Naval History and Heritage Command

The Navy Department Library

Related Content

The Battle of Savo Island August 9th, 1942 Strategical and Tactical Analysis

Part I

Cover image - The Battle of Savo Island August 9, 1942 Strategical and Tactical Analysis

PDF Version [129.3MB]

The Battle
Savo Island


August 9th, 1942

Tactical Analysis.

U. S. Naval War College


Prepared By
Department of Analysis
Naval War College

Commodore Richard W. Bates, USN (Ret)
Commander Walter D. Innis, USN

Table of Contents

Foreword i-ii
Table of Contents iii-xxii
Zone Time xxiii
Principal Commanders xxiv-xxv
Introduction xxvi
Brief Narrative of the Battle of Savo Island xxvii-xxx
Chapter I The Strategic Area 1-4
General Discussion 1-3
  (b)  Tulagi-Guadalcanal Area 3
  (c)  Weather 4
Chapter II Japanese Arrangements 5-14
  (a)  Japanese Command Relations 5-6
  (b)  Information Available to Japanese Commanders 7-8
  (c)  Japanese Land and Tender Based Aircraft 8-10
  (d)  Japanese Search and Reconnaissance 10-11
  (e)  Japanese Disposition of Naval Forces 11-13
    (a) Disposition at 0652 August 7th 12
  (1) Support Force and Chokai 12
  (2) Escort Force 12
  (3) Auxiliaries 25th Air Flotilla 12
  (4) Submarines 12
  (f)  Japanese Tasks Assigned 13-14
  (g)  Japanese General Concept 14
Chapter III Allied Arrangements 15-42
  (a)  Allied Command Relations 15-19
    Tasks CINCPOA 15
    Tasks COMSOPAC 15
    Tasks COMSOWESPAC 16
    Admiral Nimitz Becomes CINCPOA 16
    Vice Admiral Ghormley Becomes COMSOPAC and COMSOPACFOR 16
    General MacArthur Becomes COMSOWESPAC 16
    Command Relations Between Australian and American Naval Officers 17
    Joint Chiefs of Staff Assign New Missions--Task ONE to COMSOPAC 17
    COMSOPACFOR is Designated as Task Force Commander for Task ONE 17
    COMSOPACFOR Issues Operation Plan 18
    Commander Expeditionary Force (CTF 61) Issues Operation Plan 19
  (b)  Information Available to Allied Commander 19-21
  (c)  Allied Land and Tender Based Aircraft 21-25
    (1) South Pacific (SOPAC) 21-23
    (2) Southwest Pacific (SOWESPAC) 23-25
  (d)  Allied Search and Reconnaissance 26-31
    (1) South Pacific (SOPAC) 25-27
    (2) South Pacific 27-31
  (e)  Communications Arrangements Between COMSOPAC and COMSOWESPAC 31-32
  (f)  Allied Deployment of Naval Forces 32-37
    (1) Approach to Guadalcanal-Tulagi Area 32-34
    (2) CTG 61.1 Operates his Carriers 34-35
    (3) Approach TG 61.2 (TF 62) (Amphibious Force) 35
    (4) Deployment of SOWESPAC Submarines 35-36
  (a) Deployment of S-44 35-36
  (b) Deployment of S-38 36
    (5) Deployment of Allied Forces at 0652 August 7th 36-37
  (a) TG 61.1 Air Support Force 36
  (b) TG 61.2 (TF 62) (Amphibious Force) 36-37
  (c) Submarines 37
  (g)  Composition of Forces and Tasks Assigned 37-40
    (1) Composition of Forces 37-38
  (a) TG 61.1 (Air Support Force) 37
  (b) TG 61.2 (TF 62) (Amphibious Force) 37-38
    (2) Tasks Assigned 38-40
  (a) TF 61 (Allied Expeditionary Force) 38
  (b) TG 61.1 (Air Support Force) 39
  (c) TG 61.2 (TF 62) (Amphibious Force) 39
  (h)  The Allied Plan 40-42
  (i)  General Summary 42
  (a) Operations of Japanese Cruiser Force 43-50
    Tulagi is Attacked by Allied Forces 43
    Commander Outer South Seas Force Estimates the Situation 43-44
    Commander 5th Air Attack Force Makes Additional Searches and Launches Two Air Attack Groups 45-46
    Commander Outer South Seas Force Organizes Cruiser Attack Force 46
    Rabaul is Attacked 47
    Japanese Air Attack Hits Allied Shipping at Tulagi-Guadalcanal 47


  Commander Outer South Seas Force Decides to Command Cruiser Force 48 
  Japanese Air Attack Hits Allied Shipping at Tulagi-Guadalcanal 48 
  Commander Outer South Seas Force Receives Intelligence Summary about Allied Forces 48-49 
  Commander Cruiser Force Departs with Cruiser Force for Tulagi 50 
  (b)  Movements of Japanese Submarines 51 
CHAPTER V ALLIED OPERATIONS (0652 - 2400 August 7th) 52-71 
  (a)  Operations of CTF 62 (Commander Amphibious Force) 52-54 
    Zero Hour Set for Guadalcanal and Tulagi 52 
    Allies Land on Guadalcanal, Tulagi and Gavutu 52 
    TF 62 Struck by First Japanese Air Attack 53 
    TF 62 Struck by Second Japanese Air Attack 53 
    CTF 62 Becomes Concerned Over Japanese Capabilities 53 
    CTF 62 Requests CTF 63 to make Additional Searches 53-54 
    CTF 62 is Erroneously Informed by COMSOPACFOR that Enemy Submarine is Nearby 54 
    CTF 62 is Informed by COMSOWESPAC of Contact on Enemy Cruiser Force North of Rabaul and of contact on Six Unidentified Ships in St. Georges Channel 54 
  (b)  Operations of Allied Screening Group 54-61 
    (1) Night Disposition 55-56 
  (a) AUSTRALIA Group 55 
  (b) VINCENNES Group 55 
  (c) SAN JUAN Group 55 
  (d) RALPH TALBOT and BLUE 56 
  (e) Remaining Destroyers 56 
    (2) CTG 62.6.1 Instructions 56-57 
    (3) Discussion of Night Disposition 57-61 
  (c)  Operations of CTF 61 (Commander Expeditionary Force) 61-64 
    CTF 61 Remains in SARATOGA 61 
    Communications for Air Operations 61-62 
    CTF 61 Hears of Carrier Type Bombers Over Guadalcanal and Re-estimates Situation 62 
    Directs Air Operations South of Cape Henslow for August 8th 62-63 
    Discussion of This Order 63-64 
    Intercepts CTF 62's Request to CTF 63 for Additional Air Search 64 
    Receives CTF 62's Summary of Operations for August 7th 64 
  (d)  Operations of CTG 61.1 64-68 
    Weather Conditions 64 
    WASP Air Support Group Destroy Japanese Seaplanes at Tulagi 65 
    Provides Combat Air Patrol 65 

- v -

    Provides Air Search and Reconnaissance 65-66 
    Hears that Carrier-Type Bombers Have Attacked Transports and Re-estimates the Situation 66 
    Discussion of New Day Launching Position 67 
    Suffers Aircraft Losses 67 
  (e)  Operations of Allied Submarines 68-69 
    (1) Operations of S-38 68-69 
    (2) Operations of S-44 69 
  (f)  Operations of CTF 63 (Commander Aircraft South Pacific Force) 69-70 
  (g)  Operations of Commander Allied Air Forces SOWESPAC 70-71 
    Sights Japanese Cruisers North of Rabaul 70 
    Contact Reports Delayed 11.5 Hours 71 
    Sights Six Unidentified Ships in St. Georges Channel 71 

(0000 August 8th to 2400 August 8th)
  (a)  Operations of Commander Cruiser Force 72-78 
    Launches Search Planes from Cruisers 72-73 
    Sights Lockheed Bomber at 1020 73 
    Recovers Search Planes 74 
    Receives Report of Situation Tulagi - Guadalcanal from Pilot of AOBA 74 
    Decides Carry Out Night Attack 75 
    Issues Instructions for Night Action 75-76 
    Receives Report of Japanese Bombing Attacks on Allied Shipping at Tulagi - Guadalcanal 77 
  (b)  Operations of Commander Fifth Air Attack Force 78-81 
    Launches Air Searches 79-80 
    Launches Air Attack 80-81 
  (c)   Operations of Japanese Submarines 81-82 
    (0000 August 8th to 2400 August 8th)  
  (a)  Operations of CTF 62 83-91 
    Informs TF 62 Enemy Submarine Might Enter Area That Day 83 
    Receives Word of Small Japanese Surface Forces Heading South 83-84 
    TF 62 is Attacked by Japanese Planes Damaging GEORGE F. ELLIOTT and JARVIS 84 
    Orders GEORGE F. ELLIOTT Sunk 85 
    Intercepts CTF 61 Dispatch Recommending Withdrawal of Carriers 85 
    Received Contact Report that Two Enemy Destroyers, Three Cruisers and Two Seaplane Tenders or Gunboats are Headed South 86 


    Also Receives Contact Report of Two Japanese Submarines Headed Towards Tulagi 86 
    Estimates the Situation and Decides No Enemy Night Attack 87 
    Calls Conference of CTG 62.6 and Commanding General, First Marine Division 87 
    CTG 62.6 Departs Western Screen and Fails to Notify Commander VINCENNES  Group 87-88 
    CTF 62 Holds Midnight Conference 89-90 
    RALPH TALBOT Sights Enemy Plane and Broadcasts Warning 80 
  (b)  Operations of CTF 61 (Commander Expeditionary Force) 91-95 
    Launches Air Search 91 
    Recommends Withdrawal of Carriers 91-95 
  (c)  Operations of CTG 61.1 (Air Support Force) 95-97 
    Launches Air Search 95 
    Launches Additional Air Search 96 
  (d)  Operations of Allied Submarines 98 
    (1) Operations of S-38 98 
    (2) Operations of S-44 98 
  (e)  Operations of CTF 63 (Commander Aircraft South Pacific Force) 98-100 
    Launches Routine Searches 98 
    Results of Searches 98-99 
  (f)  Operations of Commander Allied Air Forces SOWESPAC 100-103 
    Launches Routine Searches 100 
    RAAF Hudson Contacts Japanese Cruiser Force at 1025 100 
    RAAF Hudson Contacts Japanese Submarine I-121 100 
    Contact Report Delayed in Transmission to CTF 62 101 
    RAAF Hudson Contacts Japanese Cruisers at 1101 101-102 
    Contact Report Delayed in Transmission to CTF 62 102-103 
  (a)  Forces Engaged 104 
    (1) Allied Force 104 
    (2) Japanese Cruiser Force 104 
  (b)  Strength and Weakness Factors 104-105 
    (0000 August 9th to 0132 August 9th)  
  (a)  The Approach 106-109 
    Forms Cruising Disposition 106 
    Receives Intelligence of Allied Forces From Cruiser Plane 106 
    Alerts Battle Stations 107 
    Sights Allied Destroyer BLUE 107 
    Superiority of Japanese Night Naval Detection 107 
    Maneuvers to Enter Iron Bottom Sound 107-108 
    Enters Iron Bottom Sound 109 


  (0000, August 9th to 0132, August 9th)  
  (a) Operations of BLUE 110-11 
  (b) Operations of CHICAGO Group 111 
  (c) Operations of VINCENNES Group 111-112 
  (d) Operations of RALPH TALBOT 112 
  (e) Operations of SAN JUAN Group 112-113 
  (0132, August 9th to 0150, August 9th)  
  (a) Action with CHICAGO Group 114-120 
    CHOKAI Sights the JARVIS 114 
    CHOKAI Sights the CANBERRA and CHICAGO 115 
    CHOKAI Sights the VINCENNES Group 115 
    CHOKAI Fires Four Port Torpedoes at CANBERRA and CHICAGO 116 
    COMCRUDIV SIX Sights CHICAGO Group 116 
    YUNAGI Decides to Attack JARVIS 117 
    CHOKAI Commences Firing 118 
    Japanese Cruiser Planes Drop Aircraft Flares 118 
    FURUTAKA Fires Guns and Four Port Torpedoes 118 
    Changes Course to 050°(T) 119 
    AOBA Fires 8-inch Battery and Three Torpedoes at CANBERRA 119 
    CRUDIV EIGHTEEN Changes Course to the Northeast 120 
    YUNAGI Reverses Course 120 
    Cruiser Force Breaks up into Two Groups - Eastern and Western 120 
  (b) Approach to the VINCENNES Group 121-124 
      KAKO Fires Three Torpedoes at CHICAGO 121 
      KAXO Fires on CANBERRA 121 
      FURUTAKA Fires One Torpedo at PATTERSON 121 
      FURUTAKA Fires at CANBERRA 122 
      CHOKAI Fires Four Torpedoes at VINCENNES 122 
      CHOKAI Changes Course to 069°(T) 123 
      FURUTAKA Fires Three Torpedoes at CANBERRA 123 
      Summary of Torpedo Firing Against CHICAGO Group 123 
      Position of Ships at 0150 124 
  (0132 August 9th to 0150 August 9th)  
  (a) Action of CHICAGO Group with Japanese Cruisers 125-138 
      Conditions of Readiness 125 
      Weather Conditions 126-127 
      (1) Action by CANBERRA 127-129 
         Sights Four Enemy Torpedoes and Two Enemy Ships 127 
         Is Heavily Hit by Gunfire 128 
         Sights Additional Torpedoes Which Miss 128 


      Commanding Officer Killed and Ship Disabled 129 
      Damage Received 129 
    (2) Action by CHICAGO 129-133 
      Sights Aircraft Flares Over Area X-RAY 129 
      Sights Three Enemy Ships 130 
      Fails to Notify Command 130 
      Sights Three Additional Torpedo Wakes 131 
      Hit by Torpedo (KAKO's) 131 
      Observes Gunfire Flashes 132 
      Fires Star Shells to Port 133 
      Hit by Enemy Shell 133 
    (3) Action by BAGLEY 133-135 
      Sights Number Unidentified Ships 134 
      Observes Enemy Salvos Landing Short of CANBERRA 134 
      Fails to Report Contact 134 
      Turns to Port and Fails to Fire Starboard Torpedoes 134 
      Observes Aircraft Flares Over Area X-RAY 134 
      Fires Port Torpedoes at CRUDIV EIGHTEEN 134-135 
      Turns Left to Scan Passage Between Guadalcanal and Savo Island 135 
    (4) Action by PATTERSON  135-138 
      Sights Enemy Ship (FURUTAKA) 135 
      Goes to General Quarters 136 
      Broadcasts Contact Report by TBS and Blinker 136 
      Turns to Port and Attempts to Fire Starboard Torpedoes But Fails to do so 136 
      Broadcasts Contact Report by TBS 137 
      Opens Fire on CRUDIV EIGHTEEN 137 
      Hit by One Enemy Shell 137 
      Succeeds in Hitting YUBARI 138 
  (b) Operations of VINCENNES Group 138-145 
    (0132 August 9th to 0150 August 9th)  
      Conditions of Readiness 138 
      Weather Conditions 139 
      Information Available to Commanding Officers 139 
      Commander VINCENNES Group Estimates the Situation 140 
      Concern Over Possible Submarine Attack 141 
      Sights Aircraft Flares 141 
      WILSON and HELM go to General Quarters 142 
      Commander VINCENNES Group Re-estimates the Situation 142 
      PATTERSON's Warning Message Received by QUINCY, VINCENNES and WILSON 143 
    QUINCY goes to General Quarters 143 
    VINCENNES goes to General Quarters 144 
    QUINCY Observes Enemy Ships 144 
    Commander VINCENNES Group Increases Group's Speed to Fifteen Knots and Awaits Developments



    Gunnery Officer ASTORIA Requests Unsuccessfully That General Quarters be Set 145 
    VINCENNES, QUINCY and ASTORIA are Illuminated by Enemy Searchlights 145 
  (c) Operations of BLUE 146-147 
      Unaware of Passage of Japanese Cruiser Force 146 
      Sights Four Flares and goes to General Quarters 146 
      Endeavors to Report Enemy Planes to Officer-in-Tactical Command 147 
  (d) Operations of RALPH TALBOT 147 
      Hears PATTERSON's Contact Report and goes to General Quarters 147 
  (e) Operations of CTG 62.6 (AUSTRALIA) 147-149 
      Decides to Remain in Transport Area X-RAY 148 
      Fails to Inform CTF 62 or Commanders VINCENNES and CHICAGO Groups 148 
      Sights Aircraft Flares Vicinity Area X-RAY 149 
      Observes Gunfire From AUSTRALIA Group as well as in Direction VINCENNES Group 149 
  (f) Action by SAN JUAN Group 149-150 
      Sights Aircraft Flares 150 
      Observes Firing 150 
      Hears PATTERSON's Warning 150 
      Goes to General Quarters 150 
  (a) Actions Between Japanese Eastern Group and VINCENNES Group 151-162 
      Commander Cruiser Force Changes His Objective 151 
      CHOKAI Sights ASTORIA 152 
      CHOKAI Commences Fire on ASTORIA, AOBA on QUINCY, KAKO on VINCENNES Employing Searchlights 152 
      KINUGASA Fires at CANBERRA 152 
      Japanese Salvos are Short 153-154 
      Japanese Searchlight Technique 154-155 
      AOBA Fires Second Salvo at QUINCY 155 
      KINUGASA Fires at HELM 155 
      CHOKAI Fires Second Salvo at ASTORIA 155 
      KAKO Fires Second Salvo at VINCENNES 156 
      AOBA's Third Salvo Hits QUINCY 156 
      KAKO's Third Salvo Hits VINCENNES 156 
      CHOKAI Fires Third Salvo at ASTORIA 156 
      KINUGASA Continues Firing at HELM 156 
      AOBA's Fourth Salvo Hits QUINCY 156 
      KAKO's Fourth Salvo Hits VINCENNES 157 
      KINUGASA Fires Four Torpedoes at CANBERRA 157 
      CHOKAI Fires at WILSON 157 
      CHOKAI Fires Fourth Salvo at ASTORIA 157 




    AOBA's Fifth Salvo Hits QUINCY 157-158 
    KAKO Continues Hitting VINCENNES 158 
    KINUGASA Fires at VINCENNES 158 
    AOBA Fires at HELM 158 
    AOBA's Sixth Salvo Hits QUINCY 159 
    AOBA Divides Fire Between VINCENNES and QUINCY 159 
    CHOKAI's Fifth Salvo Hits ASTORIA 159 
    KAKO Hits VINCENNES 159 
    CHOKAI Hits ASTORIA 160 
    KINUGASA Fires at VINCENNES 160 
    KINUGASA is Hit by PATTERSON 160 
    Commander Cruiser Force Closes Enemy and Observes that VINCENNES Group had Sunk 160 
    CHOKAI Falls Behind 160 
  (b) Action Between Japanese Western Group and VINCENNES Group 162-167 
    Changes Course to 000°(T) 164 
    Observes CANBERRA Sinking 164 
    FURUTAKA Opens Fire on QUINCY 164 
    COMCRUDIV EIGHTEEN Sights the VINCENNES Group 164-165 
    TENRYU Opens Fire on QUINCY 165 
    FURUTAKA Shifts Fire to VINCENNES 165 
  (c) Action of YUNAGI with JARVIS 166-167 
    Opens Fire on JARVIS 166 
    Breaks off Action 166 
   (0150 August 9th to 0200 August 9th)  
  (a) Engagement of VINCENNES Group with Japanese Eastern Group 168-203 
    VINCENNES Group is Illuminated by Searchlights 168 
    Commander VINCENNES Group Estimates the Situation 168-169 
    Decides to Maintain Course and Speed 169 
    Discussion of This Decision 169-171 
    (1) Action by VINCENNES 171-178 
      Trains on KAKO But Fails to Open Fire 171 
      Observes Enemy Salvos Landing 171-172 
      Fires Star Shells to Illuminate Enemy 172 
      Directs Destroyers HELM and WILSON to Attack 172 
      Is Heavily Hit 173 
      Fires First Salvo 173 
      Commander VINCENNES Group Re-estimates Situation 174 
      Decides to Close Enemy 174 
      Discussion of this Decision 174 
      Fires Second Salvo and Hits KINUGASA 175 


        Commander VINCENNES Group Decides to Turn Away 176 
        VINCENNES is Torpedoed and Slows Down 176 
        Commanding Officer Endeavors to Change Course to Port 177-178 
        Changes Course to Right and Fires Third Salvo 178 
        (2) Action by QUINCY 178-187 
        Commanding Officer Estimates Situation 178-179 
        Considers Searchlights Friendly 179 
        Directs Gunnery Officer Fire on Searchlights 179-180 
        Observes Enemy Salvos 180 
        Unable set Condition One Promptly 180 
        QUINCY is Hit 182 
        Maneuvers to Follow VINCENNES 183 
        Observes Japanese Western Group 184-185 
        Fires First Salvo 185 
        Continues to be Heavily Hit 186-187 
        (3) Action by ASTORIA 187-198 
        Gunnery Officer Requests General Quarters be Set 187 
        O.O.D. Fails to Call Commanding Officer 187-188 
        Observes Salvos Land off Port Side 188 
        Gunnery Officer Urgently Requests General Quarters be Set 189 
        Obtains Radar Range on Japanese Cruisers 189-190 
        Gunnery Officer Opens Fire 190 
        First Salvo 190 
        Quartermaster Sounds General Alarm 191 
        Commanding Officer Called 191 
        Fires Second Salvo 192 
        Commanding Officer Appears on Bridge and Orders "Cease Firing" 192 
        Discussion of this Order 192-193 
        Commanding Officer Observes ASTORIA Straddled and Hit 194 
        Orders "General Quarters" and "Commence Firing" 194 
        Fires Third Salvo 195 
        Hits CHOKAI With Machine Gun Fire 195 
        Commanding Officer Estimates Situation 195-196 
        ASTORIA is Heavily Hit 196 
        Fires Fourth, Fifth, Sixth and Seventh Salvos 196-197 
        (4) Action by HELM 198-200 
        Fires One Salvo 199 
        Estimates the Probable Situation 199 
        Ordered to Attack 199 
        Observes Own Cruisers on Fire 199 
        Changes course to South 199 
        (5) Action by WILSON 200-203 
        Observes Unidentified Ships Illuminating and Firing on VINCENNES Group 200 


        Opens Fire 201 
        Observes VINCENNES Group Under Heavy Fire 201 
        Slightly Damaged by CHOKAI Gunfire 202 
  (b) Operations of CHICAGO Group 203-210 
      (1) Action by CHICAGO 203-207 
        Fails to Appreciate Objective 203-204 
        Opens Fire on YUBARI 204 
        Hits TENRYU 204 
        Sweeps With Searchlights 205 
        Observes Gun Action in Direction of VINCENNES Group 205 
        Determines Extent of Damage Received 206 
      (2) Action by CANBERRA 207 
      (3) Action by PATTERSON 207-209 
        Proceeds Independently to Eastward 207 
        Discovers Had Failed to Fire Torpedoes 208 
        Opens Fire on KINUGASA 208 
      (4) Action by BAGLEY 209-210 
        Passes Under Stern of CANBERRA 209 
        Decides to Continue Scanning Operations 210 
  (c) Operation of Radar and Anti-Submarine Screen 210-211 
      (1) Operations of BLUE 210-211 
      (2) Operations of RALPH TALBOT 211 
  (d) Operation of CTG 62.6 211 
        Observes Firing 211-212 
        Believes VINCENNES and CHICAGO Groups Coordinated 212 
         Decides Station AUSTRALIA Between Enemy and Area X-RAY 213 
         Decides Rendezvous Destroyers on AUSTRALIA 213 
  (e) Operations of SAN JUAN Group 213-214 
  (0200 August 9th to 0220 August 9th)  
  (a) Action Between Japanese Eastern Group and VINCENNES Group 215-223 
      CRUDIV SIX Separates From CHOKAI 215 
      Discussion Thereon 215 
      KAKO and KINUGASA Shift Fire to ASTORIA 215 
      KINUGASA Fires Torpedoes at Area XRAY 215-216 
      AOBA Opens Fire on ASTORIA 216 
      KAKO Fires Four Torpedoes at ASTORIA 216-217 
      AOBA Shifts Fire to QUINCY 217 
      CHOKAI Observes ASTORIA Being Hit 217 
      CHOKAI Opens Fire on ASTORIA 217 
      CHOKAI Hit by Three Shells From QUINCY 218 
      KAKO Fires Two Additional Torpedoes at ASTORIA 218 
      AOBA Observes QUINCY Attacking Formation 218 
      KAKO Fires on QUINCY and WILSON 219 
      CHOKAI Illuminates VINCENNES and Opens Fire 219 


     AOBA Fires One Torpedo at QUINCY 219 
     KINUGASA Illuminates QUINCY and Opens Fire 219 
     CHOKAI is Hit by ASTORIA 220 
     Japanese Extinguish Searchlights and Cease Firing 220 
     Commander Cruiser Force Estimates Situation Decides to Withdraw 220 
     Discussion of this Decision 220-222 
  (b) Action Between Japanese Western Group and VINCENNES Group 223-227 
     FURUTAKA Shifts Fire to VINCENNES 223 
     TENRYU Fires Six Torpedoes at QUINCY 223 
     YUBARI Fires Four Torpedoes at VINCENNES 223 
     TENRYU Fires on QUINCY 224 
     YUBARI Torpedo Hits VINCENNES 224 
     TENRYU's Torpedoes Hit QUINCY 224 
     TEURYU Opens Fire Momentarily on WILSON 224-225 
     YUBARI and FURUTAKA Cease Firing 226 
     TENRYU Illuminates RALPH TALBOT and Opens Fire 226 
     FURUTAKA Opens Fire on RALPH TALBOT 226 
     RALPH TALBOT is Lightly Hit 227 
     FURUTAKA and TENRYU Cease Firing 227 
     YUBARI Illuminates RALPH TALBOT and Opens Fire 227 
     FURUTAKA Re-opens Fire 227 
  (c) Operations of YUNAGI 228 
  (a) Action by VINCENNES 229-235 
       Fired on by FURUTAKA 229-230 
       Is Heavily Damaged 230 
       Fires Fourth and Fifth Salvos 230 
       Is Torpedoed by One Torpedo 231 
       Is Fired on by YUBARI 231 
       Engineer Force Secures Engine and Fire Rooms 231 
       Observes Two Destroyers Ahead 231-232 
       Gunnery Officer Informs Commanding Officer Battery is Out of Action 232 
       Considers He is Being Fired on by Friendly Ships 232 
       Continues to be Heavily Hit by FURUTAKA and YUBARI 234 
       Hoists U.S. Ensign 234 
       FURUTAKA and YUBARI Cease Firing 234 
       Is Fired on by CHOKAI for Several Salvos 235 
  (b) Action by QUINCY 235-243 
       Fires Star Shells 236 
       Is Hit by AOBA 237 
       Is Fired on by TENRYU 237 
       Commanding Officer Estimates Situation 237-238 
       Decides to Close Enemy 238 


       Fires Second Salvo 238 
       Is Hit by Two Torpedoes From TENRYU 238 
       Fires Third Salvo and Hits CHOKAI 239 
       Is Heavily Hit 240 
       Turret II Explodes 240 
       Commanding Officer Dies 241 
       Relief Attempts Beach Ship 241 
       Command Situation Within QUINCY in Utter Confusion 242 
       Is Hit by One Torpedo 242 
       Enemy Firing Ceases 242 
  (c) Action by ASTORIA 243-249 
       Is Being Heavily Hit by AOBA, KAKO and KINUGASA 243 
       Fires Eighth and Ninth Salvos 243 
       Unable Increase Speed 245 
       Fires Tenth Salvo 246 
       Is Fired on by CHOKAI 246 
       Fires 1.1 inch Guns and Hits AOBA 246-247 
       Fires Eleventh Salvo 247 
       Enemy Fire Slackens 247 
       Combat Effectiveness at this Time 248 
       Fires Twelfth (and last) Salvo and Hits CHOKAI 248-249 
  (d) Action by HELM 249-250 
       Prepares Torpedo BAGLEY 249 
       Reverses Course to Rejoin VINCENNES Group 249-250 
       Heads for Rendezvous North of Savo Island 250 
  (e) Action by WILSON 251-252 
    Shifts Fire to AOBA 251 
    Is Fired on by TENRYU 251 
    Ceases Firing 251 
    Avoids Collision with HELM 251 
    Resumes Fire on CHOKAI 251-252 
    Ceases Fire and Heads for Savo Island 252 
  (f) Action by CHICAGO 253-254 
    Heads to Westward to Investigate Firing 253 
    Fires Star Shells 253 
    Contacts BAGLEY 253 
    Fails to Report Activities to CTG 62.6 253-254 
  (g) Action by CANBERRA 254 
  (h) Action by BAGLEY 254-255 
  (i) Action by PATTERSON 255 
  (j) Action by BLUE 255-256 
       Contacts Small Schooner 256 
  (k) Action by RALPH TALBOT 257-259 
       Is Temporarily Illuminated by YUNAGI 257 
       Is Illuminated by TENRYU 257 
       Is Fired on By Tenryu and FURUTAKA 257 
       Considers Enemy Ships Friendly 257 
       Received One Hit 258 




       FURUTAKA and TENRYU Cease Firing at Ralph Talbot 257 
       Is Illuminated and Fired on by YUBARI 258 
       Is Fired on by FURUTAKA 258 
       Opens Fire on Yubari 258-259 
       Is Heavily Hit 259 
       Fires Three Torpedoes at YUBARI 259 
       FURUTAKA Ceases Firing at RALPH TALBOT 259 
  (l) Operations of CTG 62.6 (AUSTRALIA) 260 
       Awaits Information 260 
       Orders Destroyers to Concentrate 260 
  (m) Operations of San Juan Group 260-261 
  (0220 August 9th to 2400 August 9th)  
  (a) Withdrawal of Japanese Cruiser Force 262-266 
       (1) Operations of Commander Eastern Group 262-263 
          Continues Retirement 262-263 
       (2) Operations of Japanese Western Group 263-265 
          YUBARI Fires on RALPH TALBOT 264 
          Reforms Formation 264 
          YUBARI Receives Slight Damage 265 
          YUBARI Ceases Firing 265 
          Continues Retirement 265 
          Rejoins Eastern Group 265 
       (3) Operations of YUNAGI 266 
  (b) Operations of Japanese Cruiser Force 0340 to 0958 266-268 
       Assumes Original Column Formation 266 
       Prepares for an Attack 267 
       Considers Himself in Safe Waters 267-268 
  (c) Separation of Japanese Cruiser Force 0958 268 
  (d) Operations of Commander Japanese Cruiser Force 0958 to 2400 268-271 
       (1) Operations of Commander Bismarck Island Group 268-270 
          Retires Through Bougainville Strait 268-269 
          Estimates Situation 269-270 
       (2) Operations of Rabaul Group 0958 to 2400 270-271 
          Retires Through Bougainville Strait 270 
  (a) Operations of Commander Fifth Air Attack Force 272-279 
       Receives Word of Night Action at Savo 272 
       Estimates Situation 272 
       Conducts Searches 273-275 
       Launches Air Attack Group 276 
       Probable Operations Air Attack Group 276-277 
       Hears of Retiring Allied Cruiser 277 
       Re-Estimates Situation 277-278 
       Decides Attack Retiring Allied Cruiser 278 


  (b) Operations of Japanese Submarines 279 
  (0220 August 9th to 2400 August 9th)  
  (a) Operations of CHICAGO Group   280-289 
    (1) Operations of CHICAGO   280-283 
       Commanding Officer is Interrogated by CTG 62.6 and Replies   280 
       Discussion of His Operations   280-281 
       Commanding Officer is Interrogated Again by CTG 62.6 and Replies   281 
       Opens Fire on PATTERSON Which Returns Fire   282 
       Ceases Firing   282 
       Joins SAN JUAN Group   283 
       Rejoins TF 62   283 
       Summary of Shells Fired and Hits Received   283 
    (2) Loss of CANBERRA   283-285 
       Abandons Ship   284 
       Is Fired on by SELFRIDGE in Attempt to Sink   285 
       Is Fired on by ELLET by Mistake   285 
       Is Sunk by ELLET on Orders   285 
       Summary of Shells and Torpedoes Fired and Hits Received   285 
    (3) Operations of BAGLEY   285-287 
       Removes Survivors from ASTORIA   286 
       Discovers Many Additional Survivors Still on Board ASTORIA   286 
       Summary of Shells and Torpedoes Fired and Hits Received   287 
    (4) Operations of PATTERSON   287-289 
       Stands by CANBERRA   287 
       Removes Survivors   288 
       Engages CHICAGO, No Damage   288-289 
       Returns Area XRAY   289 
       Summary of Shells and Torpedoes Fired and Hits Received   289 
  (b) Operations of VINCENNES Group   289-297 
    (1) Loss of VINCENNES   289-290 
       Abandons Ship   289-290 
       VINCENNES Sinks   290 
       Bureau of Ships Comments on Loss   290 
       Summary of Shells Fired and Hits Received   290 
    (2) Loss of QUINCY   290-291 
       Abandons Ship   290 
       QUINCY Sinks   291 
       Bureau of Ships Comments on Loss   291 
       Summary of Shells Fired and Hits Received   291 
    (3) Loss of ASTORIA   291-296 


       BAGLEY Removes Some Survivors   292 
       Commanding Officer Endeavors Salvage   293 
       Taken Under Tow by HOPKINS   293 
       Towing Operations Discontinued   294 
       Abandon Ship   294 
       ASTORIA Sinks   294 
       Bureau of Ships Comments on Loss   295 
       Summary of Shells Fired and Hits Received   295 
    (4) Operations of HELM and WILSON   295-297 
       Rescue Survivors from VINCENNES, QUINCY and ASTORIS   296 
       Summary of Shells and Torpedoes Fired and Hits Received   297 
  (c) Operations of Radar and Anti-Submarine Screen   297-301 
    (1) Operations of BLUE   297-299 
       Contacts JARVIS   298 
       Rescue Survivors from CANBERRA   299 
       Summary of Shells and Torpedoes Fired and Hits Received   299 
    (2) Operations of RALPH TALBOT   299-301 
       Continues Engagement with YUBARI   299 
       Fires Torpedo at YUBARI   300 
       Makes one 5-inch Hit on YUBARI   300 
       Ceases Firing   300 
       Heavily Damaged Requires Assistance   300 
       Summary of Shells and Torpedoes Fired and Hits Received   301 
  (d)  Operations of CTG 62.6 in AUSTRALIA   301-309 
    Continues to Remain in Area XRAY   301 
    Queries His Groups as to the Battle   301 
    Receives Indefinite Replies   301 
    Queries Commander CHICAGO Group   302 
    Receives Laconic Reply   302 
    Receives CHICAGO's Amplifying Report   303 
    Informs CTF 62 of Battle   303 
    Queries PATTERSON   303 
    Receives Ambiguous Reply   303 
    Discovers His Destroyers are Concentrated at Wrong Rendezvous   304 
    Receives Report CHICAGO Heading Towards XRAY   304 
    Learns that CANBERRA is Heavily Damaged   304 
    Receives Orders Destroy CANBERRA   305 
    Directs COMDESRON FOUR Destroy CANBERRA if Cannot Retire by 0730   305 
    Concerned About Night Action   305 
    Informs Australian Commonwealth Naval Board of AUSTRALIA's Condition   306 
    Hears of Damage to ASTORIA, QUINCY and RALPH TALBOT   306 
    Endeavors Obtain More Information   307 


    Learn from FULLER of Loss of QUINCY and VINCENNES and Damage to ASTORIA   307 
    Broadcasts Air Raid Warning   307 
    Hears of Further Damage to RALPH TALBOT   307 
    Learns Full Details of Night Battle   308 
    Retires with TF 62 to Noumea   308 
  (e) Operations of SAN JUAN Group   309-312 
    CTG 62.4 Receives No Reports   309 
    Observes Ships on Fire and Apparently Sink   309 
    Observes Action Between CHICAGO and PATTERSON   310 
    Directs Ships of TG 62.4 Resume Daylight Screening Stations   310 
    Prepares for Battle in Transport Area   310-311 
    Receives Air Raid Alarm from CTG 62.6   311 
    Detects by Radar Japanese Reconnaissance Plane   311 
    Rejoins Squadron YOKE as Screen   311-312 
    Retires with Squadron YOKE   312 
  (0000 August 9th to 2400 August 9th)    
  (a) Operations of CTF 62   313-322 
    Advises CTF 61 Will Have to Retire TF 62   313 
    TF 62 Gets Underway   314 
    Hears That CHICAGO Has Been Torpedoed   315 
    Learns That CTG 62.6 is not in Battle   315 
    Intercepts Dispatch Authorizing Retirement TG 61.1   315 
    Hears CANBERRA is Heavily Damaged   315 
    Decides Retire TF 62   316 
    Discussion of this Decision   316 
    Learns RALPH TALBOT Heavily Damaged   316 
    Becomes Concerned About Unloading Operations   317 
    Decides Remain Tulagi Area with TF 62   318 
    Advises CTF 61   318 
    Advises CTG 62.6 Concerning Damage to ASTORIA, QUINCY, RALPH TALBOT   318 
    Receives Contact Report on Unidentified Planes   318 
    Discusses Unloading Operations with Commanding General, First Marine Division   319 
    Reports Progress of Action to CTF 61   319 
    Requests Admiral Fletcher's Plan   319 
    Air Raid Warning Causes Discontinuance of Unloading   319 
    Learns Full Details Night Battle   320 
    Orders Unloading Continued   321 
    Learns Will Obtain No Air Support From CTF 61   321 
    Re-estimates the Situation   322 
    Decides Retire   322 
    TF 62 Retires   322 


  (b) Operations of CTF 61   323-327 
    Hears Flash Reports About Night Battle   323 
    Receives Orders COMSOPACFOR Authorizing Retirement   323 
    Commences Retirement   324 
    Hears Additional Flash Reports   324 
    Takes No Action   324 
    Receives CTF 62's Request for His Plan   325 
    Receives Word that SARATOGA's Searches had been Negative but had Located JARVIS   325 
    Hears that CTF 62 Plans Retire   325-326 
    Advises COMSOPACFOR of Losses and Damages in Night Battle   326 
    Continues Retirement   326-327 
  (c) Operations of CTG 61.1   327-330 
    Weather Conditions   327 
    Receives Flash Reports of Night Action   327 
    Conducts Planned Search Operations   327-329 
    Locates JARVIS   328 
    Fails to Collect Information Concerning Night Battle   328-329 
    Continues Retirement   330 
  (d) Operations of Allied Submarines   330-331 
    (1) Operations of S-38   330-331 
       Sinks Japanese Transport Meiyo Maru   330 
    (2) Operations of S-44   331 
  (e) Operations of CTF 63   331-334 
    Conducts Planned Search Operations   331 
    Misses Japanese Cruiser Force   332 
    Launches Air Attack Against Shipping in Rekata Bay   333 
  (f) Operations of Commander Allied Air Forces SOWESPAC   334-335 
    Conducts Planned Search Operations   334 
    Contacts CRUDIV SIX   334 
    Contact Report Not Received in SOPAC   335 
       (1) Allied Operations   337 
       (2) Japanese Operations   337 
        (3) Situation as of 2400 August 10th   337-338 


  1. Relation Between Strategy and Tactics 343 
  2.  Importance of Surprise 344 
  3.  Correct Location of the Commander of an Expeditionary Force 344-345 
  4.  Necessity for Making Every Effort to Accomplish the Objective 345-346 
  5.  Importance of Correctly Reporting Enemy Damages 346-347 
  6.  Discussion Concerning Division of Forces 347-348 
  7.  Capabilities vs. Intentions in Planning 348-349 
  8.  Importance of Setting Conditions of Readiness Promptly 349-350 
  9.  Influence of Technological Advantages on Naval Operations 350 
  10.  Commander Should have Operational Control over Shore-Based Aircraft Assigned 350-351 
  11.  Necessity for Providing Land and Tender-Based Aircraft Adequate in Numbers and in Training 351-352 
  12.  Fundamentals in Planning 352 
  13.  Necessity for Promptly Broadcasting Contact Reports 352-353 
  14.  Officer-in-Tactical-Command Should be Informed of the Various Changes in the Situation 353-354 
  15.  Advisability of Providing Battle Plans 354-355 
  16.  Importance of Advising Command of Changes in Officer-in-­Tactical Command 355 
  17.  Employment of Commanding Officer as Group Commander as well not Recommended 355-356 
  18.  Gunnery Effectiveness Stems from Gunnery Training 356-358 
  19.  Necessity for Improvement of Professional Judgment in Command 359-360 
  20.  Functions of a Carrier Covering Force 360-361 
  21.  Importance of Damage Control Training 361-362 
  22.  Necessity for Maintenance of Reliable, Rapid and Secure Communications 362-363 
  23.  Importance of Correct Identification and Recognition 363-364 
  24.  Importance of Mobile Logistics Support in the Operating Area 364-365 
  25.  Task Organizations Should be Flexible 365 
  26.  Tactical Voice Radio Discipline Should be Maintained 365-366 
CHAPTER XXIV Combat Appraisal of the Japanese Commander Cruiser Force 367-369 
  * * *  
Appendix I Organization of Southeast Area Force at the Time of the Battle of Savo Island 370-371 
Appendix II Organization of the South Pacific Force at the Time of the Battle of Savo Island 372-374 
Appendix III Organization of Southwest Pacific Forces Which Assisted SOPAC Operations at the Time of the Battle of Savo Island 375 
Appendix IV Summary of Japanese Damage 376 
Appendix V Summary of Allied Damage 377 


  Disposition of Usable Japanese Shore and Tender-Based Aircraft as of 2400 August 6th, 1942
  Disposition of Allied Shore and Tender-Based Aircraft as of 2400 August 6th, 1942 22 
  Disposition of Allied Shore and Tender-Based Aircraft as of 2400 August 7th, 1942 70 
  Disposition of Allied Shore and Tender-Based Aircraft as of 2400 August 8th, 1942 100 
  Japanese Command Relations, August 8th
  II  Allied Command Relations, August 8th 16 
  III  Communication Between SOPAC and SOWESPAC 31 
  IV  Night Disposition of Allied Screening Force 55 
  Action With CHICAGO Group, 0143 to 0144 118 
  VI  Action With CHICAGO Group, 0144 to 0145 119 
  VII  Action With CHICAGO Group, 0145 to 0146 120 
  VIII  Action With CHICAGO Group, 0146 to 0147 121 
  IX  Action With CHICAGO Group, 0147 to 0148 122 
  Action With CHICAGO Group, 0148 to 0149 123 
  XI  Action Between Japanese Eastern Group and VINCENNES Group, 0150 to 0151 152 
  XII  Action Between Japanese Eastern Group and VINCENNES Group, 0151 to 0152 153 
  XIII  Action Between Japanese Eastern Group and VINCENNES Group, 0152 to 0153 155 
  XIV  Action Between Japanese Eastern Group and VINCENNES Group, 0153 to 0154 156 
  XV  Action Between Japanese Eastern Group and VINCENNES Group, 0154 to 0155 158 
  XVI  Action Between Japanese Eastern Group and VINCENNES Group, 0155 to 0156 159 
  XVII  Analysis of Courses of Action 2 and 3 169 
DIAGRAMS     Follows 
  "A"  Strategic Area Chart 377 
  "B"  Air Searches and Movement of Forces, 6 August 378 
  "C"  Air Searches and Movement of Forces, 7 August 381 
  "D"  Air Searches and Movement of Forces, 8 August 385 
  "E"  Japanese Approach Past Radar Pickets, 0000 to 0132 Chart "D" 
  "F"  Action With CHICAGO Group, 0132 to 0150 387 
  "G"  Opening Plan of Action With VINCENNES Group 0150 to 0200 389 
  "H"  Final Phase of Action with VINCENNES Group, 0200 to 0220 394 
  "I"  Withdrawal of Japanese Cruiser Force, 0220 to 0240 Chart "H" 
  "J"  Composite Track Chart Chart "I" 
  "K"  Air Searches and Movement of Forces, 9 August 398 



This analysis of the Battle of Savo Island was prepared by the Naval War College. It is based on information from both Allied and Japanese sources which is wider and more complete than that available to writers on this subject up to this time. It endeavors to maintain, at all times, the viewpoint of the Commanders of the units involved on both sides.

Complete information from all sources was not available to this analysis. This was especially true concerning Japanese information. Unfortunately, sufficient translators were not available to provide all of the additional translations which the progress of the analysis indicated were desirable. New facts and circumstances, therefore, may come to light, from time to time, which may change some of the analyses produced herein.

In view of the critical nature of this analysis an effort has been made in certain important situations to place the critic in the position of the Commander in order to obtain the latter's point of view. In employing this system it is realized that although the critic can often succeed in placing himself sufficiently near the position of the Commander for any practical purposes, in many instances he may not succeed in doing so.

Because of the nature of this Allied defeat and the numerous controversies which have arisen concerning it, a complete background has been provided. In addition, when the time comes to analyze the other battles in the struggle for Guadalcanal, this material will be available.

The Battle of Savo Island was a real test of existing Allied and Japanese night tactical concepts as well as of the combat ability in night action of the various Commanders on both sides. The pages of history have invariably revealed defects in command in similar situations, and it would have been surprising had such defects not appeared in this action.

As a result of battle lessons learned, and ass quickly applied, the ability of the Navy to conduct warfare steadily improved during the course of the war. As time went on the lesson so often forgotten - that the test of battle is the only test which proves the combat ability of Commanders - was relearned. The ability or the lack of ability of the various Commanders in the art of war became apparent. Valor alone was shown to be insufficient, for valor is not an attribute of only one race, but is an attribute and a heritage of many races. The indispensable qualification for command, the art of war, was shown to be the ability in combat to apply the science of war to active military situations.

The present senior officers of the Navy are well aware of the reasons for changes in established doctrines and in the development of new ones. But this cannot necessarily be said of the Commanders of the future, who very probably will be inexperience in command in war.


Finally, all comments and criticisms are designed to be constructive. By indicating what appear to be sound and unsound decisions, and the apparent reasons for arriving at them, it is hoped to provide earnest thought among prospective commanders and thus to improve professional judgment in command.


[Note: Table of Contents moved to beginning of project, pages iii - xxii.] 





Principal Commanders

Commander Screening Group, CTF 62.6       Rear Admiral V.A.C. Crutchley, RN
Commander Australia (Chicago) Group       Captain Howard D. Bode, USN
Commander Vincennes Group       Captain Frederick Reifkohl, USN
Commander San Juan Group, CTG 62.4       Rear Admiral Norman Scott, USN
Commander Amphibious Force, CTF 61.2 or CTF 62       Rear Admiral Richmond K. Turner, USN
Commander Allied Expeditionary Force, CTG 61; also CTG 61.1.1       Vice Admiral Frank J. Fletcher, USN
Commander Aircraft South Pacific Force (COMAIRSOPAC), CTF 53       Rear Admiral John S. McCain, USN
Commander Southern Pacific Area (COMSOPAC)
and Commander Southern Pacific Naval Forces (COMSOPACFOR)
      Vice Admiral Robert Ghormley, USN
Commander in Chief, Pacific Fleet (CINCPAC)
and Pacific Ocean Areas (CINCPOA)
      Admiral Chester W. Nimitz, USN
Commander Allied Air Forces Southwest Pacific Area (COMAIRSOWESPAC)       Major General George C. Kenney, (Air Corps) USA


Commanding General South Pacific Area (COMGENSOPAC)       Major General Millard F. Harmon, (Air Corps) USA
Commander Allied Naval Forces Pacific Area (COMNAVSOWESPAC)       Vice Admiral Fairfax Leary, USN
Supreme Commander Southwest Pacific Area (COMSOWESPAC)       General Douglas MacArthur, USA
Commander Cruiser Force       Vice Admiral Gunichi Mikawa, IJN
Commander Cruiser Division SIX       Rear Admiral Aritomo Goto, IJN
Commander Cruiser Division EIGHTEEN       Rear Admiral Mitsuharu Matsuyama, IJN
Commander 26th Air Flotilla       Rear Admiral Sayayoshi Yamada, IJN
Commander Southeast Area,
Commander 11th Air Fleet and
Commander Outer South Seas Force
      Vice Admiral Nighigo Tsukuhara, IJN
Commander in Chief, Combined Fleet       Admiral Isoroku Yamamoto, IJN



The Battle of Savo Island is an action of singular interest to students of naval history for several reasons: It was the first occasion in which Japanese and Allied naval forces had engaged in night battle since the Allies had assumed the offensive; it was a serious tactical defeat to the Allied forces, and finally it was a classic example of a powerful night raid by surface forces with the attendant features of surprise and withdrawal. During the early operations of 1942 Japanese and Allied surface forces had clashed in several minor night actions but no night surface actions had occurred between cruiser forces of approximately equal strength.

