed that

any Allied offensi~e would be directedagainst, the airfields in the southern por-

tion of the island with a possible subsidi-il ry action in the Buka area. Troop

streng! h elsewhere was proportionate to

CONTINUING THE PRESSURE

the .Japmwse estimate of the All its’ abilityto hit each anm. 13011gainville’s defensewm based on the prem iw that a landinganywhere on the is] and could be met by atransfer of ground troops nnd a counter-manding by an amphibious grol~p.

Responsible for the defense of Bougain-rille and the adjacent, islands was an oldxclversary, I.ieutenant Generzl HarrryoshiHyakut ake,’ who had commanded the,Japanese forces on Guadalcanal. Thegeneral lmd apparently lost little prestigewit h the Z7nperY”a7~~tafl thro~lgll the de-feat, for he was still ill command of theSmenteenth .4 rmy. His forces, however,had not shared his fortlme. ‘1’he 2d Divi-sion was almost wiped out at Grradalcanal,and the 38th Di&iorL had lost heavily atGuadalcana] and New Georgia. His soleremaining division, the 6thj commandedby Lieutenant General Masatane Kanda,was still in fighting shape. Two regi-ments, the W%? and 45th, were neartop-strength, but the third regiment of theclivision, the lJtA lnfc~?itv~/. had beenbadly mauled in the Central Solomons.To this division, Hyakutake could adddetachments of SIW,F units, plLMIhe scat-tered remnants of other infantry regimentswhich were, trickling into 130ugainvil]eafter the vritl~drawal from the CentralSolomons.

The .411ied estimate of the .Tapaneseships in the immediate vicinity ( Buka,

Bougainvillea) was 2 cruisers, 8 to 10 de-stroyers, 21 personnel transports, and 12

submarines, plus a variety of smallercraft. The Imperial Navy also had ahealthy reserve of warships at Truk andRabaul. Air support in the Northern

“ Miller, Reduction of Rabaul, p. 235n. Earlyintelligence reports translated Hyakutake’sgiven name variously as Siekichi or Harukichi,

173

Solomons was believed to be about 160fighters, 120 dive bombers, 120 mediumbombers, and 39 float planes.’

There were definite signs in Septemberthat the, Japanese expected an assault onRorrgainville. Despite increasing air at-trit ion, p] ane strength at 13uka, Kara,Kahili, and Ballale remained fairly steadyas the Japanese replaced their losses. Air-fields were improved and expanded despitesteady pollnding by Allied bombers, andsupply routes to the island were main-tained in spite of losses incurred throughharassment by Allied patrol bombers andtorpedo boats. In late October, as the.411ies completed their plans for attack,even the ]ong-neglected west coast of Bou-gainvil le was given some attention by theJapanese. Intelligence photos for thefirst time revealed evidence of military ac-tivity near Empress Augusta Bay. Someminor construction with a few scattereddefensive installations were discoveredbehind Cape Torokina, but since theimprovements were ]imited and no addi-tional troops appeared to have been movedinto the area, the Allies remained con-vinced that the Japanese had not alteredtheir basic defensive plans and that theyhad not awakened to the dangers inherentin an undefended coastline.

HALSEY’S DE(’lSION ‘“

Before a final decision was made on thedirection of the SoPac attack, reconnais-

‘ ThirdFlt OPlan 14+3, Annex A, dtd 150ct43

( COA, NHD) .‘o Unless otherwise noted, the material in this

sertion is derived from : Com130Pac i3@&?VarD; T?!irdFlt A-orrRept; IIIPhibFor AR;IMAC AR-I; HLstDiu Acct; Rentz, Bouga irwille

(U’Ld t/Le .Vo?’thmn ~O10WtOn8; &fOriSOn, Breaking

tlLe Bis?fLarclc8 Barrier; Miller, Reduction of

Raba u1; Halsey and Bryan, Ha~8eg’8 Story.

174 ISOLATION OF RABAUL

sance efforts to obtain every possible scrapof information and intelligence about thevarious projected landing sites continued.AS at Guadalcanal and New Georgia, priorknowledge of the Bougainvillea area waslimited to sketchy reports from former res-idents, planters, medical officers, schoonermasters, and missionaries. Althoughthese report~ were valuable in regard togeneral conditions and physical improve-ments in certain areas, few facts of mili-tary significance were obtained. As be-fore, personal reconnaissance by trainedobservers was required to accumulate thenecessary detailed geographic and hydro-graphic information upon which to base adecision. The first intelligence efforts cov-ered the entire Northern Solomons-SantaIsabel, Choiseul, the Treasurys, Short-hands, ancl Bougainvillea. Later, as someislands were eliminated and the choice oftargets mwrowed to either tile east or ~vestcoast of Bougainvillea, reconnaissance ac-tivities were concentrated on the Kieta andEmpress Augusta Bay areas.

Initial combat intelligence was gatheredfrom air reconnaissance and submarine pa-trols. Aerial photography was limited be-cause of unfavorable weather, enemy air in-terference, and lack of fighter pla~,e es-corts. Submarines moved’ in close to theisland t.o shoot pictures through raisedperiscopes, but this practice was hazardoussince the outdated hydrographic chartsthen available failed to show the exact,location of dangerous coral outcroppingand reefs known to exist. Besides, thepictures did not reveal much except a good

profile shot of Bougainvilleas ruggedpeaks. In time, patrols from submarines,torpedo boats, and seaplanes slipped ashoreto scout various areas, and the inforn~a-tion gained from physical reconnaissance

and personal observation was added toConlSoPac’s growing file.

Two patrols, dispatched to Bougain-villeain September after Halsey announcedhis two-part alert, helped the SoPac com-mander decide on the final choice of ob-jectives. One Marine-ATavy team, with anAustralian officer and four natives asguides, remained four days in the Kietavicinity, prowling the northeast coast ofthe island during the night and spendingthe daylight hours underwater in the sub-marine USS Gccto. Considerable Japanesetroop activity was observed; and despiteseveral close scrapes from patrolling en-emy barges, the group measured beachdistances, took depth soundings, andscouted the area inland. On 28 September,the patrol boarded the submarine for thelast time and returned to Guadalcanal.The patrol’s report was generally unfavor-able,ll indicating that the harbor had manyreefs and coral outcropping, and that thearea inland was not, suitable for airfieldssince the Japanese had apparently givenup on Kieta.

Another patrol similarly organizedlanded from the submarine USS GuardfLsh

near the Laruma River in north EmpressAugusta Bay. Here the terrain was found

to be fairly solid with thick bush and adense rain forest inland. Reluctant toarouse the ,Japanese to any Allied inten-tions, the patrol studied Cape Torokinathrough binoculars and took photographsthrough telescopic lens. The long-rangeexamination revealec] a narrow. beach stripsome 10,000 yards long with the expectedcoastal swamps inland. Tidal range in thebay was moderate, about 3~2 feet. A coco-

“ IIIPhibFor Rept of Reconnaissance of theNortheast Piougainville Coast, 2.3-27 Sep43, n.d.(Bougainvillea AreaOpFile, Hist13r, HQMC).

CONTINUING THE PRESSURE 175

nut, grove on Cape Torokina looked like afavorable spot for an airfield, the patroldecided, since the area appeared dryenough and long enough to support afighter plane strip. Unable to obtain asoil sample of the area~ the patrol did thenextbest thing andbrought back soil froma similar area. The scouts then turnedback tothe Laruma river and headed intothe bush in a -wide circle that ended fourdays later in a rendezvous with the Guard-fish in ~tsinima Bay, some distance to the

north. The only enemy si@ed were a. lone

sentry on post near the Laruma river,

and a number of Japanese reconnaissance

planes flying patrol duty over Empress

.lugusta Bay beaches.”

The endeavor had one big dividend.

Jvhile waiting for the patrol, the subma-

rine commander checked his position and

discovered that the navigation chart then

in use was about seven miles in error in its

location of Cape Torokina. Undetected,

this one factor mi@t well have jeopar-

dized any future operations. Too, the soil

sample returned by the patrol was de-

clared favorable for construction of an

airfield.

The prospect of landing in a lightly de-

fended area close to an acceptable airfield

site appealed to Halsey and his SoPac

planners. To confirm the patrol’s recom-

mendation, a number of low-level aerial

reconnaissance fl@ts were made by

IM~C staff members. Their quick

glimpses of the Cape Torokina area con-

vinced them that the operation could be a

success. On 1 October, Halsey notified

Mac~rthur that Cape Torokina was the

main objective and that D-Day would be

“ IMAC Reconnaissance Rept of Empress Au-gusta Bay area, 23–26 Sept43, dtd 10ct43 (Bou-gainvillea AreaOpsFile, HistBr, HQMC).

1 November. The Sol$TesPac commanderexpressed his complete agreement with thisdecision.’3

After 12 days of planning, Halsey’sComSoPac headquarters issued the orderswhich outlined the missions of the sea, air,and ground forces under the admiral’s

command. Specifically, ComSoPac di-rected the III Amphibious Force toseize and hold (on D-minus five days) theTreasury Islands as a staging area andadvanced naval base, prior to establishinga beachhead on D-Day in the northernEmpress Augusta Bay area for construc-tion of airfields and another advancednaval base. The ultimate aim was stran-gulation of enemy operations in southBougainvillea and preparations for furtheroffensives against Rabaul.

Code names selected by ComSoPac forthe planning phases were DIPPER, de-noting the entire Northern Solomons op-eration, and CHERRYBLOSSOM, theEmpress Augusta Bay area. The Treas-ury Islands phase of the operation waslabeled GOODTIME. Later the codename DIPPER was applied to the Bou-gainvillea operation and the island, whilethe Treasury Islands landing retained theGOODTIME designation.

Selection of Cape Torokina as a land-ing site despite its disadvantages was tac-tically and logistically sound. The loca-tion fitted well into the plans for neutral-ization of Rabaul by air, and a beachheadon the weskrn side of 130ugainville madelogistic support much easier. Moreover,the attack was aimed at a weak point inthe Japanese defenses, thus avoiding a di-rect assault on main enemy defenses in thesouthern and eastern coasts of the island.Success of the venture depended upon the

‘a ComBoPa.c Oct& WarD.

176 ISOLATION OF RABAUL

ability of the Allies to beat back any de-termined sea and air offensive by the Jap-anese during the critical stages of landingand estnhlishin,g a beachhead. ComSoPacplanners admitted that a strong enemyreaction was highly probable, but thisthreat was accepted as a ca]cukted risk.The projected operation was ‘(no betterdemonstration of the firmly held, but attimes sorely tried, belief in the Allied su-periority over the enemy in the SouthPacific.” “

Less obvious in the choice of EmpressAugusta Bay was the fact that the CapeTorokina plain, bordered by the naturalobstacles of the I.arLll]a River to the north-

west, the mountains inlancl, and the Toro-kina river to the southeast, fell into anideal defensive area about six miles deepand eight miles long which could be de-fended by tile Allied forces then available.The location, too, was believed so isolatedfrom known Japanese dispositions by thenature of the island’s terrain that at, leastthree months would be required before astrong force moving overland could seri-ously threaten the beachhead. In short,large bodies of reinforcing enemy troopscould come to the area only from the sea,and the Thircl Fleet felt confident that itcould handle that. threat. The Allies haclno desire to capture the entire island—thesize of Bougainvillea and the rough terrainprecluded SLICII zrnbitions — but two in-

fantry divisions could hold the (lape Toro-

kina area against any enemy forces in the

immediate area or likely reinforcements.

These facts clid little to increase enthu-

siasm for such bold plans. Sent iment was

mixed — some optimism, some hesitatio]l,some reluctance — but with D-Day lessthan a month away, all hands bent, to the

“ IM.4G’ AR–?, p. 2,

task of preparing for the assault on theTreasurys ancl Bougainvillea.

AMPHIA’10 UA~ PL.4NNZNG ‘5

The planning team that clirectecl thepreparations for the Bougainvillea opera-tion was essentially the same commandlineup that out 1ined the ATew Georgialttack-with one important change. Asbefore, Admiral Halsey retained personalcontrol of the proceedings> dividing hisattention between completion of the Cen-tral Solomons campaign and the develop-ment of a new operation. From hishelclquarters in Noumea, Halsey coordi-

nated the planning activities of AdmiralFitch on Espiritu Santo and AdmiralWilkinson and General Harmon onGuadalcanal. A new member of the plan-ning staff was Lieutenant General Alex-ander A. Vancleg-rift, who had replacedGeneral Vogel as commanding generalof the I Marine Amphibious Corps, thecounterpart to the Army’s ground forcecommand under Harmon. .4s such,Vanclegrift, held administrative respon-sibility over practically all Marine Corpspersonnel in the South Pacific, the ex-ceptions being land-based air units un-der General Mitchell’s command (Marine.iir, South Pacific) and ships; detach-ments. The lMAC staff had not partici-pated in the New Georgia planning sincethe bulk of the troops were to be furnished

“ Unless otherwise noted, the material in thissection is derived from: ComiSoPac S’ep-Ott-j3WarDs; IIIPhil)For AR; CTI? 31 WarDs, Se~0ct43, n.d. ( COA, hTHD ), hereafter CTF .31 WarDwith month ; IM.4C AR-1; ThirdlW NarrRept;lIi.~tDi?> .4cct; Rentz, Bougainvillea and the North-frn Solomons; Isely and Crow-1, Marines and.Arnphibiows War; Morison, Breaking t7Le Bis-marclcs Barrier; Miller, Redaction of Rabaul.

CONTINUING THE PRESSURE 177

by Griswold’s XIV Corps, but Halsey’sfirst planning directive designated theNorthern Solomons as an IMAC assign-ment. As a result, the Marine commandbecame tactical as well as administrative.

Vandegrift, who had won a Medal ofHonor for his leadership of the 1st MarineDivision at Guadalcanal, was the first ofthree IMAC commanders to participate inthe 130ugainville operation. After com-pleting preliminary plans, Vandegrift wasrelieved by Major General Charles D. Bar-rett cm 15 September and was en route tothe United States to become the 18th Com-mandant of the Marine Corps when hisreturn was abruptly sidetracked. 10 Gen-eral Barrett had suffered a cerebral hemor-rhage in an accidental fall, and his un-timely death left IMAC planning in mid-air. The operational order for IMAC’Spart in the seizure of the Treasurys andBougainvillea had been drafted before Bar-rett’s death, and Vandegrift resumed com-mand on 18 October in time to signthe completed order “-noting, meanwhile,

that the entire concept of the operationhad changed within the space of onemonth.lx Vandegrift completed the plan-ning and witnessed the start of the opera-tion before relinquishing command of

‘0 Vandegrift was informed in January that hisnomination to succeed General Holcomb as Com-mandant hi](l been approved by Admiral King:lnd the Secretary of the A’avy. The general,accompanied by Colonel Gerald B. Thomas, hadreached Penrl Harbor when his retention in theSouth P/icilic was requested by Halsey to King1hrough Ximitz. (len Alexander A. V&ndegriftinterview by HistBr, G–3, HQMG, dtd 5Aug59( WWI1 OpHistFile, HistBr, HQMC ), hereafterT’andcorift ~?ttf’WieU).

“ BGen James Snedeker ltr to CMC, dtd14May48 ( 130ugainville Monograph CommentFile, HistBr, HQMC ).

“ ~7undrgrif t Intt?rvieu~.

IMAC to Major General Roy S, Geiger on9 h’ovember, well after the success of thebeachhead was assured.

Even as the Northern Solomons plan-ning shifted through a number of changesbefore the final draft emerged, so did thetask orgmization assigned to the IMAClanding force. Of the three Marine divi-sions in the Pacific in 1943, only the 3d wasavailable for the 130ugainville operation.The 1st Marine Division was scheduled byMacArthur to spearhead the CapeGloucester beachhead on New Britain, andthe 2c1Marine Division lmd been shifted tothe Cent ral Pacific for the C~ilbert 1s1anclsassault. To augment the 3d Marine Divi-sion and IMAC troops, the Army’s 25thDivision and the 1st Marine Raider Regi-ment were tentatively assigned to the130ugainville venture; but, as the CentralSolomons campaign wore on, both or-ganizations were committed to action farbeyond original plans, with the result thatneither was available for the NorthernSolomons.

A number of provisional units wereformed from scattered battalions, andthese—with the later addition of the 37thInfant ry Division—were assigned toIMAC. Vandegrift’s command eventu-ally included his own headquarters and

corps troops, the 3d Marine Division, theArmy’s 37th Division, the 8th BrigadeGroup of the 3d New Zealand Division,the 3d Defense Battalion, the 2d MarineRaider Regiment (Provisional), the 1stMarine Parachute Regiment, the Army’s190th Coast ~lrt illery Regiment (.4ntiair-craft ), and varied naval small craft, con-struction, and communication units. TheMarine pzrachute regiment, then on l’ells

1,avella2 was designated the corps’ reserve

force. In area reserve were several coast

178 ISOLATION OF RABAUL

artillery battalions on Guadalc anal andthe .lrmy’s .Imerical Division, then in theFiji Islalltls. This latter clivision, how-ever, coulcl be committed only on Halsey’sapproval.

After the target had been defined inlate September, Halsey established hisoperational chain of command v-ith hisamphibious force commander, AdmiralWilkinson, in charge of the entire 130 LI-

g-itj ll\~ille-Tl.e:lsllrys expedit ion. Wilkimson maintained control of Task Force 31(III Amphibious Force ) with the groundforce and transport grollps as subordinatecommands. As clirectecl by Halsey, Wil-kinson would cent inue. to conmlwlcl allforces afloat and ashore until t}le ]anclingforce commander was ashore ancl had in-dicated that he w-as able to take commanc{

of tl~e Wolmd forces.lg The 1~1:~~ com-

mancler, i~t tirst Elarrett ancl then ~Tande-

gri ft, was responsible for the scheme of

manellver ashore at, both Cape Toroltina

ancl the Treasurysj ancl was to exercise

comman(l over all units asliorej whether

;\llied forces, Jlarine corps, .Irmy, ~’a~-y,

or grolmcl echelons of :air ltnits. .1s LetNew Georgia, the actual employment of

“ This con(+pt of comnmnd evolved durinz theGllad:llc:lnal mlIap:lign nftcr disagreenlent be-tween ~’ancle~rift and the commander of the.~mphibious Force, South I’nc4itic, Rear AdmiralTllrner. 31:l]ly J]>lv:[] ofi{.prs ~{,nsidered thefol’(.~s ashore as ;In c.xtellsi[}u of the forces:Iflo:]t, :~(llllillistr:ltively a]ld militarily suborcli-IInte to the amphibio~ls comm:tnder at all times.T-all(legrift sllcccssfn]ly fmlltellded th:lt the l:ln,l-iug force c{jlllulander \vas more experienced :indbetter qualified in ~ronncl ol)erations and sholddh:iye u]l(livi(le{l rt,sl)f)msillility on(e the tr(l(~lx\vere :lshf)re, I,t(’t)l Fr:ll)lc (). Hough, lIaj VerleE. I,udwig, and Henry I, Shnw, ,Jr,, PcarZ Har-

?Ic)r to {; lladc(7mt )lal-lfi.xtor~t of ~. A’. Mc!rinc

(’orp.! Opfratio)/, s iit TI”o/lf7 TI~(/). 11, v. 1,(Wash-ington : HistBr, G–3, HQMC, 1958), p;). 240–241.

aerial support was to be under the opera-tional control of air officers. This airechelon, Comnlander Aircraft hTorthern,Solonlo]~s ( (xom.iir XorSols) was a subor-dinate commancl of Com.lirSols and wasuncler the di l’ect ion of Ilzrine Brig<lclierGeneral Fielcl IInrris. His tasks il~cludedthe respollsibi]ity for active air clefense oftho Bougainvillea region ( Torokina and theTreasurys) as well as operational controlof all supporting aircraft entering thisregion. Harris’ responsibilities also in-cl~~decl establishment of an air warningsystem for both Torokina and the Treas-urys and organization of air support con-trol procedures for both areas.

