Mel Maas spent the afternoon of 8 August with Fletcher, as frantic alerts alternated with dead calm. He joined the deliberations that led to the recommendation to withdraw the carriers one day early. At 1745, a half hour before sunset, Maas went below to eat. The three carriers cruised thirty miles southeast of Cape Henslow (that is, eighty-five miles southeast of Lunga) and had not already pulled out, as commonly asserted. Noyes, in tactical command, prescribed the customary nighttime dogleg to clear the immediate area, in this case southwest of San Cristóbal, from where TG-61.1 could retire, if Ghormley approved, or return north and again at dawn on 9 August support whatever part of Turner’s Amphibious Force lingered in the invasion area.1
By leaving Maas missed an important event. Ens. Robert B. Stahl, one of the notorious eavesdroppers in the Sara’s sky forward, recalled that during the afternoon he heard someone say on the flag bridge below: “Cruisers?” Fletcher and the staff hurriedly conferred, but to Stahl’s dismay they shifted out of earshot into flag plot.2 The commotion arose from an 1837 message transmitted from Canberra that placed, at 1025, three cruisers, three destroyers, and two seaplane tenders or gunboats off the southeast coast of Bougainville, some 320 miles northwest of Lunga.3 The enemy held course 120 degrees, speed fifteen knots, and therefore could have come a good long way. The position and composition corresponded well with information Fletcher received about 0630 from reports of the submarine S-38 observing St. George’s Channel sixty miles south of Rabaul.4 Years later Yeoman Tom Newsome, Fletcher’s combat talker, vividly remembered the evening conference in flag plot. Fletcher wanted to attack immediately, if feasible. Ramsey assured him that the carriers could not search and conduct night strikes, except in bright moonlight, which was not the case that night. Otherwise the risk of losing the planes was too great, especially with a big carrier battle in the offing. Fletcher released the strike crews. He explained in 1947 to Baldwin, “Of course the carriers would have been of no value in a night action.” A longtime surface warrior who had yet to fight a surface battle, Fletcher naturally was “confident that the cruisers could look out for themselves in that kind of action.” His trust proved sadly misplaced.5
Fletcher later commented, “At the time my decision to withdraw was made it was not known that enemy forces were approaching,” a fact confirmed from the Maas papers. Yet the Savo debacle spawned a persistent myth that the carriers knew by mid afternoon in plenty of time to strike the cruiser force, but, inexplicably, Fletcher chose not to attack. It appears that when the 1025 message was actually received about 1845, the Saratoga alerted Torpedo Eight for possible launch. One TBF pilot told correspondent Clark Lee about a few cruisers and destroyers detected to the north. “Get your life jacket and helmet. Maybe we’ll be taking off soon to hit those babies.” However, VT-8 stood down. Lee subsequently placed this exchange in the afternoon. On 10 August a VF-5 pilot repeated a rumor in his diary: “To cap matters off, on the 8th, we knew from scout planes’ reports that Jap surface forces were coming into our area. At one point they were only 150 miles away, with no plane protection at all. But no dive bombers or torpedo planes were sent out, nothing was done.” After sundown on 8 August when Fletcher actually learned of the existence of the force, no one knew where it was. At no time during daylight on 8 August was it within 150 miles of the carriers. In 1943 Captain Riefkohl (Vincennes) and Capt. William G. Greenman (Astoria) stated to Admiral Hepburn that they heard warnings of the cruiser force sometime that afternoon, but the Astoria’s communications officer declared that was not so. Crucial in the creation of this myth was the lengthy delay in forwarding the late morning contact report. It seemed incredible it did not reach the carriers much earlier than it did. Exhaustive investigation by Hepburn, as well as postwar analyses by Admiral Dyer and Commo. Bruce Loxton, RAN, confirmed the first inkling of the enemy cruiser force reached the Amphibious Force and the carriers only after sundown—far too late for Fletcher, at least, to do anything about it.6
Thus on the evening of 8 August Fletcher, while awaiting Ghormley’s response, held to the plan of shifting the carriers south but retained the option of returning to Guadalcanal. At 1857 in the fading light the carrier task units formed a column and swung southeast at fifteen knots. Newsome recalled no dissension regarding the withdrawal or anxiety over the Screening Group, especially against what was thought a much weaker surface force. Even with the expected imminent departure of the Santa Cruz landing force and the escorted transports, Crutchley should have five heavy cruisers (including the Astoria and Vincennes), a light cruiser, and nine destroyers, all equipped with radar. Illustrating the lack of concern that evening in the Saratoga, Maas dined and played darts with some of the staff until 2000, then he repaired to his cabin to write his situation estimate. He did not learn until the next day of the sighting report of cruisers logged on the evening of the eighth.7