The battle resulted from the seizure by the Allies of the Tulagi-Guadalcanal area of the lower Solomon Island sin order to protect the Allied lines of communications to Australia and New Zealand. These lines were being menaced by the expansion of the Japanese in that direction.1

In order to check this expansion and to contain the Japanese within the already occupied area, limited countering moves had been undertaken by the Allies. In the Battle of the Coral Sea, May 7th, the Allies had prevented entirely by carrier air action, the capture of Port Moresby by sea. In the Battle of Midway, June 3rd-6th, the Allies had inflicted, almost entirely by air action, such losses on the Japanese in aircraft carriers as to force the abandonment by them of planned advances into New Caledonia and the Fijis. However, within the Solomon Islands, the Japanese had continued to advance.2

On May 2nd they had overrun the island of Tulagi where they were attacked by an Allied carrier task force on May 4th, just before the Battle of the Coral Sea. On July 4th they had commenced building an airfield at Lunga Point, Guadalcanal Island. Actually, they were building this airfield, as well as others in the Solomon Island chain, primarily for the protection of Rabaul which was to be one of the strong points of the Japanese southern perimeter.2

With the Allied seizure of Tulagi-Guadalcanal, the Japanese realized that failure to dislodge the Allies might seriously affect Japanese strategy in that area. They, therefore, put into force energetic counter-measures designed to dislodge the Allies form their precarious foothold. These counter-measures which began on August 7th were highlighted on the early morning of August 9th, by the night action known as the Battle of Savo Island.


1. U.S. Navy at War 1941-1945, Official Report by Fleet Admiral Ernest J. King, USN, page 49.
2. Campaigns of the Pacific War, USSBS 1946, page 105.


A Brief Narrative of the Battle of Savo Island

The directive of the Joint Chiefs of Staff, issued by COMINCH to CINCPAC on July 2nd, 10942, put into effect the Allied plan of denying the New Britain-New Guinea-New Ireland Area to the Japanese. This plan assembled an Allied force in the South Pacific area under COMSOPACFOR for the purpose of seizing, in its first phase, the Santa Cruz Islands, Tulagi and adjacent positions.

The Allies planned the above operations with the utmost secrecy. However, the Japanese Commander in the area knew that the Allies were reconnoitering frequently the Tulagi-Guadalcanal area, and that during August, the intensity of this reconnaissance had increased. He correctly assumed that the Allies would counterattack Japanese positions in the Solomons in the near future, though not as early as August 7th. However, he had not taken any special precaution, either of air search or of reinforcing outlying bases.

Meanwhile, COMSOPAPCFOR had organized an expeditionary force consisting of an Amphibious Force and an Air Support force, the latter composed of three single carrier task groups, to seize Tulagi and Guadalcanal. This force was to be supported by land-based support aircraft operating from islands in the SOPAC Area, as well as from SOWESPAC air bases.

The Allied Expeditionary Force successfully landed on Tulagi and Guadalcanal Island at daybreak on August 7th, having achieved complete surprise.

The Japanese immediately decided that it was necessary to drive the Allies out and employed their land-based air forces and their limited surface forces for this purpose. They conducted air attacks on the Allied transports and cargo ships at Guadalcanal during two successive days. On August 7th they launched two air attacks; on August 8th they launched one air attack. All of these attacks seriously interfered with the unloading operations, but they were otherwise ineffective.

Meanwhile, on August 7th, the Japanese Commander at Rabaul decided to attack the Allied shipping at Guadalcanal with surface ships as well as with aircraft. He therefore assembled a cruiser force consisting of five heavy cruisers, the Chokai, Aoba, Kako, Kinugasa, and Furutaka; of two light cruisers, the Tenryu and Yubari; and of one destroyer, the Yunagi, and at 1628 dispatched it, commanded in person by himself, to the Tulagi-Guadalcanal Area to destroy the Allied transports and cargo ships there. While this cruiser force was assembling off Rabaul, it was sighted at 1231 by allied bombers from SOWESPAC. This contact report reached COMSOPAC at 2400.

As this cruiser force moved south toward Guadalcanal, it was first sighted at about 1930, August 7th by the Allied submarine S-38 stationed off Cape St. George, New Zealand. The submarine was forced to submerge.


Upon surfacing at about 2010 the submarine incorrectly reported the composition of the force as two small and three larger vessels. This report was received by COMSOPAC at 0738, August 8th.

Some three hours later or at 1025, August 8th, the cruiser force was sighted by an R.A.A.F. plane operating from Milne Bay, and was incorrectly reported as three heavy cruisers, two destroyers and two seaplane tenders or gunboats, Later on the same day, it was sighted at 1101 by another R.A.A.F. plane which reported it as two heavy or light cruisers and one unidentified vessel. The first report was received by COMSOPAC at 1845; the second report at 2136. The long delay in transmitting these reports denied COMSOPAC the opportunity of requesting an immediate attack by COMSOWESPAC planes. The Allied high command at Tulagi-Guadalcanal, incorrectly, decided that these ships were bound for Rekata Bay, there to set up an air base from which to bomb the Allied shipping in Iron Bottom Sound, commencing on August 9th.

Commencing at sunset each evening the transports and cargo ships of the Allied amphibious force were covered by a night screening disposition designed to meet Japanese surface ship attacks from all entrances to Iron Bottom Sound. The disposition was composed of an Eastern and Western Screening Group. The Western Screening Group which fought the Battle of Savo Island consisted of two anti-submarine and radar pickets, the Blue and Ralph Talbot, stationed west and north, respectively of Savo Island, and of two cruiser-destroyer support groups. One support group, the Australia Group, consisting of the Australia, Chicago, Canberra, Patterson and Bagley guarded the passage northeast of Savo Island.

During the night of August 8th at about 2055, the Commander of the Western Screening Group, in the Australia, left his group under the tactical command of the Commanding Officer, Chicago and stood into the Guadalcanal Area to attend a conference with Commander Amphibious Force and the Commanding General, First Marine division.

At this conference which was held at 2326, August 8th, Commander Amphibious Force decided to retire his force the following morning, because the carrier Commander had decided to retire. This would leave the amphibious forces without air cover or air support.

However, as many necessary supplies had not been landed, the Commander Amphibious Force decided to unload as much of the supplies as was possible throughout the night, and an effort was made to accomplish this.

After this conference, the Commander Western Screening Group remained in the Guadalcanal Area with the Australia, this reducing the Australia group by one heavy cruiser.


The Japanese Cruiser Force entered Iron Bottom Sound at about 0132, August 9th, having passed the Allied destroyer Blue without being discovered. At 0236 this force surprised the Australia Group and aided by aircraft flares dropped by its own cruiser planes, opened fire with both guns and torpedoes. Within a matter of minutes, the Canberra had bee heavily. Hit and was out of action (it was finally by Allied destroyer torpedoes at 0800); the Chicago had been torpedoed in the bow and had had her speed considerably reduced; the Bagley had fired four torpedoes ineffectively, and the Patterson had broadcasted the contact before engaging the two Japanese cruisers. The Patterson was lightly damaged in this akin, none of the sips of then Chicago Group inflicted abs important damage on the Japanese ships.

During this phase of the action the Japanese Cruiser force broke up into three groups. One consisting of the, Aoba Chokai, Kako, and Kinugasa continued on towards the Vincennes group at twenty-six knots. It caught the Vincennes Group at 0100 by surprise and, employing both guns and torpedoes, succeeded in heavily damaging the three cruisers of that group. In accomplishing this the Chokai group passed under the stern of and along the starboard side of the Vincennes Group. Another group, consisting of the Tenryu, Yubari, and Furutaka passed along the port side of the Vincennes Group and employing guns and torpedoes increased the heavy damage caused by the first group. Both of these Japanese groups intermittently employed searchlights rather than star shells for illumination. As a result of the above damage, the Quincy sank at 0238, the Vincennes at 0250 and the Astoria at 1215. The third Japanese group consisting of the destroyer Yunagi turned away from the cruiser action and ineffectively engaged I a gun battle wait the retiring allied destroyer, Jarvis. Jarvis was sunk later in the day by air attack at about 1800, August 9th.

After this action, Commander Cruiser Force, because, because of a fear of Allied carrier-based planes, decided to retire rather than to remain for the purpose of destroying the Allied shipping at Tulagi and Guadalcanal. As he retired he engaged the Allied destroyer Ralph Talbot with the Tenryu, Yubari, and Furutaka.

The Japanese Cruiser Force, except for several hits, escaped from the entire action almost entirely unscathed. In its retirement, it succeeded in avoiding Allied search planes so that it received no damage from Allied air attacks. However, it was discovered by the allied submarine, S-38, on August 10th as a result of which the Kako was torpedoed and sunk.

The effects of this battle were serious insofar as the Allies were concerned. There was a shortage of heavy escort and bombardment ships in all sea areas and as the Allies were preparing to conduct the North African Invasion, as well as to conduct offensive operations in the Aleutians and to continue those already underway in the Solomons, the


loss of four heavy cruisers and one destroyer was immediately felt in all areas.

The Japanese on the other hand grossly exaggerated the Allied losses in surface ships and informed the world and the Japanese people by press and radio of their claims.


Chapter I
The Strategic Area


The strategic area involved in the combat and support operations of the Solomon Islands offensive extended from the Equator southward to the Tropic of Capricorn and from the 180th Meridian westward to the coasts of Australia and New Guinea. This over-all area, besides being entirely tropical, was oceanic. The total land mass of the small islands east of Australia and New Guinea is no greater than n the area of Celebes--or to use a more familiar example, no greater than the area of any one of the states of Washington, Missouri or Oklahoma.

The outstanding characteristic of this area is the chain of small islands extending from New Guinea to New Caledonia through the Bismarck Archipelago, the Solomon Islands, Santa Cruz Islands, New Hebrides and Loyalty Islands. These islands define the eastern boundary of the Coral Sea, including the Solomon sea north of the Louisiade Archipelago. As a series of potential air and sea bases, they constitute a barrier between Australia and North America.

This strategic area was adaptable to limited offensive operations. On the one hand, the Japanese could slowly extend themselves deeply into the South Pacific through the island chain to threaten the life-line from Australia and New Zealand to the United States. On the other hand, the Allies were provided with a route of advance toward Japanese strong points in the Western Pacific. They were also provided with an extremely advanced position from which to effect attrition upon the limited war potential of Japan. Consequently the strategic area became a theater of combat.

The strategic area was divided into two approximately equal areas by the boundary between the Southwest Pacific area and the South Pacific area along meridian 159 degrees East Longitude. This boundary passed through middle of Santa Isabel Island in the Solomons, and passed about 35 miles west of Guadalcanal Island. The Guadalcanal-Tulagi target area, wherein was fought the battle of Savo Island, lay wholly within the South Pacific area.

The threat of a large-scale overseas attack by the Japanese upon Australia and New Zealand through this area had been temporarily removed by the battle of the Coral Sea in May 1942 and the Battle of Midway in June of that year. However, the Japanese had continued to advance slowly thereafter by piecemeal penetration overland southward along the coast of New Guinea and by sea southward through the Solomons.

The relative positions of the Japanese and the Allies within this strategic area in August 1942 provided the Japanese with the more favorable military situation. The Allied positions in the south Pacific Area


were but lightly held, and consisted chiefly of airfields under construction which had been hurried to a degree of readiness for limited use in the days just prior to August 7th, 1942. These airfields were located on New Caledonia, Efate, Espiritu Santo, and in the Fijis. The Japanese positions in the Bismarck Archipelago, with their principal advanced base at Rabaul, were firmly held. Their positions at Kavieng, Rabaul, Las, Salamaua, and Buka were less firmly held and were being strengthened by a major effort in the development of airfields.* In improving these positions the Japanese had extended their occupation of New Guinea as far south as Buna, had occupied Florida Island in the Solomons, and had moved into the north-central shore love Guadalcanal Island on July 4th where they had immediately begun the construction of an airfield, harbor facilities, and other installations at large at Lunga Point.

It was apparent to the Allied High command that the firm establishment of the Japanese at Tulagi and Florida Island and its airfield at Guadalcanal would seriously hamper, if not prevent the Allies from establishing themselves in Espiritu Santo and Santa Cruz,** and that therefore the Japanese positions in the Solomons were a threat to vital links in the Allied communications. It was also apparent that the Allied positions relative to the Japanese were potentially less menacing.

The positions of the Japanese and the Allies, relative to the Guadalcanal-Tulagi area in air miles.

    Maramasike Estuary, Malaita (A) ***  90  miles 
    Rekata Bay, Santa Isabel (J) 120 
    Kieta, Bougainville (J) 340 
    Ndeni, Santa Cruz Islands (A) 365 
    Buka Island (J) 392 
    Rabaul, New Britain (j) 540 
    Milne Bay, New Guinea (A) 565  
    Espiritu Santo (A) 580 
    Kavieng, New Ireland (J) 665 
    Efate (A) 715 
    Port Moresby, New Guinea (A) 745 
    Plaines des Gaiac, New Caledonia (A) 790 
    Noumea, New Caledonia (A) 875 
    Townsville, Australia (A) 970 
    Nandi, Fiji Islands (A) 1155 

There were no allied harbors in the area suitably equipped to effect major repairs to damaged ships. The docking facilities at Pearl Harbor were about 4000 miles away via Fiji. At Sydney, which was about 2000 miles away from Savo Island via Noumea, there were repair facilities for


* ComSoWesPac Dispatch 081018 which is Part Five of Dispatch 081012, July 1942
** COMINCH Dispatch 102100, July 1942
*** (A) means Allied - (J) means Japanese


cruisers and smaller craft. In an emergency, smaller harbors in the area could be used for repairs, although they were not equipped with supporting activities or base facilities. In particular, these were Saint James Bay, Espiritu Santo Island, and Noumea, New Caledonia. The best harbor site in the Solomons was at Tulagi on Florida Island, but this position was exposed tom enemy attack and was, prior to the landing on august 7th, in Japanese hands.


The Battle of Savo Island was fought in the body of water lying between Guadalcanal and Florida Island which became known among Allied Forces as Iron Bottom Sound. This sound is the southeastern extremity of the Slot (New Georgia Sound). Then 100 fathom curve off Guadalcanal and Florida Islands are 20 miles apart at the western end of this Sound and close on each other to a distance of one mile to form the Sealark channel at the eastern end of the Sound.

Savo Island lies between the western extremities of Florida and Guadalcanal Islands, and is an approximately round volcanic island with a diameter of 4 miles and with its highest peak at 1675 feet. The 100-fathom curve is nowhere more than 1200 yards distant from the shore of Savo Island, allowing an restricted navigation closed to its shores. Savo Island divides the western approaches to Iron bottom sound into two wide passages. The southerly passage between Savo Island and Cape Esperance on Guadalcanal is seven miles wide. The northerly passage between Savo Island and Florida Islands is seven miles wide.

At its eastern end, Florida Island is thirteen miles north of Guadalcanal Island and numerous shoal patches and reefs lie within the 100-fathom curves of each island on either side of Sealark channel. Through the fouled ground off Florida Island runs the Niggela channel. Through the shoals off Guadalcanal al runs the Longo channel which is three or four miles wide with depths of eighteen to thirty fathoms.

Strong tide rips are caused by the numerous shoals and irregularities in the bin the bottom between the eastern end of Florida and Guadalcanal Islands. The tidal currents set to the westward and to the eastward along the coasts of Florida and Guadalcanal Islands, following the trend of the coast line and attaining the velocity of two knots at springs.

The general depth of water in Iron bottom Sound is 200 to 400 fathoms between Lung Point on Guadalcanal and Tulagi Harbor. Tulagi Harbor is situated midway along the southern coast of Florida Island and is the principal port of the Solomon Islands. Westward of Tulagi Harbor there are no off-lying dangers, and the 100-fathom curve runs parallel with the coast about three miles offshore.


In the Solomon Islands area there are two marked climatic seasons.


The northeast monsoon prevails from November to March. The southeast trade wind usually establishes itself by April and lasts until October. Winds are strongest during the season of the southeast trades, averaging 10-20mph. Average wind velocities are only rarely interrupted by the passage of a tropical cyclone.

It is hot throughout the year. Average monthly temperatures range from 81 degrees (F) to 84 degrees (F). Highest temperatures are recorded from November to April.

Throughout the year, average cloudiness amounts to about 5 tenths cloud cover. The cloudiest months are January, February and March. Apart from rare cyclones, other tropical depressions cause overcast skies of intermediate and high clouds. Lower type clouds prevail with local thunderstorms and rain squalls.

Thunderstorms normally occur over land during the afternoon and over the water in the early morning. Rain squalls may occur during any season of the year. Rainfall is abundant. The months of January, February and March are the wettest. At Lunga Point, Guadalcanal the average rainfall for July is 2.70 inches. At Tulagi, average rainfall is slightly greater.

Fogs at sea level are unknown. Reduced visibility is due to heavy downpours of rain, haze, or low clouds. A particularly heavy rain squall may reduce visibility to 100 yards over a limited area. Haze is prevalent during the season of the southeast monsoon. It may render objects indistinct in visibility of 15-20 miles.*

On a cloudy night in Iron Bottom Sound, visibility has been observed to be extremely low. Rain squalls seem to form over Savo Island almost every night of the year, causing heavy rain about 2330 and clearing about 0200 as they drift slowly southeastward.** On the night of August 8th-9th, this type of bad weather did not clear up. This will be shown in the discussion of the battle itself.


* All of the above is from War Department publication SSC-677, survey of the Solomon Islands, March 15, 1943, Confidential.
** CTF 62 personal letter to Captain R.C. Parker, USN(Ret), Office of Naval History, Navy Department, Washington, D.C. dated 1948.


Chapter II
Japanese Arrangements


All of the Japanese Fleets, including the Naval Air Fleets, excepting the China Seas Fleets, were under the Command of the Commander-in-Chief of the Combined Fleet. The Combined Fleet consisted of the mobile (mission) fleet which could operate anywhere in any area, and of the localized (area) fleets which were responsible for and restricted to certain geographical areas. The mobile fleet constituted the main striking force of the Combined Fleet. The area fleets were normally defensive in character and were generally unable to take any strong offensive action without assistance from the mobile fleet.* The mobile fleet consisted of the Main Body (First Division FIRST Fleet), the Advance Force (SECOND Fleet), and the Striking Force (THIRD Fleet),** the area fleets consisted of the Northern Force, later called Northeast Area Force (FIFTH Fleet) based at Horomushiro; the Inner South Seas Force (FOURTH Fleet) based at Truk, the Outer South Seas Force (EIGHTH Fleet) based at Rabaul and the Southwest Area Force (Combined Expeditionary Fleet) based at Surubaya.** In addition to the mobile fleet and the area fleets, the Combined Fleet consisted of the Base Air force (ELEVENTH Fleet) base on Tinian and the Advanced Expeditionary Force (SIXTH fleet) which was composed of submarines and was based at Kwajalein.** All tactical titles employ the term "Fleet" with the single exception that the title "Combined Fleet" was both administrative and tactical (Plate I)***

Prior to July 14th, 1942 the South seas Force was responsible for the defense of the Central and southeastern Pacific Area, including the Marshalls, New Guinea, and the Solomon Islands. However, due to the increasing importance of the Central and southeastern Pacific Areas and to the fact that there was an ever-increasing threat of an attack on the Solomons by the Allies, the Japanese High Command decided to split the South Seas Area into two areas, one, an Outer South Seas Area, the other, an Inner South seas Area. This split became effective on July 14th, 1942.****

The dividing line between the two areas thus formed was established as the line bearing 280°(T) from the juncture of the equator and Long.


* Japanese Naval Organization, Change No. 11 to ONI 49, page 5.
** Enclosure, Submitted October 26th, 1945 by Rear Admiral Nakamura, IJN, in answer to USSBS Memorandum No. NAV-1, October 10th, 1945.
*** Japanese Naval Organization, Change No. 11 to ONI 49, page 33.
**** War Diary, 8th Fleet, August 7th, 1942, WDC Document 161259.


150° E. Commander Inner South Seas Force was assigned the responsibility for the defense of the area north of this line and Commander Outer South Seas Force for the defense of the area south of this line. In order to provide for unity of command, it was directed that in the event of an enemy attack in the above areas, the Commander ELEVENTH Air Fleet would, in case of necessity, exercise over the FOURTH, SIXTH, AND EIGHTH Fleets.*

Vice Admiral Gunichi Mikawa activated the EIGHTH Fleet on July 14th. On July 16th he hoisted his flag in the Chokai. On July 23rd he was assigned the Chokai and Cruiser Division SIX and became Commander Outer South Seas Force. On July 27th he formally assumed responsibility for the New Guinea and Solomon Islands operations.*

All land based air forces in this operation wee naval and were assigned to the FIFTH Air Attack Force** (25th Air Flotilla), which was based at Rabaul and was a subdivision of the ELEVENTH Air Fleet. The Commander of this attack force retained operational control over his component and its and cooperated with the Commander Outer South Seas Force. He remained at Rabaul to direct its activities. No Army air forces were assigned at this time.***

Vice Admiral Mikawa was in command of the Outer South Seas Force until about 1415, August 7th when he was relieved by Vice Admiral Nishiso Tsukahara, who was the Commander Base Air Force and the Commander ELEVENTH Air Fleet. The latter admiral acted as Commander Outer South Seas Force which included the Base Air Force, the Inner and Outer South Seas Forces, and a portion of the Advanced Expeditionary Force.*** In this latter capacity he was the Immediate-Superior-in-Command of Commander EIGHTH Fleet.

It therefore happened that at the time of the Allied landings on Guadalcanal and Tulagi on August 7th, Vice Admiral Mikawa was both Commander Outer South Seas Force and Commander EIGHTH Fleet.** After he had been relieved by Vice Admiral Tsukahara he had the primary responsibility of securing a victory with his surface force.

From the above, it is evident that, although the Japanese command structure had room for confusion, there was no confusion in command in the battle of Savo Island. All forces wee engage naval forces, and all were under the direct command of naval officers in a clearly defined chain-of-command.


* War Diary 8th Fleet, August 7th, 1942, WDC Document 161259.
** Records 25th Air Flotilla, CIG Document 74629, May 12th 1947.
*** War Diary 25th Air Flotilla, August 1942, WDC Document 161730.


Plate 1: Japanese Command Relations on 8 August 1942 chart - The Battle of Savo Island August 9, 1942 Strategical and Tactical Analysis





Just prior to the Allied landing in the Solomons, Commander South Seas force had considerable general information concerning the strength of the Allied forces in the south Pacific area, and was generally aware of their capabilities.

He knew that there had been considerable Allied air and naval activity in the general area of the Coral Sea and that there was a possibility of an early engagement with a powerful enemy force.*

He was aware of the Allied construction of and build-up of bases in New Caledonia and New Hebrides. Interpreting this activity as a preparation for a counterattack upon Japanese advanced positions, he correctly assumed that the most likely area of allied counterattack would be the southern Solomons.**

He knew that the Allied forces had reconnoitered the new base at Guadalcanal-Tulagi frequently by air and that the number of planes raiding that area had suddenly increased during August.***

Informed by his intelligence that a powerful Allied force had sortied from Hawaii on August 2nd, he suspected that this force together with other Allied forces already in the southwest Pacific Area, would attempt to attack the Outer South Seas Area in the near future. He understood that this combined force would include three carriers (including converted carriers), and a number of cruisers.**** This Japanese intelligence was reasonably correct, except that the "powerful Allied force" had sortied from Fiji on August 1 rather than from Hawaii on August 2nd, and was at that date much closer to the Solomons than the Japanese expected.

Regardless of how powerful this allied force, he believed that he might be able to intercept and destroy it by decoying it within range of his land-based naval air power.*** He felt that he might accomplish this even without carrier task force assistance from the Combined fleet.**** This concept of destroying American surface forces by land-based planes was not new. The Japanese had held the same concept at the battle of Midway when they endeavored to lure the American carrier force within


* War Diary 8th Fleet, August 7th, 1942, WDC Document 161259.
** War Diary 8th Air Attack force, August 1942, WDC Document 161730, and CRUDIV 6 Detailed Battle Report #8C, August 7th-10th, 1942, CIG Document 86927, June 27th 1947, page 2.
*** Southeast Area Naval Operations, Part 1, document 40427, page 10, Historical Division, U.S. Army, July 31st, 1947.
**** War Dairy 8th Air Attack Force, August, 1942, WDC Document 161730 and Commander 8th Fleet Estimate of situation, August 7th, 1942, CIG Document 73846, May 7th, 1947, page 1.


range of Japanese land-based air power at Wake. It showed a confidence that their land-based air power was more effective then American carrier-based air power--a confidence that was not borne out throughout the war.

He knew that the enemy had recently assembled great air strength in the New Hebrides area and he considered that the enemy was planning to invest Guadalcanal before the Japanese air strength moved up.*

In anticipation of this latter possibility, he had recommended, in his capacity of commander EIGHTH Fleet, that a planned withdrawal of a cruiser division from Rabaul to Truk be modified, and that it should be sent to Kavieng or to Queen Carola Harbor.* This action was taken, and that CRUDIV SIX was held in the Kavieng Area as a support force of the Outer South Seas Force Area

He knew that larger numbers of enemy submarines were beginning to infest the waters of Japanese Southeast Area Force, necessitating the diversion of additional escort forces to this area. He considered it probable that this enemy submarine activity would become increasingly intense.**

He knew that CINCPAC was in the Hawaiian Area and that COMSOWESPAC was located in Brisbane. There is however, no indication that he was aware of the formation of SOPAC which was soon to undertake and direct an assault in the Guadalcanal-Tulagi Area.***

Finally, it is highly probable, in view of his visit to Truk (July 25th-26th) and his discussion there with the CINC Combined Fleet concerning the operation in the Outer South Seas Area,* that he was familiar with and generally concurred with the Imperial Headquarters opinion that from the extent of enemy preparations particularly in carriers, that any enemy counterattack would probably be no more than a reconnaissance landing.****


All Japanese aircraft employed in the operations in the Solomons, incident to the Battle of Savo Island, were naval aircraft of the 25th Air Flotilla. The Commander of this air flotilla, with headquarters at Rabaul, was directing activities of the FIFTH Air Attack Force, Base Air


* War Diary 8th Fleet, August 1942, WDC Document 161259.
** CRUDIV 6 Detailed Battle Report, Solomons Naval Action, August 7-10th, 1942, WDC Document 160997.
*** War Diary 5th Air Attack Force, August 1942, WDC Document 161730.
**** Southeast Area Naval Operations, Part I, Document 40427, page 10, Historical Division, U.S. Army, July 31st, 1947


Force.* The 25th Air Flotilla constituted that part of the Base Air Force which was located in the southeast area. Its composition on August 6th is shown on Table 1, and totaled 113 planes of various types. The units involved in this organization were the TAIMAN air Group, the SECOND Air Group, the YOKOHAMA Air Group, the FOURTEENTH Air Group, and the FOURTH Air Group.* Two special duty units (surface ships capable of serving as tenders for reconnaissance and fighter seaplanes) were attached to the 25th Air Flotilla. These were the seaplane carrier Akitsushima, Special Duty Unit ONE, and the aircraft transport Mogamigama Maru, Special Duty Unit TWO.* On August 7th and 8th, thee 25th Air Flotilla was expanded by the addition of planes of the MISAWA Air Group.**

The land-based planes of this flotilla, as yet unequipped with radar, were disposed between the Vunakanau and Lukanai Airdromes at Rabaul. The seaplanes were disposed at both Rabaul and Tulagi. The Tulagi planes were stationed there to protect the Guadalcanal Air Base wile it was under construction.**

Additional airfields throughout the Solomons were needed by the Japanese and for the purpose of building them Naval construction battalions were placed under the command of the Base Air Force.** Two airfields had been rushed to completion, the Guadalcanal Air Base which was finished on August 6th and an airfield at Buka Passage, which was ready to operate as a base on August 8th. Neither of these two bases had been assigned air groups at the time of the Battle of Savo Island. The Commander ELEVENTH Air Fleet and Commander 25th Air Flotilla had dissuaded Commander Outer South Seas Air Force from his desire to move up land attack units immediately upon completion of Guadalcanal Air Base for the reason that there was inadequate air strength, as yet, at Rabaul to permit such disposition.***

Until land bases were available, suitable harbors for basing patrol and fighter seaplanes were needed. The Japanese bad selected harbors for seaplane bases at Faisi, Shortland Island, Buka Passage, Giso Island, Kieta (Bougainville Island), and at Rekata Bay (Santa Isabel Island). No air activity was noted in these harbors by COMSOWESPAC air reconnaissance on August 5th, 6th, 7th, or 8th.

The employment of Navy land-based aircraft, other than Army aircraft, in the Southeast Area was in accordance with Japanese practice. It was necessary, in part, because Japanese Army air groups were not trained in the conduct of Joint Operations with naval air forces.****.


*Strength and disposition of 25th Air Flotilla on August 7th 1942, CIG Document 74629, May 12th 1947, and War Diary of 25th Air Flotilla, August 1st-31st, 1942, WDC document 161730.
**Records 25th Air Flotilla, Item 21D, Section 3, June 25th to August 6th, 1942, WDC Document 160155.
***Southeast Area Naval Operations, Part I, Document 40427, page 10, Historical Division , U.S. Army, July 31st, 1947.
****Japanese answers to 08586 Military Division Questionnaire #1.


Since the commencement of the war, naval aircraft had been highly successful in spear-heading Japanese naval attacks. The Japanese now expected their naval aircraft to be equally successfully in the Solomon operations*


The Japanese did not detect the Allied Expeditionary Force as it was advancing northward along the meridian of 159° E. Longitude toward Tulagi and Guadalcanal on August 5th and 6th. No searches through the Solomons sector were flown on those two days from Rabaul, reportedly because of bad weather.** COMCRUDIV SIX reported that weather also prevented the Japanese Reconnaissance seaplanes at Tulagi from locating the Allied Force.*** There is reason to believe this statement since the area the Allied Force traversed in its approach to the objective was not searched by the Allied B-17's either on August 5th and 6th.**** (The movement of the Allied Force is shown on diagram "B" and passed through SOPAC Search Sector I). However, weather did not prevent the Allied carriers in the Allied Support force from conducting routine local security patrols on these same two days.*****

The Japanese were caught by complete surprise on August 7th by the Allied Expeditionary Force, even though the High Command had directed their units in the South Pacific on August 5th to exercise strict caution.****** Although he had been advised by his air commanders that his air strength at Rabaul was inadequate,** he nevertheless gave considerably less attention to the Solomons sector than he did to the New Guinea sector. This resulted in inadequate searches in the Solomons sector prior to the Allied landing on August 7th.

The land attack planes at Rabaul were heavily engaged in almost daily bombing attacks on the Allied air bases at Port Moresby. Additionally, Commander FIFTH Attack Force was employing a few land attack planes for reconnaissance in connection with the New Guinea offensive operations******* and for reconnaissance flights to locate possible base sites


* War Diary, 6th Fleet, August, 1942, WDC Document 161259.
** Southeast Area Naval Operations, Part I, Document 40427, page 10, Historical Division, U.S. Army, July 31st 1947.
*** CRUDIV 6 Detailed Battle Report #6, Solomons Naval Action, August 7th-10th, 1942, WDC Document 180997.
**** War Diary COMAIRSOPAC (CTF) 63, August 1942.
*****War Diary Saratoga, August 1942.
****** War Diary 1, "A Brief History of World War II(B) No.1 December 1941 to March 1943", page 12, Supreme Commander of the Allied Powers, AFIS Document WW 16266.
******* Records 25th Air Flotilla, Group 21, Item 21D, Section 3, June 25th to August 6th, 1942, WDC Document 160155.


The Louisiades and D'Entrecastraux Islands.* He relied largely on the seaplanes of the YOKOHAMA Air Group, based at Rabaul and Tulagi, for reconnaissance of the Solomons Sector and of the approaches to Tulagi from the southeast.* His search planes at Tulagi, as was pointed out in the previous section of this analysis, were sent there to conduct reconnaissance for the protection of the Guadalcanal Air Base.* It is probable, therefore, that his reliance on the search planes operating out of Tulagi gave the Commander Outer South Seas Forces a confidence not warranted by the situation, and that this confidence contributed greatly to the surprise achieved by the Allies.

The routine Japanese air searches scheduled for August 7th were to be conducted by seaplane only. Those at Rabaul were to cover a sector between bearings 100° (T) to 130° (T) to a radius of 700 miles,** those at Tulagi were to cover an unidentified sector to the southward of the Solomons. The latter seaplanes were caught on the water at dawn by Allied carrier planes and were destroyed before they could take off.***

The contact report of the Allied Expeditionary Force received from Tulagi Air Base at 0652, August 7th, precipitated an immediate change in Japanese search plans. As will be shown later, Commander Outer South Seas Force requested the FIFTH Air Attack Force to augment the seaplane searches by the employment of land attack planes based at Rabaul. The Japanese thereby demonstrated a capability of increasing very considerably the number of aircraft employed for searching the Solomons. It is unfortunate for them that these augmented searches were not initiated on August 5th when the warning came from the High Command.


Just prior to the surprise Allied invasion of the Guadalcanal-Tulagi area the Japanese forces of the Southeast Area were disposed in accordance with the strategical situation as it appeared to the Japanese commanders at that particular time. The EIGHTH Fleet was operating from Rabaul as a principle naval base. Since Rabaul was at this time under the constant harassment and surveillance of Allied air power based in Australia and staging through Port Moresby, Commander EIGHTH Fleet was inclined to favor the more northerly harbor of the area, Kavieng, as an anchorage for his surface forces. As a consequence, on August 1st, he dispatched the heavy cruiser Chokai, the flagship of the EIGHTH Fleet, and Cruiser Division SIX, consisting of the Aoba, Kako, Kinugasa and Furutaka from Rabaul to Kavieng, there to remain until the morning of August 7th.****


* Records 25th Air Flotilla, Group 21, Item 21D, Section 3, June 25th to August 6th, 1942, WDC Document 160155.
**Japanese Search Plans, August 7th-9th, 1942, CIG Document 74632 May 12th, 1947.
***Strength and Disposition of 25th Air Flotilla on August 7th, 1942. CIG Document 74529 May 12th, 1947.
****War Diary CRUDIV 6, August 1942, WDEC Document 160997.




as of 2000 August 8th

Base Taiwan Air Group 2nd Air Group Yokohama Air Group 14th Air Group 4th Air Group Misawa Air Group Total
  Type Zero Shipboard Fighter (ZEKE 11) Type 2 Land Recce Plane Type Zero Shipboard Fighter (ZEKE 11) Type 99 Carrier Bomber Type 97 Flying Boat (MAVIS 11) Seaplane Fighter Type 2 Large Flying Boat Type 1 Land Attack Plane (BETTY 11) Type 1 Land Attack Plane (BETTY 11)  
RABAUL 24** 2 15** 16 6   2     65
              32 *** 32
TULAGI         7 9       16
TOTAL 24 2 15 16 13 9 2 32 0 113

Total number of Usable Planes Between 0000 August 7th and 2400 August 9th

  Type Zero Shipboard Fighter Type 2 Land Recce Plane Type 99 Carrier Bomber Type 97 Flying Boat Seaplane Fighter Type 2 Large Flying Boat Type 1 Land Attack Plane
0000 Aug. 7 39 2 7 13 9 2 32
0000 Aug. 8 37 2 7 6 0 2 36
0000 Aug. 9 34 2 7 6 0 2 27
0000 Aug. 10 34 1 7 6 0 2 24


*None of these aircraft were equipped with radar.
**Based at Rabaul and assigned to Lae and Buna, New Guinea, and Tsui Rumi, New Britain areas as fighting progressed.
***Nine land attack planes of this air group reported on August 7th and eight on August 8th, making a total of seventeen additional planes.

NOTE: The above tables are based on CIG Intelligence Report 74629 of 12 May 1947, "Strength and Disposition of Air Flot. 25 on 7 August 1942" and on WDC 160140 "Detailed Battle Report No. 8 of the Fifth Air Attack Force."


He himself remained at his headquarters in Rabaul.

(a) The disposition and composition of the principal units of the EIGHTH Fleet at 0652, August 7th, the time Commander EIGHTH Fleet received the report of the initial sighting of the Allied Amphibious force (TF 62) by the Japanese at Tulagi, is summarized as follows:

(1) Support Force and Chokai

The heavy cruisers Chokai, Aoba, and Kako were en route to Seeadler Harbor, Manus Island, Admiralty Islands. The heavy cruisers Kinugasa and Furutaka were en route to Rabaul. All five cruisers had just cleared the southern entrance to Steffen Strait on routine passage from Silver Sounds.*

(2) Escort Force

(a) The light cruisers Tenryu and Yubari and the destroyer Yunagi of DESDIV TWENTY-NINE [unreadable] at Rabaul.*

(b) The destroyers Oite, Yuzuki of DESDIV Twenty-Nine and the destroyers Mitsuki, Uzuki, Yatoi, and Mochisuki of DESDIV THIRTY were engaged in escort duties. DESDIV SEVENTEEN was ordered to report to commander Outer South Seas Force on August 5th but had not yet arrived at Rabaul. As these destroyers did not partake in the battle of Savo Island, they are dropped from further discussion.

(3) Auxiliaries of 25th Air Flotilla.

The seaplane carrier AKITSUSHIMA and the aircraft transport Momigama Maru were at Rabaul.**

(4) Submarines

(a) SUBRON SEVEN was located about as follows: RO-33 in Gulf of Papua, RO-34 off Australian East Coast, I-121 and 122 at Rabaul and I-123 at Truk.***

(b) SUBRON THREE, which reported from the SIXTH Fleet to the EIGHTH Fleet on August 7th, were on station as follows: I-11 off the


*CMDDIV 6 Detailed Battle Report #8, Solomons Naval Action, August 7th-10th, 1942, CIIG Document 86927, June 27th, 1947.
**Strength and disposition 25th Air Flotilla, August 7th,1942, CIG Document 74629, May 12th 1947.
***War diary 8th Fleet "Movements of Naval forces under this Command", August,1942, WDC Document 161259


Southeast coast of Australia; I-174, and I-175 off the east coast of Australia; I-169 off the New Hebrides, and the I-138 which had remained at Sasebo, Japan.* SUBRON THREE is dismissed from further discussion, since all boats were ordered on August 7th to return to Japan for rest and refitting. None of them were employed in the Japanese counterattack at Tulagi.


The tasks assigned the EIGHTH Fleet by the Japanese were in general:

(a) To support the Japanese program of consolidating the network of outposts protecting the Rabaul position, which often required invasion operations.

(b) To counter successfully the general anticipation of the Allied Offensive against the area.**

The first of these assigned tasks has significance only because the Japanese consolidation of Japanese outposts protecting Rabaul had precipitated the Allied amphibious operations against Tulagi and Guadalcanal. Since the Battle of Savo Island resulted directly from the second assigned tasks, it is that task with which this discussion is concerned.

Commander Outer South Seas Force knew that his objective was protective in nature. He also knew that the principal offensive weapon in his area was the land-based airpower of the FIFTH Air Attack Force (Base Air Force) at Rabaul. This weapon would be most useful to him during high visibility. He considered that its full capabilities could only be exploited against strong enemy sea forces if the latter were within striking distance; and concluded that he might best accomplish this by employing some of his surface forces as decoys to lure the enemy forces within range of such air power.***

He knew that his own offensive power lay primarily in his heavy cruiser strength--the Chokai and CRUDIV SIX--plus any light cruisers and destroyers which could be made immediately available from the convoy escort force. He felt that during high visibility this surface force would be a secondary weapon to the Base Air Force units at Rabaul and would be vulnerable to Allied carrier-based bombers. But, during night or


*Reports of Meritorious Action SUBRONS 3 AND 7, August 10th, 1942. CIG Document 74834 May 12th, 1947.
**Records CRUDIV 6, WDC Document 160997, August, 1942, War Diary 8th Fleet, WDC Document 161259, July-August,1942; Southeast Asia Naval Operations, Part I, Document 40427, page 3; and the Campaigns of the Pacific War, USSBS 1946, page 105.
*** War Diary 8th Fleet, WDC Document 161259.


low visibility, he felt that his cruiser force became his principal weapon. The smallness of his force gave him little concern, for he was confident of his ability to defeat Allied surface forces in night action.

In addition to his surface forces, Commander Outer South Seas Force had four submarines of SUBRON SEVEN, three of which were available in his area. These sub-surface units might provide some assistance in attacking enemy ships.

The above Japanese capabilities, when considered in comparison with the strong Allied carrier and amphibious support forces sent to Tulagi and Guadalcanal, were limited indeed. The confidence expressed by Commander Outer South Seas Force therefore seems to have been somewhat optimistic. The discrepancy between his appreciation of the assigned objective and his actual degree of involvements lay in his underestimation of the Allied strength. The Japanese commanders on various echelons estimated that the expected Allied attack on Tulagi and Guadalcanal would be no more than a reconnaissance landing.*


From the preceding discussion, it is apparent that the conception of Commander Outer South Seas Force as to the measures he would probably pursue, should Allied forces attempt to penetrate his area, were:

(a) During daylight, to employ his land-based aircraft as a principal weapon and to employ his surface forces as a secondary weapon possibly as decoys to lure the Allied forces into range of the above aircraft.

(b) During night or low visibility, to employ his surface forces as the principal weapon.

(c) As opportunity presented, to employ his submarines in support of both air and surface attacks.

This concept for the defense of the Outer South Seas Area would have been sound, if adequate air and surface forces as well as adequate bases, more particularly air bases, had been provided, especially in the Solomon. Commander South Seas Force might then have operated his surface ships under the protection of land-based air power. However, as has been pointed out, such adequacy of forces and bases were not available at this time, with the result that any surface ship attack launched by Commander Outer South Seas Force would necessarily be forced to operate in areas under dispute without adequate air cover. In addition, the shortage of planes and bases from which to operate them, combined with bad weather and some overconfidence, permitted an American landing at Guadalcanal-Tulagi to be achieved by surprise.


* Japanese Southeast Area Naval Operations, Part I, U.S. Army Historical Division, Document 40427, July 31st, 1947.


Chapter III
Allied Arrangements

(a) Allied Command Relations

The entire Pacific area had been designated as an area of U.S. strategic responsibility. This area had been divided into three large areas: the southwest Pacific, the southeast Pacific, and the Pacific Ocean; the latter being further subdivided into the North, Central, and south Pacific areas.

The boundaries of the south Pacific area and those of the south Pacific area [sic] are of importance in the study of the Battle of Savo Island as it was within these area that the operations in connection with this action were confuted. The pertinent portions of the northern and eastern boundaries of the southwest Pacific area were from Long. 130 degrees East along the Equator to 159.degrees East, thence South.* The South Pacific area was bounded on the west by the Southwest Pacific area and on the north by the Equator.

The Pacific Ocean Area was under the command of Commander-in-Chief (CINCPOA) who had directed, among other tasks, to--

(a) Protect essential sea and air communications.
(b) Prepare for execution of major amphibious offensives against positions held by Japanese initially to be launched from South Pacific and Southwest Pacific areas.*

CINCPOA was charged with the direct exercise of command of combined armed forces in the North and Central Pacific Area (COMSOPAC) who exercised command of combined forces which were at any time assigned to that area. The Commander South Pacific Area was also Commander South Pacific Force. (COMSOPACFOR).Under COMSOPACFOR were all of the naval forces of the Allied nations in the South Pacific Area.

The tasks listed above for CINCPOA were assigned, somewhat modified by that commander, to COMSOPAC for execution. These modifications more specifically delineated COMSOPAC'S responsibilities. The tasks were:

(a)Protect the essential sea and air communications.
(b) Prepare to launch amphibious offensives against positions held by Japan.**


*Joint Directive, U.S. Joint Chiefs of Staff, War Diary COMSOPACFOR July 1942 (COMINCS to CINCPAC 022100 July 1942).
**Organization of South Pacific Area, Lecture by Capt. T.H. Robbins, Jr., USN, Army-Navy Staff College, January 13th, 1944.


The Southwest Pacific area was under the command of a Supreme Commander who had been directed, among other tasks, to:

(a) Check enemy advance toward Australia and its essential lines of communication by destruction of enemy combatant troop and supply ships, aircraft, and bases in..........the New Guinea, Bismarck-Solomon Island region.
(b) Support the operations of friendly forces in the Pacific Ocean Areas and in the Indian Theater.
(c) Prepare to take offensive.*

The Supreme Commander Southwest Pacific Area (COMSOWESPAC) had as his naval commander a United States naval officer who was designated Commander Southwest Pacific Force (COMSOWESPACFOR) and who was vested with all powers customarily granted to the Commander-in-Chief of Fleets. Under this commander were all of the naval forces of the Allied Nations in the Southwest Pacific Area.** It is of interest that when task forces of the Southwest Pacific Area operated outside that area, coordination with forces in the new operating area was to be effected by the Joint Chiefs of Staff or by the combined Chiefs of Staff as appropriate.***

CINCPOA was Admiral Chester W. Nimitz, who on May 8th had assumed command of all land, sea, and air forces in the Pacific Ocean Area except the land defenses of New Zealand. His headquarters were at Pearl Harbor. COMSOPAC was Vice Admiral Robert L. Ghormley, who on June 19th had assumed command of the South Pacific Area and the South Pacific Force. His administrative headquarters were at Auckland, New Zealand; his operational headquarters were established on board the ARCONNE at Moumen on August 1. COMSOPAC was empowered to appoint commanders of task forces in his areas. COMSOWESPAC was General Douglas MacArthur, who had assumed command of that area on April 18th. His headquarters were at Brisbane, Australia.

Thus it is evident that unity of command for military operations in the South Pacific Area existed at this time. COMSOPAC was in command of all operations within this area. This was a marked improvement over the command organization which had been in existence at the Battle of the Coral Sea just three months earlier wherein COMSOWESPAC had not been in command of the naval forces operating within this area. Unity of command for this operation was affected by moving the boundary between the Southwest Pacific and the South Pacific areas from 165 degrees E. Longitude to 159 degrees Longitude-a change which pleased Tulagi, Guadalcanal and other islands


*SecNav Secret Letter (SC) A16-3(28) April 20th, 1942, encl. (A).
**SecNav Secret Letter (SC) A16-3(28) April 20th, 1942, page 1.
***Ibid., enclosure (A).


Plate 2: Allied Commmand Relations on 8 August 1942 chart - The Battle of Savo Island August 9, 1942 Strategical and Tactical Analysis




east of 159° E. within the South Pacific Area. Since these islands were the areas in which the Battle of Savo Island was fought, all operations within these areas were under COMSOPAC.

Any support rendered to COMSOPAC by COMSOWESPAC or vice versa was to be by cooperation.

The command of combined operations with Australian forces was as follows: If carrier units were involved, the senior American naval officer would be in command because of the nature of carrier operations; otherwise when the naval forces of the two powers were operating together and no carrier operations were involved, the senior officer of either power would be in command.**

Such were the command organizations for the Pacific, South Pacific and Southwest Pacific areas. However, in order to understand more clearly the various factors which culminated in the Battle of Savo Island it appears wise at this time to explain the command situation which obtained in the South Pacific Area with relation to its newly assigned mission - to seize and occupy Santa Cruz Islands, Tulagi, and adjacent positions in order to assist in seizing and occupying the New Britain-New Ireland Area.* This mission was designated as Task ONE.

The Joint Chiefs of Staff directed CINCPAC to designate the task force commander for this task. However, COMINCH gave CINCPAC little discretion in the matter, for he stated to CINCPAC, "It is assumed Ghormley will be made task force commander at least for Task ONE which he should command in person in the operating area." He stated in the same dispatch that the commander for Task ONE "should also have a conference with MacArthur."* As a consequence CINCPAC on July 9th designated COMSOPACFOR as the Task Force Commander for Task ONE and directed him to exercise strategic command in person in the operating area which was interpreted to be initially the New Caledonia-New Hebrides Area. CINCPAC at Pearl Harbor retained over-all command.***

The Joint Chiefs of Staff directed that direct command of the tactical operations of the Amphibious Forces was to remain with the naval task force commander; that COMSOWESPAC was to support the operations of Task ONE by providing for the interdiction of enemy air and naval activities westward of the operating area, which was evidently interpreted as westward of the South Pacific Area; and that COMSOWESPAC was to attach the necessary naval reinforcements and land-based air support.* CINCPAC advised COMSOPACFOR that as a consequence , COMSOWESPAC was making available


*Joint Directive, U.S. Joint Chiefs of Staff, War Diary COMSOPACFOR July 1942, (COMINCH to CINCPAC, 022100 July, 1942).
** War Diary CINCPAC, April 16th, 1942.
***Letter of instructions CINCPAC, Serial 0161W July 9th, 1942.


certain surface and air forces and authorized COMSOPACFOR to apply directly to COMSOWESPAC for any additional forces required.*

In accordance with his designation as Commander for Task ONE, COMSOPACFOR issued an Operation Plan wherein he formed two task forces to carry out his mission. One of these task forces he designated as TF 61, the Expeditionary Force; the other as TF 63, the Shore-Based Aircraft. TF 61 was composed of three lesser task forces; one, the Air Support Force, consisting of three Pacific Fleet carrier task forces, TFs 11,16 and 18; one, the naval forces transferred by COMSOWESPAC and termed TF 44; and one, the South Pacific Amphibious force, TF 62. TF 63 was composed of all land-based and tender-based aircraft attached to the South Pacific force, and of the aircraft temporarily attached and basing on islands in the South Pacific Area. He designated Vice Admiral Frank Jack Fletcher, USN, CTF 11, who was commander of the Pacific Fleet Striking Forces, a force consisting entirely of carrier task forces, as CTF 61 and Rear Admiral John Sidney McVain as CTF 63. He directed that coordination between these two forces was to be through cooperation, although he directed CTF 63 to supply aircraft support to CTF 61 on call.**

CTF 61 then issued an Operation Plan in which he assigned command of the Air Support Force (61.1) to CTF 18 as CTG 61.1 in the Wasp, and command of the Amphibious Force (61.2) to the Commander South Pacific Amphibious Force, CTF 62 in the McCawley. He retained command of TF 11 which for this operation became TG 61.1.1 and continued to fly his flag in the Saratoga. To 61.2 was composed of the following: TF 62; three heavy cruisers, one light cruiser and six destroyers from the South Pacific Command; and TF 44 less one heavy cruiser.*** TG 61.2 thereafter employed the designation TF 62 and organized his command accordingly.