For tl~e.Treasury 1s1ands phase of theoperation, Wilkinson acidecl another eche-lon to his commancl, dividing his taskforce into a h-orthern Force for the Em-l)ress .iugustn Bay landings anc~a South-

ern Force (Task Force 31.1 ) for the diver-sionary landings. Wilkinson retainedcommand of the. Nort hem Force ( actu-ally, a name designating the main units ofTask Force 31) and placed Admiral Fortin conmland of the Southern Force. TheTrensllrys lttncling force, comprisingmostly troops from the 8th Brigade Groupof the 3cl New Zealand Division, was conl-nlal~clecl by 13rigaclier P~. A. Row, unclertlw general clirection of the IMAC conl-n~ancler.zo

By nlicl-October, all s~lbordinate echel-ons of Task Force 31 and IMAC had is-sued operational orclers, ancl the cliversesea and air clenlents under Third Fleetcommancl had been assignecl general n~is-siolls ill support ot’ Wilkinson’s task forcein the ll{>u~,~in\~ille-Tre:~s~lr~~s venture.T~an(l-basetl air units of (’om.fi rSols

“ The l}l;lllnillg nnfl cf)n(lucl of the Tre:wurys{}lwri[tiotl is related in the following chapter.

CONTIA’UING THE PRESSURE 179

(TaskForce33) wereorde.redto continuethe general lllissions of reconnaissance andCIestrllction of ellem~ ships and aircraftwith the added duties of providing aircover and sllppol”t for the land and seafc)rcesill~-olve(lilltlle X’ortllerl~Sololllollsassault. Rear Admiral Frederick C.Sllerl~lal\’sc:lrriem (Task Force3S) weredirwtecl to support, the expedition by airstrikes at Ruka find Ilol~is airfields whichwere beyond the effective range of fighterplanes from the nowconlpleted airfieldsin the hTewGeorgia area.

Cruisers and destroyers of Rear .id-miral l[errill’s Task Force 3!) were to fl~l--nisll ])rotection for the all~l~hibiolls forceas well as bonlbarcl ene]lly illstallatio]ls inthe 13uka-130nis and Shortlands area be-fore tile operation. Task Force 72, thesl~bmat-ines lmder the colllmand of C’ap-tain ,Jwnes Fife, ,Jr., was to carry ollt offensive recollnaissnnce lj) issions north oftile Bollgainville area.

Final planning for tbe nctllal assallltsa~ Cape Torokina aucl the Treasurys wasfacilitate(l by tl~e move of 131.lC l]et~cl-quarters from Noumea to Guadalcanal,lvhcre the :unpllibiolls fol.ces were in tl’ilill -

ing. Yzn(legrift’s conullalld post was es-

tablished k a coconut ~ro~-e near Tetere,~LISt n s~~ort distance from the hEVl(lquaI-

ters of the IIIPhibFor and the 3d lIarineDi~-ision bivouac area. The close proxim -it~ of tile tl~ree m:ljor hcaquarters respm-sible for w.mldillz{ ing tl~e efforts of theforces involved eawl collsiderab]y tileproblen~s that :lJ’OSP dllrin: (IIe l:lst lf-eelfsof preparation. Ancl problenls arose-many of them. In effect, the Northern,Solomons opera timls l)a(l Iwcoll]e t\\-oo11-(’I’:lti 011S—eilCll One with L’C’Sllltilllt pl”Ol)-

lems of transport atiol~, reinforcrnlent,and resupply.

Tactical limitations in launching the130ngainTille operation became apparentenrly.zl In fact, tl~e entire ~entllre l~orethe title of “Shoestring No. 2,7’ a referenceto the general pallcity of means lvllichcharacterized the Guada]canal ol)erationill 1942.ZZ Early in :iugust, Halsey llaclrevie]ved tile shil)ping sitllation ill t]w.Solltll Pacific and estimated that he wouldneed six additional troopships to completeI)reparatiolls for tlie next operatic)ll, de-claring tl~at at least 14 .iP.~s and 6 ~~l<Ais~vould be required from D minus 15 to Dpl(~s 5 for the assault on 130ug-ainl’ille.z3Other cmlunitments, particularly the Cen-{ral Pacific campaign, hnd claimed all buteight .lP~\s and four AKAs from theSoPac area; and this limited amount ofamphibiol~s shipping was the specterw]lich haunted the Iloucgainville prepara-t ions. The acllte shortage seriously re-stricted the number of troops ancl suppliesthat could be transported in the initialassaldt and prohibited a rapid buildupat tile objective. Flwther, any loss ofsl~ips collld not, be replaced immediate-ly: :LllC~> since these h2 ships were abare minimum for transporting an as-sault division with the IIecessary artillery,itit’ service personnel! engineels, an(l heavyequiplllent to collst rtlct al~d maint aiu anairfield, the sinking of any of tllese ships

“ When Vaudegrift bPgan initial planning, heenu)hasized that his sublnitted tentative Plansfor operi~tions in the N-orthern Solomons en-visioue(l wnduct under optimum conditions, andndde(l that it was “reaclily foreseeable that themeans required may exceed the resources a~7ail-al)le in this are; l, P(lrticularly in respect to trans-ports and lauding craft. ” CG. IMAC ltr toCt)mSoPnC, dt(l Hw43 (Bougainvillea ~rea@s-File. HistTlr, IIQ31C).

“ 310rist)n, llr(,al:i)l~ th (, Bismarcks Ilarriw,P, 288.

“ ComSoPac .luq43 WarD.

180 ISOLATIOA7 OF RABAUL

dnring the beflchbeacl phase would seriolls-

ly threaten the ent ire operation. The defi-nite possibility of quick and effective airretalifltion by the rllemy pronlptecl a deci-

sion by Wilkinson that i he slow-movingLSTS under his conlnmncl wollld not beriskecl cluring the early ph:lses of theoperation.

The glaring contrmt between the comduet of the (%mdalctinfil landing ~nd theNew Georgia operation ]\-fisenl])hasizeclin the concept of the Rollgainville assalllt.Tranclegrift had the experience nnd tacti-cal foresight required for sllch an opera-tion. In Wilkinson he follnd a partnerwith a good grasp of amphibiolls sllpport

of operations ashore. Together t]le tlvo

commanders were w good team, and theplans they e~olved were prflrtical applica-tions of the ~vailablc~ me~ns to the situa-tion. Gllndal(’nnal Ilnd ken 111 fissallltfollowe(l by the establishment of :1 pro-tective perimeter thrown abont a captllretlairfield. The Japanese had dashedthemsel~es zgxinst this perimeter, sldTer-ing ruinons caslmlties in the process. ?iTewGeorgial ]~o~~-ever,]Ilc~ l>t~elltile rel.erse—n lancling, then a protracted overl:lncl at-tack thflt had berll a tiring, I)loocly smashagainst a ,Jap:lnrse perinleter. G il-enthese examples, ~~andegrift ancl Wilkin-son were (leterminecl tlmt the tact icalerrors of Xew C~eorgiI wolllci not be repeated. .It Bougainvillea, the Marines11’olllcl lane{ first, estal)lis]l ill)(l expand

a beachlleld, :In(l-wllm tile figl~tillgwirlened into extenderl land operations-the Army would take o~er. Airfielcl con-struction would con2mence upon landing.with completion of nir strips expecte(l illtime to help clefelld xga inst any deter-mined assccult,by the Jmpanese.

In order to take advantage of the exten-sive stretch of beach north of Cape Toro-ki)ln, n simlllt mncolw landing of all troops:~n{l snl)p] ies owr mllltiple benches wasschedulecl. This wonld allow tile vitallyilnportant ships to drop anchor, nnloacl,al~d depart witl~out undue clelay in the ob -je~ti \“earel. To further rec~uceunloadingtime, 111 car~o sl~ips WO1llC{be restrictedtO S]lOrt ]O:lCk, :lllCi aSSalL]t tl”OO]>S TYOLllC~

be diverted to hasten the un]oading ~ctiv-

it ies. Twel~-e hc:~chcs were to be employeclin the assaldt, eleyen extending northwestfrom Cape Torokina tow:~rd the LarumnRiver :111(1tlle tv-e]ft 1~locate(l on tlle north-ern ( inl~er) shoreline of Purllata Islancl,the l:trger of two islets lying off CapeTorokina.

Tl~ree assillllt lll~its of four l:~ndingtei~]ns each were to lancl abremt over these12 be:~c]les. (j]](. t:lsl< unit was to lnncl in

the ri,zbt sector of the beach area with aseconcl task unit lancling in the left sector.Tl~e thircl task nnit was to be divided, twolillldill: teams landing in each sector.Erich infantry unit wccs to overrun anyenenly defenses al~d seize a broacl but sh~l-low beachlleacl. Reconnaissance was to bestarted immediately to the front andfranks, whi]e Ullloacling was completed.Beach ant] antiaircraft defenses were to bepreparecl ilnnlcdi ately to insure protectionfronl a possible el~el~~ycounterl:ancling orthe expected enenly air attacks.

Ihrring unforeseen difficulties, flt leccst13.900 n]ml wol~]d go ashore in the D-Day

landings at Cape Torokina. Unloadingtime for the troops :lncl the 6,200 tons of

embnrkecl supplies and equipment was setat no mole than six hours. NTOfloating

reserve for the Rougainville :lSS2LLlt was

plannccl, since the enemy situation ashoreclicl not seem to warrant this risk of em-

CONTINUING THE PRESSLTRE 181

lmrkeci troops zncl sllpplies. Initill planscalled for the beachhead to be reinforcedand resupplied by fi>’e LST echelons fivedays apart, each ec]le]on bringing in an-other 3,000 troops and 5,300 tons of cargo.Vamiegrift, npon resllmption of theIM.LC c!OllHllW@ Obj&Cte<l to this slow”

builclul) and insisted that reinforcement ofthe beachhead proceed at fin nccelerat cclrcrte. He proposed that the 37th Divisionfollow the 3c1Division ashore no later thanD plus 7.” This change in plans was

agreed upon. The IMAC genera] also as-sumed responsibility for beachhexd logis-tics~ and, in planning tile rapicl unloatlil~gsequence, Vandegrift assi~ned about one-Ilalf of his combat Marines the tempol’arytask of getting supplies ashore. Dnringthe initial stages, these Marines wouldwork in tile holds of the ships ancl on thebeach as it shore p~rty before returningto their units for combat operations.

Such preparations were not idle ges-tures. Once ashore, tl~e IM.IC troopsfacecl the prospect of havil]s an extrwnely

tenuous supply chain cut behind them byenemy counteractions. To forestall anyelnergeucy later, N’ilkinson decided to use211 available snips in snpply operationsbefore the landing. .fccordin~]y, Tn a-tempt ]vas made to bring a ;10-day le}-el of

all classes of snpply f orwcrrd to Vella I.a-vella, consi(lelably closer to tile objectivearea than tl~e st:igillg-stormge-bi~~oll:~c

areas on Guadalcaual or in the Russells.

Thus, a steady flow of supplies to thebeachlwa(l CO1lICIbe n~oyec{ by the meansavailable-LSTs, LCTS, APDs, or evenDDs. Tl~e area picke(l was Ii[lravai 011the east coast of ~~ella Lavella, then stillunder attack by 25th Divisio]l and Ne\v

“l”a?~dc~rift Intcrcicw.

Zeal znd troops. The landing at Rurcrvai,however, was bombed with effectiveness bytlw ,Tapallesel and tl~e inability to clivert asldticient nulllber of LSTs from the 3Tew

Geor~ia Sltpply chain ultimately resulteclin a stockpile considerably smaller thnnthat propose(l-zbout 10-days’ level, infact. Cape Torokina, IMAC later dis-coverecl, l~as outsi(le the range of smallcraft from T’ells Lavella, a fact which fur-ther reduced tl~~ value of the 131AC sup-ply stat ion at Rnravai. The end resultJvas thclt, despite seemingly adequatepreparations, the supply problem re-mained a major obstacle throughout theoperational planning.

Equally vital to the operation’s successwas the speed with which airfields couldbe constructed. Although the earlier re-

connaissance patrols had locatecl a possible

airfielcl site, the OCICISwere still high thattl~is area n)ight prove too swanlpy forquick completion of a strip. At least threeI)omber fiel{ls ~ncl two fighter strips werecleemed essential to threaten Rabaul fromIlougainville, nnd the problenl of locatingsuitable areas iLs well as having a sufficientnumber of naval construction battalionsassiamled to the task were monunlentalworries. The airfield annex to the IMACoperation order called for two strips tobe established as SOOIIas possible 2’ Withthe remaining fielcls to get underway asthe beachhead widened. To assist 17ande-grift, an experienced engineer staff wasolganized within IMAC, this group conl-

prising Marine ancl 5T:avyofficers who haddirected air[iel(l cwl]strll(’i ion LInder conl-

hmt cmn(litions :It Gua(laleanal and hTew

“ 11[.ic’ ( )1)0 x-o. 1, .Anx F), IMAC Airfiel(l

1’1:111, dt(l 13C)(t43 ( P,ougainville .h-eaOIEFile,IIistllr, IIQ31C’), her(’after IX.IC OpO Xo. 1.

. . . .

182 ISOLATION OF RABAUL

Georgia” —a situation likely to be re-peated on Bougainvillea. Since the. CapeTorokina area was still relatively un-scouted, xnd the existence of coral a debat-able question, plans were made to usepierced planking for all fields. This addedz fresh burden to the limited shipping, butprovided insurance that runvvays could befashioned for at least local air support.27The proposed fighter runway, located inthe plantation area at Cape Torokina, wasto be about 4,500 feet long and about 600feet wide. A bomber strip inland about2,000 to 7,500 yards would require an ex-tensive amount of clearing. Even thoughswamps were prevalent, drainage was be-lieved possible.’* The remaining fieldswould have to wait for the widening ofthe beachhead.

Another facet of the landing—navalgunfire support—also received close atten-tion during the last. weeks of preparation.

Coordination between the amphibiousforce and the landing force had improvedmarkedly since the first days of Guadal-canal, but many imperfections unfortu-nately still existed in the tire support ofground forces by ships’ guns. The useof naval gunfire by Allied ships at Guadal -canal had little effect on the progress ofthe initial landings, and bombardrne.nts tocover a landing force in the Central Solo-mons had been impromptu affairs; ~.e., re-turning the fire of the Munda batteriesduring the Rendova landing, and theshelling of the Enogai garrison duringthe Marine raider operations at RiceAnchorage.

“ Col Francis M. M{Alister ltr to CMC, dtd2!)May48 ( Bougainvillea Monograph Comments,HistBr, HQMC).

‘7 IIIti.2“1.W.4COpo No. 1.

&Tew enthusiasm, however, had beengenerated for naval shore bombardmentsafter Kolombangara’s capture. The is-land’s fortified defenses were still in-tact after the Japanese withdrawal, andIIIPhibFor was granted permission totest new gunfire procedures on these bunk-ers and pillboxes. One of Wilkinson’s de-stroyer squadrons did the shooting, withall available IMAC and 3d Marine Divi-sion gunfire liaison otlicers and spottersaboard to witness the demonstration. zoThe destroyers pounded the beach defensesin a simulated pre-H-Hour strike beforeone destroyer moved inshore to act as aspotting ship for omcall missions. At theconclusion of the firing, snips’ officers andthe observers went ashore. The prepma-tion fires, they found, had blasted thebeach areas, and the adjusted firing mis-sions had knocked out other bunkers andfortifications.’” As a result of this con-vincing display, IMAC officers were eagerto give naval gunfire support a prominentpart in the assanlt plans for 130ugainville.

To retain the element of surprise, nopre-D-Day shelling of Cape Torokina wasscheduled. This further increased thenecessity for effective gunfire which wouldknock out beach defenses before the Ma-rines went :Ishore and provide as well forquick fire support to reduce any undis-covered Japanese defenses which mighthold up the assault waves long enough forthe expected enemy counteraction tothreaten the landing force. The IMACgunfire officers initially made plans for asupport group of about 4 heavy or light

“ Col Frederick 1’. Henderson, “Naval GunfireSupport in the Solomon Islands CamPaim,”MS, 19754 (Bougainvillea AreaOpsFile, HistBr,HQMC), pp. 51–52, hereafter Henderson, “NavalGunfire Support.”

‘0 Ibid., p. 52.

CONTINUING THE PRESSURE 183

cruisers and at least 10 destroyers; gl butin light of the multiple commitmentsof the Third Fleet, this was a prohibitiverequest. Eventually, four destroyers,none of them practised in shore bombard-ment, were scheduled for pre-H-Hour andpost-H-Hour gunfire support. In alltrut hf (llness, IM.kC liaison officers wereunhappy with the amount of assistanceoffered by the IIIPllibFor, but acceptedthe preferred support with the realizationthat Wilkinson’s force had other missions,too.

The final gunfire plan 32positioned threedestroyers (Sigouwtey, Anthony, Wads-~to~~fi,)on the southeast flank of the ]and-

ing area, with the fourth ship (Terry)stationed on the opposite or extreme north-west flank. These ships were to open fireat 10,000 yards before closing to 3,000yards for close support and on-call mis-sions. Target designation was to be takenfrom a photo mosaic of the Cape Torokinacoastline, photographed from about 1,000yards off-shore by a low-flying plane. Af-ter one final rehearsal in the New Hebri-des area where the four ships tried toapproximate area, range, bearings, andmaneuvers as close as possible to thoseexpected on D-Day$ 33the destrover sqll~d-

ron joined the assault echelons for themove to the objective.

FINAL PRZ?PARA TIONS 34

Major General Allen H. Turnage’s 3dMarine Division was a well-trained, albeit

% Ibid., p. 58.

“ ~TP’ OPO 3T0. AIS43, dtd ltWct43, Arm A( Bougainvillea AreaOpsFile, HistBr, HQMC ) ;Henderson, “Naval Gunfire Support,” pp. 61–62.

“ ComDesDiv 90 AR of 1Nov43, dtd 14Jan44(COA, NHD) .

WUnless otherwise noted, the material in thissection is derived from: IZIPhibFor AR; IMA(7

inexperienced, outfit, a fusion of a nunl-ber of combat teams and supportingtroops. Three infantry regiments-the3d, 9th, and 21st Marines—together withthe 12th Marines (artillery) and the 19thMarines (engineers and Seabees) formedthe nucleus for the division which wasfirst assembled as a unit in May 1943.To these major commands was added a

number of service and support groups-tanks, special ~veapons, nmphibian trac-

tors, motor transport, signal, and medicalbattalions. With a background of ex-

tensive jungle warfare training in Samoaand Guadalcanal, the 3d Marine Divisionwas fully expected to be capable of meet-ing the rigors of the 130ugainvilleoperation.