Such information Fletcher personally received the balance of that night did nothing to dispel the initial impression that an immediate surface threat to the Screening Group was unlikely. The latest intelligence still rated naval strength in the Rabaul region at just three or four heavy cruisers and four light cruisers, plus a destroyer squadron. Air searches seemingly accounted for most of them. At 2047 Leary advised that at 1101 another aircraft placed two heavy cruisers, one light cruiser, and a smaller ship only seven miles from 1025 search contact. Very likely both planes spied the same force. The other aircraft sightings that morning and early afternoon (but that Comsowespac only reported that evening) amounted to a few ships strung out from Buka to Shortland. That certainly fitted Leary’s notion, radioed after 2230, that the enemy was occupying Buka and Bougainville in strength. McCain’s negative search summary finally arrived about midnight. Turner’s evening 7 August situation report (only received, as previously noted, by Ghormley and Fletcher late on the morning of the eighth) had requested special scouting against an approach of forces from westward. The Enterprise’s afternoon search, extending 260 miles west and 220 miles north from western Guadalcanal, certainly helped fill that lack. At 2155, in one of the few McCawley messages to reach its addressees straightaway, Turner predicted the so-called seaplane tenders mentioned in the earlier aircraft sighting report were bound for Rekata Bay on Santa Isabel, 130 miles northwest of Lunga, where they could use “torpedo planes” to menace the Amphibious Force. The soundness of that assessment will be discussed below. Turner desired McCain to bomb Rekata the next morning. Fletcher knew the Enterprise afternoon search sighted no activity at Rekata or nothing at all for that matter, but if the supposed seaplane group kept advancing at fifteen knots, it would not have yet come far enough to be sighted. McCain replied at 0254 on 9 August that his planes would check out Rekata as early as possible. B-17s would bomb about noon, and PBYs would be on call for a night torpedo strike.8
At 2330, 8 August, after passing beyond southern San Cristóbal, the three carrier groups angled southwest for ninety minutes, then at 0100 wheeled northwest toward Guadalcanal about 115 miles distant. At 0330 Fletcher finally received the message Comsopac composed nearly five hours before. “Recommended withdrawal [of the] carriers [is] approved.” Ghormley differed with the critics who argued, “Fletcher could have stayed longer.” He called it “a question of judgment,” where “hindsight is better than fore sight.” Obviously, “The enemy could arrive in force and catch our Task Force short of fuel,” so Fletcher’s request “had to be considered very seriously.” Ghormley accorded “the man on the spot” the benefit of the doubt, because he “knew the situation in detail; I did not.” Ironically, Fletcher told Callaghan on 27 July that his boss would probably enjoy a superior appreciation of the overall situation and thought Comsopac could better assess the impact of withdrawing the carriers a day early.9
Ghormley’s dispatch also resolved the deteriorating fuel situation. The Kaskaskia and Platte, escorted by the Clark and hopefully the Perkins, were to depart Nouméa at dawn on 9 August, cut east around New Caledonia, and head northwest toward Espíritu Santo at fifteen knots. They should join TG-61.1 during daylight on 10 August as it came south to meet them. Comsopac’s earlier plan called just for the Platte and Clark to start north on the forenoon of the ninth. The Perkins dinged a propeller coming into Nouméa on the eighth, and if Fletcher also needed the Kaskaskia, he would have to provide the escort. Now with TG-61.1 low on oil, both fueling groups would go north and keep at least six miles apart, so the same sub could not target both oilers. Should the Perkins not participate, as was likely, the two oilers were to draw near the Clark for “maximum protection.” Although pleased to have both oilers because he needed that additional fuel, Fletcher worried that they might have to be entrusted to only one destroyer.10
At 0400 as TG-61.1 drew within seventy-five miles of Guadalcanal and 135 miles of Tulagi, Fletcher led the carrier column around to the northeast, then at 0430 swung them southeast into echelon to make for the fueling rendezvous. Because the formation would loosen considerably, he provided a position for reassembly after dawn. Having seen the carriers safely onto their new heading, he retired to his sea cabin on the flag bridge. The stress of the last several days also took its toll on him. Again it must be emphasized that nothing Fletcher learned that night altered his belief that Turner had pulled out most, if not all, of the transports according to plan, leaving just TF-44 and the cargo ships, and that the first stage of Watchtower was ending on a high note.