This designation of CTF 11 as the Commander Expeditionary Force with the title CTF 61 was a necessary requirement of this operation. This was because CTF 61 was the only combat trained carrier task force commander within the command, and it was felt that it was more important to retain him within the carriers than to give him freedom of action to go where he desired. This kept him far away from the scene of action.

In this connection experience has shown that it is generally wiser for the Supreme Tactical Commander to place himself within the amphibious force during landing operations and within the covering force if action with enemy surface forces is imminent. This enables him to keep himself continuously informed of the constantly changing situation and permits him to employ his communications freely once contact with the enemy has been made. This was the practice in later operations such as OKINAWA where the fleet commander was in one of the older cruisers. He was thus


*Letter of Instructions CINCPAC, Serial 0151W, July 9th, 1942.
**COMSOPACFOR Opera t ion Plan 1-42 Serial 0017, July 10th, 1942.
***CTF 61 Operation Order 1-42, July 28th, 1942.


able to move freely to the position of paramount interest without appreciably weakening the strength of the force to which he was nominally assigned*.

It will be noted from the above that CTF 61 was also CTG 61.1.1 as well as Commander Striking Forces Pacific Fleet and CTF 11; that CTF 62 was also CTG 61.2 and that many of the other commanders bore several designations, most of which had no connection with the assigned operation. Such a multiplicity of diverse titles could not have but been confusing to the commanders concerned, as well as to the subordinate commanders throughout whatever echelon.

(b) Information Available to Allied Commander

COMSOPAC realized from indications that the Japanese wee planning to extend, to the South and Southeast, the control then held on most of New Guinea-New Britain-Northern Solomons Area, and were therefore consolidating and improving their positions there.**

He knew the location of most, if not all, of the Japanese airfields and seaplane bases in the theater of operations and had daily reconnaissance information of their employment. He knew that the Japanese were developing three airfields, one at Lunga Point, Guadalcanal Island; one at Kieta, Bougainville Island, and one at Buka Passage, Buka Island. He further knew that the Japanese had seaplanes a Rabaul; Rekata bay, Santa Isabel Island; Faisi, Shortland Islands; Kieta, Bougainville Island; Gizo Island; Tulagi harbor, Florida Island; and Buka Passage.**

He believed that all three airfields were in satisfactory operating condition and that Kieta, in particular, was receiving personnel and equipment.** Actually the Buka airfield was in an advance state of construction, being completed on August 8th; the construction of the Kieta airfield had been stopped, presumably in July, and the airfield at Lunga Point was well advanced, being completed on August 6th.

He believed that there were four heavy cruisers of Cruiser Division 6 and three light cruisers of Cruiser Division 18 in the Rabaul-Kavieng Area.*** This was approximately correct. However, there but two light cruisers there--the additional cruiser was a large heavy cruiser, the Chokai.


* Statement by Admiral Raymond A. Spruance, USN, CINCPAC's Chief of Staff at the time to Commodore Richard W. Bates, USN (Ret.), Head of Dept. of Analysis, Naval War College on June 1st 1949.
** COMSOPAC Serial 0017 July 16th, 1942 Appendix III and CTF 62 Operation Plan A3-42-Ser.0010 Annex EASY, July 30th, 1942, Part (1).
*** CINCPAC Serial 0151W to COMSOPAC, July 9th, 1942, Reference "C" Information on Enemy Forces and Positions in SOWESPAC Area up to July 10th, 1942.


He believed that there were thirteen destroyers of DESRONs SIX and THIRTY-FOUR in the area.* this was partially correct in that there were eight rather than thirteen destroyers. All eight destroyers were from DESRON SIX and all appeared to have been assigned to escort duty.

He believed that there were about fifteen submarines in the Bismarck-New Guinea-Solomons Area.* This was markedly incorrect as there were but ten submarines in the entire Outer South Seas Area at the time of landing. Of these, but two, from SUBRON SEVEN, appear to have been in the Bismarck-New Guinea-Solomon Area.

He knew quite correctly that there were two Av's and three or four XAV's in the Rabaul area.

He knew that the Japanese air strength in the Bismarck-New Guinea-Solomons Area was about sixty VF, sixty VB and thirty VP planes--a total of 150 on July 30th. This estimate, at least for August 6th, was about 30% too high, due perhaps in part to losses suffered from Allied air strikes on Rabaul and on operational losses. Actually there were present on August 6th about forty-eight VF, forty-eight VB and fifteen VP. He believed that about eight of the above fighters, equipped with pontoons as float planes, and about seven to ten of the patrol planes were operating in the Tulagi area.** His estimate of the planes in the Tulagi area was reasonably correct.

He realized that the Japanese capability of striking with land-based air power at Allied Forces in the Guadalcanal Area was real and that attacks could be expected.

He knew that the strength of the striking forces which the Japanese could bring to bear against his operations in the Guadalcanal-Tulagi Area had been greatly decreased by the Japanese losses during the Battle of the Coral Sea and the Battle of Midway.*** However, he realized that there could well be still strong Japanese forces in or near the area.

He was aware of the air strength in the Marshall-Gilbert Islands Area and in the Truk-Ponape Island area, and the Japanese capability resulting therefore, of quick reinforcement of the Bismarck-New Guinea-Solomons Area.**** It eventuated that the reinforcement was not accomplished until after the Battle of Savo Island.


*CTF 62 Op. Plan A3-42, Serial OC10, Annex BASY, July 30th, Part (1), para. (c).
**Ibid., para. (d).
***CINCPAC Serial 0151W to COMSOPAC, July 9th, 1942, para. 2.
****CINCPAC Serial 0151W to COMSOPAC, July 9th, 1942, Reference "C". Information on enemy Forces and Positions in SOWESPAC Area up to July 10th, 1942.


He realized that despite the Midway and Coral Sea losses, a definite Japanese capability existed of supporting Japanese land-based air with carrier-based air operating from carrier task groups. His air searches were therefore designed to discover such carrier forces before they could reach effective striking positions.* This latter concept seems to have been the greatest motivating consideration in all of the planning in the conduct of searches in both SOPAC and SOWESPAC and in the deployment of forces. Actually there were no Japanese carrier task groups within the area and none appeared until the Allied beachheads had been reasonably secured.

Finally, he knew that the existing Japanese reconnaissance operated in depth. He therefore felt that surprise was improbable.** Actually, although Japanese searches and reconnaissance did in fact operate in depth, there appear to have been no Japanese searches or reconnaissance on the day of the Allied approach, so that the Allies did, in fact achieve surprise.

(c) Allied Land and Tender Based Aircraft

(1) South Pacific (SOPAC)

The land and tender-based aircraft in the South Pacific Area has been organized as has been pointed out earlier into a command under Rear Admiral John Sidney McCain, U.S.N. (COMAIRSOPAC). The units of this force were employed chiefly for the protection of SOPAC bases and the essential sea and air communications. AIRSOPAC included Army, Navy, Marine Corps, and Royal New Zealand Air Force units.

The administration of these units was divided. As regards training in particular, COMAIRSOPAC was charged with the training of Navy and Marine Corps aviation, whereas the Commanding General of U.S. Army Forces in the South Pacific Area (COMGENSOPAC)--a command assumed by General Millard F. Harmon, U.S. Army Air Corps on July 26, 1942--was charged with the training of all Army Air Corps units. The Royal New Zealand Air Force units present in SOPAC Area were under the unity of command structure of the particular base at which they were located, and this placed them under a U.S. Army officer in every case.

The tactical operations of all SOPAC aircraft for the support of the Allied amphibious offensive which led to the Battle of Savo Island were placed under the command of COMAIRSOPAC whose tactical title was Commander Task Force 63 (CTF 63). CTF 63 availed himself of the advice of COMGENSOPAC and, owing to the wide dispersion and dissimilar composition


* CTF 61 Dispatch 290857, July, to CTF 63.
** COMSOPAC-COMSOWESPAC Joint Dispatch 081017, July 1942, which is Part Four of Dispatch 081012, July 1942.


of the air organization, allowed the latter a certain amount of operational discretion.

Table 2 shows a tabulation (by numbers and locations) of the planes available to CTF 63 on August 6th--the day prior to the initial American landings in the Solomons, and the day that the Allied Expeditionary Force approached within range of the Japanese land based aircraft. Of all the planes listed in Table 1, the only ones suitable for search and attack missions over and beyond the objective were the Army B-17s and the Navy PBY-5s. The limiting ranges of the other aircraft restricted their roles to defense of bases, air coverage for surface units within range of those bases, and routine anti-submarine patrols. For this reason theses aircraft did not participate in operations directly involved in the Battle of Savo Island, and are dropped from further discussion.

In the Solomons offensive, the B-17 was the better search plane over Japanese held positions in the islands where enemy fighters might be encountered. The B-17 had a speed advantage over the PBY-5--30 knots faster cruising, and 45 knots faster maximum speed. The B-17 had better combat qualities as a result of its self-sealing tanks and its 8 flexible 50-calibre gun mounts. As a bomber, the B-17 was effective only to about 600 miles radius. This short range made it necessary to operate it from the furthest advanced airfield, and placed an urgency upon CTF 63 to complete the airfield at Espiritu Santo by August 1st.

The PBY-5 was better suited to cover the wide ocean areas through which Japanese surface units could approach the target area, and in which Japanese fighters were not likely to be encountered. The advantage of the PBY-5 over the B-17 in these areas was due to its greater range - in the ratio of 3600 miles to 2000 miles for economical cruising , and in the ratio of 800 miles to 600 miles radius as a bomber.*

Both the PBY-5 and the B-17 were equipped with radars of the ASE model; but little advantage was gained from these radars other than navigational assistance. These radars provided a search beam of about 15 degrees spread normal to the line of flight on both sides of the plane, designed to be effective to about 25 miles range. A homing beam was also provided, ahead of the plane to an effective range of about 15 miles. Since the pilots did not consider the lateral search beams reliable but placed their faith in the homing beam ahead, these early radars provided little more than the equivalent of the reliable limits of human vision in clear weather. They did, however, make it possible to conduct night searches in clear weather as effectively as day visual searches. Since rain squalls showed up on the radar scopes much like "sea return", these radars were about 50% effective in bad weather, day or night.** Their


*Airplane characteristics, Naval War College, June 1942.
**Tactical use of Radar in Aircraft, Radar Bulletin No. 2, COMINCH July 29th, 1942.


Table 2
Disposition of Allied Shore and Tender-Based Aircraft
As of 2400, August 6, 1942

Base 11th
VP Sqdns
11, 14 & 23
Marine Corps
VMO 251
  B-17 B-26 P-39 PBY-5's Hud-
Espiritu 5*     10**       3***   3 21
Efate 5****     4       13   3 25
New Caledonia 10 10 38 2 6     16   3 85
Nandi 12 12 17 6 12 3 9       71
Ndeni       6*****             6
Samoa               18 17 10 45
Tongatabu     24             6 30
Total 32 22 79 28 18 3 9 50 17 25 283


* 1 of 6 B-17s of 98th Squadron lost east of Espiritu Santo due navigational error.
** 1 of 11 PBYs failed to return from search Sector III.
*** 1 of 4 F4F-3A of VMF-212 crashed at sea August 6th. Replaced from Efate.
**** 2 of 7 B-17s of 26th Squadron lost previously, one August 4th, one August 5th.
***** 1 of 7 PBYs lost at sea August 6th.




Chief value to pilots was derived from their ability to pick up land about sixty miles at sea, depending on the altitude of the terrain, and assisted greatly in fixing their navigational positions

The total of thirty-two B-17s in the South Pacific Area on August 6th were assigned to the four squadrons (the Twenty-Sixth, Forty-Second, Ninety-Eighth and Four Hundred and Thirty-First Bombardment Squadrons) of the ELEVENTH Bombardment Group.* The ELEVENTH Bombardment Group had been formed in Hawaii in mid-July and designated as the Mobile Air Force, Pacific Ocean Area. As such, its disposition anywhere within the Pacific Ocean Area rested with the discretion of the Joint Chiefs of Staff, who determined upon its employment in the South Pacific Area in support of the Allied offensive in the Solomons.**

The disposition of these aircraft on August 6th (Table 1) was the deployment ordered for Task ONE in the Solomons by CTF 63 Operation Order 1-42 as follows: five B-17s of the Twenty-Sixth Bombardment Squadron at Espiritu Santo; ten B17's of the Ninety-Eighth Bombardment Squadron at Efate and five B-17s of the Forty-Second Bombardment Squadron on New Caledonia at Koumac and Plaines Des Gaiace airfields; and a reserve of twelve B-17s of the Four Hundred and Thirty-First Squadron at Nandi, Fijis. This disposition was not hard and fast, and heterogeneous groupings of planes resulted from the rotation forward of reserve aircraft replacements for damaged and lost B-17s, and the flexibility with which airfields could be employed in case of bad weather emergencies.

The twenty-eight PBY-5s in the South Pacific were disposed on August 6th (Table 1) as follows: six planes shore based at Nandi, Fijis; four planes shore-based at Havannah Harbor Efate; two planes shore-based at Ile Nou, Noumea; ten planes based on the tender Curtiss in Segond Channel, Espiritu Santo; and six planes based on the McFarland at Graciosa Bay, Ndeni. The aircraft tender, Mackinac was en route on August 6th from Noumea to Maramasiko Estuary to establish an advance seaplane base to which nine PBY-5s would be moved up on August 7th.

(2) Southwest Pacific (SOWESPAC)

The land-based aircraft in the southwest Pacific Area included units of both the U.S. Army Air Corps and the Royal Australian Air Force. This combined organization was called Allied Air Forces, Southwest Pacific. Its headquarters were in Brisbane, Australia, in the same building as the headquarters of the Supreme Commander, Southwest Pacific. Major General George Kenny, Air Corps, U.S. Army assumed command of the Allied Air Forces on August 4th--only three days before the Allied amphibious landings in the Solomons.***


* COMAIRSOPAC War Diary, July 1942 also U.S. ARMY IN WORLD WAR II, Guadalcanal: The First Offensive.
** U.S. ARMY IN WORLD WAR II, Guadalcanal: The First Offensive, page 27.
*** Army Air Forces in the War Against Japan 1941-1942, page 134.


This air command was divided into four sub-area commands, but only those units of the Allied Air Forces, North Eastern Area with headquarters at Townsville, Australia,* were involved in operations in New Guinea and in the Bismarck designed to checks the advance of the Japanese.

The 19th Bombardment Group, based at Townsville had been designated as a Mobile Air Force in July--a Southwest Pacific counterpart to the 11th Bombardment Group the Pacific Ocean Area. These two Mobile Air Forces in the Pacific Theater were disposed on August 6th against the Japanese advance south of the Equator. The 19th Group, with approximately the same strength as the 11th Group in SOPAC but with rarely more than twenty B-17ís in commission,** was the chief offensive weapon against the Japanese base at Rabaul. His attack groups suffered an attrition rate over Rabaul of about 20% per month.***

In order to reach Rabaul with B-17s from Townsville, Commander North Eastern Area had to stage through the advanced air base at Port Moresby. His bombers usually avoided Japanese air attacks on Port Moresby by arriving after dark, and preparing for the next day's mission during the night. They took off in the early morning for Rabaul, and followed a route along the New Guinea coast for 40 or 50 miles to gain sufficient altitude to cross the Owen Stanley Mountain range at approximately 7000 feet. They discovered in these operations that the equatorial weather proved to be as dangerous as enemy fighters.****

Commander North Eastern Area had pushed the construction of additional fields in the vicinity of Port Moresby in order to obtain dispersal of his bombers and to base defending fighters in the vicinity to ward off Japanese air attacks. By August 6th, he was able to base his B-17s at Port Moresby, under the protective cover of about forty P-39s of the 35th Pursuit Group.

He had also pushed the construction of the Fall River field at Milne Bay, New Guinea. This field would not take heavy bombers but was useful for reconnaissance planes. On August 6th, he moved a detachment of five Hudsons from the 32nd General Reconnaissance Squadron of the Royal Australian Air force to the Fall River field from Port Moresby. This detachment was to reconnoiter the Northwestern Solomons thereafter.***** He provided for the air defense of Fall River by basing there the 75th and 76th R.A.A.F. Fighter Squadrons equipped with P-40 aircraft.


* Headquarters, Allied Air Forces--Operations-Instructions No. 18, July 31st, 1942.
** Army Air Forces in the War Against Japan 1941-1942, page 124.
*** THE ARMY AIR FORCES IN THE WORLD WAR II, Plans and Early Operations, Note 27, page 723.
**** Ibid., page 480.
***** Letter from Major-General Harry J. Maloney, U.S.A., Chief, Historical Division to President Naval War College, October 11th, 1948.


Other air groups in the North Eastern Area were: the 3rd Bombardment Group (17 B-25ís) and the 22nd Bombardment Group (about 54 B-26s total, but only 26 planes available). Since these groups were employed on shorter range attack missions incident to the SOWESPAC operations and did not contribute directly to the support of the SOPAC offensive operations, they will receive no further attention.*

(d) Allied Search and Reconnaissance

(1) South Pacific (SOPAC)

CTF 63 planned the Allied searches from South Pacific bases to detect any enemy interference in that portion of the Coral Sea which lay east of Longitude 158 degrees East. He designed his air searches to cover both the Allied operations within the Tulagi-Guadalcanal Area and the Japanese approaches thereto from the northward as far as the range of his search aircraft would permit. A primary objective of these air searches was the detection of any Japanese carrier striking group which might enter the South Pacific from the direction of Truk or the Marshall Islands.**

In order to extend the range of his searches to the north as the Allied Expeditionary Force moved forward to its objective, he deployed his seaplane tenders progressively forward to establish more advanced bases for his seaplanes. This deployment consisted of moving the Curtiss with two patrol planes of VP-23 from Noumea to Segond Channel, Espiritu Santo, where search operations were commenced on August 5th; then of advancing the McFarland with seven patrol planes (five PBY-5s of VP-11 and two of VP-14) from Noumea to Graciosa Bay, Ndeni, where search operations were commenced on August 6th; and finally, after the landings at Tulagi and Guadalcanal, of advancing the Mackinac from Noumea to Marmasike Estuary, Halaita Island. The MACKINAC provided CTF 63 with a seaplane base in the southern Solomons as far north as Tulagi. Search operations were commenced from Marmasike Estuary on August 6th, employing the nine remaining patrol planes of VP-23 (one of the original ten had been lost on August 6th) which had been operating from the Curtiss in Espiritu Santo. The Curtiss, in turn, received two patrol planes advanced from Havannah Harbor, Efate, and placed in commission one plane of VP-11 which had been undergoing repairs in the Curtiss, and thereafter supported three search planes.

CTF 63 had deployed his seaplane tenders in such manner that the rear-most one, the Curtiss, served as a centrally located command center for Task Force 63. It provided headquarters for CTF 63, COMGENSOPAC, the 11th Bombardment Group (CTG 63.2), and the Curtiss search detachment (CTG 63.3). It also served as a communications ship for the air operations of the field at Espiritu Santo.


* THE ARMY AIR FORCES IN WORLD WAR II, Plans and Early Operations, pages 478, 479, 480.
** CTF 61 Dispatch 290857, July 1942.


CTF 63's land-based searches were flown by seaplanes based ashore at Nandi, Fiji, and by B-17's located at the airfields of Espirutu Santo, Efate, and Koumac (New Caledonia). The B-17s were employed in Sectors I and II. As can be seen from Diagrams "B","C", and "D" these sectors existed over the southern Solomons where Japanese fighter aircraft were to be expected. CTF 63's employment of the B-17s over the islands and the PBY-6s over the open ocean exploited the advantages of each type, as analyzed in the previous section.

On August 6th, the day before the Allied Expeditionary Forces reached the Tulagi-Guadalcanal target area, CTF 63 conducted the searches shown on Diagram "B". At sunrise, planes took off to search to the following ranges: Sector II, 750 miles from Espiritu Santo; Sector IV, 630 miles from Efate; Sector VI, 700 miles from Nandi. Sector I, the range of which was to be 800 miles from Koumac, was not searched, probably because of bad weather. All other sectors were reported searched with negative results.

At about noon, August 6th, additional patrol planes took off to search Sector III to 700 miles range from Espiritu Santo and Sector V to a range of 650 miles from Ndeni. CTF 63 was complying with the request of CTF 61 that the planes searching Sectors III and V on August 6th arrive at the outer limit of search at sunset, and search the return leg by radar.** These searches were so timed as to prevent the possibility of Japanese striking groups approaching undetected from Truk to the outer limits of Sector V, or from the Marshalls to the outer limit of Sector III by sunset August 6th, from which positions they could advance during darkness to launch an air attack on the Allied Amphibious Force just as it reached its objective at sunrise on August 7th.

The searches in Sectors III and V were reported searched with negative results with the exception that no report was received from the plane searching the western-most subsector of Sector III, which plane failed to return.***

The basic problem of the Solomons offensive as visualized and enunciated by COMSOPAC was: "the protection of surface ships against land-based aircraft during the approach, the landing attack, and the unloading" at the target area.**** He had assigned to CTF 63 the task: "cover the approach to, and operations within the Tulagi-Guadalcanal Area by search. Execute air attacks on enemy objectives as arranged with Commander Expeditionary Force (CTF 61). Render aircraft support on call". **** Consequently, CTF 63's searches quite properly were made to support CTF 61, and to permit the concurrent employment of South Pacific aircraft for


* War Diary CTF 63 August 6th, 1942
** CTF 61 Dispatch 2909857, July 1942
*** War Diary CTF 63, August 6th, 1942.
**** COMSOPACFOR Dispatch 112000, July 1942 addressed to COMINCH, CINCPAC, AND COMSOWESPAC.
***** COMSOPAC Operation Plan No. 1-42, July 16th, 1942.


attack missions against the Japanese. With this double support role, CTF 63 conducted but one search of his sectors daily.

CTF 63's searches were not fully effective in providing the protection to the Expeditionary Force which COMSOPACFOR considered so paramount, and the cooperation of SOWESPAC land-based air units was necessary. For this reason the Joint Chiefs of Staff had directed that COMSOWESPAC would interdict enemy air and naval activities westward of the operating area, and CINCPAC had authorized COMSOPACFOR to discuss such operations* directly with COMSOWESPAC. As a consequence, COMSOPACFOR directed CTF 63 to make the arrangements with COMSOWESPAC relative to the coordination of aircraft scouting.**

(2) Southwest Pacific (SOWESPAC)

SOWESPAC search and reconnaissance missions conducted in the North Eastern Area prior to August 5th were all flown from Port Moresby. The searches covered the approaches to Milne Bay and to Buna. The reconnaissance missions took in the Japanese installations at Lae, New Guinea, and those around New Britain and New Ireland. These search and reconnaissance flights were chiefly in support of the operations of SOWESPAC in flights were chiefly in support of the operations of SOWESPAC forces in New Guinea.*** Commander Allied Air Forces, North Eastern Area had conducted but one reconnaissance flight in the Solomons during this prior period -- he reconnoitered the Tulagi-Guadalcanal Area on August 1st,*** the day the Allied Expeditionary Force departed from the Fiji Islands.

It was on August 5th that SOWESPAC air units began their operations to support the South Pacific offensive. COMSOWESPAC had previously arranged with COMSOPACFOR that his air forces in the North Eastern Area would coordinate their operations to assist the offensive in the South Pacific by searches commencing on August 5th (D-2 Day) and extending through August 11th (D+4 Day). SOWESPAC searches would extend to the limits of range of the aircraft employed. They would cover the water areas which lay to the southeast of a limiting line drawn from Madang (New Guinea) to the Kapingamarangi Islands and northwest of another limiting line extending from Tagula Island to the easternmost point of New Georgia. The eastern limit of his searches was the 158th meridian of East Longitude, extending northward from New Georgia. Commencing on August 5th, SOWESPAC aircraft were prohibited from operating east of 158°-15'(E) between the Equator and Latitude 150°(S) missions were requested by COMSOPACFOR.****


* CINCPAC ltr. of Instructions to COMSOPAC, July 9th, 1942, Serial 0151W, page 3.
** COMSOPAC Operation Plan No. 1-42, July 16th 1942.
*** Allied Air Forces, Southwest Pacific, Reconnaissance Reports for August 1st-5th, 1942 inclusive.
**** COMSOWESPAC Dispatch 191034, July 1942.


By referring to Diagram "B", it can be seen that the area of SOWESPAC air searches covered the Bismarck and Solomon Seas and the Pacific Ocean approaches to these waters. COMSOWESPAC gave instructions that the entrances to the Coral Sea, from the north and from the east, were to be given particular attention by his search planes.

The eastern limit of SOWESPAC searches was determined by mutual agreement between COMSOPACFOR and COMSOWESPAC. CTF 63 had suggested to COMSOPACFOR that SOWESPAC aircraft be requested to search west of 1580 East Longitude.* It is presumed that CTF 63 preferred to extend his own searches as far as possible to support the operations of the Allied Expeditionary Force, since he had been charged by COMSOPACFOR with the responsibility of providing (and arranging with SOWESPAC) for adequate coverage. Certainly he could be more assured of receiving contact reports from planes under his own control then from SOWESPAC planes which were merely cooperating. CTF 63 informed COMSOPAC that his own searches were disposed to isolate the Coral Sea east of Longitude 158°(E),* and that for increased effectiveness his searches overlapped that meridian to the westward by an average of 120 miles.* COMSOPACFOR informed COMSOWESPAC of CTF 63's plans for air searches and suggested the eastern limit of 158° East for SOWESPAC searches.** COMSOWESPAC concurred in this suggestion.*** The reason for his acquiescence to the penetration of the Southwest Pacific Area (the eastern boundary of which was 159° East Longitude) seems to have been revealed in his dispatch to COMSOPAC a week earlier, wherein he had pointed out that: "all available aviation in this area is subject to actual limitations of range....."****

On August 5th, five Hudsons of the R.A.A.F. 32nd General Reconnaissance Squadron, based at Port Moresby, reconnoitered Buka, Kieta and Bougainville Strait, and then returned to base thereafter on the recently constructed field at Fall River in Milne Bay.***** On succeeding days, the Hudsons searched an area extending through the northern Solomons as far south as the easternmost tip of New Georgia Island, which area was referred to as Reconnaissance Area B.

Commander Allied Air Forces, North Eastern Area, also initiated Reconnaissance Areas C, D, and E on August 5th. Each of these areas were searched on August 5th and subsequent days by one B-17 (or one LB-30, an armed air transport version of the B-24) operating from Port Moresby.*** The geographic boundaries of these reconnaissance areas are not known definitely, but by plotting the time and position of each contact reported, it has been possible to approximate the tracks of the searching planes on Diagrams "B", "C", "D", and "E". Reconnaissance Area C appear to have


*COMSOPAC (CTF 63) Dispatch 220737, July 1942 addressed to COSOWPAC
** Dispatch 230250, July 1942, addressed to COMSOPAC
*** COMSOPAC Dispatch 260955, July 1942, addressed to COMSOPAC.
****COMSOWESPAC Dispatch 191034 July 1942 addressed to COMSOPAC.
***** Allied Air Forces, Southwest Pacific, Reconnaissance Report for August 6, 1942


extended from Vitiaz Strait along the south coast of New Britain past St. George's Channel to the Feni Islands, thence on the return it included Green Island and the Solomon Sea. Reconnaissance Area D conformed approximately to the line between Madang and the Kapingamarangi Islands and out across the western portion of the Bismarck Sea in a north-south direction. Reconnaissance E was a photographic coverage of the ports of Rabaul and Kavieng and a search of the sea areas enroute.

COMWESPAC'S search and reconnaissance operations were superimposed upon on his offensive air attack missions in support of the SOPAC force. His air offensive was directed, commencing on D- Day, primarily at interdicting Japanese air operations in the Rabaul Area, denying refueling of Japanese planes at Buka enroute to Tulagi, and smothering the Japanese air power based at Lae and Salamaua by periodic attacks in order to prevent it from reinforcing the Japanese air strength at Rabaul.

COMSOWESPAC had directed his aircraft to be prepared to strike hostile naval targets discovered in the North Eastern Area.* This directive "to be prepared to strike" (rather than "to strike") naval targets as well as naval targets is of significance in this study, since it so happened later that the Japanese Cruiser Force (which attacked the Allied Forces at Savo Island) passed with impunity southward through the Solomons on August 8th and northward through the Solomons on August 9th, within a 600 mile radius of Port Moresby during twelve hours of daylight on each day.

COMSOWESPAC'S directive is here analyzed since he had been directed by the Joint Chiefs of Staff to interdict hostile naval targets as well as hostile air operations as a part of his supporting role for Task ONE. On August 10th, the day after the Battle of Savo Island, the Army Chief of Staff queried COMSOWESPAC as to "the degree of success you are locating and attacking Japanese surface craft."** COMSOWESPOAC replied: "Had arranged with Ghormley for missions if called, but have had no requests".*** It is revealed in the light of this statement that the basis for COMSOWESPAC'S directive merely "to be prepared to strike hostile naval targets was the agreement reached between himself and COMSOPACFOR.

The arrangements made between these two commanders were: (a) that Southwest Pacific air units would concentrate primarily upon interdicting Japanese air operations against the Allied forces in the Tulagi area, and (b) that SOWESPAC aircraft would operate against hostile naval targets only if COMSOPACFOR made specific requests for such attacks direct to COMSOWESPAC Area.****


* COMSOWESPAC Operation Instructions No. 14, July 26th, 1942, and Annexure (A) to Allied Air Forces Operation Instruction No. 18, July 31st 1942.
** Radio No. 658, CM-OUT 2795, C of St. to CINC, SWPA, August 9th 1942 (August 9th 1942 (August 10th (-) 11 Zone Time).
*** Radio No. C-246, CM-IN 3795 CINC, SWPA to C of St., August 11th, 1942.
**** COMSOWESPAC Dispatch 191034, July 1842, addressed to COMSOPACFOR.


The reasons why COMSOPACFOR, the designated commander whom the Joint Chiefs of Staff had charged with the responsibility for the execution of Task ONE, would enter into such an agreement are nowhere set down. It is known however, that he was convinced that COMSOWESPAC did not have adequate air strength, for he stated to COMINCH: "I consider means now prospectively available SOPAC sufficient for accomplishment Task ONE provided SOWESPAC Area be furnished sufficient means for interdiction hostile aircraft activities based on New Britain-New Guinea-Northern Solomons Area." In the same message he quoted a previous dispatch to the effect that: "The air forces in sight for the Southwest Pacific Area are not adequate to interdict hostile air or naval operations against the Tulagi area."*

However, despite this inadequacy, the above agreement is not believed to have been sound in all particulars. COMSOPACFOR was in the position of having to rely upon the cooperation and support of SOWESPAC air units to locate and interdict Japanese surface forces in the approaches to Tulagi from Rabaul, since this area lay almost wholly within the SOWESPAC Area. Should not SOWESPAC aircraft have attacked automatically a strong surface ship formation, such as the Japanese Cruiser Force, heading in the direction of Tulagi? The Allied Air Forces North Eastern Area, significantly enough, had launched promptly and automatically a flight of B-17's from Port Moresby on August 2nd to attack an aircraft carrier falsely reported to be in Rabaul Harbor.** Is it not logical therefore to consider that COMSOPACFOR should have insisted that COMSOWESPAC direct his air forces to interdict large and powerful naval forces located in his area which were obviously making advances southward through the Solomons?

It is clearly evident that the planning and the operation of the Allied forces were concerned chiefly with the enemy capability of attacking by air, either from land bases, sea plane tenders, or from aircraft carriers. COMSOPACFOR had expressed this concern to COMINCH when he stated: "I wish to emphasize that the basic problem of this operation is the protection of surface ships against land-based aircrafts".* Had COMSOPACFOR considered more fully the enemy capability of attacking with surface ships which could approach through the SOWESPAC Area under low visibility conditions, he might have exercised his own responsibility, rather than to have depended upon the cooperation of SOWESPAC forces, and taken measures to insure that the threat of an enemy surface ship attack was met before it reached Savo Island. Perhaps then he would have been alerted to the need for late afternoon searches in addition to the early morning searches in Sectors II and IV.


*COMSOPAC Dispatch 112000, July 1942, to COMINCH.
**Allied Air Forces, Southwest Pacific, Operations Report for August 2nd, 1942.


Plate 3: Battle of Savo Island, Diagram of Communications Between SOPAC and SOWESPAC - The Battle of Savo Island August 9, 1942 Strategical and Tactical Analysis




Perhaps also he would have provided for the air coverage of the approaches from Rabaul and thus have precluded the necessity for CTF 62 later to request a special search of this area.*

(e) Communication Arrangements Between COMSOPAC and COMSOWESPAC

The Communication Plan 1-42 (Annex "D" to CTF 63's Operation Plan 1-42) provided, among other circuits, a communications Net "E" established between those air bases from which both long range air searches and heavy bomber strikes could originate. This included Espiritu Santo, Efate, and Noumea in SOPAC and Townsville and Port Moresby in the North Eastern Area of SOWESPAC. This circuit was a rapid and positive means for the wide dissemination of operational information and intelligence, and was in effect an Air Operational Intelligence Circuit (AOIC) as now employed in the naval communication service.

This Communication Plan Provided an additional radio circuit known as Net "C" between the air bases ashore, the tenders, the task group commanders and all reconnaissance and bombardment aircraft in the SOPAC Area. It did not provide an arrangement whereby SOPAC air bases, or task group or force commanders would receive messages or contact reports from SOWESPAC aircraft in flight. This circuit was in effect a task force common, although it was not so designated.

COMSOWESPAC Area's air communication plan was promulgated by the Allied Air Forces in Signal Instructions, Annexure "B" to Operation Instruction No. 18. These instructions provided for the establishment of special point-to-point watches at Port Moresby and Townsville to listen on the AIRSOPAC (TF 63) Net "E". These instructions further provided in paragraph 3 (c):

"All signals originated in North Eastern Area and addressed to South Pacific Forces are to be routed via Headquarters Allied Air Force (Brisbane). Additionally, when urgency demands, they may be routed via Port Moresby or Townsville on the above point-to-point series."

Net "C" also was guarded by the three Allied Air Force base stations in the North eastern Area - Fall River, Port Moresby and Townsville - for the purpose of being able to communicate with aircraft (in flight) of the South Pacific Force, should such aircraft desire to communicate. Listening watches were maintained and transmitters were ready to reply, but no provision was made for the initiation over Net "C" of messages from SOWESPAC bases to SOPAC bases


* CTF 62 Dispatch 070642, August 1942, addressed to CTF S3.


The procedure governing SOWESPAC Air Force reconnaissance missions was set forth in the Signal Annex to Operations Instruction No. 2, dated May 23, 1942. Contacts made at sea were to be reported immediately by transmission at the target. The reconnaissance plane making contact was to remain in the vicinity of the sighted target until recalled or forced to retire, sending MO's and the plane's identifying call. The Air Force ground station to which the contact was transmitted was to repeat the entire message in acknowledgement.

The contact codes and ciphers to be used by reconnaissance and bombardment aircraft of the South Pacific Area were issued to the Air Command Headquarters, Townsville and to the Naval Officer-in-charge, Port Moresby. Arrangements were made with the latter that North Eastern Area air units have access to the codes.

A study of the above communication plans for the air forces in both the SOPAC and SOWESPAC Areas reveals that adequate means were provided for the prompt transmittal of any contact report to the commands concerned so that immediate and positive action might be taken. CTF 63 made arrangements with COMSOWESPAC that all search reports would be immediately rebroadcast on the respective circuits of the air command in each area.* However, the employment of the means in SOWESPAC in practice, did not exploit the full capabilities of rapid and effective communications, thus causing long delays in the transmittal of vital information. It will be shown later that contact reports in the SOWESPAC Area followed the echelon of command before being broadcast to SOPAC forces.

Fleet broadcast schedules were the primary means of delivering contacts made in SOWESPAC Area to naval units in the SOPAC Area. The two primary fleet broadcast schedules that were available to deliver vital information to forces in the area of operations were the Canberra (VHC) "BELLS" broadcast, which was copied by the SOWESPAC naval units that were involved in the operation (old Task Force 44), and the Pearl Harbor (NPM) "HOW" Fox broadcast, which was copied by the SOPAC naval units.

(f) Allied Deployment of Naval Forces

(1) Approach to Guadalcanal-Tulagi Area.

A large concentration of naval forces was first assembled under the command of COMSOPACFOR in the Fiji Islands in late July 1942. CINCPAC had made available for the execution of Task ONE three Striking Forces of the Pacific Fleet: (1) Task Force 11, flagship Saratoga; (2) Task Force 16,


* Letter to President, Naval War College, from Rear Admiral M.B. Gardiner, USN (Chief of Staff to COMAIRSOPAC, CTF 63, in August 1942), October 20th, 1948.



Flagship Enterprise; and (3) Task Force 18, Flagship Wasp. COMSOWESPAC had transferred the cruisers and destroyers of Task Force 44 to COMSOPACFOR. These joined at Koro Island, Fiji's with South Pacific Area Amphibious Force (Task Force 62) in which the First Marine Division was embarked, to form the Expeditionary Force, TF 61.

After conducting rehearsals at Koro Island from July 28th to July 31st, the Expeditionary Force (TF 61) sortied on July 31st for Tulagi. CTF 61 headed to the westward for Longitude 159°-00'(E) Latitude 16°-30'(S), passing south of Efate through the New Hebrides en route. At 1200, August 5th, he headed his Expeditionary Force northward along the meridian of 1590 East Longitude and remained in cruising disposition until 1600, August 6th. This track is shown on Diagram "B".

The Allied approach commenced at this latter time. CTG 61.2 (CTF 62) placed in effect his Operation Plan A3-42 and directed the Amphibious Force (TG 61.2) to take the Attack Force Approach Disposition AR-3 and to proceed to the assigned transport areas (Area XRAY at Lunga Roads and Area YOKE at Tulagi), complying with the courses and times specified in Attack Force Approach Plan AR-11. According to this plan, the Amphibious Force continued north to reach position Latitude 09°-50'(S), Longitude 159°-00' (E) at 2235, August 6th, at which time course was to be changed at 040°(T) to close Savo Island at a speed of twelve knots. In the execution of this plan, this change actually was made at 2250.

The Air Support Force, TG 61.1 broke off from its cruising position (relative to TG 61.2) at 1830 by changing course to 305°(T) and increasing speed to twenty-two knots to pass through Point ABLE and Point BAKER (shown on Diagram B) seventy-five miles to the west of the meridian 159° East Longitude. Point BAKER was to be reached by 0030, August 7th at which time TG 61.1 was to change course to 090°(T) to reach Point VICTOR at dawn, the dawn aircraft launching position.

At 0300, the Amphibious Force (TG 61.2), in a position fifteen miles southwest of Savo Island, deployed into two groups: Group XRAY which proceeded to Lunga Roads off Guadalcanal, and Group YOKE which proceeded to Tulagi. Both groups were in Iron Bottom Sound at dawn and arrived at their objectives shortly after sunrise, August 7th.

This deployment at dawn--with the Amphibious force (TG 61.2) in Iron Bottom Sound and the Air Support Force (TG 61.1) operating in its support in the area south of Guadalcanal Island -- established the strategic disposition of Allied forces with which this study of the Battle of Savo Island is concerned, for it was this disposition that the Japanese Cruiser Force had to meet on the 8th and 9th of August. In conjunction with this disposition of surface ships of TF 61, COMSOPACFOR had the land and tender - based aircraft of Task Force 63 deployed and COMSOWESPAC had his land-based air disposed, in the manner already described. In addition, COMSOWESPAC had deployed two submarines in the Bismarck Area for


reconnaissance and attack patrols.

(2) CTG 61.1 Operates His Carriers

CTG 61.1 operated each of his three carrier striking forces as a separate group, rather than a single task force of three carriers as was done in later operations. He formed them into a disposition approximating an equilateral triangle with the Saratoga in TG 61.1.1 (TF11) at the apex as guide, followed five miles on his starboard quarter by TF 61.1.2 (TF16), and five miles on his port quarter by TF 61.1.3 (TF18).* In this manner these groups remained within mutual supporting distance and visual signal distance of one another.

CTG 61.1's reasons for not combining his task groups into one Task Force of three carriers were:

(a) His belief that protection of a carrier task force under air attack could best be accomplished by the separation of the carriers into groups containing only one carrier each, as was done at the Battle of Midway. At this early date, maneuverability was given almost equal importance with anti-aircraft fire in defense of a task force. It was felt that separation would reduce the risk of collision which would otherwise exist in a tight formation when the carriers were taking independent evasion action. In August 1942, the lesson derived from both the Battle of the Coral Sea and the Battle of Midway--namely, that an independent single-carrier task force could be readily penetrated by Japanese attacking planes--had not been adequately evaluated. In later actions, as the volume of anti-aircraft fire of a task force increased in proportion to the increased number of anti-aircraft guns and improved fire control, the importance of maneuverability decreased and the necessity for independent freedom of action for each carrier obtained to a far less degree.

(b) The necessity for obtaining air space for rendezvous and break-up of carrier air groups. At this time the realization had not yet evolved that fighter defense is made easier and more economical by concentrating ships and by controlling a spot defense rather than a dispersal of fighter strength in defense of separated targets.

While TG 61.2 was heading for its anchorage off Tulagi and Guadalcanal, TG 61.1 was operating about seventy miles to the southwest of Tulagi and was generally steaming on a southwesterly course at thirty knots while launching aircraft, since the wind was from that direction and was very light. This task group, commencing at 0530, had been providing air cover for TG 61.2 and air strikes for assaulting enemy positions


* War Diary Enterprise August 1942.


at Guadalcanal and Tulagi. At about 0625 sixteen fighters (launched by the Wasp) destroyed all of the Japanese aircraft based at Tulagi - seven Type-9 flying boats and nine seaplane fighters -- without suffering any losses whatsoever.

(3) Approach of TG 61.2 (Amphibious Force)

The approach of TG 61.2 was blanketed by very favorable weather conditions, in that there were clouds and sufficient light rain to cloak the advance of the force. At the same time there was sufficient moonlight to facilitate taking bearings and making the necessary and prescribed changes of courses.* At 0406, Amphibious Group XRAY changed course to 120°(T) to proceed south of Savo Island directly to Lunga Roads. At 0500, Amphibious Group YOKE, in position four miles north of Savo Island, changed course to 120°(T) to make the final approach to Tulagi. At 0614, Fire Support Group MIKE commenced shelling designated targets at Tulagi and at 0617 Fire Support Group LOVE opened fire on Japanese positions at Guadalcanal. Group YOKE arrived at the transport area off Tulagi at 0637 and Group XRAY arrived off Guadalcanal Beach at 0650. As a consequence of these weather conditions, and of the failure of the Japanese to locate TF 61 on August 6th, the advances of the Amphibious Groups on Tulagi and Guadalcanal were effected with complete surprise, and were first reported by the Japanese at Tulagi at 0645. This message was received by Commander Outer South Seas Force at 0652, which time has been accepted in this analysis as the initial contact between Allied and Japanese forces. At this time CTG 61.2 set the hour for landing on Guadalcanal as 0910, August 7th.*

(4) Deployment of SOWESPAC Submarines.

COMSOWESPAC deployed two submarines in the New Britain, New Ireland area during this operation. These were the S-38 and the S-44 which operated independently.

(a) Deployment of S-44.

The S-44 departed Brisbane at 0930, July 24th and headed for a patrol station off Bougainville Island on the assumed Japanese traffic route between Rabaul and Tulagi where she arrived at 0830, July 30th. Sea conditions were poor so the Commanding Officer headed for a patrol station off Cape St. George. He arrived on station at about 0350, August 1st where he noted some merchant shipping. He remained off Cape St. George until morning when he headed up the east coast of New Ireland arriving off North Cape at about 0800, August 4th. Here he encountered considerable merchant shipping, but was unable to close it. At 1845,


* War Diary CTF 62, August 1942, page 5.


August 6th he headed around the west end of Hanover Island and commenced cruising eastward along the south shore to the entrances of Steffen and Byron Straits where he arrived at about 0700, August 7th*

(b) Deployment of S-38

The S-38 departed Brisbane, Australia at 0930, July 28th and headed for her patrol station off the entrance to Wide Bay, New Britain where she was directed to cover the assumed Japanese traffic lane between Rabaul and Gona, New Guinea. She arrived on station at 1817, August 4th, and at 0300, August penetrated Wide Bay, but discovered no evidence of Japanese activity. The Commanding Officer then headed for a patrol station off Cape St. George, the southern tip of New Ireland, where he arrived at 0610, August 6th and commenced patrolling again.**

The decisions of the Commanding officers of the S-38 and S-44 to change these patrol stations on their own initiative indicated a correct appreciation of their objective which was the destruction of Japanese shipping and their positions at 0652, August 7th, were just short of being very fruitful against the Japanese Cruiser force enroute to Tulagi which passed out the south entrance of Steffen Strait at 0650, August 7th and later passed very close aboard the S-38at 1945, August 7th.

(5) Deployment of Allied Forces at 0652, August 7th.

At 0652, August 7th - the time at which Commander Outer South Seas Force received the initial contact report from Tulagi - the various Allied forces deployed in his area were located in the following positions:

(a) TG 61.1 Air Support Force

This force was located sixty-eight miles on bearing 240°(T) from Tulagi, heading southeast.

(b) TG 61.2 (TF 62) Amphibious force

(1) Amphibious Group YOKE was in the vicinity of Tulagi. Fire Support Group MIKE had commenced shelling the Japanese positions at Tulagi at 0614, and Transport Squadron YOKE had anchored at 0637. The screening group at Tulagi remained underway, the Chicago and Canberra operating with the Bagley and the Henley, Helm, and Blue providing an anti-submarine screen for the transports.

(2) Amphibious Group XRAY was in the vicinity of Guadalcanal Beach. Fire support Group LOVE had opened fire at 0617, and Transport


* Third War Patrol Report, S-44, July 24th, 19421 to August 23, 1942.
**  Seventh War Patrol Report S-38, July 28th, 1942 to August 22nd, 1942.


Squadron XRAY had anchored at 0650. the underway anti-submarine screen for the transports consisted of the Selfridge, Jarvis, Mugford and Ralph Talbot. The Australia and Hobart, screened by the Patterson, remained underway.

(c) Submarines

(1) The S-38 was patrolling an eighteen mile line parallel to SW Coast of New Ireland in the vicinity of Cape St. George and eight miles off shore.

(2) The S-44 was cruising submerged about three miles south off the coast of New Hanover en route to Steffen Strait.

(g) Composition of Forces and Tasks Assigned

(1) Composition of Forces

The Expeditionary force, TF 61, was a very powerful force consisting of two almost entirely separate forces; one, the Air Support Force, and the other an Amphibious force. The composition of these forces are set forth below.

(a) TG 61.1 - Air Support Force

(1) TG 61.1. (Pacific Fleet Task Force 11)

Saratoga (36 VF, 36 VB, 18 VT) 1 CV
Minneapolis, New Orleans 2 CA
Phelps, Farragut, Worden, McDonough, Dale 5 DD

(2) TG 61.1.2 (Pacific Fleet Task Force 16)

Enterprise (36 VF, 36 VB, 16 VT) 1 CV
North Carolina 1 BB
Portland 1 CA
Atlanta 1 CL (AA)
Balch, Maury, Gwin, Benham, Grayson 5 DD

(3) T.G. 61.1.3 (Pacific fleet Task Force 18)

Wasp (27 VF, 28 VB, 6 VT) 1 CV
San Francisco, Salt Lake City 2 CA
Lang, Sterrett, Aaron Ward, Stack, Laffey, Farenholt 6 DD

(b) TG 61.2, Amphibious Force (TF 62 plus SW Pac TF 44)

(1) TG 62.1, Transport Group XRAY

Fuller, American Legion, Bellatrix, McCawley (F)
Barnett, Geo. F. Elliott, Libra, Hunter Liggett
Alchiba, Fomalhaut, Betelgeuse, Cresent City
President Hayes, President Adams, Alhena 15 AP


(2) TG 62.2, Transport Group YOKE
Neville, Zeilin, Heywood, President Jackson 4 AP
Colhoun Gregory, Little, McKean 4 APD

(3) TG 62.3 Fire Support Group LOVE
Quincy, Vincennes, Astoria 3 CA
Hull, Dewey, Ellet, Wilson
4 DD

(4) TG 62.4 Fire Support Group MIKE
San Juan 1 CL (AA)
Monssen, Buchanan 2 DD

(5) TG 62.5, Minesweeper Group
Hopkins, Trever, Zane Southard, Hovey 5 DMS

(6) TG 62.6 Screening Group
HMAS Australia, HMAS Canberra, USS Chicago 3 CA
HMAS Hobart 1 CL
Selfridge, Patterson, Bagley, Blue, Ralph
, Henley, Helm, Jarvis, Mugford 9 DD

(7) TG 62.7 Air Support Control Group
Fighter Director Group in Chicago
Air Support Director Group in McCawley
Air Support Director Group (Standby) in

(8) TG 62.8, Landing Force (1st Marine Division)
Guadalcanal Landing Group
Tulagi Landing Group

(2) Tasks Assigned

The tasks assigned the Allied naval forces were, in part:

(a) TF 61, Allied Expeditionary Force.