The missions assigned the 3d MarineDivision were an” extension of lMACtasks: capture or destroy enemy forcesin the area: establish a beachhead in theCape Torokina area about 2;250 yards in-land and about T,350 yards wide to includethe two small islands offshore; commenceselection of airfield sites and construction

of air strips; establish long-range radarpoints and an advance naval base to in-clude operating facilities for torpedoboats: expand the beachhead on corpsorder.

After Turnage had been alerted to themajor role of the division in the f orthcom-ing landings,35 the task organization as-

AR-I; 3d MarDiv AR, 1-11 Nov43, dtd lSDec43,hereafter .!?d MarDiv AR; H&?tDiv Acct; Rentz,Bouguinville and the Northern Solomona; Iselyand Crowl, Marines and .4mpitibious War;

Miller, Reduction of Rabaut; Morison, Breakingthe Bi.!imarck8 Barrier; Aurthur and Cohlmia,$d ,MarDivHist.

= CG, IMAC Itr of Instm to CG. 3(1 MarDiv,dtd 27 Se@3.

691–360 0—63—13

184 ISOLATION OF RABAC1,

signed to the division grew rapidly withthe addition of a number of corps troopsancl provisional battalions, some of themstill in the process of forming. The majorattachments were the 2d and 3d RaiderBattalions, now joined in a provisionalregiment, and the 3d Defense Battalion.As expected, the combat experience ofthese corps and division forces varied.Elements of the 2d Raider Battalion hadraided W&in Isltnd in 1942 before goingto Guadalcanal, and the 3d Defense Bat-

talion was battle-tested at Tulagi andGuadalcanal. But the rest of the IMACforce—with the exception of the 37th Di-vision—were without prior experience.The Army division, however, had beenblooded in the Munda campaign.

Early in October, the diverse elementsof IMAC and the 3d Marine Division were

assembled at Gundzlcxnal where the tmsk

of welding them into landing teams began.

The three task units were based on thereinforced 3d and 9th Ilarines and the

3d Defense Battalion. Task Unit A-1,four landing teams comprising the 3dMarines and the 2d Raider Battalion, wasassigned six beaches in the vicinity of CapeTorokina. Task Unit .4-2 (9th Marines)was scheduled to land over five beaches onthe left (northwest ) flank, with the Marine3d Raider Bzttalion, attached to the

9th Mmrines, to go ashore on Puruata Is-

land. Task IJnit A–3, the antiaircraftbatteries of the 3d Defense Battalion andsllpporting troops, was to land behind theassault. troops in each sector. Each taskunit included artillery, air liaison and

signal personnel, engineers, and naval baseconstruction troops. .4s directed, everylanding team was self-sustaining and self-supporting until the division as a whole

coulcl be consolidated behind the contem-plated force beachhead line.3s

The 21st Marines plus the remainingartillery units and supporting troops wereformed into task units which would bebrought to the objective area after D plus3 clays. By 7 November, all elements of the3d Marine Division would be ashore withthe last increments of the IMAC head-quarters troops slated for arrival by 15November. The first unit of the 37th In-fantry Division-the 148th CombatTeam—was to be unloaded starting 7hTovember with the remaining combattewns-the 129th and the 145th—sched-uled to be on Bougainvillea by 22 Novem-ber. In all, an additional 13,000 troopsand another 26,672 tons of cargo were tobe brought to the Torokina beachhead toreinforce and resupply the assault ele-ments.37 Originally these echelons wereto be transported by LSTS and APDs, butlater Allied sea and air victories permittedthe continued use of the larger APAs andAKAs in reinforcing the Bougainvilleabeachhead.

With only limited shipping space avail-able to the combat troops, the assaultechelons carried only enough supplies—three units of fire, B-rations, and fuel—tocontinue operations ashore for 10 days.The rest of the allotted cargo space con-tained additional ammunition, rations,fuel, engineering tools, and equipmentwhich could be unloaded quickly if thesituation permitted. The heavier equip-nlel~t and facilities materiel would be

36CofS, 3d MarDiv ltr to CO, 3d Mar, dtd16Ck.t43.

3’ IM.4C OpO No. 1; IM.\C, Movement of TaskOrganization, dtd 290ct43 ( Bou~ainville Area-Ol,sPile, IIistIlr, HQMC ) ; NorFor LoadingOsNos. X)1-43 throuxh 309+3, dtd 120ct–15hTov43in lIIP1! ifiFor CJct-No@$ WarDx.

CONTINUING THE PRESSURE 185

brought ashore in later echelons. TheIMAC troops making the initial landingswould carry a haversack only; knapsacksand blanket rolls were to be carried ashorewith organizational gear. Officers ofIMAC anti the 3d Mm-ine Division orderedall seabags, cots, and mattresses to be storedat Guadalcanal; ancl these were neverembarked.

The three assault task units and equip-ment were embarked at Guadalcanal.Tzsk Unit A–1 went on board four trans-ports on 13 C)ctober for rehearsals atEfate, and then stood by at Espiritu Santoto await the assembly date. Task Unit&2, after embarking 18 October on theremaining four transports, also rehearsedthe landing operation for four days atEfate before heading back to the Guadal-canal area for the rendezvous. The thirdtask unit continued training and rehears-als at Guadalcanal until 26 October, atwhich time the troops and equipment weretaken on board the four .kKAs.38 The re-hearsals indicated that, with a 2,500-yardrun to the beach and each ship restrictedto about 500 tons, uninterrupted unload-ing could be accomplished in about 41/2hours.30 JTTithallowance for time lossesby air alerts while underway and duringthe assault phme, the amphibious forcecommander was sure that troops and sup-plies could be xshore before the end of D-Day. On 30 October, the various elementsof the Northern Lancling Force—trans-ports, cargo ships, mine sweepers, and de-stroyers—began steaming toward the ren-

3’c(l, 3d }larDiv ltrs of Instrn, ~td 7C)ctA3aI)d l~()~t~3 (Bougainvillea .kreaOp~File, Hi~tBr,HQMC) .

w ComTransGru, IIIPhilJFor, Rept of I,and-ingOps, Empress Augusta Bay area, Bougain-villea Island, 1–2A”ov43, dtd 22Dec43 (Bougain-villea AreaOpsFile, HistBr, HQMC).

dezvous point off Guadalcanal for the finalrun toward Bougainvillea.

I,90LA TING THE TARGET ‘0

As Wilkinson’s transports and screen-ing elements formed into the main con-VOvto open the second phase of OperatiOII

D~PPER, the final naval bombardmentand air strikes aimed at the complete neu-tralization of Bougainvillea airfields be-gan. For nearly a month the island’s de-fenders had been on the receiving end offrequent bombing and strafing attacks asa preliminary to the actual landings. Thefinal strikes were calculated to insure neg-ligible air interference by the Japaneseduring the amphibious assault the nextday.

Although the .411ied air power was rec-ognized as greater than that which could

be mustered by the ,Japanese in the area,the fact that the Allies were moving closerto the main enemy strength was an equal-izing factor. Late September estimatesplaced the Japanese air strength in the

Northern Solomons at about 154 planes, orless than $f~ of the estimated 476 in theRabaul-Bougainville area. Against thisnumber, the Allied air command—Com-AirSols-could ante 728 planes of all

mIJnless otherwise noted, the material in thissf?ction is derived from : MarCorpsBd Study andEvaluation of AirOps Affecting the USMC dur-ing the War With Japan, dtd 31Dec45 ( HistBr,HQMC ) ; ComSoPac 0ctJ3 WarD; TMrdFltA7arrR~pt; ComAirSols StrikeComd WarD,26.Jill–19A”ov43, dtd 17Jan44 ( CO.4, NHD ), here-after i5’trike(70md WorD; ComAirSols IntelSeeRepts for 0ct43, n.d. ( COA, hTHD ) ; ZIZPhibFor.4 R; I.UAC AR–I; HistDiv Acct; SE A?xm NawOp.+III; Miller, Reduction of 17abaal; Morison,Brcoking tile 13ismarcl:8 Barrier; Rentz, Bou-

{]aiari71e ond ttt c ii’orihem S’olomons; Sherrod,.UarAirHist.

186

types. This composite group included 14Marine Corps squadrons with 181 planes,19 Navy squadrons with 274 planes, 16Army squadrons with 216 planes, and 3NTew Zealand squadrons with 57 planes.By n~id-October, at least 314 fighters and317 bombers were being- clirected againstthe enemy bases. The recent acquisitionof airfields on Sew Georgia and Vella La-vella was a decided asset to ComAirSo]soperations; for, although most of thebombers were still based on Guadalcanaland the Russells, nearly all of the fighterstrength had been moved forward withinstriking distance of Bougainvi]le. Thisincluded 31 fighters and 148 bombers atMuncla, 103 fighters at nearby Ondonga,48 at Segi, and 60 at Barakoma.

The ai,r activity over Bougainvillea hadsteadily increased as D-Day moved closer.During the month of October, ComAirSolsplanes rangecl northward on 21 of the 31days in raids that ranged from tree-levelstrafing runs by torpedo and scout bombersancl fighter planes to high-level bombingattacks by B–2% and B–25s. ih themonth neared its encl, intermittent attacksbecame almost daily occurl’ences to thebeleaguered ,Japanese defenclers. Kahiliand Kara were hit most often, 23 timesand 17 times respectively, but Buka, Bal-lale, Kieta, and the Tre.asurys also wereraked ancl cratered by .lllied planes.Choiseul, too, was ,rocked occasionally byConAirSols craft with Choiseul Bay anclthe Kakasa submarine base the favoritetargets.

By 18 october, jubilant aviators haclreported that Ballale>s airfield was “pul-verized.>? ‘I Photographs verified theirclaims by showing 98 bomb craters on therunway, 23 of them in the center of the

“ 8trihvContd WarD, 180et43,

ISOLATION OF RABAUL

strip. By the 22d, Kahili was likewise in-

operable. ~ week later, Ballale was hit.

again, and the enemy’s repair work ren-

clered ineffective. Postwar enemy records

also attest to the attention ~iven the

enemy airfields by Com~irSols. The

Japanese admittecl that, just prior to the

invasion, the airfields on Bougainvillea

were useless.

llabaul, too, w7as feeling the sting of

.\llied bombings. General Kenney’s Al-

liecl Air Forces, committed by General

Alac.lrthnr to ]encl all assistance to the

neutrcclization of Rabaul by air, roared

ont of 3Tew ~uinea bases on 12 October to

slug Simpson I1arbor and the Rabaul air-

fields with a 349-plane raicl. This blast

was e.ncorecl by an equally large attempt

six clays later, but this time bad weather

closed in over NTew Britain and only 54

bombers completed their mission. Day-

li@lt attacks by fighter-bomber groups

were ,repeatecl on the 20th, 24th, and 25th

of October. Kenney’s fliers insistecl that

the bombings had crippled Rabaul, and

optimistic reports of clamage inflicted and

enemy aircraft destroyed were relayed to

the invasion forces. Later assessments,however, discounted these evaluations.Far from being a smoking ruin, the Jap-anese stronghold with its five airfields wasstill very much operational and still a fac-tor to be reckoned within the Bougainvilleaseizure.

As a matter of record, however, the So-WesPac bombers did considerable dam-age to enemy installations and consider-ably recll~ced the. Japanese ability to strikeat the Bougainvillea assault forces. Fur-ther, this damage was accomplished onrestricted tmgets in the face of hot recep-tions by enemy fighter craft and accurateantiaircraft clefenses. Unfortunately,foul weather near D-Day prevented the

COXTINUIATG THE PRESSURE 187

Allied .kir Forces from cent inu ing their at-tacks which woulcl ha\-e f urt her divertedtl~e .Japanese attention from the Bougain-villea ]an{lings. As it turilcd out, the So-WesPac fliers Ivere able to stl’ike anothersolid blow against P~abald only after theIMAC troops ~~ere.a]ready ashore at CapeTorokina.

As part of ComSoPac’S progr:un to stunthe Bougninville clefenses dllring theinitill stages of tlm operntion, tile support-ing task forces of ~ldnlirals Merrill andSherman took UIJthe cudgel for a whackIt the ,Japanese airfields prior to D-Day.Merrill, with fo(lr light (misers and eiglildestroyers, steamed close inshol’e to tileBuka Passaxe shortly after nli(lni~llt, ;;lOctober–1 November, to r:~ke Ilukn an(l130nis airfields with a 30-nlil]llte bonlbar(l-ment. The glllltire was givel~ addecl ac-CLU’aC~ by the. spotting reports froln twoCom.\irSols plnnes overhead. The t:~skforce then retired, Imrasse(l but nllscathecl

by enemy planes. TWO hours and 60 miles1ater, the BLIka flames were still visible to

the task force..

llerrill was barely o~~t of the nei@bor-

hood before Shernmn arrived in the area

with tile carriers flarato~ctand Przkeetonj

escortecl by 2 light cruisers and 10 cle-stroyers. Undetectecl by several enemyflights, the C:lrrier planes took off and

dropped an additional 20 tons of bombs ontl~e. two airfields before the carrier taskforce retirecl nnopposed and unchallengecl..idn~iral Merrill, meanwhile, was complet-ing the second part of his mission. The

:lll’ezcly- riclcllecl airstrip at Ballale wasshelled by his tmk force, as was Faisi andseveral of the smaller islan~ls. Still uno-pposed, Menill’s force headed for Vella

IJavella to refuel ancl rearm, satisfied thatt]~e s~~ppolting bonlbardments hacl slLc-

cessflllly launc]~ed tl~e assault on I;ougaill-

ville.

CHAPTER 2

Diversionary Assaults

TREASURY ISLA,$’Z) LANDINGS 42

If the initial plans for the direct assaulton the Buin area or the Shortlands hadbeen carried out, the two small islands ofthe Treasury Group would have been by-passed and left in the backwash of thecampaign. Instead, with the change inplans to strike clirectly at Empress Augus-ta Bay, the islands of Mono and Stirlingbecame important as long-range radar sitesand torpedo boat anchorages. Moreover,in an attempt to deceive the enemy as tothe direction of the attack on Bougaimvine and convince him that the ultimateAllied aim might be the Buin area or theShorthands, the seizure of the Treasuryswas given added emphasis by being setas a preliminary to the Torokina Ianclings.To help this deception succeed, reconnais-

“ Unless otherwise noted, the material in thissection is derived from: CornSoPac Ott-iVov&3WarDs; ThirdFlt NarrRept; IMAC AR–I; IMAG

C–2 Repts, 270ct–13Dec43 ( Bougainvillea AreaOpFile, HistBr, HQMC ), hereafter IMAC C–2Rept.$; IMAC C–2 .Jnl, 270ct–27Nov43 ( Bougain-villea AreaOpFile, HistBr, HQMG ), hereafterZMAC C–2 Jut; IMAC D–2 MiscRepts ( TreasuryIs), 270ct43, dtd 10Nov43 (TreasuryIs AreaOp-File, HistBr, HQMC ) : 8 ( XZ ) BrigGru Rept onOps, TreasuryIs ( Op GOODTIME ), dtd 30Sov-43 (TreasuryIs AreaOpFile, HistBr, HQMC ) ;8( NZ ) BrigGru OpO Xo. 1, Op GOODTIME, dtd210ct43 (TreasuryIs AreaOpFile, HistBr,HQMC ) : OA’1, ComZJut 37arra.t tve XII: Hender-son, “lNaval Gunfire Support ;“ Rentz, Bouqain-

ville aufl the Northern Rolomons; Miner, Reduc-

tion of Rabaul; Morison, Breaking the Bis-

marckg Barrier.

sance patrols to the Shortkmds and diver-sionary operations on the island of Choi-seul—plus low-flying photo missions overthe Short lands—were scheduled by IMACto increase the enemy’s conviction that thefollow-up objective was the Shortlands.

This could have been a natural assump-tion by the enemy. The Treasurys areabout 60 miles northwest of Vella Lavell aancl ol~ly 18 miles south of the Shorthands.While the size of the Treasurys limited con-sideration as a major target, Mono andStirling were close enough to ShortlandIsland to cause the Japanese some concernthat, tl~ey might be used as handy steppingstones by SoPac forces. But then again,the Tremurys are only 75 miles from CapeTorokim-a fact which the Allies hopedmight be lost on Bougainvilleas defenders.

The Treasury Islands are typical ofother small islands jutting out of the seaill the Solomons chain. Mono is a thickly

forested prominence of volcanic origin,with abrupt peaks and hill masses morethan 1,000 feet high in the southern part.These heights slope gradually in an ever-widening fan to the west, north, znd eastcoasts. The shores are firm, with fe~~’swamps, and rain waters drain rapidlythrough deep gorges. The island is small,about four miles north to south ancl lessthan seven miles lengthwise.

Stirling Island to tile south is smaller,more misshapen. Fairly level, this island

is about four miles long and varies from300 yards to nearly a mile in width. There

Ma

DIVERSIONARY ASSAULTS 189

are several small, brackish lakes inland,but the island is easily traversed and, oncecleared of its covering forest, would be anexcellent site. for an airfield. Betweenthese two islands is a mile or more of deep,sheltered water+ne of the many ancl~or-ages in the Solomon Islancls to bear thename Blanche Harbor. The combinationof these features—airfield site, radarpoints, good anchorage-was the factorwhich resulted in the seizure of t]le Treas-urys as part of the Bougainvillea operation.

Early information about the islandswas obtained by an IMAC patrol whichspent six days in the Tremur-ys in Au-gust, scouting the area, observing themovement of the Japanese defenders, andinterrogating the natives. In this latterinstance, the loyal and friendly people ofthe Trezsurys were a remarkable contrastto the suspicious and hostile Bougain-villea inhabitants. Additional details werereceived from rescued aviators who foundMono Island a safe hiding place aftertheir planes had been forced down bydamage incurred in raids over Buin andthe Shortlands. This first-hand intelli-gence was augmented by aerial photo-graphs. The reports and photos indicatedthat the best landing beaches were insideBlanche Harbor, on opposite shores ofMono and Stirling. The only beachessuitable for LSTS, however, were on Monobetween the Saveke River and a smallpromontory, Falamai Point.

As limited as this information was, theamount of intelligence on the enemy dis-positions on the two islands was even moremeager. The Japanese strength was esti-mated at 135 men, lightly armed. Thesewere bivouacked near Falnmai but maiw

tained a radio stztion an(l ol)servationposts in various areas. Natives reportedthat much of the time the Japanese moved

about Mono armed only with swords orhand guns. Stirling Island was appar-ently undefended.

The 8th New Zealand Brigade Group,attached to I Marine Amphibious Corpsfor the seizure and occupation of these is-lands, arrived at Guadalcanal from NewCaledonia in mid-September. Althoughthe New Zealanders would form the bulkof the assault troops, the GOODTIMEoperation was IMAC-directed and IMAC-supported. The landing force comprisedabout 7,700 officers and men, of whomabout 1,900 were from I Marine Amphi-bious Corps support troops-antiaircraftartillery, construction battalions, signal,and boat pool personnel. Marines at-tached to the brigade task organizationincluded a detachment from the IMACSignal Battalion and an air-ground liai-son team from General Harris’ Com.&ir-XorSols headquarters.