Late on the afternoon of 8 August Turner received welcome news that the marines finally overran much of Lunga, including the all-important airfield. Also it seemed he finally got a handle on the unloading of supplies and gear. He recalled in 1950 that he alerted TF-62 “to prepare to depart early on the 9th, as the hourly unloading reports from the X-Ray Group led me to believe that the supply situation on shore would then be reasonably good.” Turner did not elaborate whether he meant the whole force of transports and cargo ships or just the transports, but after the morning of 8 August he never admitted he ever envisioned withdrawing any of them separately. After sunset he copied Fletcher’s message to Ghormley proposing to pull out the carriers. Within the hour came the tardy Comsowespac sighting report of three cruisers, three destroyers, and two seaplane tenders or gunboats discovered off central Bougainville more than eight hours before. In Turner’s retrospective account these two messages ignited incandescent anger in the McCawley, but whether that occurred immediately or actually a little afterward remains unclear. Ostensibly if the supply situation was as favorable as Turner thought or hoped it was, carrier air support during daylight on 9 August might not be necessary, for most of the Amphibious Force would have already cleared the area. In that event it also would not matter what the mixed group of ships found near Bougainville might do. They did not seem to Turner (as they did not to others) to be a surface force intent on sweeping through the invasion area that night. What really appears to have ruined Turner’s evening was word that Yoke Group’s unloading at Tulagi “was lagging,” meaning he might not get all his landing force out “early on the 9th” after all. Suddenly the message from Fletcher and the belated sighting report took on highly sinister implications.11
The venomous account Turner wrote for his administrative history called Fletcher’s recommendation to pull out the carriers “almost unbelievable,” because it “changed considerably the plans which had been arranged.” Obviously that circumstance would only have pertained if Turner had anticipated having the carriers remain on 9 August, which was, of course, what Fletcher said he would do on 27 July. Therefore Turner had every right to be perturbed when Fletcher proposed to leave early. But as detailed before, Turner’s assertions of such a betrayal collide head-on with his equally vehement insistence that Fletcher declared the carriers would not stay beyond two days. If Fletcher indeed warned he would withdraw after only two days, Turner could scarcely profess astonishment when it came to pass. Indeed, Turner succeeded in getting it both ways. He propagated the contradiction concerning Fletcher’s preinvasion pledges and the fable that he himself had always anticipated keeping the entire Amphibious Force in the invasion area for four to five days, because transports as well as cargo ships needed that long to unload. It also usefully soft-pedaled criticism of the extended delays in unloading.12
Thus with contradiction and myth, Turner tainted Fletcher’s action as “desertion” in the face of the new peril revealed by the 1025 sighting report. He deduced the enemy tenders would use the sheltered waters of Rekata Bay to employ float torpedo planes and professed to fear such strikes more than the massive land-based air attacks already weathered on 7–8 August. Although accepted without question, Turner’s assessment that the seaplane tenders themselves could execute torpedo attacks is highly suspect. Intelligence available to him amply demonstrated that Japanese seaplane tenders, unlike those in the U.S. Navy, handled only single-engine float planes that were far too light to wield torpedoes. Only by servicing flying boats could these tenders even contribute to a torpedo threat. Even so, the limited number of available flying boats would pose far less threat than the fast twin-engine land planes that already assailed the Amphibious Force. Instead a far likelier explanation for the suspected presence of seaplane tenders was to reestablish the reconnaissance capability lost with the fall of Tulagi. Yet Turner, working from an obsolete premise based on his earlier aviation experience, conveniently seized on the supposed torpedo threat from seaplane tenders to go ahead and execute his previously mentioned “prepare to depart” order and withdraw the entire Amphibious Force before sunrise on 9 August. He certainly intended to return sometime later to complete the unloading.13
Turner drew up a report as of 2000, 8 August, which, like the previous night, he inexplicably did not transmit for eighteen hours, whereas a prompt airing of his concerns might have given Fletcher a reason to stay.14 CTF-62 message 090230 (1330, 9 August) is also worth quoting in full:
Progress retarded today by beach congestion stubborn enemy resistance north side one enemy air attack and one false alarm. Desire express appreciation for warnings by coast watcher Bougainville.
Elliotts fire caused by torpedo plane diving on board fire out of control removed personnel and had ship torpedoed but she grounded on shoal and is still burning Jarvis anchored not taking water engines and boilers okeh hull bottom open from frame 30 to 55 deck damaged can make 4 to 7 knots own power will try to get her to Roses [Efate] starting tonight 14 enlisted missing 7 wounded.
Continuing discharge of cargo tonight plan temporary retirement from area tomorrow see separate dispatch [CTF-62 message 081405, noted below].
Shore Guadalcanal Kukoom and airfield occupied. Enemy troops and construction workers scattered considerable stores equipment captured. No advance yet to east. Condition airfields reported separately.
Florida area Marines hold Tulagi Gavutu and Tanambogo except for occasional snipers. Assault on Makambo planned for 2300 GCT on 9th [1000 on the tenth]. Severe enemy resistance after initial landing required use all 3 battalions of 2nd Regiment. This will delay Apricot [Ndeni] operation. Recommend use there of 7th Regiment [then in Samoa]. In view further operations CTF 62 remains on McCawley.
Thus Turner completely abandoned the idea of pulling out just the empty or near-empty transports and their escorts and leaving Crutchley’s TF-44 to protect the five partially unloaded cargo ships anchored close off Lunga. He never explained that radical departure from the original plan.