(1) On DOG Day to capture and occupy Tulagi and adjacent positions, including an adjoining portion of Guadalcanal suitable for the construction of landing fields.
(2) To defend seized areas until relieved by forces to be designated later.
(3) To call on TF 63 for special aircraft missions.*


*COMSOPAC Operation Plan I-42, July 16th, 1942.


(b) TG 61.1 Air Support Force.

(1) On DOG Day and subsequently, to cooperate with Commander Amphibious force by supplying air support.
(2) To protect own carriers from enemy air attacks.
(3) To make air searches as seem advisable or as ordered.*

(c) TG 61.2 (TF 62 plus TF 44) Amphibious Force

(1)To proceed to Tulagi in tactical support of Amphibious Force. On DOG Day to seize and occupy Tulagi and adjacent positions including an adjoining portion of Guadalcanal suitable for the construction of landing fields.
(2) To defend seized areas until relieved by forces to be designated later.
(3) On departure of carriers, to call on TF 63 for special aircraft missions.*

The Amphibious Force was composed of eight separate task groups, but since only the Screening Group and the Fire Support Groups LOVE and MIKE were involved in the Battle of Savo Island, the tasks assigned to these three groups only are pertinent. These tasks were, in part:

(a) Screening Group -- TG 62.6**

(1) To screen the transport groups against enemy surface, air, and submarine attack.

(b) Fire Support Group LOVE -- TG 62.3**

(1) In case of air attack to defend transports and troops at Guadalcanal with anti-aircraft fire, acting under the directions of CTG 62.6.
(2) To support TG 62.6 in case of surface attack.

(c) Fire Support Group MIKE -- TG 62.4.

Same as for Fire support Group LOVE, except that it operated in the Tulagi Area rather than at Guadalcanal.


*CTF 61 Operation Order I-42, July 28th, 1942. ** CTF 62 Operation Plan A3-42, July 1942.


The composition of TG 61.1, as shown above, combined extreme mobility and offensive striking power. This Air Support Force was replied upon to meet any air or surface threat that the Japanese might be able to bring against the Amphibious Force, in addition to its role of air strikes in support of the landing operations. Allied intelligence did not indicate that there were any powerful Japanese carrier striking force in the vicinity of Tulagi; but the responsible commanders realized that, despite the Japanese carrier losses at Midway, strong Japanese forces might well be brought to bear in or near the Tulagi area. Serious opposition could be expected, not only from land-based aircraft at present within the Bismarck-New Guinea-Solomons area but also from reinforcement aircraft which could be rapidly flown in from the Truk-Ponape area. A combination of attacks from both land-based and carrier-based aircraft was considered to be sufficient to counter such opposition. In addition Allied intelligence disclosed the relatively limited extent of Japanese surface strength within the area. Consequently Japanese surface forces so far reported were scarcely a match for the Allied surface ships within the TG 61.1.

TG 61.2 was, as the above listed composition shows, an extremely powerful amphibious force which possessed the capability of defeating the strongest surface forces that might be employed by the Japanese at this time. With the support of TG 61.1, it also had the means of defeating any land-based and carrier-based aircraft which might be employed against it. The Battles of the Coral Sea and of Midway had seriously decreased the carrier striking power of the Japanese Combined Fleet, and the limited number of airfields within the Solomon Islands restricted the striking power of the Japanese land-based aircraft.

The Screening Group, plus the Fire Support Groups LOVE and MIKE, had sufficient strength to carry out the tasks assigned to it. Its primary role was to defend the transport groups against enemy surface, air and submarine attack during the amphibious operation. During daylight hours, when covered by the air power of TG 61.1, it was capable of defending itself against any probable attack. During the night, when it was without the support of TG 61.1, its preponderance of power against possible Japanese surface attack was considerably lessened. However, it was capable, provided its strength was properly concentrated, of defeating Japanese surface forces known to be in the area, should such forces attempt to interfere with the landing operations during night or low visibility.

(h) The Allied Plan

This study is concerned merely with those aspects of planning for the execution of Task ONE which culminated in the clash between Japanese and allied surface forces in the night action near Savo Island. Consequently the entire plan for the Allied landings and occupation of positions in the southern Solomons has not been included. The Allied plan for


offensive action in the South Pacific originated with the Joint Chiefs of Staff and was officially made known by COMINCH and CINCPAC on July 2nd, 1942. Operations were to commence on August 1st for the accomplishment of the objective in three stages, with the task for each stage set forth as follows:

(a) Task ONE: Seize and occupy Santa Cruz Islands, Tulagi and adjacent positions.
(b) Task TWO: Seize and occupy the remainder of the Solomon Islands, Lae, Salamaua, and the northeast coast of New Guinea.
(c) Task THREE: Seize and occupy Rabaul and adjacent positions in the New Guinea-New Ireland Area.*

As a consequence of this directive, COMSOPAC and COMSOWESPAC held consultations and recommended that the operation not be initiated until adequate air strength was built up in SOPAC and SOWESPAC Areas. They stated that, in view of (1) the recently developed strength of the enemy positions, (2) the shortage of airplanes for the continued maintenance of strong air support throughout the operation, and (3) the shortage of transports and the lack of sufficient shipping that would make possible the continued movement of troops and supplies, the successful accomplishment of the operation was open to the gravest doubts. They therefore jointly recommended that Task ONE be deferred, pending further development of forces in the South Pacific and Southwest Pacific Areas. They offered an alternate plan that the Allies proceed with an infiltration process through the New Hebrides and Santa Cruz Island groups until such time as bases could be developed for the support of the three stages of the operation as one continuous movement.**

The Joint Chiefs of Staff refused to defer the operations already underway for Task ONE for two reasons: (a) that it was necessary to stop without delay the enemy's southward advance that would be effected by his firm establishment at Tulagi, and (b) that enemy airfields established at Guadalcanal would seriously hamper, if not prevent, Allied establishment of bases both at Santa Cruz and Espiritu Santo. They agreed to provide additional ship borne aircraft and additional surface forces: to increase the rate of flow of replacement aircraft and to make available for the South Pacific Area one heavy bombardment group of thirty-five planes.***

COMSOPAC and COMSOWESPAC then went ahead with the planned operation. COMSOPAC planned to accomplish Task ONE by seizing the Tulagi-Guadalcanal


*Joint Directive U.S. Joint Chiefs of Staff, War Diary, COMSOPAC for July 1942-COMINCH to CINCPAC 022100, July 1942.
**COMSOPAC and COMSOWESPAC to Joint Chiefs of Staff (COMINCH) Dispatch 081012 July 1942.
***COMINCH to COMSOPAC Dispatch 102100, July 1942.


area on DOG Day and, after it had been secured by seizing Ndeni in the Santa Cruz Islands. The purpose of these operations was to deny these positions to enemy forces and to prepare bases for Allied future offensive operations.*

The plan was simple and direct. An amphibious force of suitable strength was to seize the Tulagi and Guadalcanal area under the air support of land-based aircraft flying from land and seaplane bases within both the SOPAC and SOWESPAC Areas and from carrier-based aircraft within the Expeditionary Force. It was also to be supported against naval attack (both surface ships and submarine) by the cruiser and destroyer escorts and screening ships attached to the amphibious force.

Based on the intelligence available concerning the enemy forces within the objective area and those capable of being moved into the area in time to interfere with the landings, the plan was sound. This was particularly true, providing the factor of surprise could be achieved at the objective area. However, the Allied Commander did not expect to achieve surprise,** and relied on the coverage of his land and carrier-based aircraft and on the gunpower of his ships to defeat expected enemy counter-attacks.

(1) General Summary

The preceding discussion of the background for the Battle of Savo Island shows, in a general way, that:

(a)The Japanese effort south of the Equator was designed to expand Japanese power in the South Seas Area and to counter any Allied attack that might be made in that area. This effort was spearheaded by that portion of the land-based air power of the Base Air Force (ELEVENTH Air Fleet) which was based at Rabaul. It was supported by limited naval forces based primarily at Rabaul and Kavieng, and by submarines, all under the command of Commander Outer South Seas Force (Commander EIGHTH Fleet) whose headquarters were at Rabaul. It was supported also by reconnaissance seaplanes of the Base Air Force based at Rabaul and Tulagi.

(b)The Allied effort in the South Pacific was designed to stop the advance of the Japanese in that area and to seize advanced bases in the Solomons from which to continue further operations against the Japanese. This effort was spearheaded by an Expeditionary Force, strong in naval and air power. It was supported by land-based air power operating both from SOPAC and SOWESPAC bases. In strategic command of all forces within the SOPAC Area, including the Expeditionary Force, was COMSOPACFOR with operational headquarters at Noumea.


*COMSOPAC OpPlan 1-42 Serial 0017 of July 16th, 1942.
**COMSOPAC and COMSOWESPAC to Joint Chiefs of Staff (COMINCH) Dispatch 081017, July 1942.


Chapter IV
Japanese Reaction
0652, August 7th to 2400, August 7th

(a) Operations of Japanese Cruiser Force

The Japanese cruisers Chokai and CRUDIV SIX had sortied at 0615 from Silver Sound, where they had been basing since August 1st, and had cleared the southern entrance to Steffen Strait at 0650. At that time the Chokai and Section One of CRUDIV SIX (Aoba and Kaka) had set the course for Seeadler Harbor, Manus Island, and Section Two (Kinugasa and Furutaka) had headed for Rabaul. At 0652 each of the five cruisers intercepted the urgent dispatch from Commander Air Base, Tulagi that reported: "Enemy task force sighted!"*

The reaction of the Japanese unit commanders concerned was spontaneous and correct. The Commanding Officer, Chokai, immediately reversed course toward Rabaul. COMCRUDIV SIX, in the Aoba, cancelled the orders of Section One to proceed to Manus, and ordered the Aoba and Kako to proceed to Rabaul** at twenty-four knots, notifying Commander EIGHTH Fleet at Rabaul of his action.*** He reformed CRUDIV SIX in column and joined in behind the Chokai.

Meanwhile Commander Outer South Seas Force, who was at Rabaul, was studying the situation. He had been expecting some form of Allied attack in the Solomons Area, but not at this time. The Japanese records available to this study do not divulge his mental processes at the time, but there can be no doubt as to his serious concern over the nature of the Allied effort at Tulagi.

In the meantime, the routine air searches scheduled for August 7th were commenced. At 0700, two Type-97 seaplanes departed Rabaul to search the sector between bearings 100°(T) and 130°(T) from Rabaul to a radius of 700 miles.****

At 0725, Commander Outer South Seas Force received the amplifying report from Tulagi that he had been anxiously awaiting. This urgent dispatch had been released by Commander Tulagi Air Base twenty minutes before, and reported: "Enemy task force of twenty ships attacking Tulagi;


*CRUDIV 6 Detailed battle Report #8, Solomon Naval action, August 7th-10th, 1942, CIG Document S6927 June 27th, 1947, pages 2 and 6.
**CRUDIV 6 Signal Order 289.
***COMCRUDIV 6 Classified Dispatch #246, August 7th, 1942.
**** Japanese Search Plans, August 7th to 9th, 1942, CIG Document 74632, May 12th, 1947, page 2.


undergoing severe bombings; landing preparations underway; help requested."*

This dispatch provided Commander Outer South Seas Force with an accurate estimate of ships attacking Tulagi, since in fact there were twenty-two ships of all classes at Tulagi.** He knew, of course, from this dispatch that there must be a supporting force of Allied carriers somewhere in the vicinity of Tulagi to have carried out the bombing attacks reported. No reports were received from Guadalcanal and it is possible that Commander Outer South Seas Force based his initial estimate of the total strength of the Allied effort upon the report from Tulagi.

He shortly was handed a dispatch, sent at 0725 by Commander YOKOHAMA Air Group at Tulagi addressed to commander FIFTH Air Attack force, which reported: "All large flying boats burned as a result of the 0630 air attack."

The situation confronting Commander Outer South Seas Force was somewhat as follows: the Allied attack on Tulagi was a complete surprise since the Japanese had made no contact whatsoever on the Allied force until it struck. It was in strength, and constituted a serious menace to the Japanese positions in the Solomon Islands. If these positions were to be held by the Japanese, the defending forces at Tulagi must be given immediate help. Continuous air attacks on the allied transports and cargo ships might destroy many of them, and would have a serious delaying effect on the unloading schedule of the remainder. Anchored transports were excellent targets for Japanese submarines. But air and submarine attacks were not enough in the face of Allied screening ships and carrier-based aircraft. He concluded that a bold offensive by surface ships also was needed, and decided "to put the fleet into action immediately to destroy the enemy."***

He therefore prepared his counterattack. He had been informed by dispatch of the initiative of the Commanding Officer, Chokai and of COMCRUDIV SIX in proceeding to Rabaul. At 0749 he directed the Yunagi to make full speed and escort the Chokai into Rabaul Harbor. He also directed the Tenryu, Yubari and the Yunagi to be prepared at 1300 to depart from Rabaul to attack the Allied ships in the Tulagi area.***

He continued to receive reports from Tulagi of fierce naval bombardment, air bombing attacks, and of the initial landings until 0805 when communications ceased.*** The last dispatch received at that time reported


*CRUDIV 6 Detailed Battle Report #8, Solomons Naval Action, August 7th-10th, 1942, CIG Document 86927, June 27th, 1947, pages 3 and 6.
**War diary CTF 62, July 18th to August 31st, 1942; and Annex I to CTF 62 Operation Plan A#-42.
***Commander 8th Fleet's Estimate of the Situation regarding American Landings on August 7th, 1942, on Guadalcanal and Tulagi, CIG Document 73845, May 7th, 1947.


"The enemy was in great force; but that no matter what the odds, the garrison would fight bravely to the last man, praying for everlasting victory."*

The disruption of communications from Tulagi denied Commander Outer South Seas Force the further information he desired of the enemy. Accurate intelligence of the strength, disposition and activities of Allied forces was of primary importance in the promotion of his freedom of action. Seaplanes had already commenced search operations at 0700 in the Tulagi sector. Undoubtedly he knew, as the Allies had discovered at the Battle of Midway, that seaplanes are generally unsuitable for reconnaissance in an area where air opposition would be encountered. At any rate, additional searches were commenced at 0900, employing land attack planes rather than seaplanes.** Commander FIFTH Air Attack Force DESOPORD No. 195, ordering his search, was timed at 0910,** ten minutes later than the actual time of takeoff of the search planes.** The indications are that preliminary orders were given by Commander FIFTH Air Attack Force either by telephone or in conference, to his subordinate commander, Commander SECOND Air Group (who conducted the search) and were followed up with confirmation by dispatch orders. The urgency of the searches by land attack planes on August 7th is accented by the fact that the directives for searches on subsequent days were each given on the day preceding.

The land-attack planes from the SECOND Air Group conducted reconnaissance missions to a radius of 700 miles from Rabaul over courses and lateral distances as follows:

Plane Course Out Lateral Distance
No. 1 170°(T) 60 miles to the left
No. 2 117°(T) 60 miles to the left
No. 3 127°(T) 60 miles to the right

A study of these searches, as depicted on diagram "C", indicates that searches by plane number two and three were well designed to cover the area in the vicinity of the Tulagi landing. The search of plane number one covered in an isolated sector interposed between Tulagi and the Australian mainland from where, as a far as Commander Outer South Seas Force knew, the Allied forces may have been mounted.

The offensive of Commander Outer South Seas Force included the immediate launching of air attacks on the Allied naval forces at Tulagi. At 0900, simultaneous to the departure of the search planes, commander FIFTH Air Attack Force sent off an air attack group composed of twenty-seven


*Commander 8th Fleet's Estimate of the Situation regarding American Landings on August 7th, 1942 on Guadalcanal and Tulagi. CIG Document 73845, May 7th, 1947.
**Japanese Search Plans, August 7th-9th, 1942, CIG Document 74632, May 12th, 1947.


land attack planes of the FOURTH Air Group and eighteen Zero fighters of the TAIMAN Air Group.* At about 1045 a second air attack group, composed of sixteen Type-99 land-based carrier bombers of the SECOND Air Group, was launched to attack the Allied forces at Tulagi.** This latter attack group undoubtedly staged through the Japanese airfield at Buka Passage, since the range capability of the carrier bombers would otherwise have been inadequate for this operation.***

Commander Outer South Seas force "decided, depending on the results of the reconnaissance and counterattacks of our air force to the south, to strike the anchored enemy convoys at night and destroy them."*

With this plan in mind, he organized a naval force (hereinafter referred to as the Japanese Cruiser Force) composed of the Chokai as flagship, the four heavy cruisers of CRUDIV SIX, the light cruisers Tenryu and Yubari of CRUDIV EIGHTEEN and the only available destroyer, the Yunagi.

This plan depended upon the reports from his reconnaissance planes launched at 0700 and again at 0900. It also called for accurate information on the results of the two bombing attacks launched against the enemy, one at 0900 and the other at 1045. By 1110 the two search seaplanes had reached the 700 miles radius from Rabaul on tracks 107°(T) and 117°(T), then proceeded on their sixty miles lateral legs, and at 1131 headed back to Rabaul. Their reconnaissance had been restricted, as can be seen by referring to Diagram "C", to the area northeast of the southern Solomons. The weather southeast of Tulagi was reported to be bad, making reconnaissance impossible.**** The enemy was not sighted.****

Commander Outer South Seas force requested that a reconnaissance plane be sent to Tulagi to make direct observation of the results of the Japanese bombing attacks on the Allied ships there. At 1120, commander FIFTH Air Attack Force ordered one land attack plane from the SECOND Air Force to proceed from Rabaul at 1145 to carry out this reconnaissance mission at Tulagi and Guadalcanal.**** This plane actually took off at 1202.****

At 1203, the Chokai and CRUDIV SIX enroute from the Kavieng Area to Rabaul, sighted the Yunagi approaching the formation from a position fifteen and one-half miles away on bearing 193°(T).***** The Yunagi


* Commander 8th Fleet's estimate of the Situation on August 7th, 1942, CIG Document 73845, May 7th, 1947.
** Strength and disposition of 25th Air Flotilla on August 7th, 1942, CIG Document 74629, May 12th, 1947.
*** "Japanese Aircraft Performance Characteristics", Technical Air Intelligence Center (TAIC) Manual No.1, OpNav 16-VT #301.
**** Japanese Search Plans, August 7th-9th, 1942, CIG Document 74632, May 12th, 1947.
***** CRUDIV 6, CIG Document 86927, June 27, 1947.


effected rendezvous with the Chokai at 1224, and escorted her into Rabaul. CRUDIV SIX continued on to Rabaul separated from the Chokai and Yunagi.*

At 1239, these Japanese cruisers had reached a position about twenty-five miles north of Rabaul and sighted thirteen Allied B-17's attacking Rabaul.* The cruisers prepared to repel air attacks, and maneuvered to the westward until 1243 when the B-17's disappeared. Although this cruiser force was sighted and reported by the B-17's, it was not attacked. So far as the Japanese cruiser commanders could surmise, the airfields at Rabaul were more profitable targets for the B-17's than were the cruisers.

The Japanese air units at Rabaul endeavored to repel the B-17 attack group, employing the twenty-one Zero fighters at Rabaul.** They succeeded in destroying one B-17,*** but suffered damage to seven fighters of the TAIMAN Air Group and to one fighter of the SECOND Air Group.**

By 1307 the three Japanese land attack planes on search missions arrived at the outer limit of their sectors. They proceeded along their respective cross-legs, and at 1328 turned back toward Rabaul (where they landed at 1735). The combined reconnaissance of both land and seaplane searches detected one Allied seaplane in Task Force 63 Sector IV and the Allied ships at anchor in the Tulagi-Guadalcanal area.*** Perhaps the Japanese search plane on leg 127°(T) from Rabaul was discouraged by bad weather (reported by the Japanese in the southern Solomons and reported by Allied search planes in Sector I and II) and had turned back before he reached the Allied Air Support Force (TG 61.1) which was operating in his search sector. The Allied carriers reported a heavy overcast sky, but conducted flight operations below this overcast in good visibility.*****

At 1325, the first Japanese air attack struck at the Allied ships at Guadalcanal. This attack group encountered the opposition of Allied carrier-based fighters and anti-aircraft fire from the Allied ships. It suffered the following casualties: three attack planes and two fighters shot down, two attack planes made emergency landings, and nineteen attack planes and two fighters were damaged.** The Japanese in turn succeeded in shooting down nine Allied fighters and in dropping all bombs. No hits were made on the Allied ships, however since the bombs fell well clear of the surface targets, landing between the transports and the cruisers.*****


*CRUDIV 6 Detailed Battle Report #8, Solomons Naval Action August 7th-10th, 1942, CIG Document 86927, June 27th, 1947.
**Strength and disposition of 25th air Flotilla on August 7th, 1942, CIG Document 74629, May12, 1947.
***Operation Report, Allied Air Forces, Southwest Pacific Area for August 7th, 1942.
****Japanese Search Plans, August 7th-9th, a942 CIG Document 74632, May 12th, 1947.
*****War Diary, Enterprise, August 1942.
******War Diary, American Legion, August 1942, Report of Action off Guadalcanal, August 7th, 1942.



At 1330, the CHOKAI and YUNAGI arrived and anchored inside Rabaul Harbor. CRUDIV SIX arrived outside Rabaul at 1401 where these four cruisers conducted anti-submarine search patrols with their scouting planes until 1450 in accordance with orders from Commander EIGHTH Fleet.* Thereafter the cruisers' scouting planes were recovered and CRUDIV SIX remained underway awaiting the CHOKAI's departure from Rabaul.

At 1440, Commander EIGHTH Fleet decided to command the cruiser force in person. He therefore hoisted his flag in the CHOKAI,** and made preparations to get underway.

He realized that as Commander Cruiser Force, he would be denied the full freedom of action that he exercised at Rabaul as Commander Outer South Seas Force and Commander EIGHTH Fleet. He also realized that he would be especially limited in communications, since the CHOKAI would probably have to operate under radio silence. In making his decision to command the cruiser force in person, his reasoning therefore is assumed to have been that as the Japanese attack on the Allied forces at Tulagi must be gotten underway immediately under a competent commander, and as he was the senior and most experienced flag officer in the area, who was fully familiar with the planned operations, he should properly command it. He knew of course that Commander ELEVENTH Air Fleet was en route to Rabaul to relieve him that day and that his own status as Commander Outer South Seas Force would terminate in a matter of hours, in any event. This reasoning is considered correct.

The second Japanese air attack group, composed of sixteen carrier-type bombers, made its attack on the Allied ships at about 1500. This group had no fighter escort, for the twenty-one fighters remaining in Rabaul (after the first air attack group took off for Tulagi) were retained there for local air defense against B-17 raids. Consequently, the flight commander of the carrier bombers employed the tactic of dividing his force into two attack groups, one of which attracted the Allied fighters thereby allowing the other to make its attack unopposed. The unopposed attack group made one hit on the MUGFORD. The opposed attack group made no hits, but suffered the loss of six bombers and damage to three which made emergency landings.***

At 1521 the land attack plane (sent out from Rabaul at noon) reached the Tulagi-Guadalcanal Area, conducted his reconnaissance and made his report to Rabaul by radio. By 1530 Commander Outer South Seas Force had an intelligence summary**** compiled jointly from his reconnaissance and


*CRUDIV 6 Detailed Battle Report #8, Solomons Naval Action August 7th-10th, 1942, WDC Document 160997.
**War Diary 8th Fleet, August 7th-10th, 1942, CIG Document 74623, May 12th, 1947.
***Strength and Disposition of 25th Air Flotilla on August 7th, 1942, CIG Document 74629, May 12th, 1947.
****Japanese Search Plans, August 7th-9th, 1942. CIG Document 74632, May 12th, 1947.

- 48 -

attacking planes, which is quoted as follows:

"(a) Enemy making landings on Guadalcanal air base; three fires on the airfield.
(b) One fire in the area of the Tulagi seaplane base.
(c) Three heavy cruisers, several destroyers, and about thirteen transports off Tulagi.
(d) Several destroyers and twenty-seven transports off Guadalcanal air base.
(e) From 1320 to 1440 there were sixty to seventy enemy planes over the anchored transports.
(f) At about 1530 no enemy planes were sighted over the anchorage."*

This intelligence summary was somewhat in error as regards the composition of the Allied forces in the Tulagi-Guadalcanal Area. The number of transports reported was twice the actual. The number of cruisers and destroyers reported was far less than the number present. The reason for the errors stems from the short time spent over the target area by the reconnaissance plane at 1521. The time of takeoff, time of landing, and cruising speed of this plane** indicate that it proceeded directly to the Tulagi-Guadalcanal Area and returned promptly to Rabaul with no more than a single sweep past Tulagi and Guadalcanal. This intelligence summary reported the presence of Allied carrier-based aircraft at Tulagi but gave no information as to the location of the carriers. It did constitute, however, the results of the reconnaissance and air attack missions, upon which the plan of Commander Outer South Seas Force depended.

It is obvious that Commander Outer South Seas Force decided tentatively that his plan for a night surface attack upon the Allied forces at Tulagi and Guadalcanal was feasible, since he embarked upon it within the hour. He still would have the benefit of air reconnaissance and air attacks on the Allied forces on August 8th while he was en route to his objective.

Commander Outer South Seas Force then organized his command to function during his absence from Rabaul. The weight of evidence***


*Japanese Search P1ans, August 7th-9th, 1942, CIG Document 74632, May 12th, 1947.
**Technical Air Intelligence Center (TAIC) manual No. 1, OpNav 16-VT-#01 "Japanese Aircraft - performance characteristic."
***War Diary 8th Fleet, August 7th-10th, 1942, CIG Document 74633, May 12th, 1947, and Composition of 8th Fleet on August 7th, 1942, CIG Document 74835, May 12th, 1947.

- 49 -

indicates that the command of the Outer South Seas Force was turned over to Commander ELEVENTH Air Fleet; that the operational control of various units and the administration of EIGHTH Fleet affairs was kept ashore at Rabaul, temporarily under the cognizance of Commander ELEVENTH Air Fleet, and that Commander EIGHTH Fleet, himself, retained full command over that part of the EIGHTH Fleet which constituted the Japanese Cruiser Force. For the purpose of this analysis and to indicate the specialized nature of this specific operation, Commander EIGHTH Fleet will be known hereinafter as Commander Cruiser Force.

At 1628 Commander Cruiser Force departed Rabaul in his flagship the CHOKAI accompanied by the TENRYU, YUBARI and YUNAGI to conduct a night surface attack on the Allied ships at Tulagi and Guadalcanal.* By 1710 he had cleared the harbor and its approaches sufficiently to order a line formation, steaming in Condition of Readiness ONE.* At 1755, he ordered a zigzag plan with all ships making simultaneous turns.*

At 1805, he radioed Commander CRUDIV SIX, who had bean underway outside Rabaul, waiting to join up, the following message: "The CRUDIV SIX will place itself to the rear of the CHOKAI."**

CRUDIV SIX then executed a change of course and joined in column astern of the CHOKAI in an alert cruising disposition,** heading southeast through St. George's Channel. This cruising disposition is judged to have been the same as that ordered by Commander Cruiser Force on August 8th,*** with the CHOKAI and CRUDIV SIX in column screened against submarines by the YUBARI on the starboard bow and the TENRYU (followed in column by the YUNAGI) on the port bow. Zigzagging was not continued after the join-up of CRUDIV SIX, because of sunset and the approach of darkness.

At 1945, the Japanese Cruiser Force was detected by the Allied submarine S-38 in such a cruising disposition, headed on course 140°(T) at high speed, passing through a point eight miles west of Cape St. George.**** The Japanese ships were unaware of this contact and were not attacked because the S-38 was too close to them.****

At 2030, Commander Cruiser Force reached a position twenty miles due south of Cape St. George and ordered course 080°(T) to pass north of Buka Island during the night. His 2400 position was twenty miles distant on bearing 285°(T) from Cape Henpan, Buka Island.


*War Diary 8th Fleet, August 7th-10th, 1942, CIG Document 74633, May 12th, 1947.
**CRUDIV 6 Detailed Battle Report #8, Solomons Naval Action, August 7th-10th, 1942, WDC Document 160997, June 27th, 1947.
***COMEIGHTH Fleet Signal Order 25.
****War Diary S-38, August 1942.

- 50 -


Commander Outer South Seas Force ordered four submarines of SUBRON SEVEN to proceed on August 7th to Tulagi for observation, reconnaissance and attacks on Allied surface ships. The movements of the five units of SUBRON SEVEN on August 7th were:

(a) RO-33: En route to Tulagi from its station in the Papua Sea, where it had sunk a small freighter on August 7th.
(b) RO-34: En route from east coast of Australia to the vicinity of Port Moresby.
(c) I-121: Departed Rabaul about sunset August 7th for Tulagi.
(d) I-122: At Rabaul. (Departure for Tulagi was delayed twenty-­four hours)
(e) I-123: Departed Truk for Tulagi.

The contact made on the I-121 in St. George's Channel by the Allied submarine S-38 fixes the 2400 position of the I-121 as about thirty miles due south of Cape St. George. The midnight position of the RO-33, RO-34 and I-123 are not known.

- 51 -

Chapter V


0652, August 7th to 2400, August 7th

(a) OPERATIONS OF CTF 62 (Commander Amphibious Force)

At 0652, August 7th, the entire Amphibious Force had arrived on station at the objective area. As has been pointed out earlier, Transport Squadron YOKE had anchored off Tulagi at 0637, about one-half hour behind schedule, and Transport Squadron XRAY had reached the berths assigned to it off Guadalcanal Beach at 0650.

At 0652, CTF 62 signalled to all interested commanders and ships of TF 62 the ZERO hour (the time of the first landing in the Guadalcanal Area) would be 0910.*

At 0715 Commander Transport Group YOKE likewise signalled that H hour (the time of the initial landing at Tulagi) would be 0800.* The Marines of the Tulagi Landing Group landed unopposed at 0800 and the Marines of the Guadalcanal Landing Group landed unopposed at 0913.* This landing without opposition was an interesting and unexpected phenomenon which set a pattern for Japanese operations in defense of their islands in the Pacific war although it was not recognized at the time. Whenever the Japanese could retire as at Guadalcanal and Tulagi they did so in order to avoid the destructive effect of Allied bombardment. Whenever they could not retire as at Gavutu and Tanambogo, they stood and fought courageously and fiercely.

For the initial and subsequent landing operations, gunfire support was provided by Fire Support Groups LOVE (at Tulagi) and MIKE (at Guadalcanal) against beach emplacements and enemy installations. The planes of the Air Support Force (TG 61.1) provided air cover and facilitated the landings by strafing and bombing enemy positions. The operations of the Amphibious Force proceeded without interruption throughout the morning. After completion of minesweeping operations at 1100, CTF 62 moved the transports of Squadron XRAY closer to the shore, and anchored within 2000 yards of the beach to expedite unloading.

At about 1130, CTF 62 received a warning from CINCPAC that the Japanese air command at Rabaul had seventeen fighters and eighteen long-range bombers available which were being sent to attack TF 62.** Shortly thereafter CINCPAC warned the SOPAC forces: "Enemy submarines enroute to attack


*War Diary CTF 62, August 1942.
**CINCPAC Dispatch 062325, August 1942 to COMSOPACFOR, CTF 61, CTF 62, COMSOWESPAC.

- 52 -

Allied occupation forces at Tulagi.*

At about 1200 the Marines landed at Gavutu and encountered serious opposition which continued throughout the day. An air raid alert was broadcast which interrupted the unloading operations. Transport Squadron XRAY at Guadalcanal remained at anchor, but Squadron YOKE at Tulagi got underway. At 1325, the first Japanese air attack group, consisting of twenty-seven bombers and eighteen fighters, arrived and concentrated their full attack on the anchored transports at Guadalcanal. This attack was made by high level bombing from 8,000 to 10,000 feet and effected no damage to any ship of TF 62.** It was driven off with the loss to the Japanese of two bombers and two fighters, and damage to nineteen bombers. At about 1500 the second Japanese air attack group, consisting of sixteen carrier-type bombers tactically operating in two divisions of eight planes, attacked. The Allied fighters opposed one group and shot down six bombers and damaged the rest. The other Japanese bomber group closed the surface ships at Guadalcanal unopposed by Allied fighters, but succeeded only in attaining a hit on the destroyer MUGFORD. This hit did not prevent the MUGFORD from performing her assigned duties.**

The unloading of equipment and supplies from the transports and cargo ships had been delayed in all areas because of the nuisance effect of the Japanese air attacks which otherwise had delivered but slight damage to TF 62. The unloading at Tulagi and Gavutu had been delayed additionally by the severe fighting ashore.** Consequently, CTF 62 pressed the urgency of unloading operations throughout the night, using necessary lights.**

During the operations of August 7th, CTF 62 had been confronted with several enemy capabilities. His force had been attacked by land-based aircraft from Rabaul; it had been attacked by carrier-type bombers which withdrew to the westward; and it could expect to be harassed by Japanese submarines. He became aware of the danger threatening to his command from Japanese forces including carriers that might approach from Rabaul. He became concerned over the extent of the Allied land-based searches in that direction. He knew that the seaplane searches of CTF 63 would be extended on the next day, August 8th, in Sector IV to 650 miles northwest of Marmasike Estuary but would reach no further west than 318°(T) from that point of origin. He knew that in the direction of Rabaul, the northern limit of CTF 63's Sector II reached no further than the southern tip of Choiseul Island and that beyond that point search and reconnaissance depended upon the cooperation of SOWESPAC aircraft. He therefore sent a dispatch to CTF 63 at 1742 stating: "The plan of search for D plus One Day (August 8th) does not cover sector 290 to 318 degrees from Malaita. Southwest Pacific is responsible for this sector, but I consider a morning search by you is necessary for adequate cover."** CTF 62, in making this


*CINPAC Dispatch 062336, August 1942 to COMSOPACFOR, CTF 61, CTF 62, COMSOWESPAC.
**War Diary CTF 62, August 1942.
***CTF 62 dispatch 070642, August 1942, to CTF 63.

- 53 -

requst, demonstrated a keen appreciation of the ever-changing situation, for it was through this area that the Japanese Cruiser Force passed on August 8th without detection by the search planes of CTF 63.

Sunset was at 1816; and the screening group took stations at 1830 to protect the transports in the disposition described in detail in the next section.

At 2242, CTF 62 was informed by COMSOPACFOR that an enemy submarine had been reported in Latitude 10°-(S), Longitude 162°(E). This position was that of Lark Shoal, sixty miles east of Guadalcanal Island. This was an erroneous report, for no Japanese submarines were in the area. It is mentioned here as the reported presence of Japanese submarines probably influenced CTF 62's concept of his freedom of action.

At 2400, CTF 62 received a message from COMSOWESPAC reporting (a) the contact made by SOWESPAC B-17's at 1231, August 7th on four cruisers and one destroyer on a westerly course in a position twenty-five miles north of Rabaul, and (b) the contact made also by COMSOWESPAC B-17s on six unidentified ships in St. George's Channel on course southeast.* This dispatch was delayed in the SOWESPAC Area for almost eleven hours, as the difference between the time of contact and the time group of the message well shows. Because of the extensive surface ship traffic in the Rabaul Area, the two reports in this message may not have indicated to CTF 62 the southward movement of the Japanese Cruiser Force. But it should have served to remind him of an additional Japanese capability - that of attacking TF 62 with a force of cruisers and destroyers.


Since the Battle of Savo Island was directly concerned with the Screening Group, including Fire Support Groups LOVE and MIKE, a discussion of the disposition assumed by these groups for the night defense of the landing areas is herewith presented.

The disposition assumed is that shown in Plate IV. The night screen was composed of three main groups and of two anti-submarine and radar pickets. The three main groups were to destroy or beat off hostile surface forces which attempted to enter Iron Bottom Sound.** The anti-submarine and radar pickets were to give warning of the approach of hostile surface forces and of submarines.*** CTG 62.6 instructed the anti-submarine and radar pickets as follows: In the event of an enemy surface force being detected, immediate report was to be made. The force was to be shadowed


*COMSOWESPAC Dispatch 071219, August 1942, to all TFC's, Pacific.
**CTG 62.6 Special Instructions to Screening Group and Vessels Temporarily assigned August 1942, para. 3(b).
***Ibid, para. 6.

- 54 -

Plate IV: Battle of Savo Island, Night Disposition of Screening Force, 7 & 8 August 1942

- 54a -


- 54b -

and frequently reported; and when about to be engaged by cruisers, the destroyers in contact might be ordered to illuminate the enemy force with searchlights.*

(1) Night Disposition

The stations taken by the Screening Groups at 1830 were as follows:

(a) The "AUSTRALIA" Group, commanded by CTG 62.6, was composed of the three cruisers AUSTRALIA, CANBERRA and CHICAGO in column, distance 600 yards, and the anti-submarine screening destroyers PATTERSON and BAGLEY, which were on station about 2000 yards on either bow of the AUSTRALIA. This Group was patrolling back and forth at twelve knots in its assigned area south of a line drawn 125°(T) from the center of Savo Island and west of Longitude 160°-04'E. Its courses averaged 125°(T) in an easterly direction and 303°(T) in a westerly direction. Its patrol length averaged about twelve miles. Its western limit bore 148°(T) distant five miles from the southern tip of Savo Island; and its eastern limit bore about 345°(T) distant five miles from Lunga Point. The PATTERSON always remained on the western flank, and the BAGLEY on the eastern flank, regardless of courses in or out.

(b) The "VINCENNES" Group, commanded by the Commanding Officer of the Vincennes, was composed of VINCENNES, ASTORIA, QUINCY, in column and screened by JARVIS and HELM. This Group was underway at ten knots north of the 125°(T) limiting line from Savo Island and west of Longitude 160°-04'E. This group was steaming clockwise around the perimeter of a square five miles on each side, the center of which was in Lat. 09°-07'S., Long. 159°-57'-12"E.** Each trip around the square covered twenty miles. The corners of the square were on the north-south and east-west diagonals from the center, which was the reference point. Plate IV shows that the west corner of the square was about three miles from Savo Island, and the north corner was about five miles from the eastern limit of the RALPH TALBOT's patrol. The eastern limit of the VINCENNES patrol was the 100-fathom curve. The VINCENNES Group commenced the patrol of the perimeter of this square by passing through the west corner of this square at 2000 on course 045°(T). It changed course 90°to the right every thirty minutes with such adjustments of speed as were necessary to pass through each corner on schedule.**

(c) The "SAN JUAN" Group, commanded by CTG 62.4, was composed of the SAN JUAN, HOBART screened by MONSSEN and BUCHANAN. It was underway at fifteen knots in its assigned area east of Longitude 160°-04'E. It patrolled on a north-south line, the length of which was about ten miles.


*CTG 62.6 Special Instructions to Screening Group and Vessels Temporarily Assigned, August 1942, para. 6.
**Action Report ASTORIA, Battle of Savo Island, Serial AP37/A16-3/(00500) August 20th, 1942.

- 55 -

(d) The anti-submarine and radar pickets were on station to the westward and northward of Savo Island. The RALPH TALBOT was on patrol on a line between position Lat. 08°-59'S., Long. 159°-55'E. and position Lat. 09°-01'S., Long. 159°-49'E. Her courses were on 072°(T) and 252°(T). The BLUE was on patrol on a line between position Lat. 09°-05'S., Long. 159°-42'E. and position Lat. 09°-09'S., Long. 159°-37'E. Her courses were 051°(T) and 231°(T). The patrol seed for both destroyers was twelve knots; the length of each patrol was six and one-half miles.

(e) The remaining destroyers of TG 62.6, notably the SELFRIDGE, HENLEY and MUGFORD, remained with the transports on anti-submarine screening stations.

(2) CTG 62.6's Instructions

The night disposition was drawn up in its broad aspects by CTG 62.6, but the manner in which each of the main groups accomplished its mission within its own area was left to the discretion of its group commander. For their guidance, CTG 62.6 set forth two principles: (a) that it was essential that an enemy force be beaten off before it sighted or reached the convoy and that Allied naval forces be concentrated so as to avoid confusion in night action, and (b) that it was his aim to meet the enemy to seaward of the area between Savo Island and Sealark Channel, and the Allied force which engaged should remain interposed between the enemy and that area. He expected that the extensive Allied air reconnaissance would give warning of the approach of enemy surface units.* CTG 62.6 stated further that:

(a) If both the AUSTRALIA and VINCENNES Groups were ordered to attack the enemy, it was his intention that the VINCENNES Group should act independently of the movements of the AUSTRALIA Group so as to give greatest mutual support.

(b) It was his intention that the destroyers of each group engaged should concentrate under their particular senior officer (COMDESRON Four or COMDESDIV Seven) and attack the enemy with torpedoes and gunfire as soon as the enemy was being effectively engaged by Allied cruisers.

(c) If ordered to form a striking force, all destroyers of DESRON Four, less the BLUE and RALPH TALBOT, should concentrate under COMDESRON Four, in SELFRIDE, five miles northwest of Savo Island. In the event of contact with enemy surface units, this striking force would at once attack with full outfit of torpedoes and then maintain touch from the westward. The striking force would engage the enemy in gun action when own cruisers


*CTG 62.6 Special Instructions to Screening Groups and Vessels Temporarily Assigned, August 1942.

- 56 -

engaged, provided it was quite clear that own forces were not in the line of fire, and the destroyers must be prepared to illuminate the enemy targets for own cruiser gunfire.*

These instructions were written with the conviction that the enemy would be located by aircraft at such distance that the destroyers would have time to concentrate into a squadron organization and to rendezvous five miles northwest of Savo Island. No provision can be found in CTG 62.6's instructions for night action in the event of a surprise raid by enemy surface ships detected only after they had gotten within gun range. This was the actual situation which resulted in the Battle of Savo Island.

Each screening group commander operated independently of the other group commanders. Commander VINCENNES Group notified the other group commanders of his planned operations, but he was not in turn advised by them as to their planned operations.**

CTG 62.6 stated that the consideration that the cruisers VINCENNES, QUINCY and ASTORIA had not operated under his command before they joined him just prior to the rehearsals at Koro Island, July 28th, led him to decide to employ them as a separate tactical group. He stated that he had never had an opportunity to confer with, or even meet, the Commanding Officers of the above ships nor to issue them the standing instructions which he had issued his own task force.*** It seems somewhat dubious that no opportunity had presented itself to issue these instructions or to exercise tactically for drill purposes the forces he would have under his command at Tulagi.**** Could not the instructions have been delivered by destroyer, with explanations by both visual and voice means? It appears that ample time was available for appropriate tactical exercises while in the vicinity of Koro Island and while enroute to the objective.

(3) Discussion of Night Disposition


*CTG 62.6 Special Instructions to Screening Groups and Vessels Temporarily Assigned, August 1942.
**Personal Interview by Captain Frederick L. Reifkohl, USN, Commanding Officer, VINCENNES, recorded January 26th, 1945 by the Chief of Naval Operations, Office of Naval Records and Library.
***Report of Admiral Arthur J. Hepburn, USN (Ret) to CINCPAC, May 13th, 1943 on Informal Inquiry into the Circumstances Attending the Loss of the VINCENNES, etc. on August 9th, 1942, para. 92, page 41.
****Memorandum February 10th, 1943 by Comdr. H. B. Heneberger, USN, QUINCY to Admiral Arthur J. Hepburn, USN (Ret) relative to "Additional Information in regard to the ex-U.S.S. QUINCY" at the Battle of Savo Island, August 9th, 1942, para. 4, page 2.

- 57 -

There is considerable doubt as to the soundness of the above disposition and plan for the following reasons:

(a) The radar and anti-submarine screen consisting only of the BLUE and RALPH TALBOT was entirely inadequate. The two destroyers could have been as far as twenty miles apart when at the opposite ends of their patrol lines. They could have been as close together as six and one-half miles at the nearest points of their patrols. This was possible because the timing of their course changes was not coordinated so as to maintain a uniform distance between them. Actually at 0110, August 8th, when the Japanese Cruiser Force passed between them on its way into Iron Bottom Sound, they were about fourteen miles apart.

The two destroyer pickets were equipped with S C radars which gave a reliable range of from four to ten miles.** It was thought likely at the time that a destroyer-type ship could not approach closer to a radar ­equipped ship than eight or nine miles without being detected.** Unfortunately, the effectiveness of the radar equipment of the BLUE and RALPH TALBOT on their assigned stations near Savo Island was adversely affected by the closeness of land, and this situation necessitated the usual standard of alertness on the part of lookouts.

The anti-submarine detection range of the sonic devices of the two destroyers varied from a few hundred to about two thousand yards, depending upon the sonic characteristics of the water at the time.

It would have been wiser had at least two additional destroyers been assigned to the outer screen so that their stations would have been no more than five miles apart. In that case, any two adjacent destroyers would have been no further apart than ten miles at any time, which distance was considered within reliable radar limits. As a further refinement, had their patrols been properly coordinated, they would always have remained approximately five miles apart. Despite the threat of submarine attack on the transports, against which CTF 62 took positive defensive measures, two additional destroyers could well have been spared from the inner anti-submarine screens in the transport areas without having seriously reduced their effectiveness.

The radius from the cruiser screening groups on which the radar picket destroyers operated was insufficient. The RALPH TALBOT was not sufficiently distant from the VINCENNES Group to give adequate warning of enemy approach. With the RALPH TALBOT at the easternmost extremity of its patrol line simultaneously with the VINCENNES Group at the northern point of its square, there was only five miles distance between them. Five miles was no more


*A similar conclusion was arrived at by COMINCH in his Battle Experience Bulletin No.2 on Solomon Islands Actions August and September, 1942.
**Radar Bulletin No. 1, The Tactical Use of Radar, United States Fleet Headquarters of the Commander-in-Chief, March 1942.

- 58 -

than the limit of night visibility in the existent weather conditions. Had the optimum of radar effectiveness been obtainable, the RALPH TALBOT might possibly have given a warning about fifteen minutes before an enemy could close to within gun range of the VINCENNES Group from the northward; but "land interference" denied this degree of radar effectiveness. This analysis is equally applicable to the BLUE's radius of patrol from the AUSTRALIA Group but to a considerably less degree, since the patrol of the BLUE was approximately normal to that of the AUSTRALIA Group. COMSOPAC commented on this formation as follows:* "The orders to the Radar Guard were faulty in requiring them to 'shadow' an enemy force and report them frequently. Time and space did not permit the employment of tactics of this nature. A high speed enemy force would have arrived dangerously close to the objective before our destroyers could have instituted tracking or 'shadowing' tactics. The implied restriction on, and lack of definite instructions covering the use of searchlights by the Radar Guard was unfortunate. In the restricted waters in which the Screening Group was stationed, the underlying concept of instructions issued should have been to reveal the presence of enemy vessels in the quickest most positive manner, and this called for the use of searchlights for that purpose, if contact were made."

(b) The disposition of the heavy cruisers of the Amphibious Force in two main screening groups - the AUSTRALIA Group and the VINCENNES Group - to defend the approaches to the transport areas precluded the proper concentration of their total strength against Japanese forces threatening from the west. In addition to the lack of coordination of the planned operations of these two groups, as already pointed out, there was but little exchange of information between them. It was possible for these two groups to be as much as seventeen miles apart when the VINCENNES Group was at the northern extremity of its square and the AUSTRALIA Group was at the eastern extremity of its patrol line. It was also possible, because of unknown currents, for these two groups to be so close together when the VINCENNES Group was at the southern extremity of its square and the AUSTRALIA Group was near the western end of its patrol line as to seriously interfere with one another.

Commander VINCENNES Group had objected to this disposition by dispatch to CTG 62.6 pointing out the defect whereby the two groups could become foul of one another. He informed CTG 62.6 of his plan of patrolling clock­wise so that when the two groups were near each other he would always be heading out on course 315°(T) in order to reduce this danger to a minimum.**


*Report of Admiral Arthur J. Hepburn, USN (Ret) to CINCPAC May 13th, 1943 on Informal Inquiry into the Circumstances Attending the Loss of the VINCENNES, etc. on August 9th, 1942, para. 80(6), page 35.
**Personal Interview by Captain Frederick L. Reifkohl, USN, Commanding Officer, VINCENNES, recorded January 26th, 1945, by the Chief of Naval Operations, Office of Naval Records and Library.