On 28 September, Brigadier Row, thelanding force commander, was informedof the general nature of the GOODTIMEoperation, and planning in conjunctionwith .idmiral Fort began immediately, al-though there was only enough informa-tion available to the commanders of thetask grol~p and the landing force to for-mulate a plan in broad outline. The taskwas far from easy, for the Southern Forcewas confronted with the same logisticaland transportation problems that facedthe Empress Augusta Bay operation,

Fort and Row decided that the mainassault would be made in the area ofFalamai, where beaches were suitable forLSTS. Stirling Island would be takenconcurrently for artillery positions. Noother landings were planned; but afterRow was informed that the long-range ra-d~r wonld have to be positioned on thenorthern side of Mono to be of benefit to

190 lSOLATION OF RABAUL

TREASURY ISLANDS LANDINGSI MARINE AMPHIBIOUS CORPS

27 OCTOBER 1943

MAP 12

the Bougainvillea operation, another land-ing at Soanotalu on the north coast waswritten into the plans.

Final shipping allocation to Fort’sSouthern Force included 31 ships of sixdifferent types—8 A1’Ds, 8 1.(.’1s, 2 LSTs,and 3 LCTS for landing troops and sup-plies, 8 LCMS and 2 APCS for heavyequipment. and cargo. The limited troopand cargo capacity of this collection ofships and landing craft restricted theSouthern Force’s ability to put more thana minimum of troops and supplies ashoreinitially, but this problem was solved byreducing the strength of the brigade’s bat-

talions and limiting the number of artil-lery weapons, motor transport, and en-gineering equipment in the first echelon.The brigade’s assault units included 3,795troops with 1,785 tons of supplies andequipment. Succeeding echelons werescl~edu]ed to sail forward at intervals offive days.

The final plans, issued by Row’s head-quarters on 21 C)ctober, clirected the 29thand 36th Battalions to land nearly abreastnear Falmni Point, with the 34th Battal-ion landing on Stirling Island. Sinlul -taneously, a reinforced infantry companyaccompanied by radar personnel and Sea-

DIVERSIONARY ASSAULTS

bees would go ashore at Soanotalu in thenorth. The two battalions on Mono wouldthen drive across the island to link upwith the Soanotalu landing force whilenaval base construction got underway atStirling.

The initial landings in Blanche. Harborwere to be covered by a naval gunfire sup-port group of two destroyers, the Prhgleand Philip. Liaison officers of IMACplanned the gunfire support, as the NTewZealand officers had no experience in thisphase of operations. While the brigxdegroup expectecl to have no trouble in seiz-ing the islands, the naval support wasscheduled to cover any unforeseen diffi-culties. The gunfire plan called for thetwo destroyers to fire preparation salvosfrom the entrance to Blanche Harbor be-fore moving in toward the beaches withthe landing x’aves to take targets underdirect fire. The IIIPhibFor, however,took a dim view of risking clestroyers insuch restricted waters. The desired close-in support mission was then assigned tothe newly devised LCI ( G) —gunboatsarmed with three 40mm, two 20nlm, andfive .50 caliber machine guns-which weremaking their first appearance in combat.Two of these deadly landing craft were toaccompany the assault wav~s to thebeaches.

After one final practice landing onFlorida Island, the brigacle. group beganloading supplies and embarking troops forthe run to the target arew Admiral Fort’sSouthern Force was divided into fivetransport groups under separate conl-manders, and these groups departecl inde-pendently when loaded. The slower LSTSand LCMS left first, on the 2%1 and 2-lt.hof October, ancl were followed the nextday by the LCIS. The APDs sailed on 26

191

The Southern Force departed with amessage which delighted the hTewZealand-ers as typical of the remarks to whichAmericans at war seemed addicted: “Shootcalmly, shoot fast, and shoot straight.” 2

At 0540 on the 27th, the seven APDsof the first trmsport group lay to justoutsicle the entrance to Blanche Harborand began putting troops over the sideinto landing craft. Heavy rain and over-cast weather obscured the beaches, but thepre-assault bombardment by the Prinqleand PhiZip began on schedule. The L~SS.Eaton moved to the harbor’s mouth andtook up station as fighter-director ship asthe destroyers registered on Mono Island.The firing was accomplished without as-sistance of an air spotter, who later re-ported radio failure at the critical moment.This probably accounts for the disappoint-ing results of the preparatory bombard-ments, which proved to be of little valueexcept to boost the morale of the assaulttroops. The Pringle’s fire was later de-clared to be too far back of the beacharea to be helpful, and the bombardmentby the PhiZip left a great deal to be de-sired in accuracy, timing, and quantity.

A fighter cover of 32 planes arrived

promptly on station over the Treasurysat 0600, ancl, under this protective screen,the assault waves formed into two columnsfor the dash through Blanche Harbor tothe beaches. L’nexpectedly, enemy ma-chine gun fire from Falamai md Stirlinggreeted the assault boats as they ploughedthrough the channel. At 0623, just threeminutes before the landing craft nosed intothe beaches on opposite sides of the har-bor, the pre-assault cannonading ceased;and the two LCI gunboats-one on each

‘ Quoted in Gill(,sI)ie,XC70Zcalwid History, P.October 148.

ISOLATION OF RABAUL

flank of the assault wa~~e-took over thetask of close support for the landing for-ces. At least one Mmm twill-mount gun,several machine guns, and several enemybunkers were knocked out by the accurate[ire of tIlese.two ships. Promptly at 0626,the announced H-Hour, New Zealandtroops went ashore on Mono and Stirling.

At Falamai, the 29th and 36th Battal-ions moved inland quickly against 1ightrifle and machine gm fire, mostly fromthe high ground near the Saveke River.Casualties in the first wave were light–-one New 2kdand officer and five sailorswounded-and the second -wave had nocasualties.

The New Zeal anders began to widen theperimeter as more troops were unloaded..&t 0735, enemy mortar and medium artill-ery fire registered on the beach area, caus-ing a number of casualties and disruptingunloading operations. Both LSTS werehit, with one shil) reporting 2 dead and 18wounded among the sailors and soldiersaboard. The other ship reported 12wounded. Source of the enemy fire COUIC1not be determined. The Eaton, with Ad-mirxl Fort on board, ignored a previousdecision not to enter Blanche Harbor andresolutely steamed between the two islands.This venture ended, however, when enemy

planes were reported on the way, and theEaton reversed course to head for moremaneuvering room outside the harbor..4ssured that the air raid was a false

alarm, the destroyer returned to BlancheHarbor and added its salvos to those of the

LC1 gunboats. This fire, directed atlikely targets, abruptly ended the Japa-nese exchange.

By 1800, the two battalions had estab-lished a perimeter on Mono Island andwere dug in, trying to find some comfort

in a dismal rain which had begun againafter a clear afternoon. Evacuation ofcasualties began with the departure of theIJ3Ts. With the exception of one I. ST,which still had 34 tons of supplies aboardwhen it retracted, all ships and landingcraft had been unloaded and were on theirway back to Guadalcanal by the end of D-~kLy. The casualties were 21 h’ew Zea-landers killed and 70 wounded, 9 Ameri-cans killed and another 15 wounded.

The landings at Stirling and Soanotaluwere uneventful and without opposition.There were no casualties at either beach-head. At Stirling, the 34th Battalion im-mediately began active patrolling as soonas the command was established ashore.Tile Soanot al11 landings proceeded in asitllilarly Ilnhindered manner. A perim-eter was established quickly, and bull-dozel’s iulmediately went to work con-Struct in: a position for the raclar equip-ment which was to arrive the next day.

The fighter cover throughout the dayhad shielded the troops ashore from enemyair attacks. The escorting destroyers,however, were hit by an enemy force of 25llux1illn~ and dive bombers at about 1520,and the USS Oony took two hits. Eightcre.~vmenwere killed and 10 wounded. Thetire from the destroyer screen and thefighter Co>.er downed 12 of the enemy

l)laues. That, night the bombers returnedto pollnd the Mono Island side of BlancheIIarbor and, in two raids, killed two NewZealanders and wounded nine.

Action along the Falamai perimeter the11ight of 27 October was concentratedmainly on the left flank near the SavekeRiver, the former site of the Japanesel~e.adquarters, and several attacks werebeaten back. The. following day, patrols

moved forward of the perimeter seeking

DIVERS1ONARY ASSAU1,TS 193

the enemy, and onc reinforced companyset out cross-comltry to occupy the villageof Malsi on the northeast coast. There~l:lslittlcc olltact. .Tilp:LllesegrOulld:t~tiI’-ity on the night of tl~e28tl~\v:~s ligl~t, :~~l(lenemy air :Ictivity was liniitdto one low-level strafing attack ancl severzl quickbombing r~ids-zl 1without danlage to thebrigade group.

By 31 octobm-, the entire situation wasstable. The perimeter at F’a,lamai was se-(’lll’e~3[:llsi IIil-(1 1)(’ell occupie(l witl~o(lt O])-position, ant{ radar equipment at Soano-talu lmcl been installed ancl was in opera-tion. With the arrivxl of the secondechelon on 1 November, the New Zealande-rs bega]l an extel)sive sweep of tl~e islandto search out a]1 remaining enemy troopson the island. The going was rough intile l~igl~, rllgged mountain areas, but, by,; Novel llber, enenky stragg]ers in groups of10 to 12 had been tracked down zncl killed.The New Zealanders lost one killed andfour wounded in these n~op-up operations.

~~ndistilrbed for some tinle, tl~e perim-eter at Soanotalu WZSlater subjected to anllmber of sharp att~cks, ench one. growingin intensity. The Soanota]u force wasstruck first on 29 October by small groupsof Japanese who were trying to reach thebeach after travel ing across the island

from Falamai. These attacks continued

tl~roughout the afternoon until a finalcharge by about 20 .Japnnese was hurledhack. Construction of the radar st.a.tion

continlled throughout the fighting. Enemycontact on the next two days was light, andthe first r:~dar stntion was completed anda second one be~un.

on the l~ight of 1 November, a strongforce of about 80 to 90 .Japanese suddenlystruck the perimeter in an organized at-tack, apparently determined to break

through the New Zealand clefense to seizea hmding craft and escape the island. Theiig]lt, pllllctnated by grenade bursts andmori ar fire, raged for nearly five hours int11Pflarklless. one small group of enemypenetrated the, defenses as far as the beachbefore being destroyed by a commandgroup. About, 40 ,Japanese were killed intl~o attack. The Soanotalu defenders lostone killed and nine wounded. The follow-il~g night, 2 hTovember, another attemptby a smaller ,Japanese force was made andtl~is attack was also beaten back. Thiswns the last organized assault on the Soan-otalu force, and the remainder of theJapanese on the island were searched outand killed by the New Zealand patrolsstriking overland.

l~y 12 November, the New Zealandershad occupie(l the islxnd. ,Japanese deadcounted in the various actions totaled 205;tile Sew Zealanders took 8 l)risoners.It is ckmbtful that any Jxpanese escapedthe islancl by native canoe or swimming.In addition, all enemy weapons, equip-ment, and rat ions on the island were cap-tured. The Allied casualties in this pre-liminary to the _Bougainville operationwere 40 New Zealanders killed and 145wounded. Twelve Americans were killed

and 29 wounded.During the period of fighting on Mono

Island, activity on Stirling was directedtoward the establishment of supply dumps,

the buildin: of roads, and the constructionof advallcc naval base and boat pool facil-

ities. Althollgli several nlinor enemy air

raids darnxged installations in the earlyphases of the operation, the landing atEnlpress Augusta Bay diverted the attel~-

t ion of the enemy to that area and endedall Japanese attempts to destroy the forcein the Treasurys.

,.. .

194 ISOLATION OF RABACL

If the ,Japanese had opportunity tospeculate on the significance of the Treas-urys invasion, the problem mity have beencomplicated a few hours later by a land-i]lg of an Allied force on the northwest.const of Choiseul Island, just, 45 miles fromthe southeasterly] coast of Ilougainvil]e.‘he landing was another ruse to draw,Japanese al tention from the Treasurys,point away from the Allies’ general lineof attack, and divert, the enemy’s interest—if not effoti-toward the defense of an-other area. More specifically, the Choiseuldiversion was calculated to convince the,Japanese that the southern end of Bou-gninville was in imminent danger of attackfrom another direction. The salient factswhich the Allies hoped to conceal were thatthe rei~l objective was Empress AugustaBay, and thzt the Choiseul landing forceconsisted only of a reinforced battalion ofMarine parachute troops.

.4ctually, the raid on Choiseul was asmall-scale enactment of landing planswhich had been discarded earlier.Choiseul was considered as a possible ob-jective for the lnain h’orthern Solomons

‘ ~Tnl~ss(ythel-wi~enoted, the ~ateri~l in t,hl~set.tio]l is (lerirwl from : ComN(/PacOtt–.No?.4.7Il”fIrT); SoI]arF(}rCIC, Study of (IhoiseulISl~n[I,(~td 1~SeIM (Choiseal Are:lOIJFile, HistBr,HQMC) : IIII)lt ibFor .4 R; 1.1[.4C :11{–1, AnxQ. 1;1,1SSkI~TJ: I.lf.-!(’ (’–.2.J)L/; IM.\C 01)() Xo.2. dt<l ?20(t#3 (Choiseal .kreaOl]File, IIist13r,IIQM(’ ) : (’0, M ParnBn,PrelimRept,Op BLISS-F[JL, dtd 3XOY-K+((’hoiseul AreaOpFile, HistBr,IIQ3[C) : ~{1 Il:lr:ll;n ITnit.Jnl,270~t&lh”ov-Kl(Choiseul .\reaOpFile, Hist13r, HQMC) ; 2d1’ar:~Bn 01)() No. 1, dtd 2:30(t43 ((’hoiseul Area-opFile, IIistBr, HQMC) ; MajGen Victor A.I{rnl:lk ltr to A(’ofS, G–3. HQ31C,{Itd 170MI0,here:(fter Krulalc ltr; Rentz, Ilouqai??vilteand

t 1!<’ .\’o)’tlLfr?7 No 10)n0?7,s; Isely and Crowl,.IIIcI’[Ncsand Antph ibious War.

attack; but when the decision was madethat the Allied attack would strike directlyamidships on the western coast of Bou-gainvillea, the Cboiseul idea was dropped.Then, v-hen the suggestion was advancedby Major ,James C. Murray, lMAC Staff.Serretary, tllatl because of the size and lo-cat ion of Choi seul, a feint toward thatisland might further deceive the Japaneseas to the .\ll ies’ intentions, the diversion-ary raid was aclded to the Northern Solo-Tllons operation.

Choiseul is one of the islands formingthe eastern barrier to The Slot; and asone of tl~e Solomon 1slxnds, it shares thel~igh rainfall total, the uniform lli~b heatan(l humidity common to other islands ofthe chain. About 80 miles long and 20miles wicle nt the widest point, C%oiseulis joined by reefs at, the southern end oftwo small islands (Rob Roy aml Wagina)wllicl~ seems to extend Choiseul>s lel~gthanother 20 miles. The big island is not asrlqgged as 130ugainville and tile mountainpeaks are not as high, but Choiseul is fullyas overgrown and choked with rank. im-I)enetrable jungle and rain forest. Themount ain ranges in the center of the islandextend lo])g spurs and ridges toward t]lecoasts, thus effectively dividing the islandinto a series of large compartments. Thebeaches, where existent, vary from wide,s:tlld}~ ~lreas to narrow, rocky Shores with

liewvy foliage growing almost to the wa-

ter’s edge. (lher compartments end inhig]l, broken cliffs, pounded by the sea.

TILeisland was populated by nearly 5,000natives, most o f whom (before the war)

were under the teachings of missionariesof various faiths. With the exception ofa small minority> these natives remained

milital~tly loyal to the ~illstralian govern-ment and its representatives. As a result,

DIVERSIONARY ASSAULTS

comlwatching activities on Choiseul weregiven valuable assistance wld protection.

Combat intelligence about the islandwas obtained by patrols which scoutedvariousareas. Onegroup, lancled from aPT boat on the sol~thmest coast of Choi-seul, moved nortl~~~ard :I1oI1:TIIe Slot sideof tile island towarc] the ,Japanese bzseat Kakasa before tl~rning inland. Aftercrossing the island to the coastwatcher sta-tion at Kanaga, the lmtrol was evacuatedby a li”avy patrol bomber on 12 Septemberafter six days on the island.

Two other patrols, comprising Nlzrinr+,naval officers? and New Zealandelsj scoutedthe northern end of the island and Cl~oi-seul Bay for eight days, ( 22–30 Septem-ber) before being ~vithdrawn. Their re-ports indicated tl~at tl~e main enemystreng-th was at Kakasa where nearly 1,000,Japanese were statioue{l and Choiselll Raywhere another ;100 troops maintained abarge anchorage. Several fair airfieldsites were observed l~ear (loiseul Harbor,and a number of goo(l beac]les suitable forlanding purposes were marked. ,Japaneseactivity, the patrols noted, was general]yrestricted to Kakasa and Choisenl Bay,~

])uring the enenly evacuation of theCentral Solomons, Choiseul bridgecl thegap between the New Georgia Group andBougainvillea. The retreating ,Japanese,deposited by barges on the southern endof Choiseul ~ movecl overland along the

coast to Choisw~l Bay where tile secondhalf of the barge relay to IIougainvil]e wascompleted. This traffic was checked andreported upon by two active coastwatcl~-

ers, Charles J. Wadclell and Sub-Lieuten-

‘ IM.\C I’atrfjl Rel~t. {m Choiseol Bay, 22–:Iosqm dtd 4oct43 : 31 MarI)iv ReI]t of Patrolti) K:iki~~~,CW13Spp-X3,{It(l 16Sep43 (Bougain-villea.lreaOI)File, IIistBr, HQMC).

195

ant, C. W. Seton, Royal Australian Navyjwho nlaint aiued radio co~ltact with Gua-dalcanal.

Seton, on 13 October, reported the south-ern end of Choiseul free of Japanese, butaclded tlmt at least 3,000 to 4,000 enemyhad passed Bambatana Mission about 35miles south of Choiseul Bay. On 19 Octo-ber, the coastwatcher reported that, theenemy camps in the vicinity of ChoiseulDay and %ngigai (about 10 miles northof Bambatana, Mission) helcl about, 3,000,Japanese who were apparently waiting forbnrge, transportation to 130ugainville..Setol~ indicated that the ,Jzpanese weredisorganized, living in dispersed huts, andwere short of rations. They had lootednative gar(lens and searched the jungle forfood. Furthe,r, the Japanese were edgy.ill trails had been blocked, security hadbeen ti~hteued, and sentries fired into thej ut)gle at random sounds.’

Aft er this information was received atIMAC headqlmrters, l’anclegrift andWilkinso]l decided that a diversionaryraid on Choiseul T\-OLdClbe staged. On20 October, Liet~tena]lt colonel Robert H.

Williams, conlmanding the 1st MarineParachute Regiment., and the commandingofficer of his 2d Batttlion, LieutenantColonel Victor H. Krul :Lk,were summonedfrom Vella Lavel]a to Guada]canal. AtIMAC headquarters, Williams and Krulakconferred with Vandegrift and his staff.The orders to Kr~~lak were simple: Get

ashore on Choiseul and make as big :1demonstration as possible to convince the

‘ The .TaImneseuneasiness had an excellentLmsis. S{]b-Lieutenant Seton “had organized 25]l:~tivminto a qn:lsi-military force, armed them(.Japanese weapons) and, on 2 October, am-bnshedan [enemy] group in s l:mding mwft,kill-ing seven.” h-rufuk ltr.