At 2045 Turner summoned Vandegrift and Crutchley to the McCawley to confer in person as soon as possible. That action would cause serious consequences, later laid at Fletcher’s door. Turner stated in 1950 he needed to see Vandegrift, “Largely for the purpose of asking him to go to Tulagi to find out and report the exact situation as to the amount of cargo that had been landed.” That was a curious errand for a division commander. One wonders why he did not send a staff officer or simply direct the commander of the Yoke Group transports, Capt. George B. Ashe in the Neville, or the local marine commander, Brig. Gen. William H. Rupertus, or both, to report. Perhaps the supply situation at Tulagi seemed such a mess that Turner thought only Vandegrift’s clout could straighten it out. His “chief purpose” for seeing Crutchley was “to consult him as to his views on the tactical situation, including reports of enemy ship movements and as modified by Fletcher’s departure.”15
Vandegrift could get to the McCawley simply by taking a boat to the X-Ray transports off Lunga. Crutchley, though, was in his flagship Australia twenty miles away near Savo. Warning he might not return that night, he turned over temporary command of his own patrol group to Captain Bode of the Chicago and took the Australia to Lunga. He boarded the McCawley at 2235, where he found Turner “disturbed over the question of keeping in the area, a large number of ships” and that he “intended to withdraw the surface forces [that is, everything] the next day.” Crutchley agreed the surface force spotted that morning did not look strong enough to challenge him. The absence of additional sighting reports of ships closing Guadalcanal, as well as the Comsowespacfor dispatch regarding reinforcements for Buka and Shortland, certainly reinforced that reasonable impression. Turner explained what appears now to be the less credible threat from “torpedo carrying float planes” possibly operating from Rekata Bay.16
Vandegrift finally reached the McCawley about 2315. Himself exhausted, he thought Turner and Crutchley looked “ready to pass out.” The conference lasted forty minutes. Vandegrift stated that he did not have an accurate tally of supplies landed on Guadalcanal but confirmed Turner’s belief “we are all right” there. He could not say the same for Tulagi. Given the concern over torpedo planes at Rekata, Vandegrift conceded all transports and cargo ships should depart before dawn, but he specified certain items that must be landed yet that night. He understood the withdrawal was to take place, “Provided the military situation and logistics in the Tulagi area justified,” and that Turner would soon return with the rest of the goods. The two commanders left the McCawley just before midnight. Saying, “Your mission is much more urgent than mine,” Crutchley courteously took Vandegrift in his barge directly to the DMS Southard for the trip to Tulagi. As they parted he remarked: “Vandegrift, I don’t know that I can blame Turner for what he’s doing,” a comment that Lt. Col. Jerry Thomas, who was present, “judged to be less than commendatory.”17
Turner informed TF-62 at 0118 that all ships would retire east at 0630 and compiled a dispatch for Ghormley, Fletcher, and McCain describing his plans to pull out the Amphibious Force:
Air attacks today resulting in loss of Elliott severe damage Jarvis probably increasing tomorrow and absence air support require me to withdraw all ships temporarily from this area to avoid unwarranted loss. Separate situation summary [090230 noted above]. This to obtain cooperative measures. Stores and ammunition landed estimated adequate for troops for thirty days replenishment must be accomplished in that time.
Will pass out through Lengo Channel beginning nineteen hours Zed ninth [0600 on the tenth, sic, mistake for “eighth”?] speed 13 then course 117 past San Cristobal and toward Button [Espíritu Santo]. Request fighter and antisubmarine air cover.
The message lacked the assurances given Vandegrift that the situation at Tulagi would determine if the withdrawal occurred. The McCawley finally radioed it about 0405, and, true to form, no one received it. In the interval, dire events befell the Screening Group that, oddly enough, tarnished Fletcher’s reputation much more than those directly responsible for its defense.18
The Japanese task force that the RAAF aircraft misidentified at 1025 as three cruisers, three destroyers, and two seaplane tenders or gunboats was, of course, Admiral Mikawa’s formidable array of five heavy cruisers, two old light cruisers, and one destroyer. He aimed to strike the Allied fleet between Tulagi and Lunga after midnight on 9 August. When discovered, he loitered east of Bougainville, awaiting the return of his dawn search. Knowing he was sighted and fearing air attack yet that afternoon, Mikawa tried to confuse the watcher by reversing course. Free of shadowers after 1200, he resumed his southward advance at twenty-four knots. Passing unscathed through the upper Solomons, he was by sundown just 130 miles from Lunga and two hundred miles from Fletcher’s carriers. At 2313 he launched two reconnaissance seaplanes to scout the anchorage and drop flares when the assault force drew near. At 0025 on 9 August one float plane pinpointed a group of cruisers lurking south of Savo where he wanted to penetrate the sound. Mikawa targeted them first. The moonless sky and bad weather proved ideal.19
At night Crutchley relied on two destroyers as roving radar pickets for early warning should enemy ships pass north or south of Savo. He deployed the two cruiser-destroyer groups to block the Japanese no matter which approach they tried. Northeast of Savo, Riefkohl’s group comprised his Vincennes, the Astoria, Quincy, and two destroyers, while Crutchley’s own group with the Australia, Canberra, Chicago, and two destroyers patrolled to the southeast. Covering the back door, Scott’s group of two light cruisers and two destroyers shuttled north and south between the Yoke and X-Ray transport areas. Five destroyers, four APDs, and four DMSs screened the transports. Aground off Florida, the stricken transport George F. Elliott burned, while the hurt Jarvis, contrary to orders, limped west to retire via western Guadalcanal. As noted, Turner’s summons caused Crutchley to relinquish command of his group to Bode in the Chicago. For some reason Bode let the Canberra lead his formation. At 2345 as the McCawley conference broke up, the Ralph Talbot, the picket destroyer northeast of Savo, warned on TBS of a plane headed east over Savo. Neither Turner nor Crutchley copied this message. After delivering Vandegrift to the Southard, Crutchley regained the Australia about 0115. Given the late hour, the difficulty of effecting a night rendezvous near Savo, the planned predawn withdrawal, and the fact that he, like Turner (and Fletcher), perceived no immediate threat, Crutchley patrolled just west of the X-Ray transport area until first light.