- 59 -

At the same time he requested CTG 62.6 to give him the plan of operation of the AUSTRALIA Group for the night. CTG 62.6 failed to do so, and the result was that Commander VINCENNES Group never knew for one moment where the AUSTRALIA Group was during the night.* This objection of the Commander VINCENNES Group to the night disposition ordered provides a clue as to the reason why the Commander VINCENNES Group, as well as the Commanding Officers of the ASTORIA and QUINCY, felt that the Japanese ships which later attacked them were probably friendly.

In establishing this night disposition, CTG 62.6 had created a situation in which his forces were divided into two equal screening groups composed of three heavy cruisers and two destroyers. Each group possessed less fighting strength than the reported Japanese cruiser concentration of four heavy cruisers and three light cruisers in the Bismarck Sea area. By this division of his force CTG 62.6 made possible the chances for a Japanese surface force of approximately equal strength to concentrate its full force on each of his screening groups and knock them out in succession. It cannot be determined why he deemed it necessary or desirable to divide his forces. It would be interesting to discover how he expected to obtain mutual support between these two groups, since he had made it almost impossible in an emergency to concentrate his forces into one large group, should he have the occasion to do so.

Would it not have been wiser for CTG 62.6 to have combined the strength of his six heavy cruisers in one formation? For example, one of several possible dispositions could have been a column of six cruisers operating on a patrol line about five miles southeast of Savo Island normal to the bearing line of 125°(T) from the center of that island, screened against submarines by two destroyers in the van and two in the rear. Changes of course could have been either by column movement or by simultaneous turn. The resulting formation would have provided the following strength factors:

(a) A probable gunfire advantage with the possible crossing of the "Tee".
(b) A possibility of avoiding enemy torpedo attack by a simple ships turn of approximately 90° to the right or left to comb the enemy torpedo tracks by paralleling their most probable tracks.
(c) A reduction of confusion which might have resulted from a more complex formation in which the Commanding Officers of the American and Australian ships were maneuvered together for the first time.


*Personal Interview by Captain Frederick L. Reifkohl, USN, Commanding Officer, VINCENNES, recorded January 26th, 1945, by the Chief of Naval Operations, Office of Naval Records and Library.

- 60 -

(d) In view of the interior lines a maximum concentration in minimum time might be brought against any enemy surface forces attacking from the west.

Since but one flag officer was assigned to the western screening groups, this formation would have placed all ships under a flag officer and would have allowed the commanding officers to give their full attention to fighting their respective ships, which in a night action is of transcending importance.*

(c) OPERATIONS OF CTF 61 (Commander Expeditionary Force)

Commander Expeditionary Force, CTF 61 who was responsible for the tactical operations incident to the execution of Task ONE, remained in his flagship SARATOGA throughout the operations. While so doing he served as CTG 61.1.1 under CTG 61.1 who functioned as Officer-in-Tactical Command of the Air Support Force throughout August 7th. Accordingly, he took a more active part in the operations of the Air Support Force than he did in the operations of the Amphibious Force. So far as he knew the Air Support Force had not been located by the Japanese, and he therefore restricted radio communications to the voice transmissions necessary for the conduct of the carrier-based aircraft.

Since his position in the SARATOGA was far removed from the landing areas, he allowed CTG 61.2 (CTF 62) to conduct the amphibious operations according to plan, and to exercise individual initiative to meet the changing situation as it developed in Iron Bottom Sound and ashore with the landing force. Perhaps if radio silence in the carriers had been of leas importance, he might have taken more positive control over the amphibious operations within Iron Bottom Sound. However, the fact that he did not do so shows that he felt that he had established a state of mutual understanding with CTF 62 and that within the limits of responsibility and resultant authority CTF 62 would act in accordance with his expressed desires. This is one of the fundamental requirements of command.

The communications for the air operations were handled by the Air Support Director Group in the transport MCCAWLEY at Guadalcanal, by which CTF 62 controlled the air strike groups reporting to him from TG 61.1. This Director Group had the voice call "ORANGE BASE". A standby Air Support Director Group was set up with radio communications in the transport NEVILLE at Tulagi, and a fighter director unit with voice call "BLACK BASE" was located in the CHICAGO to control the fighter cover tor the entire Amphib­Force. CTF 61 kept himself informed of the general conduct of the operations of his principal task groups by radio interception, voice radio,


*A similar disposition was later employed at the Battle of Surigao Strait in the Battle for Leyte Gulf October 25th, 1944.

- 61 -

visual signals or message drops. The carrier planes served to receive messages from CTF 62 via "ORANGE BASE" and to deliver them to their own ships or to CTG 61.1, who in turn relayed them by visual means to CTF 61. CTF 61 could communicate with CTF 62 in similar fashion.

By the means of radio intercept CTF 61 received CINCPAC's message* at about 1030 warning him of the seventeen fighters and eighteen bombers proceeding from Rabaul to attack his forces. Within fifteen minutes he received another warning from CINCPAC that Japanese submarines had been ordered to proceed to Tulagi to attack Allied forces.**

He became immediately aware of the attacks of carrier-type bombers over Guadalcanal a few minutes before 1500, and promptly sent a visual message to CTG 61.1 (Which the latter received at 1645) suggesting a morning air search on August 8th toward Rabaul, in view of the enemy carrier reported there.***

Within the next hour he re-estimated the situation presented by the possible additional threat of a Japanese carrier in the Solomon Sea, superimposed on the already existent threat of land-based aircraft. He became very much concerned over the safety of his own carriers, and decided to retire them more to the eastward. At 1600 he released a visual message containing orders to CTG 61.1 as follows: "In view of carrier dive bombers encountered this afternoon, revise night operations to arrive at launching position south of Cape Henslow in the morning. Make early search. Bombers last seen leaving Tulagi on westerly course."**** Within ten minutes CTF 61 (or CTG 61.1.1) had received CTG 61.1's orders for flight operations on August 8th, including a planned search toward Rabaul by WASP planes.***** At 1627 he had an acknowledgment from CTG 61.1 that the latter had already ordered the search.****** In this same message CTG 61.1 advised CTF 61 that his own information indicated that the dive bombers were probably land-based planes from Rabaul which had staged through Buka or Kieta. This information was well reasoned for the Japanese dive bombers had actually come from Rabaul, staging through the airfield at Buka Passage. After receipt of this evaluation, CTF 61 did not alter his orders to CTG 61.1 to operate the carriers farther to the eastward on August 8th.

The dawn launching position for August 8th south of Cape Henslow was not well chosen for the search toward Rabaul. It was 110 miles eastsoutheast of the dawn launching position on August 7th. The search conducted by ENTERPRISE planes on the morning of August 7th (shown on Diagram C) had not


*CINCPAC Dispatch 062325, August 1942, to CTF 61, 63, COMSOPAC.
**CINCPAC Dispatch 062336, August 1942, to CTF 61, 63, COMSOPAC.
***CTF 61 Visual Dispatch 070357, August 1942 to CTG 61.1.
****CTF 61 Visual Dispatch 070500, August 1942 to CTG 61.1.
*****CTG 61.1 Visual Dispatch 070510, August 1942 to WASP, SARATOGA, ENTERPRISE.
******CTG 61.1 Visual Dispatch 070527, August 1942 to CTF 61.

- 62 -

revealed the presence of a Japanese carrier within a 200 miles radius of the more western launching position or TG 61.1. If an enemy carrier were to be located to the westward or in the direction of Rabaul, should not the search at dawn, August 8th have reached even farther and should not its point of origin have been moved farther to the west rather than to the east? A dawn search from a position about 100 miles due west or Tulagi would have reached as far north as Kieta and well to the westward into the Solomon Sea, beyond the range of CTF 63's land-based search planes. Diagram D shows that the search from the position south of Cape Henslow would merely duplicate both in time and area, the land-based searches already planned by CTF 63.

CTF 61 had received no contact reports from the planes of TF 63 during August 7th and could therefore feel reasonably confident that there was no Japanese carrier in the areas searched by the land-based airplanes. He would not know for sure, however, until CTF 63's summary came through which was usually about midnight.

Why then did CTF 61 not choose the more western position? He had already lost a number or fighters to Japanese planes. The western position would place him within range or Japanese land-based attack planes and fighters from Rabaul. He appears to have considered that possible damage or loss of one or more of his carriers was not an acceptable risk, and chose not to face attacks from land-based aircraft. Perhaps he was affected by his experiences at the Battles of Coral Sea and Midway, wherein he had lost two carriers - the LEXINGTON and YORKTOWN - to the air attacks of Japanese carrier-based planes. This probably made him circumspect about the possibilities of further reduction in the already weakened strength of his carrier forces. But was it necessary to face such an attack? Could not he have moved into the western position during the night, launched his search planes at dawn, then retired along a Point Option line toward the southeast, and recovered his search planes in a more secure position?

The need for protecting TF 62 from air attacks became even more pressing with the possible presence of a Japanese carrier. It became of primary importance to CTF 61 to discover and to destroy this carrier. He solved this problem about as follows: His primary means of search for covering the operations in progress were the land and tender-based planes of CTF 63, which commander had orders to provide him with additional support on call. He was thoroughly familiar with the extent of coverage or CTF 63's search plan for August 8th. He knew that the land-based planes in Sectors II and IV (See Diagram D) were to be abreast of Tulagi at sunrise. Since he did not make any request to CTF 63 for modified or extended searches on August 8th, but instead ordered CTG 61.1 to make the dawn search for the Japanese carrier, it is presumed that he decided to augment the inadequate extent of the land-based searches with carrier-based searches. This decision was sound. It is unfortunate that he did not implement his decision with specific orders to CTG 61.1 as to the area to be searched. For how could he effectively cover the locus of a Japanese

- 63 -

carrier's most probable launching position, as well as the probable direction of approach of an enemy surface force, unless his carrier searches extended into new and vital areas of the Solomon Sea and New Georgia Sound ("the Slot")?

At about 1745 he intercepted CTF 62's request to CTF 63 for a search on August 8th of an additional sector between 290°(T) and 318°(T) from Malaita Island. He could see from this that CTF 62 had arrived at a somewhat similar analysis of the land-based searches and had confirmed his own reasoning that they were inadequate. He noted however that CTF 62 had arrived at a different solution in that he desired to cover the blank sea area to the northwest of Tulagi. This request of CTF 62 undoubtedly met with his reasoned and silent approval.

At 2130 he received CTF 62's summary of the situation in Iron Bottom Sound and of the results of the day's operations. In this summary which was addressed to COMSOPACFOR but which was intended also for himself by intercept, he noted that CTF 62 (a) requested maximum fighter cover and two scout bomber squadrons continuously over Iron Bottom Sound on August 8th, (b) requested scouting against the approach of enemy forces from the westward* which confirmed his own previous decision, and (c) reported operations on shore proceeding satisfactorily with little opposition at Guadalcanal but with the capture of Tulagi, Gavutu and Tanambogo proceeding slowly but surely.

He weighed the information thus given, and decided his search plan and operating area for the carriers on August 8th were correct and that he would abide by the decision he had already made. It is unlucky that he did so, for hindsight shows that a search made from the western position, arrived at in this analysis, would have discovered the Japanese Cruiser Force off Bougainville Island at about 0830, August 8th.

(d) OPERATIONS OF CTG 61.1 (Commander Air Support Force)

During the day's operations, CTG 61.1 operated his force about sixty to eighty miles south of Tulagi. He was generally on a southeasterly course in order to head into the prevailing wind. Diagram "C" shows his mean track on August 7th. He had chosen his dawn launching position, Point VICTOR, because his tight flight schedule necessitated that he be near enough to his objective to insure that the air support was carried out adequately and yet far enough away to insure that his carriers obtained reasonable security from land-based air attack during these operations.

The weather conditions in the vicinity of TG 61.1 were: sky overcast, occasional squalls, sea rough, wind at dawn 18 knots from 125°(T) reducing later in the day to 5 knots, visibility poor with no horizon at dawn, flying conditions poor.


*CTG 61.2 (CTF 62) Dispatch 071030, August 1942, to COMSOPACFOR.

- 64 -

He provided air support and reconnaissance for the landing operations from all three carriers from dawn until darkness. This air support consisted in general of dive bomber attacks and fighter strafing runs on enemy positions. The same planes provided reconnaissance for the ground forces and for the location of new targets for themselves. The sixteen fighters of the WASP air support group at Tulagi at dawn destroyed seven Type-97 flying boats and seven Zero-type seaplane fighters on the water before they could take off. They also destroyed all the other Japanese seaplanes of the YOKOHAMA Air Group on the beaches and in the bushes, bringing the total to a claimed fifteen reconnaissance and nine fighters.*

Throughout the seventh, CTG 61.1 provided combat air patrol over both the transport groups and the carrier force. The patrol over the transport groups consisted of a total of sixteen fighters, continuously. These fighters were in addition to those of the air support groups over Tulagi and Guadalcanal mentioned above. The patrol over the carrier force consisted of a minimum of thirteen fighters at dawn. He increased this patrol to nineteen fighters at about 0800 after the initial dawn assault, and to twenty-four fighters at about 1100. He also maintained an anti-submarine patrol around TG 61.1 employing dive bombers as necessary in a manner similar to that employed by the Allied task force in the Battle of the Coral Sea.

He provided air search and special reconnaissance as follows:

(a) At 0600 he launched a flight of seven torpedo planes of VT Squadron EIGHT from the SARATOGA to search the northeast coast of Florida Island and to reconnoiter Marmasike Estuary. This flight succeeded in strafing Port Purvis and in bombing a village at Langa Langa Harbor, Malaita Island. Bad weather obstructed its attempt to reconnoiter Marmasike Estuary. It returned to the SARATOGA at 0942.

(b) At 0620 he launched three torpedo planes, and at 0707 five more torpedo planes from the ENTERPRISE to search a sector between bearings 270°and 030°(T) to a radius of 200 miles. The delay in launching the last five planes was occasioned by the mechanics of carrier operation. The point of origin of this search was Lat. 09°-44'S., Long. 159°-16'E., which bore 232°(T) distant seventy miles from Tulagi. He appears to have ordered this search purely for the security of the Air Support Force. The search made a contact on a small Japanese ship about seventy-five miles northnorthwest of the point of origin. The pilot of the contacting plane reported this small ship as an oiler, but it is not impossible that it was the schooner which was frequently sighted by various ships of both sides during the Battle of Savo Island and which at that time was repeatedly


*Action Report, WASP, August 14th, 1942, "Capture of the Tulagi-­Guadalcanal Area, August 7th-8th, 1942".

- 65 -

mistaken for a destroyer. These search planes returned on board the ENTERPRISE at 1039 and 1141, respectively.

(c) At 1049 he launched a second flight of eight torpedo planes of VT Squadron EIGHT from the SARATOGA. This flight succeeded in searching Marmasike Estuary, found it clear, and so reported to the MACKINAC as that seaplane tender was approaching to establish a seadrome in the Estuary. This flight returned to the SARATOGA at 1430.

At about 1030, CTG 61.1 received CINCPAC's warning of an impending air attack by seventeen fighters and eighteen bombers of the FIFTH Air Attack Force from Rabaul.* He immediately increased the combat air patrol over the carriers from nineteen to twenty-four fighters, and that over the transports to twenty-four fighters. When the Japanese aircraft attacked at 1325, the Allied fighters prevented this attack group from delivering any damage to the ships of TF 62,** and shot down three bombers and two fighters and damaged nineteen additional Japanese bombers and two fighters.***

At about 1500, CTG 61.1 received word from the Air Support Control Group (ORANGE BASE) that the fighter patrol was engaging Japanese carrier-­type bombers. Of the total of sixteen Japanese dive bombers that attacked, the Allied fighters accounted for six planes shot down and three damaged.

The presence of carrier-type bombers over TF 62 naturally posed the question to CTG 61.1 as to whether or not a Japanese carrier were present in the area. If it were, his own carrier force contained the primary means of dealing with it. He noted that the Japanese dive bombers had last been seen retiring to the westward of Savo Island. He also noted that there had as yet been no contact reports from the search planes of CTF 63. It was therefore probable that the enemy carrier, if present in the area, would be operating in the area searched by COMSOWESPAC planes. While he was re-­estimating the situation at 1555, he received a visual message from CTG 61.1.2 in the ENTERPRISE which stated: "Presence dive bombers indicates possibility enemy CV this vicinity. Suggest search."**** The same suggestion had been made also by CTF 61 in the SARATOGA;***** but this latter visual message was not received by CTG 61.1 until 1645, some thirty-five minutes after he had already promulgated his operation orders for the next day, August 8th.****** These orders included a morning search by the WASP and an afternoon search by the ENTERPRISE toward Rabaul primarily for a reported Japanese carrier.******


*CINCPAC Dispatch 062325, August 1942, to CTF 61, CTF 63, and COMSOPACFOR.
**War Diary, CTF 62, August 1942.
***Strength and Disposition 25th Air Flotilla, August 7th, 1942 CIG Document 74629 May 12th, 1947.
****CTG 61.1.2 (CTF 16) Visual Dispatch 070414, August 1942 to CTG 61.1.
*****CTF 61 Visual Dispatch 070357, August 1942 to CTG 61.1.
******CTG 61.1 Visual Dispatch 070510, August 1942 to SARATOGA, WASP and ENTERPRISE.

- 66 -

How he evaluated this situation was revealed in his message sent to CTF 61 at 1627 in which he stated: "My information dive bombers probably land-based from Rabaul via Buka or Kieta." He was correct. His order to search for a possible carrier, even though he considered the probability that there was none in the area, indicates that he was using sound judgment to cover enemy capabilities.

At 1705 he received an order from CTF 61 to revise his night operations to arrive at a dawn launching position on August 8th south of Cape Henslow.* He had previously established Point ROGER, located twenty miles south of Cape Henslow, as the position where the carrier task groups would rendezvous at sunset on the completion of flight operations on August 7th. He had already promulgated his plan for the night to retire to the westward to pass at 2330 through Point EASY, located at 100 miles southsouthwest of Tulagi. These points are shown on Diagram "c".

At 1711, he modified his night operations as follows: to cease the westerly retirement at 2330; to turn south for an hour and then to return to the eastward arriving at Point ROGER at 0600 August 8th; to commence the dawn launching at that time.**

The plot shows that Point ROGER, the dawn launching position he would use for August 8th was 110 miles farther east than Point VICTOR, the dawn launching position he had used on the previous day. He certainly realized at this time therefore that the searches he had ordered for August 8th would now fall far short of reaching into new areas. The possibility of locating a Japanese carrier to the westward was thereby considerably reduced. It was apparent that his searches would now serve no further purpose than to provide security for his own carrier force. Since he did not make a recommendation thereafter to CTF 61 to the contrary, it is presumed that he considered this to be acceptable.

His operations throughout the day cost him aircraft as follows: SARATOGA: Five fighters missing in action, one fighter and one dive bomber crashed; WASP: One fighter, one dive bomber missing in action, two fighters crashed; ENTERPRISE: Four fighters missing in action and two crashed.

His total aircraft losses August 7th were therefore:

Fighters   10 MIA (Missing in Action)
    5 crashed
Dive Bombers   1 MIA
    1 crashed


*CTF 61 Visual Dispatch 070500, August 1942, to CTG 61.1.
**CTG 61.1 Visual Dispatch 070611, August 1942 to TG 61.1.

- 67 -

At 2400, TG 61.1's position was bearing 165°(T), distant 125 miles from Tulagi as shown on Diagram "C".


(1) Operations of S-38

The Commanding Officer of the submarine S-38 on August 7th was patrolling off Cape St. George, New Ireland. At 1017 he noted that the traffic of the Japanese line of communication was about eight miles off the New Ireland coast. He therefore shifted his patrol station to that area* for the remainder of the day. At 1930 (about sunset) he set course 320°(T); and seven minutes later sighted what he thought were two destroyers, one bearing five degrees to the left and the other fifteen degrees to the right of dead ahead.* He immediately submerged and then turned right to assume the course 050°(T), normal to his patrol line of 320°(T).* Five minutes later (at 1942) the two "destroyers" (possibly the light cruisers TENRYU and YUBARI which were disposed on either bow of the CHOKAI) passed very close aboard, ahead and astern of the S-38.* About three minutes later, at 1945, at least three larger unidentified ships passed directly ahead of the S-38, one of which was so close that its wash caused the submarine to roll seven degrees.* This formation was traveling at a "very high speed", and all ships were too close to the S-38 for her to fire torpedoes.*

At 2000, as the last of these ships had passed in column, the Commanding Officer, S-38 changed course to 140°(T) to trail and track them.* At 2010, all sound had faded.*

The Commanding Officer, S-38 did not know the identity or types of these ships. He had a fair idea of the cruising formation, the course and speed; but he had mistaken light cruisers for destroyers and had counted only three heavier ships when in fact there were five. He considered this contact of sufficient importance that he surfaced and reported it by radio to COMSOWESPACFOR. The latter broadcast it at 0730, August 8th, and it was received by the responsible commanders in the SOPAC Area at 0738, August 8th** - eleven hours and thirty-eight minutes after the contact was made, but in plenty of time for consideration and evaluation by those commanders.

At 2037, the Commanding Officer, S-38 steadied on course 180°(T)*; and at 2222 he sighted a large submarine on bearing 250°(T), distant about five miles, heading on a southeasterly course at high speed.* He was unable to close the enemy submarine for attack, because the latter passed ahead rapidly and disappeared to the south.* This contact was reported and reached CTF 62 at 0700 August 8th, who in turn alerted his force to


*War Diary, S-38, August 1942.
**COMSOWESPAC Dispatch 071930, August 1942.

- 68 -


AS OF 2400, AUGUST 7, 1942

BASE 11th
VP Sqdns
11, 14 & 23
Marine Corps
                VMF 111
& 212
VMO 251
  B-17 B-26 P-39 PBY-5s Hud-
ESPIRITU     3*        4**    15 
EFATE ***5            12    22 
****9  10  38      *****16    84 
NANDI ***12  12  17  12        71 
SAMOA               18  17  10  45 
TONGATABU     24              30
TOTALS 31  22  79  28  30 50  17  25  262 


*2 PBY-5's of VP-23 transferred forward from EFATE.
**4 F4F-3a's of VMF-212 rotated forward from EFATE.
***Rotated up from EFATE and NANDI to ESPIRITU SANTO.
****1 B-17 from KOUMAC failed to return from search.
*****16 F4F-3P's of VMO-251 transferred from TONTOUTA, NEW CALEDONIA on August 7th.
******9 PBY-5's of VP-23 transferred after completing search from ESPIRITU SANTO, on August 7th.

[- 68a -]


[- 68b -]

the possibility that a submarine might enter the Tulagi-Guadalcanal Area on the eighth.* The S-38 had sighted the Japanese submarine I-121, en route from Rabaul to Tulagi to attack the Allied transports.**

(2) Operations of S-44

The S-44 was proceeding eastward along the southern coast of New Hanover Island en route to the entrance to Steffen Strait until about 0700.*** It arrived there too late to intercept the five Japanese cruisers which had sortied from Steffen Strait at 0650. It remained on this station throughout the day, guarding the exit of Byron Strait as well as Steffen Strait, but made no contacts.***

(f) OPERATIOMS OF CTF 63 (Commander Aircraft, South Pacific Force)

CTF 63 conducted air searches in support of SOPAC Operations, employing a plan somewhat similar to that flown the preceding day; but modified somewhat, as shown in Diagram "C". The principal change was that Sector IV was flown from Espiritu Santo rather than from Efate. In addition, the fan of searches flown from Espiritu Santo was realigned by minor changes in the limiting bearings of Sectors II, III, IV and v.

Sector I, which was not changed, was not searched on this day because of bad weather. However, a search was made from Koumac to the area to the southwest of Sector I, as a consequence of which one B-17 failed to return.

Sector II was to be searched by two B-17's. The search in the right half was reported negative; the search in the left half was unreported and the plane assigned to it failed to return. Both of these plans were directed to be abreast of Tulagi at sunrise**** in accordance with CTF 6l's request for such action.***** However, it is possible that the missing plane failed to arrive there, since it was not heard from. CTF 61 desired these planes abreast of Tulagi at sunrise in order to provide increased security for the Allied Expeditionary Force, TF 62, which was to arrive in the Tulagi-Guadalcanal Area by dawn.

Sectors III, V and VI were searched, as shown on Diagram "C", with negative results.

Sector IV was also searched, as shown on Diagram "C", with negative results.****** The search planes in this sector, in addition to those in


*CTF 62 Dispatch 072000, August 1942, to TF 62.
**Japanese Report of Meritorious Action, SUBRONS 3 and 7 CIG Intelligence Report 74834 distributed May 12th, 1947.
***War Diary, S-44, August 1942.
****CTF 63 Dispatch 300820, July 1942 to CTF 61.
*****CTF 61 Dispatch 290857, July 1942 to CTF 63.
******War Diary, CTF 63, August 1942.

- 69 -

Sector II, were directed to be abreast Tulagi at sunrise, and presumably complied. These six planes in Sector IV, as well as the three search planes in Sector III, proceeded late in the afternoon to base at Maramasike Estuary rather than at Espiritu Santo on completion of their search mission.* The tender MACKINAC, which served as the base for these planes, had arrived there from Noumea shortly after 1400.**

Table 3 indicates the disposition of Allied land and tender-based aircraft in the SOPAC Area as of 2400, August 7th.


Commander Allied Air Forces North Eastern Area, SOWESPAC, conducted air searches in support of the Allied invasion of the Tulagi-Guadalcanal Area on August 7th as follows:***

(a) Five Hudsons from the Fall River Field at Milne Bay reconnoitered the Solomon Islands in Reconnaissance Area "B".

(b) One B-17 from Port Moresby searched Reconnaissance Area "C"; another reconnoitered Reconnaissance Area "D"; and a third B-17 departed on a photographic intelligence mission in Reconnaissance Area "E"; but turned back just beyond the Lusancay Islands in the Solomon Sea, probably because of mechanical difficulties, and did not complete this mission.

None of the above reconnaissance revealed any significant movement of Japanese forces.

Commander Allied Air Forces North Eastern Area also conducted routine bombing attacks on the air fields at Rabaul. The first striking group, composed of thirteen B-17's from Port Moresby, sighted a Japanese cruiser force about twenty-five miles north of Rabaul. This contact was reported as one heavy cruiser, three light cruisers and one destroyer in position Latitude 03°-45'S, Longitude 151°-56'E, on course 270°(T) at speed thirty knots.*** These Japanese ships were in fact, the CHOKAI and the four cruisers of CRUDIV SIX. The B-17's did not attack the cruisers, but concentrated on their assigned objective, Vunakanau airdrome at Rabaul.

This act indicates that Japanese naval vessels were not primary targets for the bombers of the SOWESPAC command at this time. The matter had been thoroughly discussed in conference by COMSOWESPAC and COMSOPAC. They had agreed that attacks on naval targets, with the possible exception of aircraft carriers, would be made by SOWESPAC aircraft only upon specific request from COMSOPAC.


*War Diary, CURTISS, August 1942.
**War Diary, CTF 63, August 1942.
***Operations Report, Allied Air Forces, SOWESPAC Area, August 7th, 1942.

- 70 -

The report of the contact on the Japanese cruisers at 1231 was not broadcast by COMSOWSPAC until 2319,* and was not received by COMSOPAC until 2400. This delay of eleven hours and twenty-nine minutes denied COMSOPAC the opportunity of requesting an immediate air attack, should he have desired it.

On the afternoon of August 7th, the Allied Air Forces, SOWESPAC, also sighted six unidentified ships in St. George's Channel, heading out on a southeasterly course.** The report of this contact was made in the same dispatch* that reported the Japanese Cruiser Force north of Rabaul.


*COMSOWESPAC Dispatch 071219, August 1942, to CINCPAC, all CTF's Pacific and all CTF's SOWESPAC.
**Operations Report, Allied Air Forces, SOWESPAC Area, August 7th, 1942.

- 71 -



0000 August 8th to 2400 August 8th


Commander Cruiser Force continued onward after 0000, August 8th along the track 080°(T) as shown on Diagram "D". He realized that during the 8th he would be called upon to make some important decisions for which he required additional intelligence. Among these decisions was the vital one whether or not to make the planned night attack. There is no documentary evidence to indicate that Commander Cruiser Force at any time entertained doubts as to his chances for ultimate success in the forth­coming night action. However, he apparently considered that the intelligence which he had received on August 7th from the Commander FIFTH Air Attack Force, although considerable, was not fully adequate for the purposes of his planned night action. He knew that he would receive more information from Commander FIFITH Air Attack Force on August 8th. In addition to this he could obtain information more directly applicable to his own planned operations by employing his own cruiser scouting planes.

To obtain first-hand intelligence it was necessary for him to close the enemy so that his ship-based planes might be within proper range to make suitable reconnaissance of Tulagi and to search for the Allied carrier forces. He determined to close the Tulagi area to a distance as would not seriously endanger his command to Allied counteraction by air, and as would be close enough to permit him to reach his objective under the cover of darkness to execute his attack that night with an expectancy of success. Mindful of the fact that he had been sighted the preceding day off Rabaul, he desired an operating area wherein his presence, if again discovered, would not necessarily divulge his particular course of action. He hoped to confuse any Allied reconnaissance planes as to his projected plans for a night action. He therefore decided to operate in an area due east of Bougainville throughout the morning of August 8th while his cruiser planes scouted. This area was three hundred miles from Tulagi and appeared to be well beyond the range of carrier aircraft.

At 0100, he passed the northern tip of Buka Island ten miles abeam to starboard and changed course to the southeastward to parallel the east coast of Bougainville Island and to arrive at his selected dawn launching position. At 0600, about a half-hour before sunrise, as Commander Cruiser Force neared his selected operating area, he directed that one plane from each of the heavy cruisers be catapulted at 0625 to search fanwise from the point of origin as follows: the CHOKAI plane 070°(T), the FURUTAKA plane 085°(T), the KINUGASA plane 100°(T); the KAKO plane 115°(T) and AOBA plane 130°(T). The latter plane was to reconnoiter the Tulagi area. Each of the above planes was to search for the enemy to a radius of 250 miles, fly a cross-leg to the left for

- 72 -

thirty miles, and return to the parent ship.* The point of origin for the search and for rendezvous on return, was Lat. 05°-34'S., Long. 156°-20'E.

At about 0625, just a few minutes before sunrise which occurred at 0632, each of the heavy cruisers except the FURUTAKA launched one plane as directed. The latter plane was cancelled, probably because the extensive atoll and reefs of Ongtong Java on bearing 085°(T) obstructed the approach of any enemy surface force through that sector and made a search of it unnecessary. The weather at this time was satisfactory for flying and the visibility, except for a partly cloudy sky, was excellent in the vicinity of the Cruiser Force.

After launching planes, the Cruiser Force operated in two groups. The heavy cruisers operated together in a somewhat dispersed manner to facilitate the recovery of aircraft. The light cruisers and the destroyer YUNAGI remained within visual signal distance of the heavy cruisers. There were no destroyers to screen the heavy cruisers against submarines, so Commander Cruiser Force employed an inner air patrol for this purpose, flown by the cruiser planes that were not used on the search and reconnaissance mission. These tactics were employed presumably to assist in deceiving any Allied plane or submarine as to his intention, and to allow complete freedom of action for individual ships to avoid submarines located by the inner air patrol. Certainly a number of single ships maneuvering independently afforded much smaller and less vulnerable targets, and the multiplicity of their wakes presented a much more confused underwater sound condition to submarines than would the entire formation.

At 1020 one or more ships of CRUDIV SIX, while operating in the vicinity of Lat. 05°-56'S., Long. 156°-30'E., sighted a single Lockheed bomber;** and six minutes later the CHOKAI sighted this plane on bearing 116°(T).***

At 1030 the KAKO received information that another Lockheed bomber was about forty-five minutes away.* The source of the information was possibly the AOBA's plane which might have observed the R.A.A.F. Hudson which was then to the northeast in the vicinity of Tauu Islands.

At 1036 the CHOKAI lost contact with the Hudson that had been shadowing her. This plane, which was operating from Fall River, had not remained to trail the Japanese ships but instead had continued on towards its base. In view of this contact, the ships of CRUDIV SIX changed course


*War Diary KAKO, August 7th-10th 1942, "Solomons Sea Battle", WDC Document 160143.
**CRUDIV 6 Detailed Battle Report #8, Solomons Naval Action, August 7th­10th 1942, CIG Document 86927, June 27th, 1947.
***War Diary 8th Fleet, August 7th-10th 1942, CIG Document 74633, May 12th, 1947.

- 73 -

from time to time in an endeavor to confuse the enemy plane believing that they were still being trailed.

Between 1036 and 1050 every cruiser except the AOBA recovered her planes.* The AOBA plane, which had gone to the Tulagi-Guadalcanal Area did not return until later because of the greater flight distance involved in its mission.

After all planes except the AOBA's had been recovered, Commander Cruiser Force directed the cruisers to return on his flagship. As the ships were closing at about 1100, the CHOKAI observed an enemy plane bearing 330°(T),** and at 1106 COMCRUDIV EIGHTEEN observed the same plane, which he identified as a B-25*** but which was actually a second Australian Hudson operating out of Fall River. Commander Cruiser Force seems to have been convinced that his command had been shadowed by the same plane since 1020.

Some few minutes later, the CHOKAI observed that the trailing plane was closing, apparently to obtain more information. At 1110 the CHOKAI opened fire and succeeded in discouraging the Hudson, for it retired and disappeared from view at 1113.**

Sometime around 1200 the AOBA plane which had been reconnoitering the Tulagi-Guadalcanal area was recovered and the entire formation was reformed. The pilot of the AOBA plane gave a most detailed account of the Allied activities in the Tulagi-Guadalcanal area. He reported:

(a) In the Lunga Point area, one battleship, four cruisers and seven destroyers, one ship resembling an escort carrier, and fifteen transports. He also reported that the cruisers and destroyers were conducting a roving patrol and that the transports were at anchor.

(b) In the Tulagi area, two heavy cruisers, twelve destroyers and three transports all underway in the vicinity of the Tulagi light­house.****

The above intelligence was surprisingly accurate. There were, of course, neither the battleship nor the escort carrier present. There


*War Diary KAKO, August 7th-10th 1942, "Solomons Sea Battle", WDC Document 160143.
**War Diary 8th Fleet, August 7th-10th, 1942, CIG Document 74633, May 12, 1947.
***Records CRUDIV 18, August 8th, 1942, WDC Document 160984.
****CRUDIV 6 Detailed Battle Report #8, Solomons Naval Action, August 7th-10th, 1942, CIG Document 86927, June 27th, 1947.

- 74 -

were in fact eight cruisers, rather than six: twenty-four destroyers, destroyer transports and minesweepers, rather than nineteen; and nineteen large transports or supply ships rather than eighteen.

Commander Cruiser Force now re-estimated the situation. He had information making it apparent that the strength of the Allied surface forces in Iron Bottom Sound was greater than that of his own force if concentrated against him, but their division between Tulagi and Guadalcanal was a factor in his favor. He had just been located, and he presumed that the Allied reconnaissance plane had immediately reported his position. He hoped that his deceptive tactics, principally of heading northwest while the Allied plane was in the vicinity, would confuse the Allied commanders on receipt of the contact report as to his intentions. He knew the FIFTH Air Attack Force would deliver a strong air attack against the Allied forces at about noon. He could expect a summary of the results of this attack and of the land-based reconnaissance for the day at about 1530, since that was the time he had received it on the 7th. He decided to carry out his planned night attack, and at 1300 headed south at twenty-four knots to pass through Bougainville Strait at 1537.

His decision was bold; for if he had underestimated the intelligence or aggressiveness of the Allied carrier force commander, he could well have been placing his force in a position to be bombed by Allied carrier-­based aircraft before sunset. Actually Commander Cruiser Force had made a correct decision and, to use the words of COMCRUDIV SIX, "The Commander's decision to attack at night and his direction of the operation offered an opportunity of great success."*

At 1600, he had traversed Bougainville Strait and turned into "The Slot." He had not as yet received the air attack and reconnaissance summary from Commander FIFTH Air Attack Force. He had committed his force, and only two and a half hours remained before sunset. He desired more information as to the Allied strength and disposition late in the afternoon; and wished to ascertain the extent of damage done by the Japanese bombers on August 8th. He therefore for a second time, dispatched a cruiser scouting plane from the AOBA at 1612 to reconnoiter Tulagi. The pilot was briefed to cruise at two hundred and sixteen knots, to proceed on course 140°(T) to pass between Choiseul and Vella Lavella Islands, and thence to proceed to Tulagi. Commander Cruiser Force expected that this pilot would arrive over Tulagi by 1725, would obtain detailed reconnaissance information by 1750 and would return to the AOBA before the end of evening twilight at 1910.

At 1640 Commander Cruiser Force issued his Signal Order No. 25 which contained his instructions to his command for the night action. These instructions have been translated as follows:


*CRUDIV 6 Detailed Battle Report #8, Solomons naval Action, August 7th-10th, 1942, CIG Document 86927, June 27th. 1947.

- 75 -

"1. During the night the cruising disposition will consist of a main body and a vanguard. The main body will be composed of the CHOKAI, followed by CRUDIV SIX, with 1000 meters between ships. The vanguard will be stationed 3000 meters ahead of the CHOKAI; and will consist of the TENRYU and YUNAGI to port, and the YUBARI to starboard, separated by an interval of 6000 meters.

"2. Before the rush in (approach), if enemy small craft are encountered, the vanguard will check them strongly while the main body heads southward.

"3. At the time of the rush in (approach), all ships will form battle column. Order of ships: CHOKAI, CRUDIV SIX, TENRYU, YUBARI, and YUNAGI with a distance of 1200 meters between ships.

"4. In the approach the Cruiser Force will first pass south of Savo Island and will torpedo the main enemy force in the Guadalcanal anchorage; after which it will head towards the Tulagi anchorage to shell and torpedo the enemy. The Cruiser Force will then withdraw north of Savo Island. Each commanding officer will operate independently as regards guns and torpedo firings.

"5. As a means of recognition each ship will display white sleeves on each aide of the bridge. Each sleeve will be one meter in diameter and seven meters long.

"6. Speed will be twenty-four knots at the time of the rush in (approach)."*

This order was simple. It appears to have been based on gaining the factor of surprise, although surprise is not mentioned. Whether surprise were achieved or not, the attack was to be carried out. The objective of this attack was the destruction of any Allied forces encountered, but particularly the transports, as stated in Commander Cruiser Force's Estimate of the Situation made on the previous day wherein he planned "to strike the anchored enemy convoys at night and destroy them." COMCRUDIV EIGHTEEN also stated that the targets were the transports.***

Commander Cruiser Force had chosen a column formation for battle, led by his strongest ship, the CHOKAI, tapering off in strength with the heavy cruisers of CRUDIV SIX next in column, followed by the light cruisers of


*Records CRUDIV 18, August 7th-9th 1942,  WDC Document 160984.
**Commander 8th Fleet's Estimate of the Situation regarding American landings on Guadalcanal and Tulagi on August 7th, 1942, CIG Document 73845, May 7th, 1942.
***U.S. Strategic Bombing Survey Interrogation Nav. #61, Vol. I. Interrogation of Japanese Officials, Interrogation of Rear Admiral Matsuyama, IJN, page 255.

- 76 -

CRUDIV EIGHTEEN and the destroyer YUNAGI bringing up the rear. Japanese officials stated later that he employed this "single file" night battle formation because this occasion was the first that the CHOKAI, CRUDIV SIX and CRUDIV EIGHTEEN had joined forces and, up to this time, they had never engaged in maneuvers together.*

At 1654, the Cruiser Force turned to a southeasterly heading to pass between Santa Isabel and New Georgia Islands on course 120°(T), speed twenty-four knots. At this time Commander Cruiser Force directed that Condition TWO be set in all ships and that at l930 Condition ONE be set. What these Japanese conditions of readiness were is not known, but it is probable that they corresponded closely to the American Conditions TWO and ONE Easy. For, as will be shown later, the Japanese appear to have merely alerted the gun crews as battle appeared imminent.

At 1715 the seaplane carrier AKITSUSHIMA was sighted "hull down" in the vicinity of Gizo Island bearing about 140°(T).

At 1816 sunset occurred.

At 1840 Commander Cruiser Force exhorted his command, "Let us attack with certain victory in the traditional night attack of the Imperial Navy! May each one calmly do his utmost!"

At this time each ship properly disposed of all combustible materials, depth charges, and all light oil stored above the water line. All battle preparations were made, and course was set for Tulagi Strait.

At 1910 evening twilight ended.

At 1915 Commander Cruiser Force shot off signal flares to guide the AOBA plane back to the cruisers; but the plane failed to return, having been shot down over Tulagi.**

At 2100 Commander Cruiser Force received a dispatch from Commander FIFTH Air Attack Force, stating that his bombers had attacked the Allied Forces in the Tulagi-Guadalcanal Area about noon and had sunk two heavy cruisers, one large cruiser, two destroyers and nine transports; and had badly damaged one heavy cruiser and two transports, all three of which were left burning.*** This was most heartening news to Commander Cruiser


*GHQ, SCAP, Military Intelligence Section, General Staff, Allied Translator and Interpreter Section WDC Document 15685, March 28th, 1946, Night Engagement Track Chart for
Report of Sea Battle off Savo Island, August 8th, 1942.
**Track Chart #1 Annexed to CRUDIV 6 Detailed Battle Report #8, Solomons Naval Action August 7th-10th 1942, CIG Document 86927, August 27th, 1947.
***Records CRUDIV 18, August 8th, 1942, WDC Document 160984.

- 77 -

Force for his own reconnaissance plane was now two hours overdue and failed to return from Tulagi with the information he desired. It then became clear to him that his chances of making a successful night attack had been greatly increased. According to his best estimate the Allied force then numbered but one battleship, three cruisers (one of which was badly damaged) seventeen destroyers, and nine transports (two of which were burning).

The Japanese had placed great stress in their fleet training on night attack procedures, and all ships were expected to operate under night attack doctrines. One of the Japanese doctrines for dark nights was to dispatch from the attack force an illuminating unit, composed generally of two planes, to illuminate the targets when the attack force was within striking distance. Pursuant to this doctrine, Commander Cruiser Force dispatched two planes at 2313, one from the AOBA and one from the KAKO with instructions to lay a course marker to guide the cruiser force into the target area, to reconnoiter and report the disposition of Allied forces, and to illuminate the transports when the CHOKAI was twenty miles away.

At 2335, CRUDIV EIGHTEEN lighted on bearing 140°(T) a marker lamp dropped by one of the planes. This marker had been dropped with Cape Esperence bearing 140°(T), distance about thirty miles. This distance seems to have been chosen to insure that the marker was not seen by the Allied forces so that the factor of surprise might not thereby be lost. It is assumed that this marker was employed by the Japanese in their navigation, which navigation was very accurate indeed.

At 2342, Commander Cruiser Force observed a light on bearing 125°(T) in the Tulagi area. Three minutes later, at 2345, the KAKO also sighted fires of land installations in the direction of Tulagi; and finally at 2355, COMCRUDIV SIX reported a reflection of a large fire in the sky over Tulagi.

At 2400 the Cruiser Force was in Lat. 08°-43.5'S., Long. 159°-23.5'E. This position bore 320°(T), distant about thirty-seven miles from the southwesterly entrance to Iron Bottom Sound.


During the 8th, Commander FIFTH Air Attack Force was heavily engaged in attacking the Allied forces at Tulagi-Guadalcanal and in searching the most probable areas in the direction of Tulagi and beyond, primarily for the Allied Air Support Force (TF 61.1).

The search operations of this command were:*


*Japanese Search Plans August 7th-9th, 1942, CIG Document 74632, May 12th, 1947.

- 78 -

(a) At 0540 one large flying boat from the FOURTEENTH Air Group departed Rabaul on course 130°(T), flew to a distance of 700 miles, then searched sixty miles to the left, and finally returned to Rabaul where it landed at 1645 without having located TF 61.1 or any other Allied contacts. The subsector flown by this plane was that in which the Allied carrier force, TG 61.1, was operating on August 8th, as shown on Diagram "D". This plane reached its outer limit at 1003 and turned back at the end of sixty miles lateral leg at 1025. It should have passed within visual distance of the Allied carrier force at 1050.The weather conditions in the vicinity of TG 61.1 were partly cloudy .5, surface visibility unrestricted, wind ENE seventeen knots, scattered showers. Since the Japanese reports indicate their search planes flew between 6000 and 12,000 feet altitude,* it is assumed that the pilot flew above the clouds and failed to detect TG 61.1 between them.

(b) At 0636 one large flying boat from the YOKOHAMA Air Group proceeded from Rabaul on course 100°(T) to a distance of 600 miles, covered a sixty miles lateral distance to the left, and returned to Rabaul at 1600 without having made any contacts on the Allied forces.

(c) At 0700 two search groups departed Rabaul as follows:

(1) Three land attack planes from the FOURTH Air Group departed on courses 110°(T), 140°(T) and 120°(T) from Rabaul to search to a radius of 700 miles. At the outer end of their search all planes flew laterally sixty miles to the left, then returned to land at Rabaul at 1615. No contact was made with TF 61.1. The plane flying course 120°(T) reported that the Tulagi Area was covered with clouds and reconnaissance there was impossible. Confirmation of this weather condition is found in CTF 62's statement that throughout the day the sky in the Tulagi-Guadalcanal Area was partly cloudy becoming increasingly overcast during the night.** It is probable then that the pilot on the 120°(T) subsector had encountered heavy clouds that made his search ineffective. The plane on course 140°(T) turned left at 600 miles radius, possibly because of cloudy weather also.

(2) One land reconnaissance plane from the TAINAN Air Group departed Rabaul to reconnoiter the Allied forces in the Tulagi-Guadalcanal area. At 0915, when thirty miles to the northwest of Tulagi, it was pursued by three Allied fighters. It landed at Rabaul at 1400 without having observed any unusual occurrences. This time of landing at Rabaul indicates the pilot remained in the Tulagi area for about two hours. He


*Japanese Search Plans August 7th-9th, 1942, CIG Document 74632, May 12th, 1947.
**War Diary CTF 62, August 8th, 1942.

- 79 -

reported "enemy ships near Tulagi, several destroyers mostly anchored near the Guadalcanal Air Base, and about fifteen enemy fighters and bombers aloft."* This incomplete report provided little information of the Allied surface units, though it was more specific as to the Allied air cover. It is probable that the Japanese reconnaissance pilot had great difficulty in reconnoitering the area, and was forced repeatedly to take cover in the clouds.

(d) At 1445 one land reconnaissance plane left Rabaul to observe the condition of the Buka Island fighter landing strip which bad just been rushed to completion. This reconnaissance plane returned to Rabaul at 1716.

The above searches were designed by Commander FIFTH Air Attack Force to cover the entire Solomons sector and the sea areas to the eastward. This search plan covered the areas wherein the Allied carriers would probably be operating to provide adequate air cover for the landing operations at Tulagi-Guadalcanal. The searches, as laid out, should have been reasonably effective although a denser search would have been more effective in the existing weather conditions for locating the Allied carriers.

The failure of his pilots to locate the carriers must have been highly confusing to Commander FIFTH Air Attack Force. His pilots reportedly had searched the areas and discovered nothing and yet the Allied carriers were present somewhere because their fighters and bombers were seen over the target area. He determined to increase the area of his search on the following day.

The only air attack on August 8th was carried out as follows: At about 0740 Commander FIFTH Air Attack Force launched a torpedo attack group composed of approximately twenty-three land attack planes of both the MISAWA Air Group and the FOURTH Air Group to attack the Allied surface forces off Tulagi and Guadalcanal. These planes were escorted by fifteen fighters of the TAINAN Air Group. At the same time he launched another attack group, consisting of four carrier-type bombers escorted by about twenty-two fighters of both the TAINAN Air Group and the SECOND Air Group.** This attack group, although intercepted by three ENTERPRISE fighters from TG 61.1, struck its target at about 1200. As a result of the air attack, Commander FIFTH Air Attack Force reported sinking two heavy cruisers, one large cruiser, two destroyers and nine transports; and damaging and setting afire to one large cruiser and two transports. This report was greatly exaggerated. The actual damage delivered to the Allied ships was solely to one destroyer, the JARVIS torpedoed but afloat,


*Japanese Search Plane, August 7th-9th, 1942, CIG Document 74632, May 12th, 1947.
**Records 25th Air Flotilla, August 8th, 1942, WDC Document 16l730.

- 80 -

and to one transport, the GEORGE F. ELLIOTT hit and set afire by a torpedo bomber which crashed into it. The cost of this attack to the FIFTH Air Attack Force was the loss of eleven land-attack planes from the FOURTH Air Group and six from the MISAWA Air Group, all shot down; one plane made an emergency landing; and five planes were damaged. One fighter from the TAINAN Air Group also was shot down, one was missing and one was damaged. The total losses were nineteen planes, shot down or missing.* This Japanese report of losses checks reasonably well with the Allied claims of destroying at least fourteen Japanese planes.

At some time during the day, the MISAWA Air Group was reinforced with eight land attack planes. (See Table 1).