196 ISOLATION OF RABAIJI,

DIVERSIONARY ASS.4ULTS 197

Japanese that a major landing was inprogress. In addition, reconnaissancewould be conducted to detel’mine possiblesites fol a torpedo boat patrol base.

The lMAC opemt ion order, giving thecocle nanle 1)1.ISSFTIL to tile (’hoiseul di-version> was issued o]~ 2’2 October. Basedon information and l.ec~ollll~lel~cl:~tiollsfroluSeton, the Marines’ ]allding \Yas set forthe beaches in the vicinity of Voza vil-lage, about mi(lway between Choiseul Bayand Bambat ana Missiml. Thel’e thebenches were good, friendly hat iYes wouldhelp the invaclinx forces, and there re-portedly were no enemy troops. More-over, it was firmly astride the mxin routeof evacllation of the ,Japa]]ese stragglersfrom Kolombal~gara a)~d points south.After receiving fhe order: Krulak v-ent tothe airstrip on Guadalcanal, and, whilewaiting for a l)lane to take l~im bark tohis comnmnd, wrote out the operation or-der for his b~ttalion’s lancling.

This was to be the first combat operat iol~of the 2d Battalion as well as its firstamphibious venture. Althm~~h equippedand trained for special assignnlents be-hind enemy lines, these, Marines-knownm Pararnarines to their co]rll”:lcies—neT7(’rchuted into action bemuse sllitable ob-jectives were usLmlly beyon(l the rangeof airborne troops and the necessary trans-port planes were in chron icall-y short sup-ply. The 1st Parachute Battalion, how-ever, had taken Gavutu and TanambogoIslands before going to Gllzdalcmlal totake part in the defense of the airfieldthere in 1942. This battalion had thenformed the nucleus for the present 1stParachute Regiment, now consisting ofthree bntta]ions in IMAC reserve nt VellaLave]la. Each battalion, of three. riflecornpmies each, was armed with a pre-ponderance of light automatic and semi-

automatic weapons. The nine-man squadsin Lieutenant Colonel I{rulak’s rifle pla-toons carriecl three Johnson light machineguns G and six Johnson semi-automaticrifles; each company had, in addition, sixWmm mortars.

Krulak’s return to his command set offa flurry of near-frenzied activity, sincetile battalion had a minimum of time forpreparation. For the next four days, offi-cers and men worked almost around theclock to assemble equipment, make finalplans, ancl brief themselves on the taskahead. Ch the 24th, Coastwatcher Setonand two of his native guides arrived atYells I,avella to meet Kru]ak’s officersand illen and give them last-minute infor-mat ion. i~fter being briefed by Seton,Krlllak requested ai~ci was given author-ity by lMAC to operate in any directionon Choiseu], if consistent with his mission.

Eqllipment and supplies for the opera-tion were pre-sorted into four stacks; andlate on the afternoon of the 27th of Octo-ber the parachute battalion and its gear\Y:LSembarked on board eight LCMS bor-rowed fronl the Ve]la Lavella boat pool.Krulak’s three companies were reinforcedby a communications platoon, a regimentalweapons company with mortars and lightmachine guns, and a detachment from anexperimental rocket, platoon (bazookasand rockets) from IMAC. Total battalionstrength was 30 officers and 626 men, Inaddition, one naval officer accompanied thebattalion for reconnaissance purposes re-lated to the possible estabolishment of a

torpedo boat base.At dusk, when four APDs which had

just co]nplet ed the Treasury landings ar-

G“The ,Johnmn light machine gun was more:)n autwrifle th:ln a nuachinegun : more .a m,a-{,hine gun than the BAR.” Ibid.

198

rived off Vella Lavella, the troops andequipment were transferred from theLCMS to the ,VcXean, ~ro~b,y, Kilt,y, andWa~d in a qnick operation that was com-pleted in less than 45 minutes. The de-stroyer division , with the ITSS (?wnuwyacting as escort, sailed from Vella La-vella at 1921. The don?ooy)s radar wou]dlocate the hmding point in tl~e dark.

Moving in rolnmn through the night,the convoy was sigl~ted short 1y after 2300by an enm~ly snooper p] me which droppe(lone bomb, scoring a llezr miss on the lastAPD in line. Shortly before midnight,at a point son~e 2,000 yar(ls off the nortli-west coast of Choiseul, the convoy stopped,and x reconnaissance party in a rubberboat l~eadecl toward shore to scout thelanding area. A signal light was to bcshown if no enemy defenders were en-connterecl. While waiting for the signal.Krulak ordered Companies F and G intothe lanc{ing boats.

After waiting until 0019 (28 october),Company F hea(led toward the beach w’itl~Company G close behind. The operationorder had directed Company G to makethe initial assault, but the APDs haddrifted apart and the Kitty with Com-pany F embarked was closer to shore.Since no light on sl~ore was yet discm-nible,the Marines expected opposition. Thelanding, however, was uneventful, ancl thepatrol was waitin~ on shore. Observerson ship reported later that the light wasvisible at 0023, just four nlinutes after theparachute companies beg-an the run forthe beach. After setting the troopsashore, the 1anding craft immediately re-turned to tl~e transports to bring in a loadof supplies.7

‘ DesDi,\-M AR for night of 27–280ct, InitialI,illlciing of h~~~rineI’:iratroopers oil Choiseul

ISOLATION OF RABAUL

II lone enemy plane detected the ~on-zuay standing patrol duty seaward, anddropped two bombs near the ship. The(’on.ll;(ty: reluctant. to draw attention tothe landing, did not return the fire, andthe enenly l)lal~e droned away. An Alliedescort plane, :Issigned to protect the con-voy agccinst sL~chattacks, d,rew consider-able criticism, however, for not remaininglow enough to spot sl~ch bombing runs.

Two hours after arrival in the area, theconvoy reversed course and steamed backto T’ell a I.avella, leaving behind four land-ing cr:~ft, (LCP (R) ) with their crews fortl~e battalion’s use. These craft were dis-persed llnder the cover of overhangingnml~groves near the offshore island ofZinoz, and the Marines turned to moving

sL~pplies off the beach. Seton, who landedon Choiseul with the battalion, disap-peared into the bush and returned almostinlmedi ately with a group of native bear-ers. With their help, the Marines movedinto the jungle. The transfer was nonetoo soon; enemy reconnaissance planes ap-peared at dawn to bomb the area but -with-out success.

Early on the morning of the 28th, abase of operations was established on a

high jungle plateau about a mile to thenorthwest of ~~oza and outposts were setup on the beac]~ north and south of thevillage. Security was established andwire communications installed. The pla-

teau, behind natural barriers of rivers andhigh cliffs, was an ideal defensive spotand :Lnecessary base camp for the heavyradio gear with which IMAC hadequipped the parachute battalion.

Islan(i, dtd Nov-13: USS Con?ra~ AR, 27–280ct4:l, [M 25Ncw43 (Ci]okeul AreaOpFile,IIistBr, lIQMC).

DIVERSIONARY ASSAULTS 199

During the day of 2?8 October, whilethe Marines established their camp, an-other enemy tligl~t appeared and raked thebeachhead with a strafing ancl bombingattack, The effect was wasted. The Ma-rines had dispersed; their equipn~el~t hadbeen moved; and good camouflage disci-

pline h:td been observed. TOO, t]le *,a.tives had obliterated every sign of a land-ing at, I?oza and established a dummybeachlmacl several miles to tile north forthe special bel~efit of .Japanese pl anes seek-ing a target,

Informed by Seton’s .g~~ides that theMarine battalion, was bi\-ouackeci betweena barge staging-replenishment base :Iti%ngigai fibollt eight miles to the Withand an enemy outpost, at tile WarriorRiver about 17 miles to the north, K,rulalcon the morning of tlle XX 11sent. out recon-naissance patrols to the nortl~ and south.These yrollps were to locate trails, scout,any enemy dispositions, :~ndbecotne famil-iar with the area.

Kru]ak persortal]y led one patrol to-ward Sangi~ai , going overlancl toward theVagara R-i&r which was about halfwaybetween the Marine can~p and the enemybase. While part of tl~e patrol headedinland toward the hig]l ground to therear of Sangi~Ti, to sketch the approachesto the village, the Marine comander ledthe rest of tile patrol to the river. Therethe hidden Marines silently watched agroup of about 10 ,Japanese lmloading ab~rge; and sil~ce this appeard to be anexcellent, opportunity to announce the ag-gressive illtellt ions of tile Kr~llak fo[’~e,

tile Americans opeue(l fire. Seven of the.Jxpal~ese w-ere killed, an(l the barge sunk.Krul:lk’s section then returned to tlle base,follolved shortly by tile other half of thepatrol. .\fter the attack order NW issued,:1 sqllad was sent bzclc over the trail to

691–3600—63——14

the Vag-ara to hole] a landing point forKru] ak)s boats ancl to block the Japanesewho might be following t}le Marines>track. The patrol ran into a platoon ofthe enen]y about three-quarters of a milefrom the original Marine landing pointand drove the <JapaneseOfi.s

Jkt (MOOtile following morning, 30 Octo-ber, Kruhlk lecl Companies E and 1?, plusthe rocket detaclmlent, toward Voza foran .lttzc L- on San~~igai. The barge basella(l been n~arked as a target since 22 Octo-ber. To help him in his assault, andgive the impression of a larger attackingforce, Krulak requested a preparatory airstrike on reported <Japanese positionsnorthwest of the base. Estimated enemystren@~ was about 150 defenders,although Seton warnecl that SangigaicoL~ld have been reinforced easi]y fromtile Southlvest since the Marines7 landing.

I{ru]ak’s attack plans were changed at

VOZ:t, ho~~ever, since one of the four boatshad been damaged a few minutes ear]ie.rin an attack by Mlied planes, The straf-ing ended when the fighter pilots dis-covered their error and apologized to tileboat crews with a final pass a]lcl a clearlyvisible “thumbs-up>’ signal The request-ed air strike at Sangigai hit at 0610 withbetter results. While 26 fighters flewescort, 12 T1~J?s dropped a total of moretl~an two tons of bombs on enemydispositions.

Unable to use the boats for passage tothe Vagara, Krulak ordered his troops to

begin a route march overland from Voza.Seton and his native guides led the way,fol Iow-ec]by Company ~ ( Gtpt ain SpencerH. Pratt) with a section of machine guus

‘ This encounter left Krulak “in no doubtthat we needed to go at them quickly, becausethey were obviously aggressive.” K“rutalcltr.

ISOLATION OF RABAUL

ancl the rocket detachment. Company I?(Captain Robert E. Manchester) and at-tached units followed. ~~t about 1100,Japanese outpostson the l~agara openedfireon the Marinecolumn. Tlrisk returnfire from tl~e parachutists forced theenemy pickets to withdraw towardsSangigai.

Following the en~-elopment plan he hadformulated on tlie 29th, Krnlak sentCompany E along the coast] ine to Iallnchau att ark on ~San~Tiqraifrom that d irec -t ion while the remainder of the force, un-der his command, moved inland to attackfrom the high *~ound to the rear and eastof Sangigai. The assault was set for 1400,but as that, hour drew near, tl~e group inthe interior found that it N-as still a con-siderable distance from the village. Themountainous terrain, tangled closely byjungle creepers and cut by rushingstreams, slowed Krulak’s force, and, byH-Honr, the colllnm was still not in position to make its attack effort. When thesound of firing came from the {lirectiou ofSangigai village, the second force was st illmoving towards its desi~uated jump-offpoint. ,Seton’s natives> however, indicatedthat the enemy \vere just ahead.

Company E, moving along tile bezch,reached its attack position without trouble.Although the assault was delayed a fewminutes, the company opened with an ef-fective rocket, spread and mortar fire. 44sthe Marines moved fmn~ard, the Japanesedefenders lmsti]y withdrew, abandoningthe base and tile village to flee to the highground inland. The hlarine company emtered the village without opposition.”

“The Marineswith Krulak saw the first enemyposition “within a few minutes of E Company’sopening fire. In this sense, the timing wasextremely lucky. Had the enveloping columnbeen30 minutesslower, the .Tapswould ha~egot-

The enemy’s withdrawal to preparedpositions inland fitted perfectly into Kru-lak’s scheme of maneuver. The Japanesemoved from the village straight into thefire of Company F, and a pitched battlethat lasted for nearly an hour ensued. Anenveloping movement by the Marines be-hind the effective fire of light machineguns forced the Japanese into several un-coordinated lm-wai charges which re-sulted in further enemy casualties. Asthe Marines moved once more to turn theenemy’s rigl~t flank, the ,Japanese disen-gaged and about 40 survivors escapedinto the jungle. A final count showed 72enemy bodies in the area. Krulak’s forcelost four killed. Twelve others, includingKrulak and Pratt, were wounded.

Company E, possessors of Sangigai, hadbeen busy in the interim. Manchester’scompany, using demolitions, destroyed thevillage, the Japanese base and all enemysupplies, scuttled a new barge, and cap-tured a number of documents, including achart of enemy mine fields off southernBougainvil]e. The Marines then withdrewto the Vagara to board the four landingcraft (the disabled boat had been re-paired) for the return to Voza. Krulak’sforce, after burying its dead, retraced itspath to the Vagara and spent the night ina tight defensive perimeter.l” Early the

ten away from E Co into the bush. As it was,the sentence in the operation order ‘PreventenemY withdrawal into the mountains’ (}Var1)i:\ry—1600W Dee) worked out well. ” Ifiid.

‘0The ori~inal plan was for the boats to maketwo trips on the X)th, but by the time CompanyE got hack to base it was getting dark. TheImttalion executive officer cancelled the returntrip in view of the dangers of runningthe boatsalong the reef shelf at night. Krulak’s radioh:~dbroken down and so he had no way of learn-inz of this decision, :[lthough he guessed thattl]is was the case. Still, it was an anxions night.

DIVERSIONARY ASSAULTS

next morning, 31 October, the landingcraft returned to carry the parachutiststo Voza and the base camp.

Tt’ith the battalion reassembled oncemore, the Marines prepared ambushes toforestall any Japanese retaliatory attacks,and aggressive patrols were pushed outalong the coast to determine if the Japa-nese were tlmeatening and to keep theenemy off balance and uncertain aboutMarine strength. A Navy PBY landednear Voza the following day to evacuatethe wounded Marines and the captureddocuments; and, on the same day, in an-swer to an urgent, request by Krulak, 1,000pounds of rice for the natives and 250hand grenades and 500 pounds of TNTwere air dropped near Voza. Severalbrisk engagements between opposing pa-trols were reported on this day, 1 hTovem-ber, but the base camp was not attacked.Seton’s natives, however, reported thatSangigai had again been occupied by thetJapanese.

After Krulak returned to the base campon 31 October, his executive officer, MajorWarner T. Bigger, led a patrol to NukikiVillage, about 10 miles to the north. Noopposition was encountered. On the fol-lowing day, 1 NTovember, Bigger led 87Marines from Company G (Captain Wil-liam H. Day) toward Nukiki again to in-vestigate prior reports of a large enemy in-stallation on the Warrior River. Bigger’sinstructions were to move from Nukikiacross the Warrior River, destroying anyenemy or bases encountered, and thenmove as far north as possible to bring themain Japanese base at Choiseul Bay underHhnm mortar fire. Enemy installationson Guppy Island in (Moiseul Bay were analternate target.

The patrol moved past Nukiki without,opposition, although the landing craft

201

carrying the Marines beached continuallyin the shallow mouth of the WarriorRiver. Since the sound of the coxswainsgunning the boats’ motors to clear obstruc-tions was undoubtedly heard by anyenemy in the area, Bigger ordered the Ma-rines to disembark. The boats were thensent downriver to be hidden in a cove nearNukiki. Bigger’s force, meanwhile, leftfour men and a radio on the east bank ofthe river, and all excess gear includingdemolitions was cached. Mortar ammuni-tion was distributed among all the Ma-rines. -The patrol then headed upriveralong the east bank, and the Warrior wascrossed later at a point considerably in-land from the coast.

By midafternoon, the natives leadingthe patrol confessed to Bigger that theywere lost. Although in the midst of aswamp, the Marine commander decided tobivouac in that spot while a smaller patrolretraced the route back to the WarriorRiver to report to Krulak by radio and toorder the boats at Nukiki to return toVoza. In response to Bigger’s message,Kru]ak asked Seton if he had any natives

more familiar with the country north ofthe Warrior River; the only man who hadvisited the region was sent to guide the lostMarines.

The smaller patrol bivouacked at theradio site on the night of 1–2 Novemberand awoke the next morning to the reali-zation that a Japanese force of about 30

men had slipped between the two Marinegroups and that their small camp was vir-tllally sllrrounded. Stealthily slipping

past enemy outposts, the patrol membersmoved to Nukiki, boarded the boats, andreturned to Voza. After hearing the pa-

trol’s report, Krulak then radioed IMACfor fighter cover and PT boat support to

202 ISOLATION OF RABAUL

lvithdraw tIle groul) fronl 1lle ( ‘lloise~d]Iay area.

Bigger was unaware of the activity be-hind him. Intent upon his mission, he de-cided to continue toward Choiseul Bay.After determining his position, Biggerordered another small patrol to make itsway to the river base camp and rac]io arequest that the boats pick up his forcethat afternoon, 2 hTovember. This secondpatrol soon cliscovered the l)resence of anenemy force to Bigger’s rear, and wasforced to fight its way towards hTukiki.This patrol was wziting there, when thelanding craft returned to h-t~kiki.

The main force, meanwhile, followed thenew ~guide to the coast and then turnednorth along the belch toward ChoiseulBay. opposite Redman Island, a smalloffshore islet, a four-man Japanese outpostsuddenly opened fire. The Marines quicklyknocked owt this opposition, but one,Japanese escaped—undoubtedly to givethe alarm.

Because any element of surprise waslost and thinning jungle towards ChoiseulBay provided less protection and coverfor an attacking force, Bigger decided toexecute his alternate mission of shellingGuppy Island. Jungle vegetation grow-ing down to the edge of the water maskedthe fire of the 60mm mortars, so Biggerordered the weapons moved offshore. Theshel]ing of Guppy was then accomplishedwith the, mortars emplaced on the beach~vith part of the baseplates under water.The enemy supply center al~d fllel base washit with 143 rounds of hig]l explosives.Two 1arge fires were observed, one of themobviously a fuel ({ump. Bigger’s force,under return ene]lly tire, turned aroundand heaclecl back toward the WarriorRiver.

The Japanese , attempting to cut offBigger’s ret iren-mnt, landed troops frombarges along the coastline; and the Ma-rine, force W:LS~lllder attack four separatetimes before it successfully reached theWarrior River. There the patrol set upa perimeter oll the west bank and waitedfor the expected boats.

Several mel~ were in the river washingthe slime :Lnd muck of the jungle marchfrom their clothing when a fusillade ofshots from the opposite bank hit the Ma-rine force. The patrol at first thought itwas being fired upon by its own base camp,but, when display of a small American flagdrew increased fire, the Marines dove forcover. Heavy return fire from the Marineside of the river forced the enemy to with-draw. Seizing this opportunity, Biggerclirected three Marines to swim across theWarrior to contact the expected boats andwarn the rescuers of the ambush. Beforethe trio coL~ld reach the opposite shore,though, the ,Japanese returned to the fight,and only one survivor managed to returnto the il[arine perimeter.