Mikawa’s force cut between Savo utterly undetected by the westerly picket destroyer, the Blue. Keen eyesight augmented by excellent optics decisively trumped Allied radar. Discerning the unsuspecting Chicago group, Mikawa launched torpedoes about 0138 and opened fire five minutes later. Right on cue the float plane circling over Lunga dumped green flares south and east beyond the X-Ray area, effectively silhouetting the Allied ships. Bode’s two cruisers responded with utter confusion as Mikawa battered the Canberra and torpedoed the Chicago’s bow. Swinging counterclockwise around Savo, Mikawa raked Riefkohl’s Vincennes group from two sides and annihilated it. All three U.S. heavy cruisers sustained fatal damage, but the Japanese were barely scratched. Gatacre, the Screening Group operations officer, later wrote that in accordance with the plan for the split deployment, “An enemy force had encountered one cruiser group and had been deflected into the maw of the other.” Only he and Crutchley never dreamt “the maw would be closed!”20
After sweeping around Savo, Mikawa’s entire force by 0220 tore away northwest at thirty knots. He reasoned he could not reach the Allied transports until nearly first light, when he could expect a beating from carrier planes. In addition to preserving his own limited strength, a speedy retirement might draw pursuing carriers into easy range of bombers based at Rabaul. If he could reach Vella Lavella, two hundred miles northwest of the battle area, he thought he would be safe. Final reports tallied seven heavy cruisers, a light cruiser, and six destroyers either sunk or fatally damaged. Given the highly exaggerated air claims on 8 August, it is no wonder some in the Japanese high command concluded on 10 August that the Americans had evacuated Guadalcanal. Criticism of Mikawa’s failure to take out the transports mounted only after that happy event failed to come to pass.
Mikawa came and went before Turner and Crutchley knew anything of the battle other than gun flashes and flickering light from burning ships. Squalls rendered visibility toward Savo quite poor, and lightning continued to mimic gunfire. When the flares appeared, the X-Ray transports and cargo ships ceased unloading, and without orders headed east. At 0156 Turner barked to Crutchley, “Form up your force to repel enemy attack,” and twenty minutes later radioed Fletcher to alert him, “Surface forces attacking near Savo apparently on our outpost vessels.” Crutchley summoned the X-Ray destroyers into a blocking force northwest of Lunga, but they misunderstood and raced toward a designated point five miles from Savo. For an appreciable time Bode said little over TBS, and the Vincennes group kept silent. Not until 0245 did Scott hear any kind of contact report from the Chicago group. At 0307 Bode finally revealed that the Chicago had been torpedoed and was “slightly down by the bow,” and the Canberra, attended by two destroyers, was afire five miles southeast of Savo. He also warned of enemy ships “firing to seaward,” although Mikawa was long gone. A few minutes later Crutchley could provide Turner only the obvious fact of a “surface action,” and that the “situation” remained “undetermined.” Because heavy rain obscured the battle area, he knew little more as he interposed the Australia between Savo and the X-Ray transports. Turner was appalled to learn at 0330 that Comsopac approved the carrier withdrawal, and at 0345 again tried to awake Fletcher to the danger at Savo. “Surface attack on the screen coordinated with use of aircraft flares. Chicago hit torpedo. Canberra on fire.” Aware the carriers could do nothing before dawn, Turner still hoped to get the Amphibious Force moving out by then. At 0435 he cautioned TF-62, “It is urgent for this force to depart this area at 0630,” and, “If Canberra cannot join retirement in time she should be destroyed.”21
Bewilderment plagued the ships off Savo. At 0515 the Ralph Talbot reported she was badly damaged by gunfire, but unwittingly gave the impression her misfortune just happened rather than nearly three hours before. At 0525 the destroyer Patterson radioed the Canberra had abandoned ship and that she rescued her crew. About that time the Patterson briefly exchanged fire with the Chicago. By first light at 0547 Crutchley possessed only scant news from Bode and none from Riefkohl. “Situation obscure,” he alerted TF-62, “be prepared to give battle at dawn in the vicinity of the transport groups.” Not until well after sunrise did he learn from the destroyer Selfridge that the Astoria was burning and that four destroyers recovered her crew. As late as 0819 Crutchley told Turner he had nothing on the Vincennes or Quincy. In fact the Quincy sank at 0238 and the Vincennes a dozen minutes later, both with great loss of life.22