The operations of the FIFTH Air Attack Force on August 8th contributed little of direct value to the Japanese overall effort against the Allied forces at Tulagi and Guadalcanal, except to delay unloading operations of the transports. Of indirect value was the attrition of Allied fighter planes, since the loss of Allied fighters in opposing Japanese air attacks finally contributed in some measure to the reasons of CTF 61 for the retirement of TG 61.1 on August 9th.

The inaccuracy of the Japanese reports of damage inflicted by the FIFTH Air Attack Force could well have led Commander Cruiser Force into a trap, had he relied exclusively upon them. Fortunately for this Japanese commander, he employed the cruiser scouting planes of his own force for reconnaissance of the target area on three different occasions on August 8th and obtained more reliable information.


The exact movements of the Japanese submarines on August 8th are not available but can be discussed in a general way.

(a) The RO-33 was en route to Tulagi area from the Gulf of Papua, but was presumably several days from Tulagi.

(b) The RO-34 was on station off Port Moresby where it was reconnoitering Allied movements in that area.

(c) The I-121 was en route from Rabaul to Tulagi. While operating on the surface she was attacked at 1127 by an Australian Hudson which dropped two bombs, but both missed. Her 2400 position was approximately 250 miles due west of Tulagi.


*Strength and Disposition 25th Air Flotilla on August 7th, 1942, CIG Document 74629, May 12th, 1947.
**War Diary 4th Feet, August 1942, "Movements of Naval Forces Under This Command", WDC Document 160336.

- 81 -

(d) The I-122 was en route from Rabaul to Tulagi. She evidently departed Rabaul in the early evening, for at 2155 she was sighted passing through St. George's Channel on a southerly course at high speed by the Allied submarine S-38. Her 2400 position was bearing approximately 170°(T), distant thirty-five miles from Cape St. George.

(e) The I-123 was en route from Truk to Tulagi.

- 82 -



0000 August 8th to 2400 August 8th

(a) OPERATIONS OF CTF 62 (Commander Amphibious Force)

The Amphibious Force continued operations throughout the night in support of the landing forces ashore. By 0135 the Raider Battalion at Tulagi reported that it had suffered twenty-two per cent casualties and the Parachute Battalion fifty to sixty per cent casualties. Reinforcements were requested.* The unloading of Squadron XRAY was discontinued from 0230 to 0730 because of the congestion of stores on Guadalcanal Beach. The unloading of Squadron YOKE was delayed by the opposition ashore at Tulagi. The night passed without any form of interference with the Allied surface forces from the enemy.**

At 0500, the beginning of morning twilight, CTG 62.6 ordered the outer screening groups and units to return to the transport area and to resume the day screening operations. By sunrise, at 0632, the day screening dispositions had been assumed.

During this day the sea in Iron Bottom Sound was calm, there was a light breeze from the southeast, and the sky was partly cloudy becoming increasingly overcast toward night.

At 0710 CTF 62 received the report of the contact made by the S-38 of a large enemy submarine headed southeasterly at high speed in St. George's Channel at 2222, August 7th. He immediately notified TF 62 that an enemy submarine was reported near and might enter the area that day.*** CTG 62.6 in turn, ordered the destroyer minesweepers to form an anti­submarine patrol westward of the Sealark and Lengo Channels,** and established anti-submarine air patrols, employing a minimum of three cruiser planes in the air throughout the day.**

At 0738, CTF 62 received word from COMSOWESPAC that one destroyer and two unknown ships passed through Lat. 04°-58'S., Long. 152°-50'E. at 1100 the preceding day on course 140°(T) at high speed: and that, at 2000 the same day, two destroyers and three larger unknown ships also had passed through the same point on the same course at high speed.**** This latter


*War Diary CTF 62, August 8th, 1942.
**Remarks of CTG 62.6 to CTF 62, August 10th, 1942.
***CTF 62 Dispatch 072000, August 1942, to TF 62.
****COMSOWESPAC Dispatch 071930, August 1942, to all CTFs Pacific Fleet.

- 83 -

contact was the Japanese Cruiser Force en route to attack TF 62, though its composition, destination and objective could not be determined by CTF 62 from this report. He probably decided to await further reports from air reconnaissance during the day to determine whether Japanese combatant ships, or merely auxiliary ships under escort, were moving toward the Solomons.

At 0945, CTF 62 had delivered the reinforcements needed ashore at Tulagi, and the Second Marines began their landings there.*

At 1027, CTG 62.6 passed to CTF 62 the message he received from an Australian coast watcher on Bougainville Island, reporting forty heavy bombers proceeding southeast. Shortly thereafter CTF 62 ordered the transports to get underway. Both Squadrons XRAY and YOKE were formed independently and maneuvered between Guadalcanal and Florida Islands awaiting the expected air attacks.** All destroyers and minesweepers formed an anti-submarine screen for the transports, maneuvering extensively to maintain sea room for themselves while at the same time acting as anti-submarine and anti-aircraft screen.**

At 1200 TF 62 was attacked by about twenty-three torpedo planes, and by four dive bombers, escorted by fighters, which approached around the southeast end of Florida Island.* The transport GEORGE F. ELLIOTT, in Transport Group XRAY, was hit amidships and set afire by a torpedo bomber which crashed, Kamikaze style. The destroyer JARVIS was hit in the forward part of the ship by a torpedo,* which opened up a hole on her starboard side abreast gun Number Two from thirty to frame fifty-five.***

At 1207 CTF 62 notified his command to be alert for possibly one or more enemy submarines in the transport area.

At 1300 the JARVIS was towed by the DEWEY on the orders of CTG 62.6, into shoal water near Guadalcanal where she reported that, although her hull was badly ruptured, she was able to proceed under her own power.* At this time the transports were returning to their anchorages.

At 1355 CTF 62 received word that another flight of enemy planes was en route to the Guadalcanal area. He again directed that the transports remain underway and form the anti-aircraft disposition.* This air attack did not develop, so he finally directed the transports and cargo ships to return to their anchorages where they resumed unloading operations at about 1650.*•


*War Diary CTF 62, August 8th, 1942.
**Remarks of CTG 62.6 to CTF 62, August 10th, 1942.
***JARVIS Dispatch 080620, August 1942, Enclosure (G) to COMDESRON 4, Serial 005 of September 9th, 1942, COMDESRON 4 War Diary, August, 1942.

- 84 -

A little after 1500 he received word of the capture of Guadalcanal Air Base. This was gratifying news to him, since it was planned with the departure of the carriers, to provide air cover for his force by carrier fighters based ashore and operating from the captured Japanese airfield at Guadalcanal.

In the meantime, the destroyer HULL had been assisting the transport GEORGE F. ELLIOTT in fighting fire, pursuant to orders from CTG 62.6. The fires eventually reached the engineering spaces and gained in intensity, so that she had to be abandoned. At 1730, CTG 62 ordered the DEWEY to sink her. After the DEWEY had fired three torpedoes into her, she settled in shoal water off Florida Island and continued to burn, illuminating the overcast after dark.

At 1807, CTG 62 intercepted the dispatch wherein CTF 61 recommended to COMSOPACFOR the immediate withdrawal of the carriers.* Of course he had known all along that CTF 61 had planned to withdraw the carriers prior to August 10th, and the matter of air support for the forces in Iron Bottom Sound had been discussed by COMSOPACFOR by dispatch** to CTF 61 and CTF 63 on August 2nd in which he proposed that, with the departure of the carriers, air cover was to be provided by carrier fighter planes from TG 61.1 operating from the captured Guadalcanal airfield. But now, CTF 61 indicated his desire to retire the carriers at least one day earlier than anticipated. Also, the Guadalcanal airfield had been captured scarcely three hours before, and would not be immediately ready for operation of Allied carrier aircraft.

CTF 62 therefore became seriously concerned, for he clearly recognized the serious effect the loss of his carrier air cover would have on his operations. His transports and cargo ships had not been able to carry out the unloading operations according to plan because of the long interruptions caused by Japanese air raids. As yet, very little unloading had been accomplished at Tulagi because of the strong Japanese resistance ashore there. He realized that delays in unloading probably would be added to by the hasty manner in which the cargoes had been loaded, the lack of preparation, and the lack of experience in logistics work down to the lowest unit.*** CTF 62 was, therefore, concerned for the safety of his transports and cargo ships, for which strong air power was needed.


*CTF 61 Dispatch 080707, August 1942, to COMSOPACFOR.
**COMSOPACFOR Dispatch 020240, August 1942, to CTF 61 and CTF 63.
***Lecture Army-Navy Staff College, Washington, D.C., by Brigadier General C.C. Thomas, USMC, Chief of Staff, 1st Marine Division during this operation; Subjects: Analyses of the Plans and Operations for the Assault and Seizure of Tulagi-Guadalcanal, January 13th, 1944, page 6.

- 85 -

Sunset occurred at 1816, and at 1830 CTG 62.6 ordered the Screening Group, including Fire Support Groups LOVE and MIKE, to assume the Night Screening Disposition employed on the previous night. These groups took station and commenced patrolling at about 1850. After dark, the weather was overcast with occasional rain squalls and electrical storms in Iron Bottom Sound, light winds, and calm sea.*

At about 1900 CTF 62 received a dispatch over the Pearl Harbor HOW Fox broadcast schedule that two enemy destroyers, three cruisers and two seaplane tenders or gunboats had been contacted by SOWESPAC aircraft at 1025 that day in Lat. 05°-49'S., Long. 159°-07'E., on course 120°(T), speed fifteen knots. Here, at last, was further information on the reported surface ships which had stood out of St. George's Channel the preceding evening on a southeasterly course at high speed.

This report was in error, for there were no seaplane tenders or gunboats. The surface force east of Bougainville Island actually consisted of five heavy cruisers, two light cruisers, and one destroyer of the Japanese Cruiser Force. The pilot of the Australian reconnaissance plane who had made this report had erred in identification, probably because in the early days of the war sufficient emphasis was not given to proper recognition of forces. He had also erred in the number of ships present, because he had not remained over his target to properly develop the contact.

CTF 62 received word in the same dispatch that two submarines had been noted at 1127 in Lat. 07°-35'S., Long. 154°-07'E., on course 150°(T).** This report placed in his hands further information of the movements of Japanese submarines toward his area, as previously reported at 2222, August 7th, by the S-38 patrolling in St. George's Channel. This report was in error also since there actually was but one submarine, the I-121 en route to Tulagi.

These contact reports served to present a picture of continuing enemy action against him and had reached CTF 62 at the time he had been cogitating over the seriousness of the prospective situation with no carrier air support. He evaluated the report of the surface force of enemy cruisers, destroyers and seaplane tenders or gunboats off Bougainville at 1025 and decided that it constituted a seaplane base group, under cruiser and destroyer escort, headed for a suitable harbor in the Solomons. He received an evaluation by dispatch from COMSOWESPAC that the cruisers and tenders reported were probably engaged in establishing a base in the


*Action Report QUINCY, Engagement Morning August 9th, 1942, off Guadalcanal Island, by LCdr. H.B. Heneberger, USN, Senior Surviving Officer, CA39/A16-3/(004)hmc, August 16, 1942 and Action Report CHICAGO, Concerning Action Against Enemy Forces August 9th, 1942, Guadalcanal-Tulagi Area, Serial 099, August 13th, 1942.
**COMSOWESPAC Dispatch 080717 to COMSOPAC.

- 86 -

Shortlands.* From his own information, he knew that a WASP scout had shot down a seaplane north of Rekata Bay that very morning and had strafed a small surface ship in the same area. This fact, taken together with the reported enemy course of 120°(T), probably influenced him to arrive at the conclusion that the destination of the Japanese force was Rekata Bay.

The presence of a Japanese seaplane group in Rekata Bay, just one hundred and thirty miles away, constituted an additional air threat to CTF 62's transports and cargo ships. His considered opinion was that this force had the intention of operating seaplanes that would deliver torpedo attacks against his own force. Such a possibility would increase, by one or two torpedo attacks, the air raids which TF 62 would have to face commencing on August 9th. This conviction prompted him to request CTF 63 to immediately attack the seaplane tenders basing at Rekata Bay.**

The extent to which he considered the possibility of a Japanese night surface attack does not show up in any of the documents available for this study, but it is clear that he contemplated no such attack. The reported Japanese force was much weaker than the Allied combatant forces at present in the Tulagi-Guadalcanal area, and he assumed that this force would not attack the stronger Allied force, even though in night action, success does not always favor the strong. It appears that in estimating the situation, he studied the enemy commander's intentions rather than his capabilities. This was unfortunate; for a study of the forces, the operating area and relative positions, and the courses of action open to the enemy should have shown that one of the most dangerous enemy capabilities, from the viewpoint of CTF 62, would have been to attack the Allied forces at Tulagi-Guadalcanal that night.

As a consequence of the situation confronting him, CTF 62 at 2045 called a conference of the Commanding General, First Marine Division and CTG 62.6 to be held on board his flagship the MCCAWLEY, anchored in Transport Area XRAY.** He wished to know (a) whether or not sufficient stores were ashore to support the Marines in the event he retired his cargo ships and transports early, and (b) whether CTG 62.6 considered the screening ships could stick it out for one or two more days without carrier air support.* Meanwhile, he made a tentative decision to withdraw his transports and cargo ships at 0600, August 9th, the earliest hour that morning light would permit.

Upon receipt of this message, CTG 62.6 decided to close Transport Group XRAY in the AUSTRALIA. He, at 2055, directed the Commanding Officer of the CHICAGO to take charge of the AUSTRALIA Group and carry


*CTF 62 Personal Letter to Captain R.C. Parker, USN, Office of Naval History, Navy Department, Washington, D.C., 1948, and again on June 1st, 1949 in oral statement to Commodore R.W. Bates, USN (Ret), Head of Department of Analysis, U.S. Naval War College.
**Action Report CTG 62.6, August 13, 1942, para. 91.

- 87 -

out the basic plan while he proceeded to close the MCCAWLEY. He advised the Commanding Officer of the CHICAGO that he did not know whether or not he would return to the formation that night.* Unfortunately, he did not at the same time notify the Commanding Officer of the VINCENNES of his impending departure nor did he direct that officer to assume command during his absence.** Instead, he retained command within the AUSTRALIA even though he knew that he would soon be some distance away in the transport area. Apparently the possibility of night action did not seem pressing to him. This was a serious mistake, for it intensified an already loose command situation and not only left the Night Screening Groups without any overall commander in the screening areas but also left the major portion of these groups without any knowledge of his absence whatsoever.

At 2123, the Commanding Officer of the CHICAGO, which was the last ship in column prior to the departure of the AUSTRALIA, directed the formation to remain as it was with the CANBERRA leading the patrol in the manner previously ordered by CTG 62.6. The changes of course were made approximately on every even hour by the CANBERRA, employing movements.

There seem to be several reasons why the Commanding Officer, CHICAGO did not lead the group - hereinafter referred to as the "CHICAGO Group". There was a possibility that the AUSTRALIA might return earlier than expected, and he did not choose to shift positions twice in the darkness and mist. He had no personal staff to assist him, and his ships officers, like himself, were very tired from two days constant vigil and attack. It appears that he expected no enemy action that night despite the fact that he had received the dispatch at about 1900 reporting the enemy surface force east of Bougainville. His action of remaining in the rear position of his group is therefore understandable, but it was not sound. A commander must always be alert to insure for his command every possible advantage that can be obtained.

At 2136 SOPAC forces received another dispatch from COMSOWESPAC, addressed to all task force commanders Pacific Fleet, reporting an air sighting at 1101 in Lat. 05°-42'S., Long. 156°-05'E. of two heavy or light cruisers and one small unidentified ship. One cruiser was reported as similar to the SOUTHHAMPTON Class.*** No mention of this second surface contact report appears to have been made in the discussion of CTF 62's conference. It was, however, mentioned in the report of at least one of the subordinate commanders.**** If it were received by CTF 62 and


*Action Report CTG 62.6, August 13th, 1942, para. 91.
**Personal Interview of Captain Frederick Riefkohl, USN, Commanding Officer, VINCENNES, recorded January 26th, 1946, by The Chief of Naval Operations, Office of Naval Records and Library.
***COMSOWESPAC Dispatch 080847, August 8th, 1942.
*****Action Report VINCENNES, Report of Action Occurring off Savo Island (Guadalcanal-Florida Island) Area, Night of August 8th-9th, 1942, Serial 0021, August 14th, 1942.

- 88 -

CTG 62.6 it was not considered as serious threat. Perhaps CTF 62 and CTG 62.6 considered that the force reported was either a part of the cruiser force bound for Rekata Bay, the intentions of which they had already discerned, or perhaps they felt that it was merely another small force which was retiring to the north and therefore unimportant. The reported position was about seven miles to the north of the 1025 enemy cruiser position and was given by a second Australian pilot who obviously had only reported the two light cruisers of CRUDIV EIGHTEEN and the destroyer YUNAGI.

At 2200 CTF 62 received word that Tulagi, Gavutu and Tanambogo were completely in Allied hands except for a few isolated sniper positions.* At last, the unloading operations there might proceed without the tie-up from Japanese forces ashore.

At 2200 the WILSON, which had been directed by COMDESDIV SEVEN at 2015 to replace the damaged JARVIS, joined the VINCENNES Group in its night cruiser station.**

At about this time also CTF 62 directed the HOVEY to escort the damaged destroyer JARVIS through Longo Channel back to the New Hebrides. He directed the Commanding Officer, HOVEY to pass these orders to the JARVIS.

At 2325 CTF 62 held the conference on the MCCAWLEY with the Commanding General, First Marine Division, and with CTF 62.6,*** and advised them of the situation and of his tentative decision. This conference was of extreme importance to the three commanders. In the case of the Commanding General, First Marine Division, it was important because the unloading of supplies in the Tulagi Area had not been proceeding in a satisfactory manner, and the withdrawal of the transports and cargo ships for both Tulagi and Guadalcanal two days ahead of time had serious implications for the Marines ashore. He felt that the necessity for retiring these transports and cargo ships was most alarming, as their retirement before being completely unloaded would profoundly affect the entire future course of operations in the Solomon Islands.****

In the case of CTG 62.6 the conference was important because it familiarized him with the thoughts of CTF 62 and clarified the military situation. CTG 62.6 at this time was told by CTF 62 that, in his opinion, the enemy surface force reported off Bougainville was destined for Rekata Bay, possibly to operate torpedo-carrying seaplanes against the Allied


*ONI Combat Narrative, The Landing in the Solomons, page 70.
**Action Report WILSON, Action Against Enemy Surface Ships off Savo Island Night of August 8th-9th, 1942, Serial 008, August 20th, 1942.
***War Diary CTF 62, August 8th, 1942.
****Commanding General First Marine Division's Final Report on Guadalcanal Operational Phase II (from H-hour to Evening, August 9th), Serial 00204-108/333.

- 89 -

forces. CTF 62 also stated that he had requested for the next day, full-scale bombing of this force, which he felt sure would be in Rekata Bay.*

It was important to CTF 62 because he desired to discuss the situation with his senior subordinate commanders, and in particular with CTG 62.6 who at 0911 that morning had requested a rough outline of the present situation and future intentions.** Although he had already made an estimate of the situation and had arrived at a tentative decision, he nevertheless thought it not only wise but of paramount importance that the senior commanders should be allowed to comment. This action of CTF 62 in calling this conference was sound. For after the retirement of CTF 61 in the SARATOGA, he would be the senior commander in the Tulagi-Guadalcanal area and his was the responsibility for the decisions that had to be made as a result of the withdrawal of the carrier air support.

At this conference all commanders agreed with CTF 62's opinion as to the projected operations of the Japanese surface force and the probability of increased air attacks from Rekata Bay. They also agreed, however reluctantly, on the necessity of withdrawing the transports and cargo ships at first light (0600), provided sufficient supplies for the Marine operations ashore could be landed before departure. In accordance with this agreement the Commanding General, First Marine Division, went ashore immediately after the conference at 2345 to ascertain the situation as regards supplies at Tulagi. Until this information had been received, CTF 62 could make no final decision nor make any recommendation to CTF 61 or to COMSOPACFOR concerning the situation. He did not realize, of course, that the Commanding General, First Marine Division, would not return to the MCCAWLEY until about 0800 the following morning.***

At 2345 the RALPH TALBOT, which was on the northern radar picket station, observed an unidentified cruiser scouting plane flying low over Savo Island and heading east towards Tulagi. The RALPH TALBOT broadcast on TBS and TBO (the latter, a portable voice radio) the following messages "Warning! Warning! Plane over Savo Island heading east." Her Commanding Officer attempted to get this report to CTF 62 and to COMDESRON FOUR, but failed to get the message through. The report was heard by the BLUE, PATTERSON, VINCENNES, QUINCY and others. COMDESDIV EIGHT then accepted the message for delivery to CTF 62, but was not able to deliver it.

The 2400 positions of the ships of CTG 62.6 and Fire Support Groups LOVE and MIKE with relation to the center of Savo Island were:

(a) Screen -

RALPH TALBOT bearing 010°(T), distant seven miles.
BLUE bearing 270°(T), distant ten and seven-tenths miles.


*Action Report CTG 62.6, Night of Action off Savo Island, August 9th, 1942, August 13, 1942, para. 91.
**CTG 62.6 Dispatch 072211, August 1942, to CTF 62.
***War Diary SOUTHARD, August 1942.

- 90 -

(b) VINCENNES Group bearing 080°(T), distant four and one-half miles.
(c) CHICAGO Group bearing 145°(T), distant ten miles.
(d) SAN JUAN Group bearing 090°(T), distant eleven miles.
(e) AUSTRALIA bearing 140°(T), distant twenty miles.

(b) OPERATIONS OF CTF 61 (Commander Expeditionary Force)

CTF 61 on August 8th appears to have been satisfied that the forenoon operations were proceeding according to plan, for he issued no important directives during that time to either of his task force commanders. He was deeply interested in any reports concerning the enemy carrier, both in his status as Commander Expeditionary Force (CTF 61) and as CTG 61.1.1. In the latter role, it was his assigned task until noon to maintain fighters and a ready attack group of bombers in the SARATOGA, Which could be launched immediately to attack the enemy carrier, if it were reported by the search planes. He followed the progress of the morning search, which the WASP launched at 0606, with interest. After the last of the WASP search planes had landed at 1024, he learned that no enemy surface forces had been discovered, but that an enemy twin-float seaplane had been shot down north of Rekata Bay and that a small Japanese vessel had been strafed in the same area.* This gave him an indication that perhaps Rekata Bay was being employed by the Japanese as a base for seaplanes.

He did not have to accept as final the results of the WASP search; for at 1354 CTG 61.1 had launched the scheduled afternoon search from the ENTERPRISE to cover more or less the same area. At about 1510, when this search was at its outer limit, CTF 61 received reports from the ENTERPRISE that the fighter pilots over Tulagi-Guadalcanal had encountered Japanese twin-engined torpedo planes and twin-engined horizontal bombers, some of which were carrying torpedoes.**

As a consequence of these reports, CTF 61 re-estimated the situation. Without awaiting reports from the afternoon search he made a tentative decision; and at 1630 referred it to CTG 61.1 by visual signal, stating: "In view of possibility or torpedo plane attack and reduction in our present fighter strength, I intend to recommend immediate withdrawal of carriers. Do you agree?"*** To this, CTG 61.1 replied in the affirmative.*****

CTF 61 also included in the same message the following: "In case we


*WASP Visual Dispatch 072358, August 1942, to TG 61.1.
**CTG 61.1.2 Visual Dispatch 080236, August 1942, to CTG 61.1.
***CTF 61 Visual Dispatch 080425, August 1942, to CTF 61.1.
****CTF 61.1 Visual Dispatch 08051, August 1942, to CTF 61.

- 91 -

continue present operation, I believe same area should be used tomorrow as today. What do you think?"* To this query CTF 61.1 also replied in the affirmative.**

CTF 61 then after about two hours further consideration decided to recommend the retirement of TG 61.1. He therefore sent the following dispatch at 1807 to COMSOPACFOR: "Total fighter strength reduced from ninety-nine to seventy-eight. In view of large number of enemy torpedo and bomber planes in area, recommend immediate withdrawal of carriers. Request you send tankers immediately to rendezvous decided by you as fuel running low."***

Despite the reasons advanced in this dispatch for the immediate retirement of the carrier force, analysis reveals that the situation did not justify his making such recommendation at this time for the following reasons:

(a) He knew that his basic Seizure objectives were far from accomplishment. The operations at Guadalcanal of course were progressing favorably, but those at Tulagi and Tanambogo had been slowed down because of serious enemy opposition.

(b) His information of the situation in Iron Bottom Sound and ashore was insufficient to make this decision. He had little, if any, information of the logistics and other difficulties that were delaying the unloading of the transports even longer than the estimate of four days which CTF 62 had given him before August 1st. Although he was fully authorized to take any action he deemed necessary without reference to his subordinate commanders, his failure to consult CTF 62 is made conspicuous by the fact that he did consult CTG 61.1. It should have been apparent to him that the withdrawal of the carrier air support would seriously embarrass the conduct of the landing operations. Why he did not consult CTF 62, who was even more vitally concerned - although plenty of time was available to do so - is nowhere explained. Significantly enough, he had not even made CTF 62 an information addressee for the message he sent COMSOPACFOR requesting approval for retiring the carriers.

(c) He knew that the success of his seizure operations and the defense of the areas captured, would depend in a large part on local air superiority. Such local air superiority could only be gained and held by the forces afloat, since the captured Japanese airfield at Lunga Point was not as yet in operating condition. (It was not finally ready for use until about 2015, August 10th). It would necessarily remain a function of the forces afloat to maintain air superiority until the airfield was provided with sufficient aircraft and facilities.


*CTF 61 Visual Dispatch 080425, August 1942, to TF 61.1.
**CTF 61.1 Visual Dispatch 080515, August 1942, to CTF 61.
***CTF 61 Dispatch 080707, August 1942, to COMSOPACFOR.

- 92 -

(d) He knew that success in battle can rarely be accomplished without some losses, and that it was necessary for him to maintain a running estimate of the situation and to weigh any possible course of action on the basis of acceptability as to costs. For some reason he believed that he had lost twenty-one fighter planes (twenty-one per cent of his original complement of ninety-nine fighters). He considered this as an unacceptable loss which deprived him of sufficient defenses against Japanese air attacks. Actually, he had lost but sixteen fighters. His remaining fighter strength totalled eighty-three in the three carriers. Which was four fighters more than carriers had had when they entered the Battle of Midway. Therefore, it seems extremely doubtful if his losses in fighter strength were sufficiently great to justify his classifying them as unacceptable.

(e) The shortage of fuel, although not advanced by CTF 61 as a primary reason for withdrawal of the carriers, seems to have been incorporated into his request to COMSOPACFOR in such a manner as to infer it as an additional important reason. Two days of flight operations in the vicinity of Guadalcanal Island, without enemy detection or attack, had entailed steaming at speeds influenced only by the wind velocity over the flight decks: and should have affected fuel consumption no differently than routine flight operations anywhere. His screening destroyers and cruisers, unlike those in Iron Bottom Sound, had not been required to steam at high speeds under enemy attacks but had merely cruised with the carriers.

It is always incumbent upon a commander to take every opportunity afforded to conserve the fuel in his force, and the night retirements of TG 61.1 to the southward provided the occasion for steaming at economical cruising speed at least half the time. Apparently, in consideration of this fact, CTG 61.1 steamed throughout each night at fifteen knots.

A study of the war diaries of these carrier task groups during July 1942 reveals they were repeatedly refueled en route from Pearl Harbor to Fijis, and that all ships were eventually "topped off" to ninety-five per cent capacity on July 28th and 29th at area 300 miles south of Suva. The oilers PLATTE, CIMARRON, TIPPECANOE and KANAWHA had all been busily engaged in these fueling operations and replenished their own supply from chartered tankers at Noumea.

While the Expeditionary Force was en route to Tulagi, the CIMARRON effected rendezvous with the carrier force in a position sixty miles southwest of Efate Island in the New Hebrides at 1000 on August 3rd. This position and the track of the CIMARRON for the next two days of fueling operations is shown on Diagram "B". The CIMARRON "topped off" the SARATOGA and eight destroyers on August 3rd. On August 4th, she refueled four more destroyers of TG 61.1; and was released at noon to refuel the destroyers of TF 62.* After this the destroyers were refueled as necessary from the heavy ships of the various task groups.


*War Diary CIMARRON, August 1942. 

- 93 -

A check of the logs of the ships of TG 61.1 for August 8th indicates the destroyers of TG 61.1.1 averaged about seventy-five per cent of capacity: those of TG 61.1.2 about forty-two per cent and those of TG 61.1.3 about forty-four per cent of capacity; the cruisers of all groups were about fifty per cent of capacity or better, and of the carriers, only the ENTERPRISE was running low and had fuel for three more days of operationas.* Despite the statement of CTG 61.1.2 that "Fuel situation in this force becoming critical. It is estimated that the destroyers have fuel for about three days at fifteen knots and the heavy ships have little more",* the facts seem to be that although the fuel in the command was diminishing daily it was not at this time so critically low as to force retirement from the area. In fact CTG 61.1 does not appear to have originated any dispatches concerning an urgent need for refueling, and his affirmative reply to CTF 61 made no mention of fuel at all.

CTF 61's reason for wanting to retire does not appear therefore to have been the low fuel in his ships. This is borne out by his suggestion to CTG 61.1 to continue to operate on August 9th in the same area as on August 8th if the retirement request was disapproved.** Does this not indicate that he was fully prepared to operate with the fuel on hand? Does it not also indicate that he was not too sure that his contemplated request would be approved?

From the above it seems clear that CTF 61's desire to retire TG 61.1 was motivated by his feeling that it was unwise to expose his carriers to possible enemy air attack - especially by torpedo planes - and by no other consideration.*** He appears to have felt that this consideration was so important that it transcended the objectives assigned to him, including the air support that CTG 61.1 was required to provide TF 62. The fact that such a precipitous departure might seriously jeopardize the success of the entire operation at Tulagi-Guadalcanal, for which he was responsible as Commander Expeditionary Force, and might prevent the inauguration of Task TWO which the Joint Chiefs had indicated a desire to expedite**** does not appear to have been given the serious consideration it deserved.


*War Diary, CTG 61.1.2 (CTF 16), August 1942.
**CTF 61 Visual Dispatch 080425 to CTF 61.1, August 8th, 1942.
***A similar conclusion was arrived at by Admiral Arthur J. Hepburn, USN, in his report of Informal Inquiry into the Circumstances Attending the Loss of the VINCENNES, etc., on August 9th, 1942, submitted to the Commander-in-Chief, Pacific Fleet, May 13th, 1943, wherein he stated "The Japanese Commander deemed it wise to retire at high speed to avoid, if possible, the air attack to be expected from our forces at dawn. Ironically enough, the only part of our force capable of making such an attack were at the same time retiring in the opposite direction because of the same apprehension."
****War Diary WASP, July 28th, 1942.

- 94 -

It is unfortunate that CTF 61, in making his recommendation, failed to inform COMSOPACFOR fully of the delicate nature of the situation at Tulagi-Guadalcanal. Perhaps COMSOPACFOR's approval might have been delayed.

At about 1845, CTF 61 received COMSOWESPAC's dispatch reporting three cruisers, three destroyers, and two seaplane tenders or gunboats in Lat. 05°-49'S., Long. 156°-07'E.,* No information is available as to his reactions at this time. Since he did not make any changes in his plans, he seems to have felt that his recommendation to COMSOPACFOR for the retirement of Task Force 61.1, and especially so in the face of an additional threat of torpedo attacks from seaplanes based in the Southern Solomon Islands, was correct.

At about 2333, he received a message from CTF 63 reporting the results of the searches made by his command on the 8th. All sectors searched were reported as negative except Sector I and the left half of Sector II which had not been searched because of bad weather.** These areas had been covered by the WASP search planes in the morning, and again by the ENTERPRISE search planes in the afternoon so that the threat of an enemy carrier was eliminated from his considerations for the time being.

(c) OPERATIONS OF CTG 61.1 (Commander Air Support Force)

Commencing at midnight, CTG 61.1 began his return to the eastward and proceeded at fifteen knots to arrive at his designated launching position at 0600 on August 8th.***

The daylight weather conditions in the vicinity of the task force had improved considerably over the preceding day. Flying conditions were now excellent. The sky was reported as O.5 cloudy, surface visibility unrestricted, wind ENE seventeen knots, scattered showers, sea calm.

Throughout the 8th, he continued his operations: (a) in support of the Allied landing operations and the Allied surface forces in the Tulagi­-Guadalcanal Areas; (b) in providing air cover for his own task force, and (c) in providing air searches to augment the search operations of CTF 63 as requested by CTF 61.

At 0606 the WASP launched a search of twelve scout bombers from her position bearing 167 (T) distant eighty-seven miles from Tulagi and thirty-four miles from Cape Henslow, to search the sector 280°(T)-040°(T) to a radius of 220 miles. She evidently extended this search to include all of Santa Isabel Island and the New Georgia Group. The planes from this search returned on board the WASP between 0958 and 1024.****


*COMSOWESPAC Dispatch 080717, August 8th, 1942, to COMSOPAC.
**CTF 63 Dispatch 081233, August 8th, 1942, to CTF 61.
***CTF 61.1 Visual Dispatch 070611, August 1942, to TG 61.1.
****Report on Capture of Tulagi-Guadalcanal Area August 7th-8th, 1942. WASP Serial 0194 of August 24th, 1942.

- 95 -

During the progress of the morning's search CTG 61.1 maintained fighters and a ready attack group on the flight deck of the SARATOGA which he desired to launch immediately in case a Japanese carrier were located. That he was insistent that this group be ready for instant offensive action, on the receipt of a contact report from one of the WASP search planes, is divulged in the text of three dispatches sent to CTG 61.1.1 in the SARATOGA alerting him at 0820,* again at 0925,** and finally at 1029*** to the necessity of keeping fighters and the attack group ready at all times until noon in case an enemy carrier was located and a bombing attack was carried out.

At 1030, CTG 61.1 learned from the WASP that the morning search had been completed with negative results, except for the shooting down of an enemy twin-float seaplane north of Rekata Bay and for the strafing of a small craft flying Japanese colors in the same area. The WASP reported the results of this search to TG 61.1 in a visual message sent out at 1058.****

CTG 61.1 commenced retiring to the southeastward at 1200 at a Point Option speed of five knots. His retirement courses were apparently the result of his own analysis of the need for sea room for the carriers as affected by the direction of the wind, and not as the result of any directive from CTF 61.

At about 1203 a Japanese air strike came in to attack the transports. At this time, CTG 61.1 had a combat air patrol over the transports of three fighters from the ENTERPRISE and eight from the SARATOGA, a total of eleven. An additional combat air patrol of four fighters from the SARATOGA was en route to the transport area but it arrived too late. These Allied fighters succeeded in shooting down four Japanese land attack planes and one Zero fighter but did not succeed in breaking up the attack.

Between 1347 and 1354, he launched an additional air search group of fourteen torpedo planes from the ENTERPRISE to conduct a 200 miles search of sector 270°-015° from a point of origin Lat. 09°-32'S., Long. 159°-30'E., and a 200 mile search of sector 345°(T) - 090°(T) from another point of origin 09°-50'S., Long. 160°-56'E. These searches had been designated to cover adjacent areas. At 1749 twelve of these fourteen torpedo planes returned and landed on the ENTERPRISE. The remaining two landed at 1839.***** Evidently this search had been extended by the pilots while in the air, for the ENTERPRISE reported to CTF 61 that two planes had searched to a radius


*CTG 61.1 Visual Dispatch 072120, August 1942, to CTG 61.1.1.
**CTG 61.1 Visual Dispatch 072225, August 1942, to SARATOGA.
***CTG 6.1 Visual Dispatch 072315, August 1942, to CTG 61.1.1.
****WASP Dispatch 072358, August 1942, to CTG 61.1.
*****Letter August 24th, 1942, from Commanding Officer, ENTERPRISE, Concerning Operations in Support of occupation of Tulagi-Guadalcanal, August 7th-8th, 1942, Serial 0194, August 24th, 1942.

- 96 -

of 260 miles in the western sector and twelve planes had searched to a radius of 220 miles in the northern and eastern sectors, all with negative results.* A study of Diagram "D" shows that the twelve planes which searched to 220 miles reached the extremities of their search at about 1550, and the two planes which searched to 260 miles reached the extremities of their search at about 1610. The ENTERPRISE also reported that her planes had encountered heavy squalls 100 miles to the eastward and 140 miles to the northeastward, but that the area to the westward was clear.*

As a matter of hindsight, it is of interest to analyse the extent of this search in relation to the position of the Japanese Cruiser Force. That force had completed its passage southward through Bougainville Strait at 1600, had entered "The Slot", and at 1612 had changed course to 120°(T). Thus, it had come well within the radius of the ENTERPRISE search planes in the western sector. It is unfortunate, from the Allied viewpoint, that the third search plane from the west had not also searched to 260 miles radius rather than to 220 miles; for had it done so its search would have covered Bougainville Strait. This plane then should have passed over the Japanese Cruiser Force which was present at that time. As it was, by searching to but 220 miles this plane missed a probable contact by a mere thirty miles.

Throughout the afternoon, CTG 61.1 continued to provide air cover for TF 62 which consisted of approximately twelve fighters: and to provide air cover for his own force which consisted of forty-two fighters until 1300, gradually reducing in number to six fighters at 1800.** Dive bombers were employed as necessary for anti-submarine patrol.

CTG 61.1 also continued to provide from dawn until sunset the air support and the reconnaissance for the landing operations. This consisted, in large part, of attacks by dive bombers which delivered both bombing and strafing attacks on enemy positions.

After the completion of the sunset recovery of all aircraft at 1857, CTG 61.1 set the retirement course for the night as 140°(T) at speed fifteen knots. At 2330 he changed course to 230°(T).***

Total aircraft losses August 8th: Fighters, one crashed; Bombers, none.

His 2400 position on the 8th was Lat. 11°-30'S., Long. 161°-47'E.


*ENTERPRISE Visual Dispatch 081040, August 1942, to CFP 61.
**Letter August 24th, 1942, from Commanding Officer, ENTERPRISE, Concerning Operations in Support of Occupation of Tulagi-Guadalcanal, August 7th-8th, 1942, Serial 0194, August 24th, 1942; Letter August 14th, 1942, from Commanding Officer, WASP, Concerning Capture of Tulagi-Guadalcanal Area August 7th-8th, 1942, Serial 003, August 14th, 1942 and SARATOGA War Diary, August 1942, Annex "B".
***War Diary, ENTERPRISE, August 1942.

- 97 -


(1) Operations of S-38

During the 8th the S-38 was patrolling in St. George' channel along a fifteen mile run: in and out on courses 320°(T) and 140°(T), respectively. At 0242, having noted Japanese units moving out on course 140°(T), the submarine commander notified COMSUBDIV FIVE of this fact by dispatch. However, he could take no offensive against them for his port motor panel had failed and he was engaged in repairing it in a position clear of the channel. At 2100 he resumed his 320°-140°(T) patrol. At 2255 he sighted a large Japanese submarine, apparently the I-122, en route from Rabaul to Tulagi, standing out of the channel at high speed. This submarine was too far away to attack and soon disappeared.*

The approximate 2400 po1ition of the S-38 was bearing 270°(T), distant fourteen miles from Cape St. George.

(2) Operations of S-44

The S-44 was patrolling submerged off the southern entrances to Byron and Steffen Straits and made no contacts. The Commanding Officer surfaced after dark. Some time thereafter he received a message directing him to move to Area ZED, which appears to have been the area off Cape St. George, New Ireland. He therefore proceeded at about 2100 in a westerly direction around New Hanover Island.**

At 2400 he was about five miles south of New Hanover Island in a position about thirty miles west of Steffen Strait.

(e) OPERAIONS OF CTF 63 (Commander Aircraft South Pacific Force)

CTF 63 search operations on August 8th were in general very similar to the preceding day, and are shown on Diagram "D". Sectors I, II, and VI remained the same; Sector III was modified; Sector IV was advanced from Espiritu Santo to Maramasike Estuary, Malaita Island; and Sector VII was a new sector flown from Espiritu Santo. The results of these searches follows:***

Sector I was not searched because of bad weather.

Sector II was scarcely half searched. The right half was not searched because of bad weather; the left half was searched to a radius of 650 miles only rather than to the planned 750 miles radius for the same reason. Results were negative. It was through Sector II that the Japanese Cruiser Force approached Savo Island, but the Allied search, even if it


*War Diary S-38, August, 1942.
**War Diary S-44, August, l942.
***War Diary CTF 63, August 1942.

- 98 -

had been flown as planned, would not have discovered the Cruiser Force. This was because that force did not enter this sector until about 1900 August 8th, whereas the B-17 that did search was abreast Tulagi about sunrise and had reached its 650 miles extremity by 0715. Had an afternoon search in addition to the early morning search been provided, and had the planes making this search been instructed to be at the extremity of their sectors by sunset, they would probably have discovered the Japanese Cruiser Force between Choiseul and New Georgia Islands. Such an afternoon search had been placed in effect in Sectors III and V on August 5th and 6th at the special request of CTF 62 who feared an attack through that area, but was dropped thereafter. It is not clear why a search of this type was not organized in Sector II, after the Japanese cruisers were reported off Rabaul and later in St. George's Channel on August 7th.

Sectors III, VI, and VII were searched with negative results.

Sector IV was searched by the patrol planes which had been moved up from Espiritu Santo to Maramasike Estuary the preceding day. These planes operated from the tender MACKINAC.

Sector V was searched to a distance of only 600 miles with negative results. The outer fifty miles of the planned 650 miles was not searched.

There is some doubt as to whether the search of the sector 290°(T) to 318°(T) from Malaita Island, which search had been requested the preceding day by CTF 62, was actually made. CTF 63 states in his War Diary that all searches were conducted in accordance with the Basic Operation Plan and there is no entry in the Diary concerning a search of this new sector.* On the other hand, CTF 63's Chief of Staff has recently stated that this search was made by two B-17's from Espiritu Santo, which searched to a distance of about 315 miles beyond Malaita Island with negative results.** Assuming that the search was made and that the planes departed at dawn, they should have reached the outer limit of their search at about 1215. A plot of this search is shown on Diagram "D". At this time, the Japanese Cruiser Force was roughly sixty miles to the northwest. It is apparent, therefore, that this search as conducted, would necessarily have been ineffective. No special report concerning the results of this search was made by CTF 63 to CTF 62.

CTF 63, who had received not only the dispatch from COMSOWESPAC reporting the Japanese Cruiser Force off Bougainville Island, but also CTF 62's request to bomb these ships in Rekata Bay on August 8th, seems to have concurred in CTF 62's opinion as to the destination of the Cruiser


*War Diary CTF 63, August 1942.
**Letter October 20th, 1948, from Rear Admiral M.B. Gardiner, USN (Chief of Staff to CTF 63) to President, Naval War College.

- 99 -

Force, for he did not recommend other action. In fact, he directed the ELEVENTH Bombardment Group to be prepared to attack this enemy force in Rekata Bay at 1200, August 9th. He also directed the Commanding Officer, MACKINAC (CTG 63.5)at Maramasike Estuary to make a four-plane night torpedo attack on this force on August 9th.* This latter order was released at about 0254, August 9th.**

Table 4 shows the disposition of Allied land and tender-based air­craft in the SOPAC Area as of 2400, August 8th.


Commander Allied Air Forces North Eastern Area, SOWESPAC, conducted air searches on August 8th in support of the operations in SOPAC Area as follows:

(a) Three Hudsons from the Fall River Field at Milne Bay reconnoitered the Solomon Islands in Reconnaissance Area "B", and a fourth Hudson made a short reconnaissance flight over Reconnaissance Area "Siren" which appears (from previous delineation of search areas) to have included the sea approaches to the Louisiade Archipelago, the Lusancay Islands, and the approaches to Buna Roads from Rabaul.

(b) One B-17 from Port Moresby reconnoitered Reconnaissance Area "C"; another covered Reconnaissance Area "D", and a third B-17 made a photographic intelligence mission in Reconnaissance Area "E" with particular emphasis on Rabaul and Kavieng.

The reconnaissance flights of the three Hudsons in Reconnaissance Area "B" are of particular significance, since these Hudsons re-established the previous day's contacts on the Japanese Cruiser Force as it moved southward towards Tulagi. The contacts were made as follows:

(a) Hudson Flight Number A16/218 was heading up from the southeast at 1025 when he made contact on the Japanese Cruiser Force in Lat. 05°-49'S., Long. 156°-07'E· and reported it as three heavy cruisers, three destroyers, and two seaplane tenders or gunboats on course 120°(T), speed fifteen.*** This force actually consisted of five heavy cruisers, two light cruisers and one destroyer.

The pilot did not trail this contact but proceeded to return directly to his base at Fall River. This fact is revealed by the contact he made an hour later, at 1127, on one Japanese submarine, apparently the I-121 en route from Rabaul to Tulagi, in a position in the Solomon Sea 175 miles away from the cruiser contact on a direct line to Milne Bay. The Hudson pilot attacked this submarine with two 100-pound bombs, but missed.***


*War Diary CTF 63, August 1942.
**CTF 63 Dispatch 081554, August 1942 to MACKINAC.
***Operations Report Allied Air Forces, Southwest Pacific Area for August 8th, 1942.

- 100 -



AS OF 2400, AUGUST 8, 1942

BASE 11th
VP Sqdns
11, 14 & 23
Marine Corps
Total Plans

VMF 111
& 212
VMO 251

  B-17 B-26 P-39 PBY-5's Hud-
ESPIRITU 10*           20**    39 
EFATE             12    15 
10  38          67 
NANDI 12  12  17  12        71 
NDENI       5***             
SOMOA               18  17  10  45 
TONGATABU     24              30 
TOTALS 30 22 79 27 30 50 17 25 280

*Rotated up from EFATE and NANDI
**VMF-212 transferred 3 of the 4 F4F-3A back to EFATE later on August 8th, leaving 17 F4F at ESPIRITU
***1 of 6 PBY-5's at NDENI down at sea on August 8th.
****1 of 9 PBY-5's at MARAMASIKE ESTUARY struck reef on August 8th.

[1 100a -]


[- 100b -]

The procedure governing Air Force reconnaissance missions stipulated that a plane making contact at sea was to remain in the vicinity of the sighted target until recalled or forced to retire.* It is assumed that this pilot was forced to retire for cause, probably because of fuel limitations. His contact report, unfortunately, was in error as regards both numbers and recognition. These errors had a profound effect on the decisions of CTF 62 and CTG 62.6.

All evidence available for this study seems to indicate that the above two contact reports were made after the plane returned to its base at Fall River, since both of them were combined into a single dispatch. The basic instructions required that each contact be reported by dispatch as soon as made by the plane in flight.*

The dispatch which reported these two contacts was passed over the Australian Air Force circuit from Fall River to Port Moresby, and thence to Townsville to the Headquarters of the Commander North Eastern Area. At this point (or at Port Moresby), had the urgency been recognized, it might well have been broadcast on Task Force 63's point-to-point Net "E". Instead, it was passed from Townsville over the Australian Air Force land-line circuit to the Headquarters of Commander Allied Air Forces, Southwest Pacific Area, at Brisbane. It was then passed by hand to COMSOWESPAC, also at Brisbane, who passed it over the Navy (U.S.N.-R.A.N.) land-line circuit to Canberra, Australia in his dispatch 080717 (time of origin 1817, minus eleven Zone Time) for transmission over the air on the Canberra BELLS broadcast schedule. Canberra then transmitted it on the BELLS schedule to the Au1tralian Forces and to Pearl Harbor for translation on the HOW Fox schedule to the American Forces. Canberra completed its transmission on the BELLS schedule at 080737 (1837 minus eleven Zone Time). Pearl Harbor completed its transmission on the HOW Fox schedule at 080743 (1843 minus eleven Zone Time).

This dispatch was received by COMSOPACFOR, the action addresses, and by CTF 62 on the HOW Fox schedule between 1843 and 1845. It was received by CTF 62.6 on the BELLS schedule at 1837, and on the HOW Fox schedule at 1843.** Thus, this vitally important contact report was received by the responsible commanders in the SOPAC Area about eight hour's and twenty minutes after the contact was made.

(b) At 1101, the pilot of Hudson Flight Number A16/185 made contact on the same Japanese Cruiser Force in Lat. 05°-42'S., Long. 156°-05'E., and reported it as two heavy or light cruisers and one unidentified vessel. This plane was reported by the Japanese to have trailed them for some time


*Signal Annex to Operation Instruction Number TWO, GHQ, Southwest Pacific Area, April 25th, 1942.
**Report of Informal Inquiry into the Circumstances Attending the Loss of the VINCENNES, etc. by Admiral Arthur J. Hepburn, USN (Ret.), May 13th, 1943, page 10.

- 101 -

and this fact is borne out by the report of a 2000-ton motor vessel by this same pilot one hour and twenty minutes later in Bougainville Strait in a position that was but eighty miles away from this first contact.* Additional contact reports on other, but unimportant, targets indicate that he reconnoitered Gizo Island and returned to his base at Fall River.