Even as the fierce exchange continued,the Marines sighted the four boats makingfor the 11’arrior River from the sea. Anapproaching storm, kicking up a heavysl~rf, added to the clifficulty of rescue. Un-cler cover of the Marines’ fire, the lanclingcraft finally beached on the west shore, andthe Bigger patrol clambered aboard.

one boat, its motor swamped by surf,drifted toward the enemy shore but wasstopped by a coral head. The rescue wascompleted, though, by the timely arrivalof two PT boats—which came on thescene with all guns blazing.’l While the

“ One of these boats was commandedby Liemtenant tJohnF. Kennedy, USNR, later the 33thPresident of the United States, Krutak ltr.

DIVERSIONARY ASSAULTS 203

patrol boats raked the jungle opposite with20mm and .50 caliber fire, the Marinestransferred from the stalled craft, to therescue ships and all craft then withdrew.A timely rain squall helped shield the re-tirement. Aircraft from Munda and PTboats provided cover for the return toVoza.

The time for withdrawal of the battal-ion from Choiseul was near, however, de-spite the fact that Krulak’s force hadplanned to stay 8-10 days on the island.On 1 November, another strong patrol, oneof a series sent out from the ~70za camp tokeep the enemy from closing in, returnedto the Vagara to drive a strong Japaneseforce back towards Sangigai. From allindications, the ,Japanese defenders nowhad a good ideza of the size of the Krulakforce, and aggressive enemy patrols wereslowly closing in on the Marines. Seton’snatives on 3 November reported that 800to 1,000 Japanese were at Sangigai andthat another strong force was at MoliPoint north of Voza.

After the recovery of the Bigger patrolfrom Nukiki, IMAC asked Krulak to makea frank suggestion as to whether the ori-ginal plan shou]d be completed or whetherthe Marine battalion should be removed.The Cape Torokina operation was wellunderway by this time, and IMAC addedin its message to Krulak that Vandegrift’sheadquarters considered that the missionof the parachute battalion had been ac-complished. Krulak, on 3 November,radioed that the Japanese aggressivenesswas forced by their ul”gent need of thecoastal route for evacuation, and that largeforces on either side of the battalion indi-cated that the Japanese were aware of thesize of his force and that a strong attack,prob~bly within 48 hours, was likely. TheMarine commander statid that he had food

for seven days, adequate ammunition, anda strong position; but that if IMAC con-sidered his mission accomplished, he rec-ommended withdrawal.

Commenting later on his situation atthis time, Krulak remarked:

As a matter of fact, I felt we’d not possi-bly be withdrawn before the Japs cut thebeach route. However, we were so muchbetter off than the Japs that it was not tooworrisome (I say now !) The natives wereon our side-we could move across the is-land far faster than the Japs could follow,and I felt if we were not picked up on theVoza side, we could make it on the otherside. Seton agreed, and we had alreadyplanned such a move. Besides that we feltconfident that our position was strongenough to hold in place if necessary.’*

on the night of 3 November, three LCISappeared offshore at a designated spotnorth of Voza to embark the withdrawingMarines. In order to delay an expectedenemy attack, the Marines rigged minefields and booby traps. During the em-barkation, the sounds of exploding mineswere clearly audible. Much to the para-chutists’ amusement., the LCI crews nerv-ously tried to hurry embarkation, expect-ing enemy fire momentarily. Krulak’s bat-talion, however, loaded all supplies andequipment except rations, which weregiven to the coastwatchers and the natives.Embarkation was completed in less than15 minutes, and, sholtly after dawn on the4th of October, the Marine parachute bat-talion was back on Vella Lavella.

Krulak’s estimate of the Japanese in-tentions was correct. Within hours of the

Marines’ departure, strong Japanese

forces closed in on the area where the para-chute battalion had been camped. Theenemy had been surprised by the landing

= Il)id.

204 ISOLATION OF RABAUL

and undoubtedly had been duped regard-ing the size of the landing force by theswift activity of the battalion over a 25-mile f rent. Then, after the operation atEmpress Augusta Bay got underway, theChoiseul ruse became apparent to theJapanese, who began prompt and ag-gres-sive action to wipe out the Marine force.The continued presence of the Alliedgroup on Choiseul complicated the evacua-tion program of the Japanese, and, onceaware of the size of the Krulak force, theenemy lost no time in moving to erase thatcomplication.

Before the battalion withdrew, though,it had killed at least 143 Japanese in theengagement at Sangigai and the WarriorRiver, sunk two barges, destroyed morethan 180 tons of stores and equipment, anddemolished the base at Sangigai. TJn-known amounts of supplies and fuel hadbeen blasted and burned at Guppy Island.Mine field coordinates shown on the cha~captured at Sangigai were radioed to thetask force en route to Cape Torokina,vastly easing the thoughts of naval com-manders who had learned of the existenceof the mines but not their location. Later,the charts were used to mine channels insouthern Bougainvillea waters that theJapanese believed to be free of danger.

The destruction of enemy troops andequipment on Choiseul was accomplishedat the loss of 9 Marines killed, 15 wounded,

and 2 missing in action. The latter twoMarines were declared killed in action atthe end of the war.’s

‘8This is the casualty figure given by Rentz,Bougai)tciUe and the il’orthern Solornons, p. 114.Few accounts of the Choiseul attack are in ac-cord on Marine casualties. Muster rolls of thebattalion indicate 9 KIA, 12 WIA, and 5 MIA.Of those missing, four were later declared deadand one believed a prisoner of war. IMAC C-2

The effect of the diversionary attack up-on the success of the Cape Torokina oper-ation ww slight. The Japanew expectedan attack on Choiseul; the raid merelyconfirmed their confidence in their abilityto outguess the Allies. In this respect,the Japanese were guilty of basing theirplanning on their opponents’ intentions,not the capabilities. There is little indi-cation that enemy forces in Bougainvilleawere drawn off balance by the Choiseulepisode, and enemy records of that pe-riod attach little siamificance to the Choi-seul attack.

This may be explained by the fact thatthe main landing at Cape Torokina tookplace close on the heels of Krulak’s ven-ture and the, ruse toward Choiseul becameapparent before the Japanese reacted suf-ficiently to prepare a counterstroke to it.Certainly, the size and scope of the land-ing operations at Empress Aua~sta Baywere evidence enough that Choiwnd wasonly a diversionary effort.

THE JAPANE~YE 14

Enemy reaction to the Allied moves wasa bit slow. The Japanese knew that anoffensive against them was brewing; whatthey could not decide was where or when.The Seventeenth Army was cautionedagain to keep a watchful eye on Kieta and

.Jnl,4Nov43, and the report of the diversion at-tack, Operation BLISSFUL, p. 4, indicate that9 KIA and 16 WIA is correct. IZIPhibFor AR,

pp. 34, states that 8 KIA, 14 WIA, 1 MIA, and1 captured is correct. LMA C AR–I, p. 11, givesthe casualties as 8 KIA, 14 WIA, and 1 MIA.ONI, Combat Narrative XII, p. 24, gives thelosses as 9 KIA, 15WIA, and 2 MIA. The figuregiven by Rentz undoubtedlytakes into cognizancea 13Dec43 message from Coastwatcher Seton tothe effect that the bodies of two Marines, one ofthem bound as a prisoner, had been found nearthe Warrior River. Co?nfYoPac De+ WarD.

DIVERSIONARY ASSAULTS

Buka, and General Hyakutake in turndirected the 6th Division to maintain afirm hold on Choiseul as well as strongpositions in the Shortland Islands. Then,the tJapanese defenders on Bougainvilleawaited for the next developments.

After the Allied landings in the Treas-urys, the Japanese thinking crystallized:Munda was operational; Vella Lavella was

not. Therefore, the only targets withinrange of NewT Georgia were the Short-hands or Choiseu]. And based upon this

reasoning, the .illies scarcely would at-tempt a landing on 130ugainvil]e before

staging bases on Mono or Choiseul werecompleted. Reassured by this assumption,the ,Japanese relaxed, confident that thenext Allied move would come during thedark quarters of the moon—probably latein November.

With the All iecl move toward Choiseul,

the Japanese were more convinced thatthe Allied pattern was predictable. Witha firm foothold on Mono and Chois@ theAllies would now move to cut Japaneselines and then land on the southern partof Bougainvillea in an attempt to seize theisland’s airfields. Basing their estimateson the increased number of Allied airstrikes on Buka and the Shorthmds, the

Bougainvillea defenders decided that thesewere the threatened areas. .411 signspointed to a big offensive soon—probably,the Japanese agreed—on 8 December, thesecond anniversary of the declaration ofwar.

“ Unless otherwise noted, the material in thissection is derived frt)m: SE .4rea. NaoOpS-III;

Scm’ntwnth Army Ops-II; Rentz, Bougainvitlea~~dthe .Vorthwn Nolomons; Miller, Reduct@nof Rubaul; Morison, Breaking the Bismarcka

Barrier.

205

The enemy had no hint that such an un-likely area as Empress Augusta Baywould be attacked. The defense installa-tions were concessions to orders directingthat the western coast be defended, and thetroops at Mosigetta-the only force cap-able of immediate reinforcement to theCape Torokina area—were alerted only tothe possibility that they might be divertedon short notice to the southern area to de-fend against an assault there.

Japanese sea and air strength was like-wise out of position to defend against theBougainvillea thrust. Admiral MineichiKoga, commander in chief of the Com-bined Fleet at Truk, had decided earlier toreinforce Vice Admiral Jinichi Kusaka’sSoutheast .4rea P7eet and the land-basedplanes of the Eleventh Air F7eet at Rabaulso that a new air campaign could be aimedat the Allies in the South Pacific. Thisoperation, li?o, to start in mid-October,was to short-circuit Allied intentions bycutting supply lines and crushing anypreparations for an offensive. To Ku-saka’s dwindling array of fighters, bom-bers, and attack planes, Koga added theplanes from the carriers Zuikaku, Sho-kai%u,and Zuiho-82 fighters, 45 dive bom-bers, 40 torpedo bombers, and 6 recon-naissance planes.

Koga’s campaign, though, was delayed.Allied radio traffic indicated that eitherWake or the Marshall Islands would behit next, and to counter this threat inthe Central Pacific, Koga sent his fleetand carrier groups toward Eniwetok toset an ambush for the Pacific Fleet. Aftera week of fruitless steaming back andforth, the Japanese force returned to Truk,and the carrier groups moved on to Ra-haul. The Japanese admiral had at firstdecided to deliver his main attack againstNew Guinea, but the Treasurys landings

206 ISOLATION OF RABAUL

caused him to swerve towarcls the Solo- operation. The elements of the Japanesemons. Then, when .Mliecl activities be- fleet reached the area north of the Bis-tween 27 October and 1 IYovernber dwin- mamks on 1 November> just in time todied, the fleet again turned toward New head back towards the Solomons to try toGuinea to take up the long-delayed Ro interrupt the landings at Cape Torokina.

CHAPTER 3

Assault of Cape Torokina

;{. . . THE TROOP8 AREMA G.VIFI(~l?NT.7’ 1

The Northern Landing Force arrivedoff Empress Auagusta Bay for the assaultof Cape l’orokina shortly after a brightdawn on 1 November, D-Day. The ap-proach to the objective area had been un-eventful. After rendezvousing near Gua-dalcanal, the transports steamed aroundthe southern and western coasts of Ren-

‘ unless otherwise noted, the material in thissection is derived from : ComNoPac .No043 Warn;Tltir(7F7t Xarrl{rpt; IIIPII ibFor :1R; IIIPhib-For 3’0174,{Tt”arD; ComTransGru, IIIPhibFor,Rept of I.:nnrlin@ps,EmpressAugustaBay, Bou-~:lin~~illeIsland, 1—2Nov43,dtd 22Dec43 (COA,NHD) ; 11~.1C AR-I; IJfAC’ C–2 Rept.s;I,IfAC?(’–zJn7;M Marniv, Combat Rept of the 3dMarDiv in the BougainvilleOps, lhTov-28Dec43,dtd 21Milr44 (IIougainville .LreaOpFile, HistBr,HQM(<), hereafter ;?d .lIarDii ~onlbatl{ept; ,?d

.J[arJ)iv AR; M MarT)ivD–1 Jnl, 10ct–14Nov43(Ilougainville AreaOpFile, HistBr, HQMC) ; 3dMarI)iv D–2 SAR, Empress .iu.gusts Bay Opsdtd lFeb44 (130ugainvilleAreaOpFile, HistBr,HQMC), hereafter .’3(7MarI)iv D–2 SAR; 3d31:lrDiv 1)–2 .Jnl, 2SOct–28Dec43 (130ugain-ville .\reaO1)File,HistL3r,HQM~,). hereafter :;/il[arJ)iv D–2 ,Jnl; 3d MarI)iv D+ .Tnl,310ct–281)w43 (Rouxainville AreaOpFile, HistBr,HQJIC), hereafter $ftMarDio D–3 Jnt; 3d Mar-J)iv D–3 Periodi(,Repts,(?.Tu-27Dec43(Bougain-villeaAreaOp],’ile,HistJ3r,HQhIC) , hereafter 3,1MarJ)iL,r)–3 Rrpt.s;Hi,silliv .4rrt; OAT,Combot

Yarratire .1’11; He]lderson,“X”avalGunfireSulJ-port :“ Aurthur and Cohlmia, 3d .lfarJ)ioHi.st;

Rentz, T{onqainvi{!e and the Northern Eolomons;Isely and (’rowl, Marines and Am~JIIibio?ts’ l~ar;

hfiller, Reduction of Rabaul.

dova and Vella Lavella toward the Short-land Islands. ComAirSols fighter planesprovided protection overhead and destroy-er squadrons screened the flanks. Subma-rines ranged ahead of the convoy to warnof any interception attempt by the enemy.

When darkness fell on 31 October, theconvoy abruptly changed course and,picking L~pspeed, started the final sprinttoward Empress Augusta Bay. Minesweepers probed ahead for mine fields anduncharted shoals, while Navy patrolbombers and night fighters took stationover the long line of transports. Eightair alerts were sounded during the night.Each time the night fighters, directed bythe destroyers, intercepted and chased theenemy snooper planes away from the con-voy. The amphibious force, moving di-rect as an arrow toward the coast of Bou-gainvillea, was never attacked.

Nearing Empress Augusta Bay, the con-voy slowed so tl~at the final movement intothe objective area could be made in day-light. General quarters was sounded at.0500, and, after the sun came Lip, the as-sault troops on the transports could seethe clark shoreline and rugged peaks ofBougainvillea directly ahead. Only a thincloud mist hung over the island, scant con-cealment f o,r enemy planes which couldhave been waiting to ambush the amphib-ious force. The element of surprise, whichhad been zealously ~guarded during allpreparations for the offensive, apparentlyhad been retained. The conflicting re-

207

208 ISOLATION OF RABAUL

ports by Japanese snooper planesof taskforces observed at various points fromBuka to the Shorthands and Vella Lavellahad the general effect of confusing theJapanese.z

At 0545, mine sweepers and the destroy-er Wadsworth opened fire on the beachesnorth of Cape Torokina to cover their ownmine-sweeping operations. As the Wads-worth slowly closed to within 3,000 yardsto fire directly into enemy installations,the busy mine sweepers scouted the bay.Thirty minutes later, advised that nomines had been found, the transportsmoved into the area. Off Cape Torokina,each APA shelled the promontory withranging 3-inch fire before turning hard topofi- to take Pnruzta Islancl under firewith 20mm guns. i~t 0645, the eight trooptransports were on line about 3,OOOyardsfrom the beach and parallel to the shore-line. Behind them, in a similar line, werethe four cargo transports with the de-stroyer squadrons as a protective screenseaward.

On board the transports, observerspeered anxiously toward the beaches nearthe Laruma River. A two-man patrolhad been landed on Bougainvillea on Dminus 4 days (27 October) with the mis-sion of radioing information or lighting asignal fire near the Laruma if the CapeTorokina area was defended by less than300 Japanese. Concern mounted as H-Hour approached without the expectedmessage or signal. The alternatives werethat the patrol had been captured or thatthe cape area was unexpectedly reinforcedby the enemy. Because the landing waveshad been organized to handle cargo andsupplies at the expense of initial combat

‘ i3h’ Area NavOps-IZI, p. 12,

strength, any change in the enemy situa-tion at this late date was cause for worry.H-Hour, set for 0715, was postponed for15 minutes on signal from Admiral Wil-kinson, but the landing was ordered asplanned. (The patrol later reported un-harmed, citing radio failure and terraindifficulties for the lack of messages.)

Preparatory fires by the main supportgroup began as soon as visibility permit-ted identification of targets. From theirfiring positions south of the transportarea, the Anthofiy and Sigourney-andlater the Wadsworth — poured 5-inchshells into Puruat a Island and the beachesnorth of Cape Torokina. The Terry, onthe left flank of the transport area, firedinto known enemy installations on thenorth shoulder of the cape. The effectwas a crossfire, centered on the beachesnorth of Cape Torokina. This indirectfire on area targets was controlled byspotter aircraft.

Debarkation of troops began after thetransports anchored in position and whilethe pre-assault bombardment crashedalong the shoreline. The order to landthe landing force was given at 0645, andwithin minutes assault craft were loweredinto the sea, and embarkation nets tossedover the side of the transports. Marinesclambered down the nets into the boats,and, as each LCVP was loaded, it joinedthe circling parade of landing craft in therendezvous circles, waiting for the signal

to form into waves for the final run to thebeach. Nearly 7,500 Marines, more thanhalf of the assault force, were boated for

the simultaneous landing over the 12beaches.

At 0710, the gunfire bombardmentshifted to prearranged targets, and fiveminutes later the first boats from the

210 ISOLATION Oli’ RABAUL

APAs on the south flank of the transportarea started for shore. The support. shipscontinued to shoot a.t beach targets unt i]0721, when the shelling was lifted to covertargets to the rear of the immediate shore-line. As the fire lifted, 31 torpedo andscout bombers from Munda streaked overthe beaches, bombing and strafing theshoreline just ahead of the assault boats.The planes, from VMTB-143, -232, and–%X3, and lTMSB-144, were covered byVMF-215 and -221 and a Navy fightersquadron, VF-–17.3 The air strike lasteduntil 0726, cut short, four minutes by theearly arrival of the first landing craft atthe beaches.

The 9th Marines (Colonel Edward A.Craig) landed unopposed over the fivenorthernmost beacheeRed 3, Red 2, Yel-low 4, Red 1, and Yellow 3. Althoughno enemy fire greeted the approach of theboats, the landing was unexpectedly haz-ardous. Rolling surf, higher and roug-h-er than anticipated, tossed the landingcraft at the beaches. The LCWPS and theLCMS, caught in the pounding breakers,broached to and were smashed againstshoals, the beach, and other landing craft.The narrow shoreline, backed by a steep12-foot embankment, prevented the land-ing craft from grounding proper]y, andthis further complicated the landing.

Some boats, unable to get near the shorebecause of rough surf and wrecked boats,unloaded the Marines in chest-deep water.Other Marines, in LCVPs with collision-damaged ramps, jumped over the sides of

the boats and made their way to shore.In -spite of these difficulties, the battalionlanding teams managed to get ashorequickly, and, by 0750, several white para-chute flares fired by the assault troops in-

2SherrocL,MarAirHist, p. 181,

dicated to observers on board ship that thelanding was successful.