As Turner tried to fathom the extent of the whipping inflicted on the Screening Group, he heard from Vandegrift in the Neville that Tulagi, Gavutu, and Tanambogo were taken, and the last defended position, tiny Makambo Island, would be assaulted after dawn. The bad news was “practically no ammunition, rations, water or AA had been landed except that carried by troops.” Ashe only began the unloading at Tulagi at 2340, 8 August, and kept going all night despite the nearby battle and heavy rains. To Turner’s great credit he decided to remain despite the threat from aircraft, subs, and surface ships. At 0624 he told his commanders to “cancel departure,” and at 0641 had the McCawley’s jinxed radio transmit to Fletcher that he was “unable [to] depart as planned because insufficient supplies have been landed. Request [air] cover for attack on enemy surface force this area.” Neither Fletcher nor Ghormley received that message. To free up the escort, Turner authorized the Selfridge to sink the crippled Canberra, which expired at 0800. The Astoria still fought for her life out near Savo. Now Turner would see what mayhem the Japanese could cause before he got clear of Guadalcanal.23
At 0930 Mikawa exhaled a sigh of relief as he safely passed Vella Lavella. The absence of instant retaliation intensified the sting of Savo. To Fletcher’s many critics, that compounded his unnecessary and blatant error of pulling out the carriers too early. Nimitz commented on 23 August, “Absence of all the carrier task forces on the morning of the 9th permitted the enemy to make a clean getaway.” Hepburn described the enemy commander at Savo hastening away, “To avoid, if possible, the air attack to be expected from our forces at dawn,” and mordantly observed, “The only part of our force capable of making such an attack were [sic] at the same time retiring in the opposite direction because of the same apprehension.” Pye’s analysis challenged Fletcher’s failure, given the battle was only 150 miles distant, to have “planes over the Japanese ships—some of which were probably damaged—shortly after daylight.” Subsequent critics even accused Fletcher of intentionally abandoning the Amphibious Force. These charges rest on the presumption Fletcher possessed knowledge of the Battle of Savo Island he did not actually have.24
Analysis of Turner’s communications reveals the pattern of missed messages lasted through 9 August and beyond. The carriers (and Comsopac) failed to intercept nearly all the dispatches the McCawley transmitted that night. Around 0300, though, Kinkaid in the Enterprise monitored Crutchley’s message to Turner regarding a surface action but heard absolutely nothing else from that quarter the rest of the night. At that early hour, while Fletcher awaited Ghormley’s decision whether he could withdraw, the three carrier task units, deployed in a widely spaced column, steered toward the south coast of Guadalcanal. Kinkaid, the most junior carrier admiral, deplored the lack of “timely and accurate information of the surface actions” that might have enabled the carriers to strike at dawn. Turner and Crutchley, of course, knew little more of what transpired at Savo, but they were in a much better position to find out than the carriers. Given the paucity of information and the absence of any directive from Fletcher or Noyes, Kinkaid neither acted on his own nor broke radio silence to consult them. Comsopac’s approval reached Fletcher and the others at 0330, and without further ado he brought the carriers around to the southeast.25
Kinkaid did not think he understood the situation off Guadalcanal sufficiently well to attack. However by 0400, after hearing more snatches of radio traffic from Crutchley and perhaps Turner, Forrest Sherman in the Wasp decided it was absolutely essential to attack at Lunga only 120 miles distant. The Wasp still had an armed strike spotted on the flight deck. Her air group, “fresh, highly trained, qualified for night operations and eager for combat,” could make all the difference up north. Sherman beseeched Noyes to seek Fletcher’s permission for the Wasp task unit to close Guadalcanal “at maximum speed” and attack. In 1949 he related to Morison that he made “three separate but unsuccessful recommendations [to Noyes] for action along that line,” and that a staff officer made “a similar recommendation.” To Noyes, though, it was obvious that by turning away from Guadalcanal, Fletcher already made up his mind. He told Sherman that Fletcher “had all the information [they had] and that he would order us to attack if he thought best to do so,” but “meanwhile we would keep the [post-dawn] rendezvous.”26 Sherman was crestfallen, and word certainly got around the Wasp of the failure to attack. The situation strongly resembled the night of 24–25 October 1944 in the Battle of Leyte Gulf, when Vice Adm. Marc Mitscher declined to challenge Admiral Halsey’s run to the north to engage Japanese carriers off Cape Engaño. A successful strike against Mikawa initiated up to an hour before dawn was possible, but only if everything went perfectly. Noyes regretted Fletcher did nothing to help Guadalcanal. Sherman later agreed with historian Fletcher Pratt that exhaustion, “Not with the local tiredness of seventy-two hours on duty, like the crews of the cruisers, but with the accumulated fatigue of many months of the most arduous command of the war,” contributed to Fletcher’s supposed lassitude, though he was never in a position to make such a personal judgment. Sherman never actually saw Fletcher during the Guadalcanal campaign and did not know the whole situation that morning.27