The report of the contact made at 1101 was even more incomplete than the 1025 report, and did not tend to clarify an already confused situation. Why the pilot did not make a more correct report is not known. However it may have been due, in part at least, to the fact that all five Japanese heavy cruisers had been recovering planes in widely separated positions until about 1200 before joining up. It seems logical to assume that the Australian pilot failed to see the five heavy cruisers, but that he did sight the TENRYU, YUBARI and YUNAGI which latter three ships had not launched planes and had been operating as a unit apart from the heavy cruisers.

The report of this contact was received by COMSOPACFOR at 2136 or about ten and one-half hours after it had been made. Whether or not it was received by CTF 62 and CTG 62.6 remains unknown.

The fact that both of these reports from the Hudson reconnaissance planes were received at Tulagi hours after the contacts had been made, and the further fact that sunset was at 1816, once again denied COMSOPACFOR the opportunity of requesting an immediate air attack by COMSOWESPAC planes, should he have desired that such an attack be made; and denied the Air Support Force (TG 61.1) the opportunity to attack this Japanese force on the afternoon of August 8th.

The recurrence of this slowness of communications on August 8th, as on August 7th, is a repetition of a similar historical incident that occurred in World War II some two years earlier. In that instance, an R.A.F. Hudson plane of the British Coastal Command on a reconnaissance flight on the eastern side of the North Sea on-Sunday, April 7th, 1940, sighted shortly after noon a large German naval force escorting transports on a northerly course. The pilot failed to report this contact by radio and only mentioned it verbally after returning to base some four hours later. This contact was of vital concern, since it was part of the German invasion force headed for Norway. The invasion of Norway commenced on April 9th.** Although a squadron of bombers was belatedly sent out from a British east coast base, it failed to intercept this enemy force. The Commander-in-Chief, British Home Fleet, was denied information of the movement of the Germans until too late for effective countermeasures.**


*Operations Report Allied Air Forces, Southwest Pacific Area, August 8th, 1942.
**Report June 24th, 1940 from U.S. Naval Attaché, London Concerning Great Britain (Norway) Norwegian Campaign - Initial Error, Serial 690, File 907.

- 102 -

Thus, long delays in Royal Air Force communications (both British and Australian) on directly opposite sides of the globe had assisted the enemy; first, in April, 1940 in the case of the German invasion of Norway, and second, in August 1942 in the case of the Japanese attempt to defeat the Allied operations in the Solomons.

It should be apparent that the long delay in the receipt by the action addresses of the two R.A.A.F. Hudson contacts made on the Japanese Cruiser Force on August 8th, 1942, taken together with the similar delay in the receipt of the B-17 contacts made on this same force on August 7th, serve to indicate a uniform pattern of slow and ineffective communications between aircraft of the Allied Air Force in the North Eastern Area and the responsible Allied commanders in the SOPAC Area.

In spite of the fact that the two bases, Townsville and Port Moresby were linked on COMAIRSOPAC (CTF 63) Net "E" and that three air bases (the above two plus Fall River) were able to communicate on CTF 63's Net "C", these circuits were not employed by the Allied Air Forces in the North Eastern Area to broadcast the vital information contained in these contact reports. Instead, as provided in the SOWESPAC communication Plan, these "signals originated in the North Eastern Area and addressed to South Pacific Forces (were) routed via Headquarters Allied Air Force".* As was shown previously, there was the additional provision that: "When urgency demands, (signals) may be routed via Port Moresby or Townsville on the above (Net "E") point to point series."* But this provision called for a decision as to what constituted information of urgency, and such a decision would therefore have required the advice of a naval liaison communications officer at each of the bases in the North Eastern Area. Unfortunately, the Allied Air Forces were dependent upon the communication facilities and personnel of the Royal Australian Air Force (RAAF). The latter personnel had unhappily become complacent as a result of their inability to take offensive action prior to the influx of American Forces, and were slow in responding to the demands of the situation.**


*Signal Instructions, Annexure (B) to Operation Instruction No. 18, Headquarters Allied Air Forces, July 31, 1942.
**Interview on January 13th, 1950, Captain B.F. Roeder, USN, Assistant Communication Officer on Staff of COMSOWESPACFOR on August 1942, conducted by Commodore Richard W. Bates, USN, (Ret.), Head of Department of Analysis, Naval War College, Newport, R.I.

- 103 -



August 9th, 1942



    (1) SCREEN  
      BLUE (DD), RALPH TALBOT (DD) - Total: 2 DD
      WILSON (DD), HELM (DD) 2 DD
      Total forces engaged: 5 CA, 6 DD
      CHOKAI (FF)(CA), AOBA (CA), KAKO (CA),    
      YUNAGI (DD) 1 DD
      Total forces engaged: 5 CA, 2 CL, 1 DD



CHICAGO, VINCENNES, QUINCY, ASTORIA (nine 8-inch, eight 5-inch); CANBERRA (eight 8-inch, four 4-inch, eight 21-inch torpedo tubes); BLUE, RALPH TALBOT, PATTERSON, BAGLEY, WILSON, HELM (four 5-inch, sixteen 21-inch torpedo tubes).

CHOKAI (ten 8-inch, four 5.0-inch, sixteen 24-inch torpedo tubes); AOBA,, KAKO, KINUGASA, FURUTAKA (six 8-inch, four 4.7-inch, eight 24-inch torpedo tubes); TENRYU (four 5.5-inch, one 3-inch, six 21-inch torpedo tubes); YUBARI (five 5.5-inch, one 3-inch, four 24-inch torpedo tubes); YUNAGI (four 4.7-inch, six 21-inch torpedo tubes).

*Additional Allied ships were in the final phase of this night action. However, for one reason or another, as will be shown in this chapter they were not actually in the battle. These were the AUSTRALIA, SELFRIDGE, MUGFORD, ELLET. The JARVIS was disabled but passed through the combat area during the battle.
**SOURCE: ATIS Historical Division G2 General Headquarters, Far Eastern Command April 1950.

- 104 -


The following survey of pertinent strength and weakness factors of each force has been made to summarize the material for testing the feasibility and acceptability of possible courses of action.


More, and more modern destroyers (6 to 1)
More 8" guns (44 to 34)
Cruisers had more life (27 to 23.6)
Cruisers generally superior to Japanese cruisers.
Close support nearby (3 cruisers, 5 destroyers)
More torpedo tubes in destroyers (96 to 6)
More 4.0" and 5.0" than enemy's 4.7", 5.0" and 5.5" guns (60 to 34)
Forces separated so could be destroyed piecemeal.
Commanders of different groups not familiar with one another's plans.
Personnel worn out after two days in combat area.
No flag officer present.
No battle plan.
Lack of correct intelligence concerning composition enemy cruiser force.
Night attack not anticipated by high command.


More light cruisers (2 to 0)
More torpedo tubes in cruisers (58 to 8)
Better trained in night operations
Ship-based aircraft well trained for night flying.
Familiar with combat area.
Freedom of action.
More war experience.
Forces concentrated.
Non homogeneous
Enemy forces stronger than expected.
Had not operated together.

This analysis indicates that, insofar as guns, torpedo tubes, radar, design, and equipment in general were concerned, the advantage lay with the Allies; that insofar as combat experience, freshness of personnel, surprise, initiative and concentration were concerned, the advantage lay with the Japanese. Thus, the Allies had material superiority; the Japanese, morale and tactical superiority.

- 105 -



0000 August 9th to 0132 August 9th


The Japanese Cruiser Force continued onward in night cruising disposition towards Tulagi on course 120°(T) speed twenty-four knots, and all ships in condition of readiness ONE. At exactly 0000, August 9th, with the south entrance to Tulagi bearing 140°(T) distant thirty-seven miles, Commander Cruiser Force changed course to that bearing, increased speed to twenty-six knots and assumed the battle formation which he had ordered in his Signal Order No.25.* The vanguard of his cruising disposition, which was composed of the TENRYU and the YUNAGI in position 1000 yards on the port bow of the CHOKAI, and the YUBARI in a similar position on the starboard bow, immediately reversed course and took position in the battle formation.** This formation was actually a simple column with the ships in the following order: CHOKAI (FF), AOBA (F), KAKO, KINUGASA, FURUTAKA, TENRYU (F), YUBARI, and YUNAGI.

At 0025 Commander Cruiser Force sent two signals to his force. One was: "Three heavy cruisers south of Savo Island, course 290°(T), speed eighteen knots"; the other was: "Prepare to fire torpedoes!"*** The signal, referring to the heavy cruisers, evidently was a report of the AUSTRALIA Group which had reversed from a northwesterly course at 0015, ten minutes earlier, to a course of 125°(T). This information had been given to the Japanese Commander by one of the two cruiser planes which were reconnoitering the Tulagi-Guadalcanal area and the observation was made at the time the CHICAGO Group was on course 306° (T) at speed twelve knots.**** The report was somewhat inaccurate as to composition. There were actually two destroyers and two cruisers information at the time. The third cruiser, the AUSTRALIA, had departed at 2123 for Area XRAY off Guadalcanal. This report was reasonably accurate as to course (306° instead of 290°) but somewhat inaccurate as to speed (twelve instead of eighteen). However, when one considers the pilot's problem of operating over hostile surface forces, in low visibility and on a moonless night, his performance must be rated unusually creditable and reflected merit on the Japanese ship-based pilots.

Commander Cruiser Force must have been extremely gratified to receive this report, for his late afternoon reconnaissance plane had not returned.


*CRUDIV 6 Detailed Battle Report No. 8, Solomons Naval Action, August 7th-10th, 1942, CIG Document 86927, June 27th, 1947, page 1.
**Records CRUDIV 18, August 8th, 1942, WDC Document 160984.
***War Diary KAKO, August 7th-10th, 1942, WDC Document 160143.
**** USSBS Interrogation Nav. No. 83, Vol. II, Interrogation of Japanese Officials, Interrogation of Captain Kenkichi Kato, IJN, page 361.

- 106 -

This information not only tended to verify his own estimate of enemy forces, based in part upon the report of Commander FIFTH Air Attack Force, but it also indicated that he was going into action against divided enemy forces, each group of which was inferior to his own force.

The signal "Prepare to fire torpedoes" was in keeping with his signalled battle plan and it foreshadowed relatively "close action".

At 0045, he ordered battle stations alerted.* This order established a condition of readiness believed to be similar to the American Condition of Readiness ONE. A few minutes later, at about 0053, the CHOKAI sighted on bearing 162°(T) a ship resembling a destroyer.** This destroyer was the BLUE on her radar patrol station. At 0054 the AOBA, which was directly astern of the CHOKAI in column also sighted the BLUE.*** Diagram "E" shows that the range at the time of contact was 10,900 yards. It was fortunate for the Japanese that the BLUE, even with radar, failed to detect the Japanese Cruiser Force on its approach. This long-range sighting by the Japanese lookouts was merely one of many long-range and astonishingly accurate reports made by them during the action.

As will be shown later, the Japanese reports of sightings were almost always made long before similar reports were made by the Allied lookouts and were, therefore, of inestimable value to the commanding officers of all Japanese ships engaged. This superiority in night visual detection was presumably due to superiority in night glasses; to longer training in actual war conditions and to freshness of lookouts, when compared with the Allied lookouts who may have been suffering from fatigue resulting from several days of intense combat operations. It may have been due also to an unwarranted reliance on radar by the Allied ships.

Commander Cruiser Force appears to have been somewhat surprised by the presence of the BLUE. His planes had not reported her, she was patrolling across his line of approach through the south channel into Iron Bottom Sound, and she might seriously interfere with his plans if she discovered his cruiser force. He had two apparent courses of action. The first was to continue on, to endeavor to pass the destroyer without discovery and without gun action, and to endeavor to surprise the three cruisers reported to be south of Savo Island. The other was to change course to a little south of east and to endeavor to pass through the north passage into Iron Bottom Sound without discovery. He chose the latter course of action, for at 0058 he changed the flag-ship's course to 100°(T) and signalled "Enter by north entrance!"***


*War Diary 8th Fleet, August 7th-10th, 1942, CIG Document 74633, May 12th, 1947.
**USSBS Interrogation Nav. No. 83, Vol. II, Interrogation of Japanese Officials, Interrogation of Captain Kenkichi Kato, IJN, page 361.
***CRUDIV 6 Detailed Battle Report No. 8, Solomons Naval Action, August 7th-10th, 1942, CIG Document 86927, June 27th, 1947, page 4.

- 107 -

At about 0103, he sighted on bearing 159°(T) what he mistook to be another destroyer,* but what evidently was a two-masted schooner. This schooner was probably the same ship which had been reported by an ENTERPRISE plane on August 7th in a position northwest of the Russell Islands, since no other Japanese ships had been observed in the area. Because of this contact, Commander Cruiser Force now had to re-estimate his situation. It must have appeared to him that this second destroyer was part of a patrol across the entrance. Although no picket ships had been reported by his planes, nevertheless, it was a logical security measure for the Allies to take. He decided that he would gain no advantage by endeavoring to enter through the north channel which would also be patrolled. It would be wise to strike immediately. He also observed that both "destroyers" (BLUE and the schooner) were gradually withdrawing to the southwestward and had either not discovered his cruisers or had decided to take no direct action other than report his presence. He did not think that either of these possibilities was likely.** Although the action reports do not so state there is reason to believe from some of the interrogations that Commander Cruiser Force's decision to carry out his original plan to enter through the south channel may have been affected also by possible reports from his cruiser planes of an Allied cruiser group east of Savo Island in addition to the one to the south.

He therefore, at 0105, directed the CHOKAI to turn right to course 150°(T); and at 0108 he signalled "Enter from south passage!"** He continued at twenty-six knots on course 150°(T) with his ships at general quarters. He decided to withhold fire, maintain silence and trust to chance. This was a wise decision, for his command had not as yet been observed by the BLUE or by any other Allied ship. His presence was unknown.

During this time, Commander Cruiser Force made no contacts on the other Allied radar patrolling destroyer, the RALPH TALBOT. This was probably because the RALPH TALBOT, at 0100, was approximately 24,000 yards away, or slightly more than double the Japanese demonstrated range of visibility. When the CHOKAI turned southward at 0105, she was approximately 19,000 yards away and therefore was still beyond visual range insofar as the Japanese lookouts were concerned and beyond radar range insofar as she, herself was concerned.

At 0112 when due west of the north point of Savo Island, Commander Cruiser Force changed course by follow-the-leader tactics to 160°(T). He then, at 0118, changed course to 180°(T); at 0120 to 150°(T); and at 0121 to 110°(T).*** This latter course would have taken the Cruiser Force just


*War Diary 8th Fleet, August 7th-10th, 1942, CIG Document 74633, May 12th, 1947, page 2.
**USSBS Interrogation Nav. No. 109, Vol. II, Interrogation of Japanese Officials, Interrogation of Captain Toshikazu Ohmae, IJN, page 472.
***Track Chart Annexed to Records CRUDIV 18, August 7th-9th, 1942, WDC Document 160984.

- 108 -

outside the 100 fathom curve on the south side of Savo Island. Why Commander Cruiser Force desired to operate so close to Savo Island is not explained. It would appear as though the correct course, once the Allied patrolling destroyers had been evaded, would have been to head directly into Iron Bottom Sound and towards the transports off Guadalcanal. Perhaps he hoped to interfere with Allied radars by operating close to a land mass. Perhaps he was seeking the three cruisers, which had been reported as operating south of Savo Island about one hour earlier, and wished to locate them prior to his "rush in". Whatever may have been the reason Commander Cruiser Force proceeded on course 110°(T) until 0124 when he commenced running into a heavy mass of clouds which obscured the southern end of Savo Island and extended several miles to the southward. He, therefore, at 0124, changed course to 160°(T) to avoid this cloud mass. At 0126, Commander Cruiser Force signalled "Proceed independently?"* It is believed by this
signal, he meant that the CHOKAI, CRUDIV SIX, CRUDIV EIGHTEEN and the YUNAGI were to act independently of each other.** This was an unusual order, but in view of the nature of night operations, was sound provided that each of the above tactical units was to operate independently against Allied shipping, their designated objective. Otherwise, does it not seem that it would have been wiser to have kept them together? Night combat operations are very difficult and unless forces are kept together, melees often result, and it is difficult to distinguish friend from foe. This was especially true of the Japanese forces at this time, for, as has been pointed out previously, the cruisers comprising this force had never maneuvered together before.

At 0128 1/2, the CHOKAI changed course to 130°(T) and at 0132, to course 095°(T).*** She passed Savo Island at a distance of about three miles on her port beam. The other ships followed in column.


*CRUDIV 6 Detailed Battle Report No. 8, Solomons Naval Action, August 7th-10th, 1942, CIG Document 86927, June 27th, 1947.
**USSBS Interrogation Nav. No. 109, Vol. II, Interrogation of Japanese Officials, Interrogation of Captain Toshikazu Ohmae, IJN, page 472 and Nav. No. 61, Vol. I, Interrogation of Rear Admiral M. Matsuyama, IJN, page 255.
***Track Chart Annexed to Records, CRUDIV 18, August 7th-9th, 1942, WDC Document 160984.

- 109 -



0000 August 9th to 0132 August 9th

As the Japanese Cruiser Force approached the Tulagi-Guadalcanal area, the ships of TG 62.6 and of Fire Support Groups LOVE and MIKE, which were assigned patrol stations in the western entrance to Iron Bottom Sound, were cruising in their assigned patrol areas as described previously and as shown on Plate IV. Because the Allied ships, which actually engaged in the battle, were not concentrated into one group but were instead widely separated into four groups consisting of two groups of single ships, one group of four ships and one of five ships, it seems wise to discuss the actions of each of these groups separately in the order that they were encountered by the Japanese.


The BLUE was patrolling her radar and anti-submarine station. She was in American Condition of Readiness TWO. The weather was partly cloudy with a wind of four knots from the northeast. The sea was calm, no moon.* The visibility of 8000 yards, as given by the BLUE, was about 3000 yards less than the 11,000 yards demonstrated by the Japanese. This discrepancy appears to have been due to the superior ability of the Japanese to detect and recognize ships at night.

At 0000, the BLUE was headed on course 231°(T) at twelve knots and was in position bearing 270°(T) distant ten and seven-tenths miles from the center of Savo Island. At 0010 she reached the southwestern end of her patrol line and reversed course to 051°(T). She completed this leg of her patrol at 0040, When she reversed course to 231°(T). At about 0053 she was sighted by the CHOKAI bearing 162°(T) distant 8700 yards. However, the BLUE in return did not sight the CHOKAI which was closing her rapidly. At 0110 the BLUE once again reversed course to 051°(T). Between 0105 and 0120 her average distance from the CHOKAI was 13,000 yards. The fifth ship in the Japanese column, the FURUTAKA, which passed through the CHOKAI's wake about six minutes later than the CHOKAI, obtained accurate bearings on the BLUE at 0121 and at 0126. At these times, Diagram "E" shows that the BLUE and FURUTAKA were 11,000 yards apart.

During this time, the BLUE and the two-masted schooner were operating fairly close to each other - about 9000 yards - but the BLUE failed to see her either. The schooner was beyond the radius of visibility which the BLUE observers reported as 8000 yards, but it is not clear why the BLUE


*Letter August 17th, 1942, from Commanding Officer, BLUE to CTF 62, Concerning Supplemental Information to Action Report Blue (Serial 031, August 12th, 1942), Serial 035, August 17th, 1942.

- 110 -

failed to locate the schooner by radar. The BLUE's radar was operative, as is shown by the fact that about one-half hour earlier it had detected one of the Japanese cruiser planes flying over Savo Island. However, she did not make this aircraft detection until she had been alerted by the RALPH TALBOT's broadcasted warning. Whether the failure to make radar contacts was due, (a) to the interference of the land masses nearby (Savo Island and Guadalcanal), (b) to inexperience or inattention on the part of the radar operators, who were probably very fatigued after several days of intense concentration, or (c) to the simple fact that the BLUE's sole search radar was more effective against aircraft targets than against surface targets cannot be determined. Suffice it to say that the BLUE's radar was highly ineffective.

At 0132 the BLUE was on course 051°(T), speed twelve knots and was bearing 273°(T) distant nine miles from the center of Savo Island.


At 0000 the CHICAGO Group composed of the CANBERRA, CHICAGO, PATTERSON, and BAGLEY was patrolling south of the line 125°(T) from the center of Savo Island on course 306°(T) at a speed of twelve knots. The CANBERRA was leasing the cruisers in column. The PATTERSON was 2000 yards on the port bow (due west) of the CANBERRA - the BAGLEY was 2000 yards on the starboard bow (due north of the CANBERRA). At 0015 the CHICAGO Group reversed course by column movement to 130°(T). At about 0100, when Lunga Point was bearing 180°(T) distant five miles, course was reversed again by column movement to 306°(T).

Except for the report of the CANBERRA that she heard one of the Japanese planes overhead at intervals, no contacts were made.

All American ships were in American Condition of Readiness TWO. The CANBERRA was in British Condition of Readiness TWO, to be discussed later.

At 0132 the CHICAGO Group was on course 306°(T) speed twelve knots and was bearing 145°(T) distant eleven and two-tenths miles from the center of Savo Island.


At 0000 the VINCENNES Group, composed of the VINCENNES, QUINCY and ASTORIA in column in that order and screened by the HELM 1500 yards on the port bow of the VINCENNES and by the WILSON in a similar position on the starboard bow, was patrolling along the perimeter of its patrol square at a speed of ten miles. All ships were in Condition of Readiness TWO. Course was being changed every half-hour on the hour and half-hour. At midnight, at the western corner of the patrol square, the VINCENNES Group changed course from 315°(T) to 045°(T). It continued on this leg until 0030, when it changed course to 135°(T). While on this leg, Commander VINCENNES Group determined that his group was being set to the southeast, so he decided to change course at ten minutes before and twenty minutes

- 111 -

after the hour to regain correct patrol station.

As a consequence of this decision, he changed the course of the VINCENNES Group at 0050 rather than at 0100 from 135°(T) to 225°(T); after which, he returned to his emergency cabin. He planned to make the next change at 0120. However, he did not make this change, for it appears that additional checking by the Executive Officer, who had relieved him as Primary Ship Control indicated that the VINCENNES Group had been set to the eastward. The plot showed that the VINCENNES Group would have to remain on course 225°(T) for some twenty additional minutes or until 014° in order to regain the planned position in the square for that hour.

At 0132, the VINCENNES Group was on course 225°(T), speed ten knots and was bearing 094°(T) distant 18,400 yards from the center of Savo Island.


The RALPH TALBOT was patrolling her radar and anti-submarine patrol line.** She was in American Condition of Readiness TWO. At 0000 she was on course 252°(T) at twelve knots and was in position bearing 013°(T) distant seven miles from the center of Savo Island. At about 0015 she reached the western end of her patrol line and reversed course to 072°(T). She continued on this patrol until about 0045 when she reversed course again to 252°(T). At 0120 she again reached the western end of her patrol line and reversed course to 072°(T). At this moment she was at her nearest point - 13,000 yards - to the Japanese Cruiser Force which had commenced its "rush in" on course 150°(T). Actually this range of 13,000 yards was on the destroyer YUNAGI, the last ship in the Japanese formation. Neither the RALPH TALBOT nor the YUNIGA sighted each other nor did the Japanese cruisers sight the RALPH TALBOT.

At 0132 the RALPH TALBOT was on course 072°(T) speed twelve knots and was bearing 014°(T) distant 16,000 yards from the center of Savo Island.


At midnight, the SAN JUAN Group, consisting of the cruiser SAN JUAN and H.M.A.S. HOBART 600 yards behind her in column, screened by the destroyers BUCHANAN and MONSSEN on anti-submarine stations on the starboard


*Letter, August 17th, 1942 from LCdr. J.R. Topper, USN, supervisory Officer-of-the-Deck ASTORIA, to Commanding Officer concerning Action Morning, August 9th, 1942, page 2.
**Action Report RALPH TALBOT, Preliminary Report of Action, August 8th-9th, 1942, Serial A16-3/DD390/03, August 11th, 1942.

- 112 -

and port bows respectively, of the SAN JUAN, were patrolling on a line running due north-south and a mile or so to the east of the limiting line, the meridian 160°-04'E., longitude, at speed fifteen knots.*

The general surface visibility to the eastward, through the semicircle from north to south, was one mile and the visibility in the westward semicircle though very poor was somewhat better, increasing to three miles.* This group passed through no rain squalls after 0000, August 9th, although many squalls occurred in close proximity and the clouds in the area remained low.*

No significant event disturbed the routine nature of the patrol of the SAN JUAN Group between 0000 and 0132. At this latter time, this Group was near the southern end of its patrol in the vicinity of the screen around Transport Area XRAY.


*Appendix 18 to CTG 62.6 Report of First Battle of Savo Island August 8th-9th, 1942, Letter from H.M.A.S. HOBART to CTG 62.6, August 15th, 1942, "Narrative of Night of August 8th-9th, 1942."

- 113 -



0132 August 9th to 0150 August 9th


At 0132, Commander Cruiser Force in the CHOKAI commenced his run in on course 095°(T) and at about 0133, he signalled the order: "All ships attack!"* At about 0134 he sighted an Allied destroyer at a range of about 3000 yards on his port hand bearing 76°to left of base course of 019°(T), and moving slowly on an opposite course.** One minute later, at 0135, both the AOBA and KAKO sighted this destroyer,*** which was the damaged American destroyer JARVIS en route to an Australian port for repairs. At this time the JARVIS seems to have been on course 270°(T), which course took her 3000 yards south of the southern tip of Savo Island. The Japanese Cruiser Force trained its guns on the JARVIS, but withheld fire.*** Presumably, the Japanese doctrine was not to open fire until the flagship had done so. Commander Cruiser Force, mindful of the fact that neither of the previous two destroyers had opened fire, apparently decided to continue his basic plan which was to withhold gunfire until after he had fired torpedoes at the southern force.**** He must have been surprised that he had not been fired on, as yet, by any one of the three destroyers sighted. This was especially significant in the case of the JARVIS which was not only plainly visible but was well within easy gun and torpedo range. It is probable that he did not know that the JARVIS was a disabled destroyer.

It would be of interest, therefore, to know why Commander Cruiser Force did not change his plan in view of the close presence of the JARVIS for he knew that she was enemy. He felt it was highly probable that, even if the other Allied destroyers had not reported his presence, this one would do so.***** He also felt that as this destroyer did not either open fire or attempt to escape, she might be leading him into an ambush.***** Therefore, the maintenance of secrecy, by not firing his guns, was of doubtful importance since surprise was probably now lost. Finally of course he knew that a destroyer has such a concentration of destructive power as to make it highly dangerous for the ships of his cruiser force to


*Records CRUDIV 18, August 8th, 1942, WDC Document 160984.
**War Diary 8th Fleet August 7th-10th, 1942; CIG Document 74633, May 12th, 1947.
***War Diary, KAKO, August 7th-10th, 1942, Solomons Sea Battle, WDC Document 160143.
****8th Fleet Signal Order No. 25 in Records CRUDIV 18, WDC Document 160984, August 8th, 1942.
*****USSBS The Campaigns of the Pacific War, Battle of Savo Island, page 106.

- 114 -

pass within close gun and torpedo range of one of them without taking any offensive action whatsoever. Whatever may have been his motives, it so happened that he guessed correctly, for the JARVIS failed to report the presence of the Japanese ships, apparently as a result of communication difficulties. Commander Cruiser Force thereby continued to retain the factor of surprise.

At 0136, Commander Cruiser Force sighted the CANBERRA and CHICAGO in column on bearing 120°(T)* distant 12,500 yards. He then changed course to that bearing in order to close the enemy. In the resultant relative movement, the CHOKAI and the CHICAGO Group closed one another almost head on at a relative speed of about thirty-six knots. At this time the CHOKAI had still not been sighted by the CHICAGO Group. Commander Cruiser Force then gave the order "Independent firing!" From this time onward the Japanese Cruiser Force while maintaining a speed of twenty-six knots, was strung out in a loose column generally following the CHOKAI, but actually proceeding independently by divisions and firing independently by ships. Because of the weather conditions existing at the time - night and varying visibility - and because of the distance of about 1300 yards
between ships in column, it was next to impossible to maintain correct station-keeping in formation. Commander Cruiser Force had taken cognizance of this situation as shown by his signals. He was fully aware that in a night attack, maneuvering the cruiser force and controlling the target designation and fire distribution of the force from the flagship was most difficult. He also was fully aware of the fact that in a surprise attack, targets of opportunity might present themselves to the individual ships of his command.

At 0137, the CHOKAI sighted a cruiser bearing 060°(T).* This appears from Diagram "F", to have been the VINCENNES which was leading the VINCENNES Group in column formation on course 225°(T), and was heading towards the CHOKAI. The VINCENNES at this time was approximately 18,000 yards away. This is the greatest distance at which a visual sighting was reported during the action. It is not clear what conditions existed at the time of this sighting, but it is possible that the VINCENNES may have been silhouetted, or even have been momentarily illuminated, by the fires at Tulagi or by lightning. This sighting may have been facilitated by additional messages from the cruiser planes, which reported the presence of cruisers east of Savo Island.** On the other hand, because of the above information, the sighting may have been imagined. Even so, the bearing appears to have been correct.


*War Diary 8th Fleet, August 7th-10th, 1942, CIG Document 74633, May 12th, 1947.
** USSBS Interrogation Nav. No. 83, Vol. II, Interrogation of Japanese officials, Interrogation of Captain Kenkichi Kato, IJN, page 361.

- 115 -

At 0138, when at a range of 10,000 yards from the CANBERRA, the CHOKAI fired four port torpedoes at the CANBERRA and CHICAGO.* Five minutes later, or at 0143, the CANBERRA noted two torpedoes on either side of her, coming from about 345°relative and passing down the column on roughly opposite courses.** The torpedoes appear to have been fired "curved fire ahead", to have run at a speed of about forty-five knots, and to have run true. It is not clear why the CHOKAI fired torpedoes at this time for the CANBERRA and CHICAGO were in the best possible position to avoid torpedoes from ahead and offered the smallest possible target.

Perhaps Commander Cruiser Force, sceptical lest he had been reported by the Allied ships, expected them to change course by column movement within the next several minutes in order to attempt to cross his "Tee" or at least to unmask their batteries. In such case, they might turn into his torpedoes. Perhaps he wished to obtain what benefit he could from his port torpedoes. He might not be able to fire them once he had changed course to port as he had apparently already planned to do. As it was, the CHOKAI torpedoes all missed.

CRUDIV SIX followed, in general, the movements of the CHOKAI although it did not follow in the same water. At 0139, COMCRUDIV SIX in the AOBA, which at the time appears to have been on course 110°(T) rather than the 120°(T) being steered by the CHOKAI, sighted another destroyer four degrees on his starboard bow on opposite course at a "great distance".*** This destroyer was the BAGLEY which, at the time, was 9500 yards away. At 0140, COMCRUDIV SIX sighted three ships thought to be one battleship and two heavy cruisers but which were later identified as LONDON class cruisers bearing a few degrees on the starboard bow of the AOBA*** at a range of about 9,000 yards. These ships were the CHICAGO Group.

At about the same time, when the CHOKAI had closed the range to about 7,500 yards and her torpedoes were about half way to their target, Commander Cruiser Force reported sighting four ships on a mean bearing of 120°(T).* This sighting was accurate, as these four ships composed the CHICAGO Group. Actually the PATTERSON bore 130°(T) distant 5,500 yards, and was so reported by the CHOKAI at 0140 1\2.*

At 0140 the CHOKAI changed course to 090°(T).**** She made this change

*War Diary 8th Fleet, August 7th-10th, 1942, GIG Document 74633, May 12th, 1947.
**Report of Executive Officer, CANBERRA to CTF 44 (CTG 62.6) concerning Loss of CANBERRA, August 12th, 1942.
***CRUDIV 6 Detailed Battle Report No. 8, Solomons Naval Action, August 7th-10th, 1942, CIG Document 86927, June 27th, 1947.
****Track Chart No. 2 Annexed to CRUDIV 6 Detailed Battle Report No. 81 Solomons Naval Action, August 7th-10th, 1942, CIG Document 86927, June 27th, 1947.

- 116 -

Plate 5: Battle of Savo Island, Action with the CHICAGO Group, 0143 to 0144, 9 August 1942 - chart - The Battle of Savo Island August 9, 1942 Strategical and Tactical Analysis

[- 116a -]


[- 116b -]

apparently to cross the "Tee" and, at the same time, to uncover the CHOKAI's battery as well as to unmask those of the ships following astern. Commander Cruiser Force appears to have chosen this easterly course, rather than one more to the northeast, because he wished to continue to close the CHICAGO Group at a rapid rate. This change of course is considered sound. It not only improved the relative position of the cruiser force and made it possible for each ship to deliver its maximum fire against the CHICAGO Group, thus hastening the destruction of that group; but it also - and this should have been the vital accomplishment - assisted in clearing the way to the transports.

At this same time, the YUNAGI, which was the last ship in the Japanese column sighted the JARVIS about 3000 yards on her port hand on roughly opposite course, proceeding to the westward and trying to escape.* The Commanding Officer of the YUNAGI decided, inasmuch as he was far removed from the main force, to attack the JARVIS rather than to continue on with the cruisers.* He made this decision without referring the matter to Commander Cruiser Force. His decision does not appear to have been correct, for his objective was the destruction of the Allied transports and cargo ships at Guadalcanal. He was only concerned with those Allied combatant ships - those targets of opportunity - which interfered with his accomplishment of that objective.

It is only fair to digress sufficiently to point out here that, in interrogations three years later, Commander Cruiser Force's Chief of Staff stated that: "The YUNAGI had remained outside of Savo to act as a picket and to engage Allied destroyers if they attempted to follow the cruiser force in".** COMCRUDIV EIGHTEEN stated in a similar interrogation that: "One of our destroyers (there was only one) left the formation and remained outside of Savo. I think it was assigned an Allied destroyer as a target."*** Both of these statements seem improbable on analysis, since there is no record of this assignment in Commander Cruiser Force's War Diary nor did Commander Cruiser Force realize at 0145 that the destroyer which he saw leaving the formation was the YUNAGI.**** Does it not seem reasonable to assume that Commander Cruiser Force, his Chief of Staff, and COMCRUDIV EIGHTEEN would have known of the expected activities of the


*Records CRUDIV 18, August 8th, 1942, WDC Document 160984. Also Diary Ensign Nakamura, IJN, Chapt. I, 2(J), page 34.
**USSBS Interrogation Nav. No. 109, Vol. II, Interrogation of Japanese Officials, Interrogation of Captain Toshikazu Ohmae, IJN, page 472.
***USSBS Interrogation Nav. No. 61, Interrogation of Japanese Officials, Interrogation of Rear Admiral M. Matsuyama, IJN, page 255, Vol. I.
****War Diary 8th Fleet, August 7th-10th, 1942, CIG Document 74633, May 12th, 1947.

- 117 -

Yunagi at this time, and yet their later testimonies disagree? How can these later opinions expressed above, concerning the YUNAGI's assignment outside of Savo, be reconciled with the documents of the activities of the YUNAGI found in Ensign Nakamura's Diary, which account seems to have been the Japanese Navy Department's approved version and which makes no mention of pickets at all?*

Returning now to the action, the CHOKAI commenced firing with her 8-inch battery at about 0143. She did not employ her secondary battery at this time. The accuracy of her gunfire was facilitated by the fact that, just prior to this time, one of the Japanese cruiser planes dropped illuminating flares over the Allied shipping in Transport Area XRAY off Guadalcanal Beach. These flares silhouetted the CHICAGO Group and made it an excellent target** for all the Japanese ships. Allied reports show that the first flare was sighted at about 0143 and that, by about 0145, at least four flares had been dropped. Technical Japanese data shows that a Japanese parachute flare in 1942 burned about four minutes.*** The Japanese dropped flares at intervals thereafter until about 0225.**** Their gunfire was impressive, for the CANBERRA was first hit between 0144 and 0145, and by 0148 - approximately three or four minutes later - she had been hit with at least twenty-four shells.*****

Also about 0143 the FURUTAKA, which was the fifth ship in column and, therefore, in position roughly 5200 yards behind the CHOKAI, sighted the CANBERRA and CHICAGO at a range of about 9,000 yards. She immediately commenced firing her guns and launched her four port torpedoes. She made no torpedo hits. One or more torpedoes passed under the CHICAGO about four minutes later****** - or at 0147 - and appear to have been the FURUTAKA's. It is not known how many salvos were fired by the FURUTAKA or at which Allied cruiser, but, at about 0144, she commenced having serious steering


*Diary Ensign Nakamura, IJN, Chapt. I, 2(J), page 34.
** CRUDIV 6 Detailed Battle Report No. 8, Solomons Naval Action, August 7th-10th, 1942, CIG Document 86927, June 27th, 1947.
***Handbook of Japanese Naval Explosive Ordnance OpNav 30-3M August 15th, 1945, Navy Aircraft Parachute Flares, page 252, August 9th, 1942.
****War Diary CTF 62, August 9th, 1942 and Letter August 15th, 1942 from Commanding Officer HMAS HOBART to CTG 62.6, "Narrative of Night of August 8th-9th, 1942", Appendix 18 to CTG 62.6 Report of First Battle of Savo Island, August 8th-9th, 1942.
*****Reports of Commander J.A. Walsh, Executive Officer, CANBERRA, August 12th, 1942 to CTF 44 (CTG 62.6) and of Lieut. Comdr. E.J.D. Wight, RAN, Principal Control Officer, CANBERRA August 12th, 1942 to CTF 44 (CTG 62.6) concerning Loss of CANBERRA, Night August 8th-9th, 1942.
******Action Report CHICAGO, concerning Action Against Enemy Forces August 9th, 1942, Guadalcanal-Tulagi Area, Serial 099, August 13th, 1942.

- 118 -

Plate 6: Battle of Savo Island, Action with the CHICAGO Group, 0144 to 0145, 9 August 1942 - chart - The Battle of Savo Island August 9, 1942 Strategical and Tactical Analysis

[- 118a -]

Plate 7: Battle of Savo Island, Action with the CHICAGO Group, 0145 to 0146, 9 August 1942 - chart - The Battle of Savo Island August 9, 1942 Strategical and Tactical Analysis

[- 118b -]

difficulties, which were occasioned by the effect of her own gunfiring on her steering mechanism.* (See Plate V)

At 0144, Commander Cruiser Force in the CHOIAI was making a radical change to the north, either to engage the VINCENNES Group or to attack the Allied shipping at Tulagi. He settled on the new course of 050°(T) two minutes later.** Students of naval warfare will ponder over his reasons for heading in that direction rather than towards Guadalcanal Beach. For his objective, self-assigned it is true but nevertheless the objective which he had issued to his command, was the destruction of the transports and cargo ships. He had heavily damaged the principal ships which were between his own forces and the transports and cargo ships off Guadalcanal, and had thereby cleared the way to those targets. Should not his next move have been to close the transport area XRAY immediately and destroy as many of these ships as possible? He did not do this but headed towards other targets. By his decision, which was somewhat similar to that of the Commanding Officer, YUNAGI, he lost an opportunity to convert a tactical success into a strategical victory which could have been of enormous consequence to the Japanese cause.

About this time the AOBA, which had closed to a range of approximately 4200 yards on the CANBERRA, opened fire with her 8-inch battery and, in addition, fired three torpedoes from her starboard torpedo battery at that ship. Although she thought that she had made three hits on the CANBERRA,*** it seems probable that all torpedoes missed.

At about 0144, COMCRUDIV EIGHTEEN, with the TENRYU and YUBARI, also changed course to the northeast to 040°(T) and away from the CHICAGO Group.* (Plate VI)

It is not possible to determine why he made this change of course rather than to have followed the movements of COMCRUDIV SIX. COMCRUDIV EIGHTEEN stated that the plan for the command was to break up into an inner and outer group after passing around Savo Island and that CRUDIV EIGHTEEN, plus possibly the YUNAGI, was to compose the inner group.****


*Records CRUDIV 18, August 8th, 1942, WDC Document 160984.
**War Diary 8th Fleet, August 7th-10th, 1942, CIG Document 74633, May 12th, 1947.
***Action Reports from Tabular Records of Japanese Cruisers WDC Documents 160623 and 161407. Also, from Track Chart #2 Annexed to CRUDIV 6 Detailed Battle Report #8, Solomons Naval Action, August 7th-10th, 1942, CIG Document 86927, June 27th, 1947.
****USSBS Interrogation Nav. No. 61, Vol. I, Interrogation of Japanese officials, Interrogation of Rear Admiral M. Matsuyama, IJN, page 255.

- 119 -

This seems doubtful for such a plan would have been in direct opposition to Commander Cruiser Force's basic plan which directed the destruction of the transports and cargo ships. According to COMCRUDIV EIGHTEEN the basic plan was to be accomplished as follows: "CRUDIV SIX was to attack the transports at Guadalcanal while CRUDIV EIGHTEEN was to attack the Tulagi transports."* It has been suggested that the erratic movements of the FURUTAKA may have forced the change. This also seems doubtful, as the FURUTAKA, despite her steering troubles, was proceeding in southeasterly direction and toward the CHICAGO Group. It appears more likely that, having been directed to operate independently and noting that the heavy cruisers were about to engage the Allied heavy cruisers, COMCRUDIV EIGHTEEN decided that prudence required him to remove his light cruisers from that phase of the action, and so he retired in a northerly direction. Whatever may have been his reason, it appears unsound. The presence of the two light cruisers in the Japanese battle line would have increased the number of targets the Allies would have had to keep under fire, would thereby probably have reduced the volume of fire directed at any Japanese ship, would have increased the number of torpedoes available for launching, and would have increased the number of intermediate guns (ten 5.5 inch) that could have been employed against the CHICAGO Group.

At 0145 the KAKO, which was next in column behind the AOBA sighted the CHICAGO Group five degrees on her starboard bow at a range of about 4,000 yards.** If this were her first sighting, then she had been a little slow in discovering the Allied ships, possibly because she was not "closed up" in formation; or because she was not as alert as the other cruisers. It is probable that she first became alerted to the presence of the Allied ships when they were illuminated by the airplane flares.

Also at 0145, Commander Cruiser Force noted that a destroyer, which he did not recognize, bearing 290°(T) had reversed course.*** This was the YUNAGI, which by this time was swinging to her new course of about 270°(T). She had reversed course apparently without notifying the other Japanese ships of her action. (Plate VII)

As a result of the casualty to the FURUTAKA and the reversal of course by the YUNAGI, the cruiser force now fell into two groups to be known as the Eastern Group and the Western Group. The Eastern Group was composed of the CHOKAI, AOBA, KAKO and KINUGASA; the Western Group was composed of CRUDIV EIGHTEEN and was joined a few minutes later by the FURUTAKA. The YUNAGI was operating by itself.


*USSBS Interrogation Nav. No. 61, Vol. I, Interrogation of Japanese officials, Interrogation of Rear Admiral M. Matsuyama, IJN, page 255.
**War Diary KAKO, Solomons Sea Battle, August 7th-10th, 1942, WDC Document 160143.
***War Diary 8th Fleet, August 7th-10th, 1942, CIG Document 74633, May 12th, 1947.

- 120 -

Plate 8: Battle of Savo Island, Action iwth the CHICAGO Group, 0146 to 0147, 9 August 1942 - chart - The Battle of Savo Island August 9, 1942 Strategical and Tactical Analysis

[- 120a -]

Plate 9: Battle of Savo Island, Action with the CHICAGO Group, 0147 to 0148, 9 August 1942 - chart - The Battle of Savo Island August 9, 1942 Strategical and Tactical Analysis

[- 120b -]


Both of the cruiser groups were proceeding to the northward. The Eastern Group proceeded in a northerly direction to engage the VINCENNES Group. The Western Group appears to have paralleled the Eastern Group and was in a supporting position. It is doubtful if the Western Group was informed as to the plans of Commander Cruiser Force. However, this was soon cleared up, for at 0146 COMCRUDIV EIGHTEEN sighted the VINCENNES Group bearing 060°(T) distant what he thought was 6,600 yards but which was, in fact, 14,800 yards from the TENRYU (Diagram "F"). Course was changed at 0146 1/2 to 070°(T).* Apparently this sighting was made possible by the light of aircraft parachute flares which were dropped at the time by one of the two Japanese cruiser planes. COMCRUDIV EIGHTEEN reported that the glare of the illuminating flares revealed five heavy cruisers. Actually the number of ships reported was correct but the recognition was somewhat in error, possibly because of the great range, as there were, in fact, three heavy cruisers and two destroyers in the VINCENNES Group on bearing 060°(T).

Slightly before 0146, the KAKO fired three of her four starboard torpedoes at what she thought was a battleship but which was the CHICAGO bearing 145°(T), distant about 3600 yards. She then changed course to follow the AOBA which was making an "S" turn to the northeast. At 0146 she opened fire on the Allied ship. She also opened fire on a cruiser of the KENT Class, the CANBERRA, employing both her main and secondary batteries. She claimed having hit the CANBERRA on her first salvo but after that, she discovered while in her turn that most of her main battery guns could not bear any more, being up against the stops; and ceased firing.** She also claimed sinking the CANBERRA with three torpedo hits.***

Also at 0146, the FURUTAKA fired one starboard torpedo at a destroyer bearing 130°(T) distant 4500 yards.*** This was the PATTERSON. The FURUTAKA thought that she had sunk the PATTERSON, but her torpedo had missed. (Plate VIII illustrates the action during this minute.)

The FURUTAKA's position at this time was outside the wake current (and to the right) of the KINUGASA and the other ships of CRUDIV SIX, which were ahead of her. Whether this offset position was the result of the KINUGASA's turning to the right to fire her port torpedoes at 0143, or was incident to steering difficulties is not clear. But it is clear that the Commanding Officer, FURUTAKA was at this time considering the


*Records CRUDIV 18, August 8th, 1942, WDC Document 160984.
**War Diary KAKO, August 7th-10th, 1942, Solomons Sea Battle, WDC Document 160143.
***Action Reports from Tabular Records of Japanese Cruisers WDC Documents 160623 and 161407. Also from Track Chart No. 2 annexed to CRUDIV 6 Detailed Battle Report No. 8, Solomons Naval Action, August 7th-10th, 1942, CIG Document 86927 of June 27th, 1947.

- 121 -

problem as to how to rejoin in single column behind the KINUGASA, being mindful of the FURUTAKA's faster movement outside the wake current.* To complicate matters, he found himself headed on a collision course with the CANBERRA which had changed course to the northward. He also noted that the CHOKAI and the preceding cruisers of CRUDIV SIX were engaging enemy vessels (CHICAGO Group). He thereupon abandoned the idea of rejoining in column, possibly because he did not trust his steering gear in the emergencies of battle; and instead, attempted to turn to the left and extricate himself.* At 0147 1/2, he was able to change course to about 000°(T)** and commenced firing with both main and secondary batteries at the disabled CANBERRA.*** (Plate IX)

At about 0147 COMCRUDIV SIX observed a heavy cruiser sink as a result of a torpedo hit. This sinking occurred about 2000 yards broad on the starboard beam of the flagship AOBA. The heavy cruiser was the CHICAGO, which did not sink, but which suffered the loss of a small part of her bow from one torpedo hit and suffered an additional hit on her starboard side amidships which did not detonate.**** These two torpedoes apparently came from the KAKO.

Also at about 0147, COMCRUDIV EIGHTEEN sighted a destroyer of the CRAVEN class going in an opposite direction bearing about 150°(T) distant 6500 yards. This was the destroyer PATTERSON which was changing course to the westward. He immediately changed course to 110°(T), illuminated the PATTERSON with searchlights and opened fire. By 0150, he considered that the PATTERSON was sunk.** Actually the PATTERSON was not sunk, but was seriously damaged on her topside aft. Why he considered her sunk is not explained, but it is probable that PATTERSON's change of course to the south and then east affected his judgment. Neither of the Japanese ships suffered other than slight damage from the PATTERSON's guns.

At 0148, the CHOKAI fired four starboard torpedoes***** by what appears to have been "curved fire ahead" at the VINCENNES which was about 9500 yards away almost directly ahead of her. The CHOKAI was headed on course 050°(T). (Plate X)

At the same time COMCRUDIV SIX sighted the VINCENNES Group off the starboard bow of the AOBA which was swinging northward following the


*Interrogation of Captain Toshikazu Ohmae, IJN, Battle of Savo Island by Historical Branch, G2 General Headquarters, Far Eastern Command, Tokyo, August, 1949.
**Records of CRUDIV 18, August 8th, 1942, WDC Document 160984.
***Track Chart No. 2 Annexed to CRUDIV 6 Detailed Battle Report No. 8, Solomons Naval Action, August 7th-10th, 1942, CIG Document 86927, June 27th, 1947.
****Summary of War Damage to Battleships, Carriers, Cruisers and Destroyers. Bureau of Ships, September 15th, 1943, page 29.
*****War Diary 8th Fleet, August 7th-10th, 1942, CIG Document 74633, May [no year listed].

- 122 -

Plate 10: Battle of Savo Island, Action with the CHICAGO Group, 0148 to 0149, 9 August 1942 - chart - The Battle of Savo Island August 9, 1942 Strategical and Tactical Analysis

[- 122a -]


[- 122b -]

movements of the CHOKAI. The range at this time was 12,000 yards, yet the first two ships sighted were identified as one ASTORIA-class and one PORTLAND-class cruiser, and shortly thereafter the third cruiser was identified as of the ASTORIA-class.* These identifications were remarkably correct when the range and low visibility at night are considered. Actually all three cruisers were of the ASTORIA-class.