Once ashore, combat units of the 9thMarines completed initial reorganizationand moved inland to set up a perimeteraround the five beaches. Active patrol-ling was started immediately, and a strongoutpost was set up on the west bank of theLarurna River. Other Marines remainedon the beach to help unload the ‘ank light-ers and personnel boats which continuedto arrive despite the obvious inadequacyof the beaches and the difficult surf. Atleast 32 boats were wrecked in. the initialassault and lay smashed and awash alongthe beach. By mid-morning, hulks of 64LCVPS and 22 I,CMs—many of them be-yond repair—littered the five beaches.’

The landings on the six southern beaches( Yellow 2, Blue 3, Blue 2, Green 2, Yellolv1, and Blue 1) and the single beach onPuruata Island (Green 1) were in starkcontrast to the nort hem zone. Enemy re-sistance in this area wzs evident almostas soon as the boat groups from the right-flank transports came within range. The2d and 3d Battalions of Colonel George W.hfcHenry’s 3d Marines landed on thethree beaches south of the KoromokinaRiver against small-arms fire. Surf washigh but not difficult, and no boats werelost. The Marines, disembarking withoutclelay, sprinted across the narrow beacht.o take cover in the jungle. Reor~~niz:~-tion was complekd quickly, and the bat-talions startec] to dig out the small nuln-ber of Japanese defenders attempting tohold back the assault, from hastily pre-

4“Many automatic weapons were mounted onthese landing boats. These weapons were sal-va~ed by the Marines and used later to reinforcetheir normal arms when the final defensive line\~ilsestablished.” I,tGen Edward A. Craig ltrto CMC, dtd 240ct60, hereafter Craio ltr.

ASSAULT OF CAPE TOROKINA 211

pared positions. In a fev- nlinlltes, thescattered enemy in the area had been killed,and sniper patrols began n~oviug inland.Contact was established with the 9thMarines on the left, but a wide sl~anlp pre-vented linkup with the 2cl Raider 13at-talion on the rigl~t.

The raiders, Iecl by Major Richard T.Washburn, went ashore in the face ofheavy machine gun and rifle tire from twoenemy bunkers and z number of support-ing trenc]~es about 30 yards inland. J:ip-anese defenders v-ere estimated at about,a reinforced platoon. After the first SaV-age resistance, the enemy fire slackened asthe raiclers blasted the bunkers apart tokill the occupants. Otller enemy soldiersretreated into the jungle. only after thebeach area was secured did the raiders dis-cover that the regimental executive officer,Lieutenant Colonel ,Joseph W. McCaffery,had been fatally wounded while coordinat-ing the assault of combat units against theenemy dispositions.s

Extensive lagoons and swwnpland back-ing tile narrow beach limitecl reconnais-sance eflorts, and reorganization of theassault platoons and companies was hin-dered by constant sniper fire. Despitethese handicaps, the raiders pushed slowlyinto the jungle and, by 1100, had wiped outall remaining enemy resistance. RaiderCompany M, attached to the 2d RaiderBattalion for the job of setting up a trailblock farther il~land to stall :Lnyenemy at-tempt, to reinforce tile beachhead, moveclout, along the well-marked Mission Trailand was soon far out ahead of the raiderperimeter.

The 1st 13atta]ion of the 3d Marines hitthe hot, spot of the enemy defenses. As

5MajGen Alan Shapley ltr to ACofS, G–3,HQMC,dtd 130ct60,hereafter S1/aplc~ ltr.

the waves of boat groups rounded thewestern tip of I>uruata Island, they werecaught in a vicious criss-cross of machinegLUIand artillery fire from Cape Torokinaand Purl~ata and Torokina Islands.Heading towarcl the extreme right of thelanding area over beaches which includedCape Torokina, the 1st Battalion plougl~edashore straight through this deadly cross-fire. An enemy 75mnl artillery piece,which had tried emlier to hit one of thetransports, remained under cover duringthe aerial bombing and opened fire a~~inonly after the assault boats reached apoint sorn~ 500 yards offshore. Its loca-tion was such that all boats heading to-ward the beach had to cross the firinglane of this g-LuI.’

One of the first casualties in the assaultwaves was the LCP carrying the boatgroup commander. The command boat,blasted by a direct hit, sank immediately.The explosion resulted in dispersion, dis-organization, and confusion among theboat group. In a split second, the ap-proach formation was broken by landingcraft taking evasive action to avoid thexntiboat fire.

The result w-as a complete mixup ofassault, waves. A total of six boats werehit within a few minutes; only four ofthem reimaged to make the bezch. Astl~e first waves of boats grounded on thebeaches, the ,Japanese opened up with ma-chine gun and rifle fire, and mortar burstsbegan to range along the shoreline. Awithering fire poured from a concealedcomplex of log and sand bunkers con-nected by a series of rifle pits and trenches.

8Col George O. Van Orden ltr to CMC, dtd27May48 (Bougainvillea Monograph CommentFile, HistBr, HQMC) places the location of thisgun well within the limits of Blue Beach 1, thelanding area of the 1st Battalion, 3d Marines.

,.,

212 ISOI,~TION OF R~BA

.

o1-

H

...z

:JI.

ASSAULT OF CAPE TOROKINA 213

The enemy emplacements, barely aboveground level and hidden beneath the

tangled underbrush along the shoreline,were sited to cover the beaches and baywith interlocking bands of fire. The pre-assault bombardment by glmfire ships andplanes had not knocked ont tile enemyfortifications; in most cases it had not evenhit them.

The Marines, -with all tactical integrityand coordination lost, p]unged across

the thin strip of beach to take cover in thejungle. An orderly landing against suchconcentrated fire had been impossible.After the scrambling of the assault waves,units from the battalion landing team hadgone ashore where l)ossible and practicallyevery unit was out of position. Contribut-ing to this confusion was the fact that themajority of the boats hit were LCPScarrying boat group commanders. TheJapanese, correctly surmising that themore distinctive. LCPS were commandcraft, directed most of their fire on theseboats.

The initial reorganization of the ele-ments of the battalion landing team washandicapped further by the wounding andlater evacuation of the battalion com-mander, Lieutenant Colonel Leonard M.Mason. Control of boat teams was dif-ficult under the pounding of 9tlmm and“knee mortar” ‘ bursts mixed with theraking fire of machine guns and rifles.

Platoons and squads from all companieswere mixed along the bench. The original

plans directed Company A to land on CapeTorokina, but after the assault waves weredispersed and tangled by the effective fireof the Japanese 75nlm artil]ery piece, e]e-ments of Company C landed on the pronl-

7The common term for the ,Japanese50mmxrenwle discharger.

ontory. Several squads from CompanyF of the 2d Battalion also landed in thisarea and were forced to fight their wayalong the beach to reach their parent unit.Only Company B of the 1st Battalionlanded on its assigned beach. Casualtieswere fairly light, though, despite the in-tense fire of the enemy. In addition to atleast 14 men lost in the landing craftwhich had been sunk, fewer than a dozenMarines had been killed on the beach.

The 1st Battalion hesitated only a shorttime; then the extensive schooling of thepast asserted itself. Training in smallunit tactics against a fortified positiont~ow paid big dividends. Rifle groupsbegan to form under ranking men, and thefight along the shoreline became a numberof small battles as the Marines fought to.widen their beachhead against the enemyfire. As the Marines became oriented totheir location and some semblance oftactical integrity was restored, the pace ofthe assault, quickened.

Before the operation, all units had beenthoroughly briefed on the mission of eachassall]ting element, and each squad, pla-toon, and company was acquainted withthe missions of other units in the area. Inaddition, each Marine was given a sketchmap of the Cape Torokina shoreline.Small groups formed under the leadershipand initiative of junior officers and staffl~oncomrnissioned officers, and these groups,in tllrn, were consolidated under one con-mand by other officers. Bunker afterbunker began to fall to the coordinatedand well-executed attacks of these groups.As the Japanese defensive complex slowlycracked, the 1st Battalion command wasestablished under the battalion executiveofficer, and the hastily re-formed com-panies took over tile mission of the areain which they found themselves.

214

The efficient reduction of the enemy’s de-fensive position added another testimonialto prior training and planning. officersof the 3d Marine Division had studied theJapanese system of mutually prokctingbunkers on New Georgia and decided thatin such a defensive complex the reductionof one bunker would lead to the elimina-tion of another. In effect, one bunker un-locked the entire position. The quickestway to knock out such pillboxes with thefewest casualties to the attacking forcewas for automatic riflemen to place fireon the embrasures of the bunker whileother Marines raced to its blind side todrop grenades down the ventilators orpour automatic rifle fire into the rear en-trance.

By midmorning, through such coordi-nated attacks, most of the Japanese bunk-ers on Cape Torokina had been knockedout. The position cent aining the murder-ous 75mm gun was eliminated by oneMarine who, directing the assault of a rifle

.ggoup, crept Up to the bunker and killedthe gun crew and bunker occupants beforefalling dead of his own wounds. Afterthe last emplacement was silenced latethat afternoon, Marines counted 153 dead<Japanesein the Cape Torokina area.

For a while, the situation on the rightflank had been touch and go. One hourafter the landing, a variation of the time-honored Marine Corps phrase was flashedfrom the Cape Torokina beach. “The sit-

uation appears to be in hand,” was thefirst message, but a few minutes later amore history-conscious officer flashed anamended signal: “Old Glory flies on Toro-kina cape. Situation well in hand.” Themost expressive message, however, to ob-servers on board the transports was thereport from a young officer to Colonel

ISOLATION OF RABAUL

McHenry: “ . . . the troops are mag-nificent. ” 8 The Marine officers who haddirected the assault on the fortified posi-tions added sincere endorsements to thisexpression of admiration.

On Puruata Island, the 3d Raider Bat-talion (Lieutenant Colonel Fred D.Beans) landed with one reinforced com-pany in the assault and the remainder ofthe battalion as reserve and shore party.Only sporadic fire hit the boats as theyneared the island. By 0930, the raidershad established a perimeter about 125yards inland against hidden snipers andaccurate machine gun and mortar fire.The Japanese, obviously outnumbered,gave little indication of yielding, and, by1330, the reserve platoons of the battalionwere committed to the attack. The raid-ers, with the added support of several self-propelled 75mm guns attached from the9th Marines, then moved about halfwayacross the island.

Puruata was not declared secured untilmidafternoon of the following day. Atwo-pronged attack, launched by the raid-ers early on the morning of 2 November,swept over the island against only spo-radic rifle fire, and, by 1530, all Japaneseresistance on the island had been erased.Only 29 dead enemy were found, althoughat least 70 were estimated to have been onthe island. The remainder had appar-ently escaped to the mainland. Theraiders lost 5 men killed and 32 woundedin the attack,Y

83d MarRegt Jnl, 1Nov43 (BougainvilleaArea-OpFile, HistBr, HQMC).

‘ This is the numbergiven in .!?dMwDiv COTWbat Rept, p. 343. The same report, on p. 153,also gives 6 KIA and 18 WIA for this part of theoperation. In cases of discrepanciessuchas this,the report of the unit engaged is given in thisaeeount.

216

The Marines’ fight uncovered extensiveenemy defenses which were not disclosedin aerial photographs taken before theoperation. The entire headland wasringed by 15 bunkers, 9 of them facingto the west and 6 of them overlookingthe beaches on the east side of the cape.Behind this protective line and fartherinland was another defensive line of eightbunkers which covered the first line offortifications. Two other bunkers, about750 yards inland, provided additionalcover to the first two lines.

Constructed of ironwood and coconut]c)gs t~~o feet thick, t]le bunkers ~Yerebill.warked by sandbags and set low intothe ground. Camouflaged by sand andtangled underbrush, the bunkers werehard to cletect and difficult to knock outwithout flamethrowers or demolitions.Despite this, the 3cl Marines suffered fe~~casualties in destroying this defensive in-stallation. Twenty of the bunkers hadbeen eliminated by the coordinated fire anclmaneuver of individual Marines; the re-maining five were blasted apart by self-propelled tank destroyers firing 75n~marmor-piercing shells directly into theembrasures.

The enemy 75mm artillery piece sitedas a boat gun hit 14 boats during the ini-tial landings before it was put out ofaction. Only four of the boats sank. De-spite the high velocity of the shells andthe slow speed of the landing craft, the 50or more rounds fired by the enemy scoredremarkably few hits. This was attributed

to two factors: the poor accuracy of theJapanese gunners and the limited traverseof the gym. Marines found, after knock-

ing out the bunker, that the aperture inthe pillbox permitted the muzzle of thegun to be moved only three degrees either

ISOLATION OF RABAUL

way from center. This prohibited the gunfrom bringing enfilade fire to bear on thebeaches. Had this been possible, the largenumber of boats along the shoreline wouldhave been sitting targets which even poorgunners could not, miss, and the casualtiesto the landing force would have been cor-respondingly greater.

The unexpected resistance on CapeTorokina and Puruata island after thenaval gunfire bombardment and bombingwas a sharp disappointment to lMAC of-ficers who had requested much more exten-sive preparatory fires. The gunfire plan,

which was intended to knock out or stunenemy defenses that might delay the land-ing, had accomplished nothing. The .4n-thony, firing on Puruata Island, reportedthat its target had been well covered;but tile raider battalion, which had to digthe clefenders out of the emplacementson the island, reported that few enemyinstallations had been damaged.

The Wudsworth and Sigourney, firingat ranges opening at 11,000 to 13,000yards, had difficulty hitting the area andninny shots fell short of the intended tar-gets. The Terry, closest to the shore butfiring at an angle into the northwesternface of Cape Torokina, was poorly posi-tioned for effective work. &Tone of the25 bunkers facing the landing teams onthe right had been knocked out by gunfire,and only a few of the ,Japanese huts andbuildings inland were blasted by the ships’fire. The gunnery performance of thedestroyers left much to be desired, III-PhibFor admitted later. Particularlycriticized was the fact that some ships firedshort for almost five minutes with all sal-vos hitting the water. After two or threerounds, the range should have been ad-justed, but apparently the practice bomb-ardment at Efate had not been sufficient.

ASSAULT OF CAPE TOROKINA

Thelong-range snipiug at ~ape Torokinawith inconclusive results was vindicationfor the lMAC requests prior to the opera-tion that, tile clestro-yers move as close tothe shoreline as possible for direct fire.

The sa{l thing about the tyhole show, tothe t,or]man[l tlivisiou guufire I)lanners,w:lsth:lt the n]e:~msactually Tyere~lvail:~bletoA.iveus just wh:~t\ve~v;intwl,ljllt~~ereflissi-I)ated elsewhere in what we felt was fruit-less cmnnonacling,’”

Valuable lessons in gunfire support werelearned at Ilougainville that D-Day. Forone thil~g, the line of flat trajectory fire insome places passed through a fringe oftall pxlm trees which exploded the shellspremztnrely and denied direct observationof the tnrget area. Further, the ships hadtrouble seeing the shoreline through thecombination of early morning haze zndthe smoke and dust of explodinx shellsznd bombs, rising n~~inst a mountainousbitck~ronnd.

.\lt bough the enemy airfiel ds in the 130u-gainville m-es were knocked out by Ad-miral Merrill’s final bombardment and theprior action of Com AirSols bombingstrikes, the ,Japnnese reaction to the lancl-ing came swiftly. .it 0718, less thfin twoIIours after the transports nppeared offCape Torokina and about eight minutesbefore the first assault boats hit the befich,a large flixht of ,Japanese plmnes was de-tected winging toward Empress AugustaBay. The transports, most of them trail-ing embarkation nets, immedi ttely pulledout of the bay toward the sex to take eva-sive action again.

The first enemy flight, of abont 30 planes,evidently fighters from the naval carriergroups land-based in New Britain, was im

‘0Henderson, “x’aval Guntire Support,” pp61-62.

217

tercepted at about 0800 by a New Zealandtighter squadron flying cover over thebefichhead. Seven of the Japanese planesIvere knocked down, but not before a fewenemy raiders strafed the beaches anddive-bombed the frantically maneuvering.~P.% and .IKAs. Ten minutes later, sn-other flight of enemy fighters and bombersstruck the area in a determined attack, butwere turned away by the fierce interferenceof other ComAirSols planes, includingMarine fighters from TTMF-215 and VMF-221. Radical evasive tactics by the trans-ports—aided by excellent antiaircraftgunnery by tl~e destroyer screen and sav-age pulswit by the fighter cover—pre-vented tl~e loss of any ships, although theWeds ~rmth took some casualties from anear miss. Tl~e fighter cover downed eightplanes, and the destroyer screen claimedanother four raiders.

Two hours after the attack began, tl~e.~P.\s and AI<.Is returned to resume op-erations. Valuable time, however, hadbeen lost. Intruding enemy planes con-tinued to harass the transports, but un-lcmding operations kept up until about1300, when the arrival of another ]argeformat ion of about TO enemy planes putthe ships into action a~~in.

One APA, the American Legion,grounded on a shoal and remained thereduring the attack despite the persistent ef-forts of tw-o tugs which attempted to freeit. A destroyer resolutely stood guard,pumping antiaircraft fire into attackingplanes. The ship was pulled free beforetl~e air attack was driven off. As before,the :Iggressive fighter cover and heavy firefrom the destroyer screen prevented dam-wge to the amphibious force, and the shipsturned bxck to the task of urdoading.During the attacks, the Allies claimed 26enemy planes as shot down-four more

218

than the ,J~panese records indicate—withthe loss of four planes and one pilot. Forthe first day, at least, the threat of enemyair retaliation had been turned back.

ESTABLISHING A BEACHHEAD “

Ashore, the defensive perimeter nowstretched a long, irregular semi-circle overthe area from the La.rums. River past CapeTorokina, a distance of about four miles.Only the northern beaches were quiet; thearea around the cape was still being con-tested by snipers \vithin this perimeter andby small groups of enemy in the jungleoutside the line. Within this area, thelogisticsSitllat,ion was beginning to be

cause for concern.Confusion began after wrecked tank

lighters and personnel boats were brokenon the northern beaches, closing thoseareas to further traffic. When unloadingoperations began once more after the first

air raid, the northern beaches were orderedabandoned and all cargo destined for the!%h Marinss sector was diverted to beaches

south of the Koromokina River. This

change, the only move possible in view of

11Ilnless ~therlvisenoted, the material in thissection is derived from : CizcPae-CincPOA

Nov.j3 Warn; Corn SoPac A’0vJ3 J+TarD; Tltird-

Flt NnrrRcpt; IIIPh ibFor .4R; lIIPhibPor

~ov~~ Ji7arD; IMAC AI<–Z; $d .liarDiv Cow bat

Rept; 3d .l[arDiti AR; 3d MarI)iv 1)–4 Jnl 1-16xOTT# (Rougainville AreaOpFile, HistBr,HQMC!). hereafter .?dMarIJivIPl J*1Z; 3d Mar-Div ServTrps Rept of Ops, Nov–Dec43, dtd27Jan44 (Bougainvillea AreaOpFile, HistBr,HQMC), hereafter $d MarDiv S’crvTrp.s Rept; 3d3fi{rDiv SIIpply and E~ac Rept, DIPPER Opera-tion, dtd 29.Jan44 (BougainviUe AreaOpFile,HistBr, HQMC) : Hi.stDiv Acct; OATI,ConlbatNarra,tice ,YZZ; Aurthnr and Cohlmia, 3d ~far-DivHist; Rentz, Bougai?wil!e and the Northerni3010mons; Morison, Rrcakinfl tJ\ r Bisntarcks

Rarrier.