Like Noyes, Sherman, and Kinkaid, the critics believed Fletcher indeed possessed “all the information” regarding the surprise assault at Savo, but nevertheless chose to run away. Bates indignantly informed Morison that the TF-16 war diary demonstrated Fletcher knew of the Savo action as early as 0300 but dawdled until he secured Ghormley’s approval a half hour later, then “acted at once to get out of the area without waiting clarification of the surface action report.” Fully in accord, Morison discussed Sherman’s doomed proposal for the Wasp to close Guadalcanal “at high speed” and attack. He could not resist adding the phrase, “With a few well-fueled destroyers,” a distinction Sherman never made and which certainly did not exist on 9 August. Instead, Fletcher “hightailed out of the area” and betrayed Turner.28
The critics ranked that inglorious retreat the worst of Fletcher’s numerous wartime sins. Yet there is no evidence the Saratoga or TF-11 even knew until after sunrise of a fight at Savo, much less terrible defeat. Wright told Morison in 1943, “Frank Jack certainly wouldn’t have pulled out had he known what happened.” In 1952 he declared: “Neither Ghormley or Fletcher knew at this time that a major surface action had taken place near Savo.” In 1963 Harry Smith, the former flag lieutenant, explained to Dyer, “For some reason the Saratoga did not or could not copy [Turner’s] blind [not receipted] despatches sent that night, and it wasn’t until other ships sent us the news by blinker or infrared that we started to get the word about the Battle of Savo.” Fletcher’s 1947 letter to Baldwin avowed it was “shortly after daylight of the ninth we began to intercept garbled radio messages indicative of action of some kind.” Dyer quoted Fletcher’s 9 September 1942 report to the effect that news of the night battle reached “the force” at 0400 [actually at 0300], but that Fletcher told him, “It wasn’t until much later that I was awakened and given the first indication of Savo.” Fletcher recalled that he first knew between 0500 and 0600. However, George Clapp, the communication watch officer on duty, confirmed that Fletcher’s original statement to Baldwin was correct. Clapp recalled it was definitely after sunrise (0634), when he personally brought a message describing the loss of a cruiser up to flag plot. Only the duty officer was present. According to Maas’s diary, the rest of the staff was below having breakfast with him, another strong indication the Saratoga was unaware of trouble up north. Rousing the skivvy-clad admiral in his sea cabin constituted one of Clapp’s most vivid memories of the war. Fletcher reacted with profound surprise. Asked by Dyer how he might actually have responded had he truly known of Savo as early as 0300, Fletcher said, “Since we were on a northerly course, I might well have continued on it.”29
What were Fletcher’s options at about 0645 when he belatedly learned of the surface fight up north? The first consideration is the status of the carriers. The Enterprise, which Noyes’s schedule placed in strict reserve, was to have combat air patrol and a strike group ready to launch, while the Wasp handled combat air patrol and the Saratoga searched and kept F4Fs on standby. At 0600 and a half-hour before sunrise, the Sara had launched eight SBDs to sweep 175 miles northwest and routinely examine the waters west of Guadalcanal. The easternmost sector crossed just west of Lunga. Kinkaid took it upon himself at 0615, in defiance of Noyes’s orders, to dispatch two SBDs to Lunga to drop aerial photos to Vandegrift and see what transpired there. About sunrise the carrier task units, steaming southeast at fifteen knots while fifteen to twenty miles apart, established visual contact and maneuvered to resume the task group cruising formation. At that time Cape Henslow was roughly one hundred miles north, and Savo seventy-five miles beyond.30
Fletcher’s initial knowledge of the previous night’s disaster was very sketchy. As yet he did not realize the whole Amphibious Force remained behind. Nor did he know the extent of the tremendous harm inflicted on Crutchley’s Screening Group. In 1963 when Dyer asked whether anyone urged him to attack, Fletcher replied, “One or two of my staff recommended that we go back.” He declined, because if he were the enemy commander he “would have planned on all our carriers coming back and would hit them with all my land based air.” However, he admitted, “If I had it to do all over again that morning and knew about our losses, I would leave one carrier group behind to fuel, and would move two carrier groups up to attack and to continue to provide air support to Kelly Turner.” He added, though, “This did not occur to me at the time as being sound,” a conclusion fully sustained by the evidence. Obviously the pounding Fletcher endured over the years from the critics took its toll. After the war Kinkaid declared on the basis of hindsight that the carriers should have turned around and chased the retreating cruisers. It is doubtful he would have thought so at the time, when he was more pessimistic than Fletcher or Noyes about the critical fuel situation and the risk of air counterattack.31