At 0149 or about one minute after firing his torpedoes, Commander Cruiser Force in the CHOKAI changed course to 069°(T).** It can be seen from Diagram "J" that this change of course headed the CHOKAI directly towards transport Area YOKE at Tulagi. This heading brings up the question as to whether Commander Cruiser Force had now given up his first objective, the transports and cargo ships off Guadalcanal, and had decided to attack those at Tulagi, or whether he had decided next to engage the VINCENNES Group. The weight of evidence available seems to support the latter view; for, as will be shown later, he made no attempt to close the transports and cargo ships at Tulagi.

Simultaneously at 0149, the FURUTAKA which had succeeded in clearing the CANBERRA and in changing course to 000°(T) in spite of her steering difficulties, fired her remaining three starboard torpedoes at that ship at a range of 3500 yards. The FURUTAKA claimed three hits*** but it now appears that all torpedoes missed.

At 0150, Commander Cruiser Force observed the sinking of a KENT-class cruiser (evidently the CANBERRA) by two torpedo hits. Both the CHOKAI and the KAKO claimed credit for this sinking. Actually the CANBERRA, although heavily hit by gunfire did not sink at this time, but remained afloat for about four and one-half hours. The CHOKAI was only about 2000 yards from the CANBERRA when she made this observation. During this action, other Japanese ships made similar claims of sinkings; none of which sinkings, in fact, occurred at the times claimed. These discrepancies show the difficulty of making correct observations during a night action, even with well trained and well equipped lookouts.

To summarize the action from 0132 to 0150, a total of seventeen torpedoes had been fired by the CHOKAI, AOBA, KAKO and FURUTAKA at the CANBERRA and CHICAGO; and one torpedo was fired by the FURUTAKA at the PATTERSON. The Japanese thought that they had made nine hits on the cruisers; and the FURUTAKA thought that she had sunk the PATTERSON with her single torpedo. Actually, it is improbable that any torpedo hits


*CRUDIV 6 Detailed Battle Report No. 8, Solomons Naval Action August 7th-10th, 1942, CIG Document 86927, June 27th, 1947.
**War Diary 8th Fleet, August 7th-10th, 1942, CIG Document 74633, May 12th, 1947.
*** Track Chart No. 2 Annexed to CRUDIV 6 Detailed Battle Report No. 8, Solomons Naval Action August 7th-10th, 1942, CIG Document 86927, June 27th, 1947.

- 123 -

were made on the CANBERRA. This improbability was established by a Board of Enquiry subsequently held at Sidney.* Two hits were made on the CHICAGO. One accomplished some slight damage; the other, a dud, accomplished none. No torpedo hits were made on the PATTERSON.

During this period, although several of the Allied ships fired partial salvos from intermediate batteries and one destroyer fired four torpedoes, only CRUDIV EIGHTEEN reported having observed any enemy gun splashes and none of the cruisers reported having observed torpedo tracks. These results clearly show that the Allied ships were caught by surprise and fired both guns and torpedoes without having adequately solved their gunfire and torpedo problems.

The 0150 positions of the Japanese ships, with relation to the center of Savo Island, were:


The CHOKAI was bearing 130°(T) distant 13,500 yards.
The AOBA was bearing 135°(T) distant 13,000 yards.
The KAKO was bearing 140°(T) distant 13,000 yards.
The KINUGASA was bearing 147°(T) distant 13,250 yards.


The FURUTAKA was bearing 153°(T) distant 12,150 yards.
The TENRYU was bearing 151°(T) distant 10,900 yards.
The YUBARI was bearing 156°(T) distant 10,100 yards.


The YUNAGI was bearing 205°(T) distant 8,150 yards.


*Note to Appendix F to British Battle Summary No. 21, Naval Operations at the Landings in the Southern Solomons, August 7th-10th, 1942, dated 1944.

- 124 -



0132 August 9th to 0150 August 9th


At 0132, the CHICAGO Group was on course 306°(T) at speed twelve knots approaching the northern end of its patrol line. The CHICAGO, PATTERSON, and BAGLEY were still in American Condition of Readiness TWO. Which meant that one-half of the armament and all the fire control and communication stations were sufficiently manned and ready for immediate use.* The CANBERRA was in a modified British Condition of Readiness TWO.** British Condition of Readiness TWO was normally a higher condition of readiness than its American counterpart. It required that all battle stations be fully manned, but permitted a certain proportion of the personnel to rest at their battle stations, and a few of them to leave their stations to obtain food and to answer calls of nature. The American Condition of Readiness ONE EASY, employed later in the war may have evolved from this British Condition of Readiness. The modifications in British Condition of Readiness TWO, as employed by the CANBERRA, relaxed the watch to something a little more than that on the American ships present, and lessened the ability of the CANBERRA to meet a surprise attack. Half of the CANBERRA gun batteries - both 8-inch and 4-inch AA - were fully manned and ready, and the other half were relaxed with the crews sleeping in the vicinity of their battle quarters. All gunnery communications personnel were awake and sentries were alert in all shell-handling rooms.***

None of the Commanding Officers, with the exception of the PATTERSON were awake. The others were asleep on the bridge level, relaxing from the heavy grind of the operations of the preceding days preparatory to the further operations to be expected in the morning. None of them gave adequate consideration to the possibility of a night attack by Japanese surface forces. Although both COMSOWESPAC and CINCPAC had broadcast over their respective Fox schedules the 1025 contact report on the Japanese forces off Bougainville, it is apparent that this report was given no more than passing notice.


*War Instructions, U.S. Navy 1934, FTP143, Chapter VII, para. 806, page 28.
**Report of Admiral Arthur J. Hepburn, USN, (Ret.) to CINCPAC, May 13th, 1943, on Informal Inquiry into Circumstances Attending the Loss of the VINCENNES, etc., on August 9th, 1942, page 33.
***Report of Executive Officer CANBERRA, August 12th, 1942, to CTF 44 (CTG 62.6) concerning Loss of CANBERRA.

- 125 -

The weather conditions at this time were smooth sea, light easterly winds, no moon and variable visibility. The Japanese reported that the visibility, augmented by the illuminating flares was excellent to the southeast.* On the Allied side, the BAGLEY reported that the visibility was very low, and the sky heavily overcast with intermittent showers.** The CHICAGO reported electrical storms in and near the area.*** The CANBERRA reported the visibility as 4000 yards in the clear weather, but there were rain patches.**** The visibility was fairly thick in the sector between the BAGLEY and the PATTERSON which were roughly 2400 yards apart. During the early stages of the action, that is, between 0132-0150, neither of these ships sighted the other.** The differences between the Japanese and Allied reports of the weather conditions within the same area resulted from the fact that the Japanese ships were on the edge of a cloud bank which had formed to the southeast of Savo Island whereas the Allied ships were about one and one-half miles outside of the edge of this cloud bank. Consequently the Allied ships, silhouetted as they were by flares, were much more distinct to the Japanese lookouts than the Japanese ships were to the Allied lookouts.

The weather conditions were not the whole answer. Among other factors was the Japanese advantage in knowing what they were looking for and the manifested superiority of their lookouts, who were well equipped and trained for night operations, as compared with those of the Allied forces. Poor visibility alone should have had no effect on the Allied radar, but it is probable that the heavy cloud bank was ionized and sent back echoes similar to land. The visual sightings by the Japanese out-distanced the electronics detection of the Allies. It is probable that the failures of detection in the CHICAO Group were due in a large measure to the same factors which affected the BLUE as pointed out earlier. These were (a) the adverse effect that land masses had on radar operations at that early date, (b) the inexperience of the personnel in war operations, and (c) some fatigue of personnel.

The cloud bank off Savo Island also interfered with the navigation of the Allied ships. Examples of this are here given by way of illustration. The BAGLEY reported at 0144 that the left tangent of Savo Island bore 310°(T) distant one and seven-tenths miles** whereas it bore 315° distant seven miles. The PATTERSON reported at 0146 that the


*CRUDIV 6 Detailed Battle Report #8, Solomons Naval Action, August 7th-10th. 1942, CIG Document 86927, June 27th, 1947.
**Action Report BAGLEY, Night Engagement August 9th, 1942, Tulagi-Guadalcanal Area, Serial 016, August 13th, 1942·
***Action Report CHICAGO, Concerning Action Against Enemy Forces August 9th, 1942, Tulagi-Guadalcanal Area, Serial 099, August 13th, 1942.
****Report of Executive Officer, CANBERRA, August 12th, 1942, to CTF 44, (CTG 62.6) concerning Loss of CANBERRA.

- 126 -

western end of Savo Island bore north distant four miles,* when, in fact, it bore 325°distant seven and one-half miles. Actually, these bearings were taken on this cloud bank and not on Savo Island at all. Both of these destroyers reported the Japanese ships close to Savo Island when first detected. Diagram "F" shows that at 0144 the CHOKAI was no nearer Savo Island than five miles.

(1) Action by CANBERRA**

When the action commenced, the CANBERRA was leading the formation on heading 310°(T). At 0143 her port lookout sighted a ship dead ahead. The bridge personnel were not able to verity this contact; so both the Commanding Officer and Navigator were immediately called and they arrived on the bridge in a matter of seconds. (See Plate V)

The CANBERRA was soon alerted to a dangerous situation, for at 0143 1/2 she sighted four torpedoes - two on either side, passing roughly parallel to the column of Allied cruisers and on opposite course.*** These four torpedoes had been fired by the Japanese flagship, CHOKAI at 0138.****

The CANBERRA immediately went to British Condition of Readiness ONE, which alerted everyone in a manner equivalent to the American "General Quarters". She then observed two large destroyers or small cruisers on the port bow at a distance thought to be under one mile by one observer*** but 5000 yards by another.***** The armament was immediately trained on these targets.*** Actually, at 0143, the CHOKAI was crossing the "Tee" of the Allied cruiser column at a range of 4500 yards and CRUDIV SIX was coming in on a bows-on approach to the CANBERRA on course 110°(T).

The Commanding Officer, CANBERRA, without waiting for instructions from Commander CHICAGO Group changed course to port to bring his starboard battery to bear. However, after a few seconds he changed course to starboard, increased speed to full speed, and prepared to fight a normal action with his port battery. He appears to have realized that his first


*Action Report PATTERSON, Engagement with Enemy Surface Ships Night August 8th-9th, 1942, Savo-Guadalcanal-Tulagi Area, Serial 001, August 13th, 1942.
**The breakdown of the tactical command of the CHICAGO Group has made it necessary hereinafter to evaluate separately the conduct of the Commanding Officers of the several ships within that group. Each of these Commanding Officers either operated independently by choice, or was forced to do so by the demands of the situation.
***Report of Executive Officer, CANBERRA, August 12th, 1942, to CTF 44 (CTG 62.6) Concerning Loss of CANBERRA.
****War Diary, 8th Fleet, August 7th-10th, 1942, CIG Document 74633, May 12th, 1947.
*****Letter August 12th, 1942, from LCdr. E.J.D. Wight, RAN, to CTG 62.6, Concerning Loss of CANBERRA, Night August 8th-9th, 1942.

- 127 -

change of course would have forced his ship to fight a reverse action which might permit the Japanese cruisers to move directly into the transport areas; and that if he was to prevent this, it was necessary to interpose his command between the transports and the enemy cruisers. This could best be accomplished by a change of course to starboard and by an increase in speed, which he executed at 0145. (See Plate VI)

It is unfortunate that he did not advise his immediate superior in command in the CHICAGO Group of his contacts with the enemy. Whatever may have been the reason for this failure, it in no way relieved him of the responsibility for alerting the command to the presence of the enemy. The failure of the CANBERRA, as well as the BAGLEY and CHICAGO, to alert the command may have contributed in no small degree to the ensuing destruction of the VINCENNES Group.

As the CANBERRA swung to starboard, she was hit by numerous Japanese shells. This happened before she could get her guns to bear on the Japanese ships to port, and before Condition ONE could be effectively set. Although the Executive Officer stated that within one minute - that is, by 0145 - she had been hit by at least twenty-four shells,* this is not believed to have been feasible. It is considered more probable that she received her damaging hits within the next four minutes. Although she was fired on by the CHOKAI, AOBA and FURUTAKA, she was maneuvering radically and it is scarcely probable that hits in any number were made until the range was found by spotting ladders. The BAGLEY's report substantiated this conjecture by stating that the CANBERRA was not hit until about the third salvo observed.**

At about 0146, the CANBERRA lighted a spread of torpedoes approaching her. These had been fired at about 0144 by the AOBA. The Executive Officer, CANBERRA reported that the CANBERRA had been hit between the boiler rooms by one or two torpedoes.* This report has been found to be incorrect by the conclusions of an inquiry subsequently held at Sidney by the Royal Australian Navy. That Board found: "That after very careful examination of the survivors, the Board of Enquiry.....came to the conclusion that this was improbable (that there were any torpedo hits). No fuel oil whatsoever was seen around the ship after the action, nor had any survivors experienced the shock of an underwater explosion. It was assumed that the list which gradually increased from five degrees to thirty degrees in about five hours, was caused by a shell going right through the ship and penetrating the starboard side under water."*** The radical turns of the CANBERRA had saved her from being hit by the AOBA's torpedoes. (See Plate VII)


*Report of Executive Officer, CANBERRA August 12th, 1942, to CTF 44 (CTG 62.6) Concerning Loss of CANBERRA.
**Action Report BAGLEY, Night Engagement, August 9th,  1942, Tulagi-Guadalcanal Area, Serial 016, August 13th, 1942.
***Note to Appendix F to British Battle Summary No. 21, Naval Operations at the Landings in the Southern Solomons, August 7th-10th, 1942, dated 1944.

- 128 -

The CANBERRA continued to swing to starboard while being hit by the gunfire of the Japanese cruisers. However, her swing was never fully completed and she tended to straighten out on a northeasterly course. Battle damage most probably was the direct cause of this modification in the change of course.

At about 0148, the Commanding Officer was mortally wounded, and the Gunnery Officer was killed. The Executive Officer was notified that the captain was down and he assumed command.

The CANBERRA endeavored to return the enemy fire.* She may have succeeded in firing a few shots at about 0147** with her port 4-inch battery, but did not succeed in firing any torpedoes.*** These shots were not well directed. Nothing in the available Japanese reports indicates that shell splashes or torpedo wakes - let along hits - were observed at this time·

In summarizing the action between 0144 and about 0148, the CANBERRA suffered heavy damage, which disabled her and rendered her almost defenseless. Her damage in part was as follows: at least four hits on the 4-inch gun deck; at least two hits in the torpedo space; one hit on the after director; one hit in the plotting room; one hit between the guns of "X" turret; one hit on "A" turret; at least one hit in each boiler room which resulted in the loss of all steam and power and one hit on the bridge. As a consequence of 8-incb armor-piercing shells passing through the ship below the water line, the ship took a list to starboard of five degrees which slowly increased to a maximum of thirty degrees some hours later.

At 0150, the CANBERRA, without power but still with way upon her, was in a position bearing 149°(T), distant seven and six-tenths miles from the center of Savo Island.

(2) Action by CHICAGO

At 0143, the CHICAGO, in column about 600 yards astern of the CABBERRA, sighted the first of the four aircraft flares which the Japanese had dropped in Transport Area XRAY, bearing between 110° and 120°(T). The CHICAGO had also sighted, just prior to this time, two orange-colored flashes near the surface of the water in the vicinity of Savo Island,** which might possibly have been the flashes from the torpedo tubes of FURUTAKA. (See Plate V)


*Report of Executive Officer, CANBERRA, August 12th, 1942, to CTF 44 (CTG 62.6) Concerning Loss of CANBERRA.
**Action Report, CHICAGO Concerning Action Against Enemy Forces August 9th, 1942, Guadalcanal-Tulagi Area, Serial 099, August 13th, 1942.
***Report of Admiral Arthur J. Hepburn, USN (Ret.) to CINCPAC, May 13th, 1943 on Informal Inquiry into Circumstances Attending the Loss of the VINCENNES etc., on August 9th, 1942, page 20.

- 129 -

At about 0145, the CHICAGO noted that the CANBERRA was swinging to starboard. The Officer of the Deck of the CHICAGO must have assumed that the CANBERRA had turned to starboard for one of two reasons: (a) that she had sighted something which required an emergency turn, or (b) that she had suffered a steering casualty. He could normally have expected to be advised immediately in either event; but in this case he was doomed to disappointment, for the CANBERRA made no report.

However, he did not have long to wonder, for at 0146 as the CANBERRA moved to starboard, the Officer of the Deck noted two dark objects between the CANBERRA and the PATTERSON, and one other object to the right of the CANBERRA.* Viewing from the left, the first two objects were the KINUGASA and KAKO; the object to the right of the CANBERRA was the AOBA. It is assumed that the Officer of the Deck sounded "General Quarters" at this time bringing the Commanding Officer to the bridge almost immediately. Meanwhile, he continued to follow in the wake of the CANBERRA, and made preparations to fire a star shell spread to starboard in order to illuminate the AOBA. (See Plate VII)

Once the Commanding Officer was on the bridge and had familiarized himself with the unusual circumstances of the situation, is it not surprising that he did not immediately attempt to broadcast a general warning to the entire task force about the presence of the enemy ships, including any amplifying information available such as: enemy types, formation or disposition, location, course and speed? Having failed to do this, is it not equally surprising that he did not attempt to notify either Commander VINCENNES Group - who, as Screening Commander in the absence of CTG 62.6, was in command of the western screen, and therefore responsible for the coordination of the screening groups for battle, or CTG 62.6 who, as overall Screening Commander, might still be able to direct some defensive measures even though he was located for the moment in Transport Area XRAY? Similar questions could be asked regarding his failure to notify CTF 61, CTF 62, CTF 63 and COMSOPACFOR.

He not only failed in his responsibility as regards these matters but he also failed in his additional responsibilities as Commander CHICAGO Group. He, of course, knew that the CANBERRA was cognizant of the situation; for she was already maneuvering for position. But What about the BAGLEY and the PATTERSON? It was his responsibility, as Officer-in-Tactical Command of this group to lead it into battle and to give the necessary orders to his four ships in order that this might be effectively done. He did not do this but instead allowed each ship to operate independently. Perhaps he found himself unable to perform the dual function of Group Commander and Commanding Officer, or perhaps under the pressure of events he simply forgot that he was the Group Commander. His failure to exercise command properly contributed in a large degree to the unfortunate events which were to follow in Iron Bottom Sound.


*Action Report, CHICAGO, Concerning Action Against Enemy Forces, August 9th, 1942, Guadalcanal-Tulagi Area, Serial 099, August 13th, 1942.

- 130 -

While preparations were in progress to fire the star shell spread to illuminate the AOBA, the starboard bridge lookout reported a torpedo wake to starboard at 0146.* The Commanding Officer started to turn to starboard with full right rudder,* presumably in an attempt to parallel the torpedo track. But within a few seconds, and before the turn to starboard could be executed, the Main Battery Control Officer reported having sighted the bubbles of two torpedo wakes approaching from port to starboard on bearing 345° relative.* Because the first wake to starboard could not be made out from the bridge and because those on the port hand could be clearly seen, the Commanding Officer immediately turned to port with full left rudder to parallel these latter wakes.*

It seems wise to digress here to explain the several torpedo wakes by the CHICAGO. (See Plates VIII and IX)

(a) The failure of the bridge personnel to make out the wake reported to starboard confirms the fact that it was a dead wake left on the smooth sea by a torpedo that had already passed. As the CHICAGO, steaming at twelve knots, should have been crossing, at about 0146, the wakes of the CHOIAI's four torpedoes which had passed the CANBERRA a little earlier it seems clear that the dead wake was from one of these torpedoes.

(b) The bubbling wakes to port - bearing 345° relative to CHICAGO's course of 306°(T) - lay on true bearing 291°. This bearing was the approximate direction from which the FURUTAKA had fired four torpedoes at the CHICAGO Group at 0143. As a Japanese cruiser torpedo, running for four minutes at maximum speed (fifty knots)* could have reached a point a few hundred yards on bearing 291°(T) from the CHICAGO at 0147,** it is considered that the wakes to port of the CHICAGO were made by the FURUTAKA's torpedoes.

Shortly after 0147, Main Battery Control sighted one of the torpedo wakes on bearing 345° relative heading for the CHICAGO's port bow. In half a minute there was an explosion well forward under the bow. The Commanding Officer thought that this torpedo had struck the CHICAGO on the port bow.* Actually the torpedo which had been fired by the FURUTAKA had missed. The CHICAGO had been hit instead on her starboard side by two torpedoes which she had failed to observe. One of these torpedoes blew off a small part of the bow at frame Number Three;* the other struck at frame Number Ninety-five (starboard), but failed to detonate.*** The underwater damage from this torpedo did not cause any perceptible change in the


*Action Report CHICAGO Concerning Action Against Enemy Forces August 9th, 1942, Guadalcanal-Tulagi Area, Serial 099, August 13th, 1942.
**Report of Basic Characteristics of Foreign Torpedoes Design Department, Naval Torpedo Station, Newport, R.I., January 29th, 1948, page 6.
***Summary of War Damage to U.S. Battleships, Carriers, Cruisers, and Destroyers, October 17th, 1941 to December 7th, 1942, NavShips (374), Bureau of Ships, Navy Department, September 15th, 1943, page 29.

- 131 -

CHICAGO's trim at the time.*

The torpedoes which struck the CHICAGO were fired by the KAKO. This conclusion has been reached after careful study of the KAKO's target identification, her time of firing, and the range and bearing of her target, all compared with the firing times and targets of the other Japanese cruisers which fired torpedoes at the CHICAGO Group. The plotting analysis of all these Japanese torpedo tracks is shown on Diagram "F". By way of partially explaining this evaluation, the KAKO stated that her target was "What appeared to be a battleship"** - the CHICAGO Class had been mistaken previously by the Japanese for battleships. Since the Japanese had already correctly identified the CANBERRA by her three smoke pipes as a KENT Class cruiser, the KAKO's target could have been none other than the CHICAGO. This evaluation is supported by (a) the fact that the estimated range to the KAKO's target was given as 3500 meters, * which closely approximated the actual range of about 3600 yards to the CHICAGO at the same time and by (b) the fact that the KAKO's target was on a relative bearing of 30° which was within five degrees of the bearing of the CHICAGO at the time of firing.

The CHICAGO's change of course to port to parallel the observed torpedo tracks was sound, for it was an emergency measure to avoid being hit. However, it became essential for her Commanding Officer to change course back to starboard, once the immediate danger from torpedoes had passed, and to follow the CANBERRA to the eastward in order to protect the Tulagi transports from the enemy force that was moving in their direction, as well as to insure that the two cruisers were in a position of mutual support and concentration. But as the CHICAGO swung to port and was approximately paralleling the FURUTAKA's torpedo tracks, her Commanding Officer decided to swing further to port because he observed "what was then believed to be a destroyer in a position to discharge torpedoes."* He had actually sighted the FURUTAKA, which at 0147 was headed approximately toward the CHICAGO and was 4000 yards away on bearing or about 295°(T). The Commanding Officer, CHICAGO did not visualize the connection between the FURUTAKA and the torpedoes which had just passed the CHICAGO, and was taking further precautions against being hit. He steadied on course 283°(T).

At about 0147 1/2, the CHICAGO observed gunfire flashes close aboard on bearing 320° relative. She also observed gunfire flashes from two ships on her starboard bow.* The flashes to port were from the guns of the PATTERSON; those to starboard were from the guns of the TENRYU and the YUBARI. These opposing units were heavily engaging one another. It is


*Action Report CHICAGO Concerning Action Against Enemy Forces August 9th, 1942, Guadalcanal-Tulagi Area, Serial 099, August 13th, 1942.
**War Diary KAKO, August 7th-10th, 1942, Solomons Sea Battle, WDC Document 160143.

- 132 -

singular that the gunfire of these ships was the first gunfire reported by the CHICAGO; for the CANBERRA - only 600 yards away - had been under heavy fire from three Japanese heavy cruisers during the preceding three minutes.

At this time the CHICAGO fired two 4-gun salvos of star shell spreads on a mean bearing of 320° relative, set for a range of 9200 yards.* Since this bearing was in the direction of the PATTERSON, it is assumed that the Commanding Officer, CHICAGO wished to investigate the area beyond the PATTERSON to discover, if possible, the extent of Japanese penetration into Iron Bottom Sound, and the character of the enemy force.

At the same time the CHICAGO also fired two 4-gun salvos of star shells to starboard, on a mean bearing of 045° relative, set for a range of 5000 yards, to illuminate a cruiser - the AOBA - which the CHICAGO reported as being about 5000 yards beyond the CANBERRA and firing upon her. The CANBERRA was bearing 045°relative distant 1200 yards from the CHICAGO.* The range setting on these star shells should have been about 7200 yards if the 6200 yards range estimated by the CHICAGO was correct. But since the range to the AOBA was actually but 4000 yards, the range setting of 5000 yards should have proven satisfactory inasmuch as the correct firing range for a star shell burst was roughly 1000 yards beyond the target at an altitude of about 1500 feet. However, the error in ranging was overshadowed by another error in fuse-setting, for it became painfully apparent to the Commanding Officer that his star shells had failed to burst.

Just prior to 0148, the CHICAGO received her first hit by a shell which struck the starboard leg of her foremast, detonated approximately over the foreward smoke pipe and showered shell fragments over the topside. It also bent the topmast, causing interference with the radar antenna for Director One.*

At about 0149, the CHICAGO observed the PATTERSON engaging "two destroyers".* These were the light cruisers TENRYU and YUBARI. The CHICAGO incorrectly thought that the PATTERSON was employing her searchlights.* The PATTERSON was employing star shells, and it was the enemy who was employing searchlights against her.

At 0150, the CHICAGO was still on course 283°(T), speed twelve knots and was bearing 155°(T) distant 16,000 yards from the center of Savo Island.

(3) Action by BAGLEY

The BAGLEY was in her screening position about 1600 yards on the starboard bow of the CANBERRA on course 310°(T) when the action commenced.


*Action Report, CHICAGO, Concerning Action Against Enemy Forces, August 9th, 1942, Guadalcanal-Tulagi Area, Serial 099, August 13th, 1942.

- 133 -

At 0144, she sighted a number of unidentified ships slightly on her port bow which appeared to her to be very close to Savo Island at an estimated range of 3000 yards from the BAGLEY on an estimated course of 125°(T) at high speed.* These ships were the CHOKAI and the AOBA, KAKO, and KINUGASA which were astern of her. This contact proves to have been quite accurate as borne out by the plot on Diagram "F". The CHOKAI was actually bearing 303°(T), distance 2800 yards from the BAGLEY, and the three above mentioned ships of CRUDIV SIX were on course 120°(T). Coincident with making the contact, the BAGLEY also observed two or three Japanese salvos land short of the CANBERRA. After the first hit was scored on the CANBERRA, there followed a concentration of heavy fire from the Japanese cruisers.*

The BAGLEY did not report this contact to her immediate superior in command in the CHICAGO, nor did she broadcast it to all ships. This was unfortunate, for, as has been pointed out in the cases of the CANBERRA and CHICAGO, information concerning enemy contacts was of vital importance to the entire command. The necessity for reporting enemy contacts should have received priority over all other considerations.

The officer of the Deck immediately ordered "General Quarters." The BAGLEY attempted to launch a torpedo attack with her starboard torpedo-battery.* She swung "hard left" and increased speed to twenty-five knots. Her action had been so precipitous that she swung completely past the safe firing bearings before primers could be inserted.* She then decided to continue her swing in order to bring her port battery to bear.* This action of the Commanding Officer, BAGLEY in making this change of course to port rather than to starboard is questionable, for by so doing, he placed himself in the position of seriously embarrassing his own cruisers** without receiving any commensurate gain, such as decreased torpedo firing time. Should he not have realized that the Commander CHICAGO Group, in order to continue to carry out his objective, must necessarily change course to starboard, and that the BAGLEY must keep clear by also changing course to starboard? A starboard turn would automatically have interposed his ship between the Japanese cruisers and the Allied transports. At the same time it would have taken the BAGLEY out of the probable line of fire, as well as have kept her in a position to provide anti-submarine protection for the cruisers of the CHICAGO Group.

At 0145, the BAGLEY observed the flares which had been dropped over Area XRAY. (Plate VII shows the situation at this time.)

At about 0149, she steadied on a northeasterly course and fired four torpedoes in a northwesterly direction from her Number Two torpedo tube mount. Her targets were identified by her as two light cruisers of the


*Action Report, BAGLEY, Night Engagement August 9th, 1942, Tulagi­-Guadalcanal Area, Serial 016, August 13th, 1942.
**Oral Statement June 1st, 1949, by CTF62 to Commodore R.W. Bates, USN (Ret.) Head of Department of Analysis, U.S. Naval War College.

- 134 -

TENRYU Class and two heavy cruisers of the ASHIGARA Class.* Since at this time the two Japanese light cruisers - the TENRYU and YUBARI - were bearing about 310°(T) from the BAGLEY and since the two Japanese heavy cruisers - the FURUTAKA and KINUGASA - were approximately on the same bearing, it is believed that the BAGLEY fired on the mean bearing of 310°(T). These torpedoes missed probably because the two light cruisers effected a course change a minute later, and probably because of poor aiming which was due in part to the fact that the visibility to the north was poor, making the enemy ships not only very indistinct in general, but actually lost to view as the torpedoes were fired.*

After firing the torpedoes, the BAGLEY again turned left for the purpose of scanning the passage between Guadalcanal and Savo Island.* This action was of doubtful correctness, for the BAGLEY had not been released by the Commander CHICAGO Group nor had she as yet received orders to proceed elsewhere. She was still part of the cruiser screen, and should properly have maintained reasonable station until such time as she was either released, assigned to other duty, or it became apparent that the Officer-in-Tactical Command was unable to exercise effective control over the existing emergency situation. Perhaps, having received no instructions from the Commander CHICAGO Group, the Commanding Officer, BAGLEY felt that this latter condition obtained.

At 0150, the BAGLEY was swinging with left rudder from her previous firing course of 040°(T), speed 25 knots; and her position was bearing 145°(T), distant eight and three-tenths miles from the center of Savo Island.

(4) Action by PATTERSON

The screening position of the PATTERSON was on the port bow of the CANBERRA between 1500 and 2000 yards. At 0140 1/2 the CHOKAI had observed the PATTERSON on bearing 130°(T) at distance 5500 yards,** thus revealing her actual position as 1900 yards west of the CANBERRA. (See Diagram "F").

The PATTERSON herself made no contact on the enemy force until five and one-half minutes after having been sighted by the leading Japanese cruiser, during which time she advanced 2200 yards farther along her patrol course of 306°(T). At 0146 she visually sighted an unidentified ship on bearing 305°(T) - dead ahead - at an estimated distance of 5000 yards, close in to what she thought was the western end of Savo Island,** but what was actually the heavy cloud bank already mentioned in previous sections. Her contact was the FURUTAKA, whose relative position as fifth


*Action Report BAGLEY, Night Engagement, August 9th, l942, Tulagi­-Guadalcanal Area, Serial 016, August 13th, 1942.
**Action Report PATTERSON, Engagement with Enemy Surface Ships Night August 8th-9th, 1942, Savo-Guadalcanal-Tulagi Area, Serial 001, August 13th, 1942.

- 135 -

ship in the Japanese column - 5200 yards astern of the CHOKAI, or six minutes run at twenty-six knots - would have placed her at 0146 in approximately the vicinity of the CHOKAI's 0140 position. The actual range from the PATTERSON to the FURUTAKA was 3700 yards rather than the 5000 yards she reported. (See Plate VIII)

It has been suggested that the PATTERSON's screening position to the west of the CANBERRA should have made it possible for her to sight the enemy before other ships in the CHICAGO Group, particularly before the BAGLEY. Plotting analysis proves otherwise, for the BAGLEY's screening station - nearly a mile north of the CANBERRA and therefore about 2500 yards northeast of the PATTERSON - was in much better position to contact the CHOKAI at 0144 at the range of 2800 yards, as pointed out in the previous section. Under the weather conditions, PATTERSON and BAGLEY were not in sight contact with each other.

The Commanding Officer, PATTERSON - who was on the bridge when contact was made - immediately sounded "General Quarters", endeavored to notify the CANBERRA and CHICAGO by blinker signal, and passed a warning broadcast to all ships over the TBS voice radio at about 0146: "Warning! Warning! Strange ship entering harbor."* The CANBERRA was already engaged with the enemy, and although she saw the visual signaling** she was too busy to reply. It is doubtful that either the CHICAGO or CANBERRA received the voice warning.

The Commanding Officer rang up maximum speed at 0146 1/2 and altered course to port in order to unmask his starboard gun and torpedo batteries.*** This change of course is considered sound for by changing course to port the PATTERSON should have succeeded in rapidly launching her torpedoes to starboard. This was so important as to take priority over all other objectives. A change of course to starboard would probably have delayed the firing of the torpedoes as much as several minutes. It might also have caused her to foul the line of fire of her own heavy cruisers.

As the PATTERSON swung to port, the Commanding Officer directed that the torpedoes be fired when ready; but the Torpedo Officer failed to hear this order,*** and the torpedoes were not fired as the PATTERSON steadied on heading 270°(T). Consequently the FURUTAKA escaped without being hit, and was observed by the PATTERSON to turn left to a northerly course at 0147 1/2.*** (See Plate IX)

As the FURUTAKA cleared the range, the PATTERSON observed two other Japanese cruisers farther beyond on bearing 70° relative, a true bearing of 340°. She identified them as one MOGAMI Class and one JINTSU Class


*Written Statement August 10th, 1942, by Lt. (jg) C.P. Clarke, USNR, QUINCY.
**Report of Cdr. J.A. Walsh, Executive Officer, CANBERRA, August 12th, 1942, to CTF 44 (CTG 62.6) concerning loss of CANBERRA.
***Action Report PATTERSON, Engagement with Enemy Surface Ships Night August 8th-9th, 1942, Savo-Guadalcanal-Tulagi Area, Serial 001, August 13th, 1942.

- 136 -

cruiser. The plot in Diagram "F" shows that these enemy cruisers were undoubtedly the TENRYU and YUBARI of CRUDIV EIGHTEEN.

The Commanding Officer - whose voice was recognized by the RALPH TALBOT - immediately transmitted via TBS voice radio the following general alarm to the entire task force: "All ships. Warning! Warning! Three enemy ships inside Savo Island.** By this action and by his previous action at 0146 he evidenced a thorough appreciation of the urgent necessity of immediately informing the Officer in Tactical Command, the Group Commanders, and all Commanding Officers of vital information concerning the enemy. It is singularly significant that he was the only Commanding Officer in the CHICAGO Group who endeavored to inform his own Group Commander, and other interested group commanders, of the initial contact with the enemy, notwithstanding the fact that the CANBERRA and BAGLEY had both made earlier contacts.

At about 0147 1/2 while broadcasting the warning, the Commanding Officer received a report of a torpedo wake fifty yards on his starboard quarter. This is believed to have been the wake of the single torpedo which the FURUTAKA had fired at the PATTERSON at 0146.** It had missed and was running beyond the PATTERSON.

The PATTERSON then brought her guns to bear and opened fire on both the TENRYU and YUBARI. She first fired two four-gun star shell spreads. Thereafter she maintained the illumination with one gun only, and employed her remaining three guns to fire service ammunition at her targets which were at actual range of about 5600 yards.*** During this engagement she was illuminated by searchlights from both Japanese light cruisers.

Although she zigzagged at high speed she was not entirely able to avoid hits. She was hit at about 0148 in the crew shelter of Number Four gun by one shell which ignited several rounds of ready service powder and enveloped the after section of the ship in flames. This forced both of her after guns out of action and caused casualties of ten men killed, or missing, and ten men wounded. As the PATTERSON swung around to an easterly course every effort was made to extinguish the fires around guns Number Three and Four so that they might open fire to port as soon as they could bear. These efforts were sufficiently successful to permit the firing of gun Number Four, but they were unsuccessful as regards gun Number Three. Slightly after 0149 the PATTERSON resumed firing to port


*Letter August 15th, 1942 from Commanding Officer RALPH TALBOT to CTF 62, Concerning Morning Report Heard on Night of August 8th-9th, 1942, Serial 008.
**CRUDIV 6 Detailed Battle Report #8, Solomons Naval Action, August 7th-10th, 1942, CIG Document 86927, June 27th, 1947.
***Action Report, PATTERSON, Engagement with Enemy Surface Ships Night August 8th-9th, 1942, Savo-Guadalcanal-Tulagi Area, Serial 001, August 13th, 1942.

- 137 -

employing guns Numbers One, Two and Four. She claimed that she maintained rapid and accurate fire and succeeded in scoring several hits during the next minute on the rear JINTSU-type cruiser (the YUBARI).* This claim was supported in part by the CHICAGO's report that she had observed the PATTERSON score hits on the YUBARI at this time,** and in part by the YUBARI's report of receiving "some scratches (during the night action) from the light fire of an enemy destroyer",*** but the YUBARI did not indicate at what time the damage was received. The YUBARI was illuminated during this minute by PATTERSON's star shells which had been fired before gun Number Three went out of action.

All firing ceased at 0150. At this time the PATTERSON was on course 090°(T) at speed of probably twenty-seven knots; and was bearing 164°(T), distant 16,000 yards from the center of Savo Island.


During the engagement between the Japanese Cruiser Force and the CHICAGO Group, the VINCINNES Group, which should have been alerted to the situation, continued its patrol in a more or less unalerted condition at a speed of ten knots. At 0132, it was proceeding along the southwest leg of the patrol square on course 225°(T).

All ships were in Condition of Gunnery Readiness TWO but there was an important difference between them. The VINCENNES and ASTORIA manned two guns in all three turrets,**** whereas the QUINCY manned one forward turret and half of her after turret.***** In addition all nine turret guns on all three ships were loaded with projectiles but not primed. These turret guns had been loaded during threatened air attack the previous day.


*Action Report, PATTERSON, Concerning Engagement With Japanese Surface Ships Night August 8th-9th, 1942, Savo-Guadalcanal-Tulagi Area, Serial 001, August 13th, 1942.
**Action Report CHICAGO, Concerning Action Against Enemy Forces August 9th, 1942, Guadalcanal-Tulagi Area, Serial 099, August 1942.
***Action Reports from Tabular Records of Japanese Cruisers, WDC Documents 160623 and 161407.
****Action Report VINCENNES, Report of Action Between USS VINCENNES and Japanese Heavy Cruiser Near Savo Island on Night of August 8th-9th, 1942, Serial 0022, August 16th, 1942; Letter August 13th, 1942, from LCdr. W.H. Truesdell, USN, Gunnery Officer, ASTORIA, Concerning Night Cruiser Action, August 9th, 1942.
*****Memorandum February 10th, 1943, from Cdr. H.B. Heneberger, USN, Senior Surviving Officer, QUINCY, to Admiral Arthur J. Hepburn, USN (Ret.) Concerning Additional Information in regard to the EX-USS QUINCY, at the Battle of Savo Island, August 9th, l942.

- 138 -

The Main Ship Control Officer in the VINCENNES at this time was the Executive Officer;* in the QUINCY the Assistant First Lieutenant and Damage Control Officer,** and, in the ASTORIA, the First Lieutenant and Damage Control Officer.***

All cruiser Commanding Officers were fully dressed and asleep in their emergency cabins adjoining their respective pilot houses. It is assumed that the Commanding Officers of the destroyers were asleep also in view of the fact that they had experienced several long and difficult days. The fact that all Commanding Officers were asleep is not surprising. Condition of Readiness TWO was designed for a situation whenever action with enemy surface ships was probable but exact information of the enemy was not available.***** It gave about half of the ship's company an opportunity to relax. Certainly, the Commanding Officers who had been on their bridges for long hours over several days were entitled to a rest, and it was their duty to keep themselves fit for further action. The only necessity was the insurance that the watch was alerted to the developing situation. Only the Commanding Officer, VINCENNES appears to have done this. The Commanding Officers of the other cruisers seem to have over­looked this possibility with the consequence that they, as will be shown later, failed to appreciate adequately the situation when surprise occurred.

Weather conditions in the north channel area were about as follows: night dark and overcast with occasional light mists and rain; light breeze from the southeast; ceiling about 1500 feet; average visibility 10,000 yards; heavy clouds around and to the south of Savo Island; sea smooth.

The Commanding Officers of the three cruisers were aware of the contact report of the first R.A.A.F. Hudson plane on the Japanese Cruiser Force at 1025, August 8th.***** In addition, the Commanding Officer, VINCENNES was aware of the second contact report on this group made by another R.A.A.F. Hudson plane at 1100, August 8th.***** Inasmuch as all ships copied the Fox schedules, it is presumed that the Commanding Officers of the screening destroyers WILSON and HELM had received at least one of these contact report.


*Cdr. W.E.A. Mullan, USN.
**LCdr. Edmund Billings, D-V(S), USNR.
***LCdr. J.R. Tepper, USN.
****War Instructions, U.S. Navy 1934, FTP 143, Chapter VIII, "General Instructions for Vessels Underway", page 28, para. 808.
*****Action Report VINCENNES, Report of Action Occurring off Savo Island (Guadalcanal-Florida Island) Area - Night of August 8th-9th, 1942, Serial 0021, August 14th, 1942, page 2; Action Report ASTORIA, Battle of Savo Island, Serial AP37/A16-3/(00500), August 20th, 1942, pages 2 and 3; Letter August 19th, 1942, from W.R. Daniel, Chief Radio Electrician, USN, QUINCY, Concerning Loss of QUINCY to Commanding Officer.

- 139 -

The Commanding Officer, VINCENNES, who was also the Group Commander, had estimated the situation on the basis of these contacts, and had alerted himself to the enemy capability of preceding with this cruiser force at high speed and of attacking the Allied forces some time during the current midwatch.* As a result he had entered a notation of this capability in his Night Orders and had stressed the importance of being particularly alert. He did not, however, alert the other ships of his command to the possibility of a Japanese night attack. This was a fatal omission, for he was the Group Commander and it was his responsibility to insure that his command was fully informed as to the probability of night action. Perhaps the Group Commander felt that any messages received by the VINCENNES had also been received by the other ships of the VINCENNES Group and that correct action had already been taken within each ship. As will be shown later, nothing could have been further from the truth; and the result was that all ships were caught by surprise.

At this time, there was considerable concern in the ASTORIA over the fact that her forward main battery fire-control radar was out of commission. Since her after fire-control radar had also been out of commission for some time because of a shorted transformer which could not be repaired aboard, this meant that her main battery was entirely without fire-control radar. Although this placed her at no disadvantage as compared to the Japanese, who were without radar of any description, the ASTORIA nevertheless realized the value of the fire-control radars and expedited every effort toe complete repairs on the forward fire-control radar. These repairs were successfully completed just prior to contact with the enemy.**

At 0140, the VINCENNES changed course to 315°(T) in order to regain the group's planned position in the square; for as has been pointed out earlier, this group had been set to the south eastward. The QUINCY and ASTORIA followed around in column, the latter completing the turn at 0144.*** The destroyers completed adjusting their screening position after the turn at about the same time.

At 0145 - immediately after the task group had settled on the new course - both the QUINCY and the HELM endeavored to fix their positions by cross-bearings on Savo Island. The resulting fixes were considerably in error, because both ships mistook the edge of the heavy cloud-bank south of Savo Island as the actual left tangent of that island, and because both found the right tangent somewhat obscured by the haze in that


*Action Report VINCENNES, Report of Action Occurring off Savo Island (Guadalcanal-Florida Island) Area - Night of August 8th-9th, 1942, Serial 0021, August 14th, 1942, page 2.
**Letter August 13th, 1942 from LCdr. W.H. Truesdell, USN, Gunnery Officer, ASTORIA to Commanding Officer Concerning Night Cruiser Action, August 9th, 1942.
***Action Report, ASTORIA, Battle of Savo Island, Serial AP37/A16-3/(00500), August 20th, 1942.

- 140 -

direction. Evaluation of data available from all sources indicates that the QUINCY's fix was about 6700 yards in error in the direction of 345°(T) from her actual position, whereas the HELM's fix was about 3200 yards too far to the northwest. These errors in piloting are discussed here to indicate the navigational difficulties which beset the Commanding Officers of ships forced to operate in restricted areas under combat conditions and under poor visibility. Radar has considerably reduced these difficulties since August 1942, but has not fully eliminated them.

Some of the ships of the VINCENNES Group evidenced more concern at this time with the danger of submarine attack then with the possibility of surface ship attack.* As a consequence, when a distant underwater explosion was felt at 0144 the ASTORIA, as well as other Allied ships which had felt the same explosion, believed that enemy submarines were being depth charged. This was incorrect, for not only were there no submarines being depth charged at this time, but no Japanese submarines had as yet arrived in Iron Bottom Sound. The disturbance noted was most likely caused by the explosion of the CHOKAI's self-destruction torpedoes after their run past the CHICAGO Group.

At 0146, the HELM, WILSON, VINCENNES and QUINCY observed the aircraft flares which had been dropped by a Japanese plane in Area XRAY to silhouette the transports. Some of these ships, notably the VINCENNES, QUINCY and WILSON, also observed gunfire. The Allied ships reported the illumination as highly effective. In particular, the VINCENNES reported having sighted four flares in the direction of Florida Island, and also reported having observed a group of ships (apparently the CHICAGO Group) being silhouetted by a "great display of light" southeast of Savo Island. The VINCENNES was able to see clearly enough to estimate the course of the CHICAGO Group as northwest, and therefore parallel to the course of the VINCENNES Group.** The ASTORIA alone did not observe the flares, silhouettes, or gunfire at this time, and fully five minutes were to elapse before the Supervisory Officer of the Deck observed them.** There is no doubt that the strenuous employment of this and other ships in the past forty-eight hours had impaired somewhat the efficiency of the various ships' companies through fatigue, but this placed no disadvantage on the ASTORIA as compared to the other ships. It is evident from the statements of ASTORIA personnel that the primary attention in a number of stations was diverted to other considerations than that of maintaining a sharp lookout in all sectors. The attention on the bridge was directed specifically to station-keeping after the completion of the turn,*** in order to take full advantage of the


*Action Report, ASTORIA, Battle of Savo Island, Serial AP-37/A16-3/(00500), August 20th, 1942.
**Letter August 14th, 1942 from Executive Officer VINCENNES to Commanding Officer Concerning Night Action, August 9th, 1942, off Savo Island.
***Letter August 17th, 1942 from LCdr. J.R. Topper, USN, Supervisory Officer of the Watch, ASTORIA to Commanding Officer Concerning Action on Morning August 9th, 1942, page 1; also Letter August 17th, 1942 from Lt. (jg) N.A. Burkey, Jr., USNR, ASTORIA to Commanding Officer concerning Action on August 9th, 1942, page 1.

- 141 -

anti-submarine screening destroyers. The Gunnery Officer was intent with the radar operator upon testing the repaired radar, using the ships ahead in column as targets.* Lookouts were scanning the sky, searching for the Japanese plane heard overhead,** and in some cases may have been looking out of their own sectors. Certainly, not the least important source of information at night or in low visibility under combat conditions is the prompt detection and reporting of visual sightings. Accordingly, Commanding Officers should take measures to insure that alert and keen lookouts are maintained in all sectors during such circumstances, lest they be surprised.

In reaction to the sighting at 0145, WILSON and probably HELM ordered Condition of Readiness ONE. The Commanding Officer in VINCENNES was notified at some of the occurrences observed at this time, and hastened to the bridge.

Immediately after his arrival on the bridge, the Commanding Officer - who as Group Commander, was the Officer-in-Tactical Command of the VINCENNES Group - discussed the circumstances of the moment with his officers and ascertained the situation from them. He determined that the position of the VINCENNES was about three and one-half miles east of Savo Island; although his actual position based on Diagram "F" was seven miles away on bearing 098°(T) from the center of that island, the error being attributable, of course, to the difficulty in obtaining accurate bearings because of low visibility. He observed three or four star shells, and he observed a ship to the southwest which he believed was firing star shells toward the southeast. This observed ship appeared to be under fire from another ship which was about thirty degrees toward the southeast.*** The action thus observed by the Group Commander was that taking place between the Japanese light cruisers of CRUISER DIVISION EIGHTEEN and the PATTERSON (which was firing the star shells actually toward the northwest). The Group Commander saw no other ships, no heavy gunfire, nor any searchlights.*** While he continued his discussions and observations, he decided at about 0147 that the CHICAGO Group had contacted an enemy destroyer. He estimated that this destroyer was a part of a Japanese diversionary plan designed to draw the VINCENNES Group out of position so that the Japanese Main Attack Force might approach via the north passage, and pass through his sector in order to attack the transports.***


*Letter August 13th, 1942, from LCdr. W.H. Truesdale, USN, Gunnery Officer, ASTORIA to Commanding Officer Concerning Night Cruiser Action, August 9th, 1942.
**Letter August 17th, 1942, from LCdr. J.R. Topper, USN, Supervisory Officer of the Watch, to Commanding Officer, ASTORIA Concerning