ISOLATION OF RABAUJJ

the difficult surf conditions, led to furthercomplications because the beaches in the3d Marines’ sector were already crowded,and the coxswains on the landing crafthad no instructions regarding where thesupplies should be dumped.

The landing beaches in the 3d Marines’sector were hxrdly an improvement. Fewhad any depth, and from the outset it wasapparent that the swampy jungle wouldstall operiat.ions past the beaches. Theonly means of movement was laterallyalong the thin beach, and the gear alreadystacked along the shoreline was causingcongest ion along this route. The difficultterrain inland made the formation ofdumps impractical, so all carg-o was placedabove the high water mark and some de-gree of orderliness attempted. Despitethis, the 9th Marines lost much organiza-t ional property and supplies, most ofwhich was never recovered.

As cargo and supplies mounted on thebeaches assigned to the 2d and 3d Elat-talions of the 3d Marines, the. bulk cargowas diverted further to Puruata Island.The landing craft were unloaded within afew hundred yards of the battle betweenthe raiders ancl the small but determinedgroup of defenders emplaced there. Al-most 30 percent of the total cargo carried

by the 1!2transports was unloaded by 1130of l)-l)ay, and this figure was extended

to almost 50 percent completion by thetime that the APAs and AKAs had to de-

part the area for the second time. Thecargo remaining on board was varied;some ships had unloaded all rations but1ittle ammunition. Other transports hadunloaded ammunition first and were ju*starting to move the other supplies.

TVhile the combat troops ashore pre-m~red to defend the newlv won beachhead,, .

ASSAULT OF CAPE TOROKINA

the transport groups proceeded to unloadas rapidly as the air attacks, loss of boats,and elimination of a number of beacheswould permit. lly 1600 on D-Day, onlythe four northernmost transportetheones most affected by the boat mishaps andthe unsuitable beaches-still had cargo onboard.

The quick unloading of the other fiveAP.is and three .~KAs, despite the inter-ruptions, was a reflection of the measurestaken by Admiral Wilkinson and GeneralVandegrift to insure rapid movement ofsupplies ashore. Embarked troops oneach APA and AKA had been required tofurnish a complete shore party of about500 men. During the unloading, 6 officersand 120 men remained on board ship toact as cargo handjers w]lile a further 60were boat riders to direct the supplies tothe proper beaches. Another 200 Marinesstayed on each beach to help unload thelanding craft. The remaining personnelwere used as beach guides, vehicle drivers,cargo handlers, and supervisors.12 The3d Service Battalion, augmented by sup-porting troops—artillerymen, engineers,military police, signal men, tank men, com-municators, and Seabees-forrned the bulkof these working parties. In some in-stances, these supporting troops were notreleased to their units until several daysafter the beachhead had been established.In all, about 40 percent of the entire land-ing force was engaged initially in shore

party activities.By late afternoon, each landing team re-

ported its mission accomplished. In theabsence of any identifiable terrain featuresin the interior, the landing teams had beendirected to extend the beachhead certain

“ IMAC Adn]inO No. 1, dtd 150ct43 in IMAC

.411-I.

219

distances, and, by the end of D-Day, eachof the battalions was established in arough perimeter along the first of theseclesignated Inland Defense Lines. Thedivision front lines extended into the jun-gle about 600 yarcls near the Laruma Riverand about 1,000 yards in front of CapeTorokina. Although the 1st Battalion, 3dMarines, in the area of the cape plantation,and the 3d Raiders on Puruata Island wereStill receiving occasional sniper fire, theremainder of the perimeter was quiet anddefense was not, a special problem.

There was, however, still congestion ontl~e shoreline. In order to bring some or-c]er out of the near chaos on the beachesan(l to redlwe the paralyzing effect of then~ountains of supply piled helter-skelter,additional Marines from the combat forceswere detailed as labor gangs to sort thesupplies and haul them to the front-lineunits. This placed a double-burden onsome units who were already near half-strength by the assignment of troops tothe shore party work.

.In additional problem, late on D-Day,was the correlation and coordination of thedefensive positions and missions of themaIly assorted and unrelated supportingnnits which had landed during the day.These included echelons of artillery, anti-aircraft artillery, and seacoast defensennits. The 12th Marines, decentralizedwith m battery attached to each landingteam, was in varying stages of readinessfor defense of the beachhead. Battery B,in the 9th Marines area, was in positionby early afternoon but was so engaged inc:~rgo hauling that the first L’e~LleStSfOr afiring mission could not be completed.Otber batteries were also in position bythe end of D-Day, and several had firedregistration shots and were available for

interrnittent fires during the first night.

220 ISOLATION OF RABAUL

The remaining batteries were ready forsupport missions the following day.

Selection of positions in most areas wasdifficult. The battery supporting the 2dRaider Battalion was forced to move in-land about 100 yards through a lagoonbefore a position could be located. TWOamphibian tractors ferriecl the guns andmost of the xmnlunition across the water,and the artillel’ymen transferred the re-maining ammunition from the beach to thegun position by rubber boats. This bat-tery registered on Piva Village by airspot, and the next day fired 124 rounds onsuspected enemy positions in the vicinityof that village.

Antiaircraft batteries (90mm) and theSpecial Weapons Group of the 3d DefenseBattalion landed right behind the assaultunits. Advance details of the seacoastdefense battery also moved ashore earlyand immediately begzn seeking suitab]epositions to mount the big guns. Afterthe first air raid on the morning of D-nay,the remaining antiaircraft guns of theMarine defense battalion were hurriedashore so that protection of the beachheadcould be increased as soon m possible. Bynight fal 1 of D-Day, 20 Mknrn Nlns, 8N)mm guns, and the .50 and .30 cali-ber mzchine guns of the battalion wereintegrated into the defense of the perime-ter nnd were ready for action.

As nightfall approached, tbe frontlineunits sited all weapons along fixed linesto coordinate their fire with adjacent units,and nll companies set,up an all-around de-fense. Supporting units on the beach alsoestablished small perimeters within thisdefensive 1inc. There was to be no unnec-essary firing and no movement. Marineswere to resort to bayonets and knives whenneeded, and any .Japanese infiltrators wereto be left unchallenged and then elimi-

nated at daybreak. An open-wire tele-phone watch was kept by all units, and ra-dios were set to receive messages but nogenerators were started for transmissions.

The night passed as expected—Marineshuddling three to a foxhole with one manawake at all times. A dispiriting drizzle,which began late on D-Day afternoon,continued through the night. Japaneseinfiltrators were busy, and several briefskirmishes occurred. An attack on acasualty clearing station was repulsed bygunfire from corpsmen and wounded Ma-rines; and one battalion command post, di-rectly behind the front lines, was hit by anenemy patrol. l’he attackers were turnedback by the battalion commander, execu-tive officer, and the bat talien surgeon whowielded knives to defend their foxhole.

While the Marines ashore had busiedthemselves getting ready for the first nightof defense of the beachhead, the transportgroups proceeded with the unloading de-tails. At 1645, the transports were ad-vised to debark all weapons, boat pool per-sonnel, and cargo handlers and leave thearea at 1700. The four transports stillwith supplies aboard (the A zchhz, .4mer-ican Legion, Hwnte~ Liggett, and theCrescent City) were to keep working untilthe final moment and then leave with therest of the transports despite any Marineworking parties still on board.

Admiral Wilkinson, aware that the situ-ation ashore was well under control, haddecided that all ships would retire for thenight and return the next day. In eventof a night attack, the transports in Em-press Augusta Bay would be sitting ducks.The admiral felt that his ships could notmaneuver in uncharted waters at night,and that night unloading operations werenot feasible. The admiral had anotherreason, too. An enemy task force of four

ASSAULT OF CAPE TOROKINA

cruisers and six destroyers was reportedheading toward Rabaul from Truk, andthese ships, aft er one refueling stop, couldbe expected near Bougainvillea later thatevening or early the next morning. Theamphibious force, as directed, moved outto sea for more protection.

At 2300 that night, 1 h’overnber, thefour transports which were still to be un-loaded were ordered to reverse course andhead back toward Empress Augusta Baywhile the rest of the transports continuedtoward Guadalcanal. The four trans-ports, screened by destroyers, rebwlatedtheir speed and dirwtion so as to reach theCape Torokina area after daybreak. Ashort time laterl alerted to tile fact, that alarge enemy fleet was in the area, thetransports headed back toward Guadal-canal again.

Admiral Merrill’s Task Force 39, afterthe successful bombardment of Buka andthe Shortlands -which opened the Bou-gainvillea operation, had moved north ofVella Lavella to cover the retirement ofthe transport group. At this particu]xrtime, Merrill’s concern was the conditionof his force which had been underway for29 hours, steaming about 766 miles atnear-maximum speed. .41though thecruisers were still able to fight, the fueloil snpply in the destroyers was belowthe level required for anything but sm~llengagements at moderate speeds.

So, while Merrill’s ciwisers waited, oneof the two destroyer divisions in the taskforce turned and headed for New Georgiato refuel. That afternoon, 1 November,while an oil barge was pumping oil intothe destroyers at maximum rate, the re-port of the Japanese fleet bearing downon 130ugainville was received. The de-stroyers, impatient to get going, hurriedthrough the refueling.

221

At 1800, all destroyers raced out of KulaGulf to rejoin Merrill. The 108-miletrip was made at 32 knots, although theengines of two of the destroyers were onthe verge of breakdown. By 2330, theships joined Merrill’s cruisers south of theTreasurys, ancl the entire task force headedtoward Bougainvillea where it interposeditself between the departing transportsand the oncoming enemy fleet. Alliedpatrol planes had kept the attack forcelll~der surveillance all day, and, by night-f i~l1, the direction of the ,Japanese shipswas well established. If the Allied think-ing was correct, another trap had beenbaited for the Japanese. The enemy,guessing that, the same task force that hitBuka had provided the shore bombard-ment for the Cape Torokina landing,might be lured into assuming that thefighting ships were now low on fuel andammunition and had retired with thetransports. If that was the enemy as-sllmptioll, then Merrill was in position fora sllccessful ambush.

Moving slowly to leave scant, wake forenemy snooper planes to detect, Merrill’sforce WM off 130ugainvil]e by 0100, 2 No-vember, i~l~dbeginning to maneuver intoposition to intercept the enemy fleet. Atthat time, the enemy was about 83 milesdistant. Merrill’s basic plan was to stopthe enemy at all costs, striking the ,Jap-anese ships from the east. so that the seaengagement would be deflected toward the

west, away from Bougainvillea. Thiswould give his ships more room to maneu-ver as well as allow any damaged ships toret ire to the east on the disengaged sicle.Further, Merrill respected the ,Japanesetorpedoes and felt. that his best chanceto divert the enemy force and turn itback-possibly without loss to his owm

222 ISOLATION OF RABAUL

force—was by long-range, radar-directedgunfire.

The naval battle of Empress AugustaBay began just 45 miles offshore from thebeachhead whose safety depended uponTask Force 39. Merrill-s cruisers openedfire at 0950 at ranges of 16,000 to 20,000yards. The enemy fleet, spread out overa distance of eight miles, appeared to be inthree columns with a light cruiser and de-stroyers in each of the northern and southe-rn groups and two heavy cruisers andtwo destroyers in the center. Detectionwas difficult because, with the enemy sospread out, the radar on Merrill’s shipscould not cover the entire force at one

time.The enemy’s northern force was hit

first, the van destroyers of Task Force 39

engaging this section while the rest of theAmerican ships turned toward the center

and southern groups. As planned, the att-ack struck from the east. Task Force39 scored hits immediately, drawing short

and inaccurate salvos in return. The

Japanese, relying on optical control ofgunfire, lighted the skies with starshells

and airplane flares; but this also helpedTask Force 39, since the enemy’s flashlesspowder made visual detection of the Jap-anese ships almost impossible withoutlight.

The two forces groped for each otherwith torpedoes and gunfire. In the darknight, coordination of units was difficultand identification of ships impossible.The maneuvering of Merrill’s task unitsfor firing positions, as well as the franticscattering of the enemy force, spread thebattle over a wide area, which further in-

creased problems of control and identifica-tion. on at least one occasion, Task Force

39 ships opened fire on each other beforediscovering their error.

In such confused circumstances, estimat-ion of damage to either force was almostimpossible, although some of the .kmericandestroyers believed that their torpedoeshad found Japanese targets, and otherenemy ships were believed to have been hitby gunfire. In the scramble for positionsto take new targets under fire, two destroy-ers of Merrill’s force scraped past. eachother with some damage, and several otherclose collisions between other destroyerswere narrowly averted. One Americandestroyer, the Foote, reported itself dis-abled by an enemy torpedo and two otherdestroyers were hit by gunfire but re-mained in action. The only cruiser dam-aged was the Denver, which took three8-inch shells and was forced to disengagefor a short time before returning to thefight.

By 0332, Task Force 39 was plainly inpossession of the field. The enemy force,roukd in all directions, had ceased firingand -was retiring at high speed. Merrill’scruiser division ceased firing at 0349 on

one last target at ranges over 23,000 yards.This ended the main battle, although theTF 39 destroyers continued to scout thearea for additional targets and disabledenemy ships. At daybreak, TF 39 was

reassembled and a flight of friendly air-craft appeared to provide escort for itsretirement. The Foote was taken under

tow and the return to Guadalcanal started.The Merrill force believed that it had

sunk at least one enemy light. cruieer andone destroyer and inflicted damage on anumber of other ships. This estimate waslater found correct.’3 In addition, the

1’ SE Area NavOps—ZIZ, p. 14.

ASSAULT OF CAPE TOROKINA 223

Japanese also had several ships damagedin collisions.

Task Force 39 was struck a few hourslater by a furious air attack from morethan 70 enemy planes, but, the Japanesemade a mistake in heading for tl]e cruisersinstead of the destroyers guarding the dis-abled Foote. The hex~-y anti~irmaft fireand the aggressive protection of the Coln-AirSols fighter cover forced the enemyplanes away. The air cover shot down 10planes, and the ships reported 7 enemyaircraft, downed. Only one Americancruiser, the Montpelier, was hit by bombs

but, it was able to continue. Nlile tlie airbattle raged, the amphibious force’s trans-ports reversed course once more and re-turned to Cape Torokina without interfer-ence and completed the unloading. Thesea and air offensive by the ,Jzpznese hadbeen stopped cold by the combined action

of (lm-ISoPac’s air and sea forces.

THE JAPANEgE “

To the Japanese defenders, the suddenappearance of a number of transports off

Cape. Torokina on the morning of 1 No-vember came as something of a shock. AllJapanese plans for the island had dis-counted a landing north of Cape Torokinabecause of the nature of the beaches andthe terrain. If the .411ies attacked thewestern coast of Bougainvillea, the enemythought the logical place would be south-ern Empress Aubgusta Bay around CapeMutupena. ,Japanese defensive installa-

“ Unless otherwise noted, the material in thissection is derived froul : Ilf..l (YAR-1; 3d MarI)i17

CwthbatRrpt; Iflf.4 (1 C–2 .Jnl; .!?d .lfarDiv AR;

SE AT(YZ ?ia~;Ops—111: S’m.x>ntccnth Army Ops—

11; [JSSBS, C’ampaiflns; I“SSBS, l~/tcrroflation.s;JIi.91Div Acct.

tions, of a limited nature, were positionedto repel an Allied landing in this area.

But the small garrison in the vicinity ofCape Torokina, about 270 men from thefi?dCompany, 2Vd Regiment, -with a regi-mental weapons platoon attached, was welltrained. From the time that the alarmwas sounded shortly after dawn on 1 No-vember, the Japanese soldiers took up theirdefensive positions around the cape andprepared to make the invading forces payas dearly as possible for a beachhead.

The invading Marines found the island’sdefenders dressed in spotless, well-presseduniforms with rank marks and service rib-bons, an indication that the Cape Torokinagarrison was a disciplined, trained forcewith high morale., willing to fight to thedeath to defend its area. But after thefirst day, when the. ,Japanese were knockedout of the concentrated defenses on CapeTorokina, the enemy resistance was almostnegligible. A wounded Japanese sergeantmajor, captured by Marines the secondday, reported that the understrength gar-rison had bem wiped almost out of exist-ence. The prisoner confirmed that theJapanese had expected an attack on Bou-gainvillea for about three days—but not atCape Torokina.

With the notice of the Allied operationsagain% Bcmgainville, all available Japa-nese air power was rushed toward Rabaul,

and Admiral Kusaka ordered the inter-ception operations of the .Southeaat AreaFleet (the Ro operation) shifted from

New Guinea to the Solomons. Because allplanes of the IstAir h’qtuzdron and addi-tional ships were already en route to Ra-baul, this action placed the entire mobilesurf ace and air strength of the CombinedFleet under the direction of the comman-

der of the Southeast Area Fleet.

224 ISOLATION OF RABATJL

The protests of some commandersagainst the use of surface vessels in thearea south of New Britain-which waswell within the range of the area domi-nated by the planes of ComAirSols-w,erebrushed aside. Combined Fleet Z?ead-qwmtens. convinced that, this was the lastopportunity to take advantage of the stra-tegic sitl~ation in the southeast, was deter.

mined to strike. a decisive blow at the Al-lied surface strength in the Solomons anddirected Kusaka to continue the operation.

After the battle of Empress AugustaBay, however, the defeated .Tapanese re-tired from the area with the realizationthat combined sea and air operationswere difficult with limited air resources,especially “in a region where friendlyand enemy aerial supremacy spheresoverlapped broadly.” 15

The Seventeenth Army, charged withthe actual defense of Bougainvillea, tookthe news of the Allied invasion a bit moreblandly:

In formulating its operation plan, the Sev-enteenth Army planned to employ its mainforce only on the occasion of an army inva-sion in the southern or northern region, orthe Kieta sector. Therefore, at the outsetof the enemylanding in the vicinity of Toro-kina Point, the Seventeenth Army was lack-

ing in determination to destroy the enemy.The army’s intention at that point was onlyto obstruct the enemy landing. 16

There were many avenues of obstructionopen to the Japanese, despite the fact thatthe Alliecl sea and air activity probablydiscouraged the enemy from many aggres-sive overtures. Deceived originally as tothe intentions of the Allies, the Japaneseapparently remained in doubt for sometime as to the strategical and tactical im-portance of the operations at Cape Toro-kina. The enemy could have counter-manded or prevented extension of thedefensive positions and occupation of theprojected airfield sites by shelling or airbombardments. But nune of these coursesof action were initiated immediately orcarried out with sufficient determinationto jeopardize the beachhead seized by thelMAc forces.

The chief threat to the Cape Torok@aperimeter seemed to be from the rightflank. Operation orders, taken from thebodies of dead Japanese at Cape Torokina,indicated that forces in the area southeast

of the Cape could strike from that direc-

tion, and it was to this Si& that the lMAC

forces ashore pointed most of their combat

strength.

“ SE Area NavOp.s-11, p. 18. ‘“ Neven.teentl&Army Op&—ZII, p. 103,