Closest to when Fletcher actually decided to continue withdrawing, his 9 September report to Ghormley more reliably indicated his thoughts at that time. He conjectured therein that if the carriers had been closer to Guadalcanal, “In the position they would have been, had they remained in the area” (that is, twenty-five miles instead of more than one hundred miles south), it was “barely possible that with definite information” a morning search might have found the enemy force. However, he would have required “sufficient fuel available to operate indefinitely at high speed,” which TG-61.1 did not possess. The first thing that evenhanded hindsight demonstrates is that Fletcher needed that “definite information” well before sunrise for any chance to catch the triumphant Mikawa. If first undertaken at 0730—about the earliest any planes could have departed after Fletcher actually became aware of Savo—an attack mission of 275 miles could have only reached the lower New Georgia group. Even that would have required the carriers to make a twenty-five-knot sprint (at a ruinous price in oil) to shorten the return flight by seventy-five miles and also exposed them “much further to the northwest.” By 0645 Mikawa was nearly three hundred miles northwest of TG-61.1. When this hypothetical strike could have first reached southern New Georgia, he would have been 115 miles beyond. Fletcher’s report warned that carriers could not “normally” strike ships that attacked at Guadalcanal at night and “retired at high speed four hours or more before daylight.” By 9 September beleaguered Cactus discovered that central truth on its own. Mikawa’s great gamble paid off.32
In the absence of a clear call for help, which TG-61.1 certainly never received, Fletcher believed the fight off Savo had been a night action and was already concluded. In that event his planes could do nothing. To have the carriers loiter off Guadalcanal would delay his fueling, which he must complete to be ready to battle enemy carriers, and also lengthen his exposure to air attack. Moreover the Sara’s search could be expected to report promptly if a sea battle raged off Guadalcanal. Fletcher broke radio silence to raise Turner, but without success. Perhaps the one fair criticism is that before going farther from Guadalcanal, Fletcher did not try harder to learn more what had occurred there. In this context it is again useful to recall what Sherman told Morison in 1949: “It was not until after the night action that the situation became critical.”33
Having decided he must press on to the fueling rendezvous, Fletcher waited six hours before the carriers finally drew out of reach of land-based bombers. At 0838 Belconnen relayed a warning from coast watcher Read in northern Bougainville that a formation of aircraft passed overhead at 0807 going southeast. At 0839 Noyes forwarded a message the Wasp intercepted between the McCawley and a Saratoga search SBD: “Appreciate knowing Admiral Fletcher’s plan by message drop.” Fletcher kept trying unsuccessfully to get into direct radio communication with Turner. The Saratoga search returned at 0930 with negative contacts except for the damaged destroyer Jarvis limping southwest of Guadalcanal. At 1009 Noyes provided Fletcher the opening paragraph (the rest was not recovered) of Turner’s 0105 order to retire (first transmitted at 0405): “Air attacks today resulted in loss Elliott severe damage Jarvis probably increasing tomorrow and absence air support require me to withdraw all ships temporarily from this area to avoid unwarranted loss.” Maas reported to the Sara’s flag bridge about the same time Fletcher received confirmation from Mason at Buin of at least ten bombers headed southeast. Thus he might expect the strike by noon. Rabaul loosed seventeen torpedo-armed land-attack planes and fifteen Zeros. Only if carriers and battleships could not be found were they to hit transports. A search plane radioed at 1000 that amphibious ships remained at Tulagi, but snoopers never came closer than sixty miles to the carriers.34
Fletcher, however, did not know he had not been sighted. Still ten miles apart, the carriers continued to prepare their air defense. The pair of Enterprise SBDs that flew to Guadalcanal saw no enemy ships or evidence of a surface action. At 1114 Fletcher resumed tactical command of TG-61.1 and ordered Noyes, “Double combat air patrol and standby for air attack.” It is possible Slonim’s radio intelligence team in the Saratoga intercepted traffic between Rabaul and its searchers that indicated something was up. At 1155 radars discovered a large bogey, perhaps “many aircraft,” fifty miles northwest and closing. Twenty F4Fs hastened to intercept but advised at 1215 the bogey was a B-17. After the search failed to turn up carriers, the Japanese strike pounced on what was first thought to be a hurt battleship leaking oil southwest of Guadalcanal. Ship identification had not improved since Coral Sea, for the quarry was luckless Jarvis, which splashed two planes before going down with all hands. The disappointed Japanese thought they sank a British Achilles-class light cruiser and regretted not finding anything bigger.35
It was only while Fletcher awaited the reported strike that he finally accumulated enough bits and pieces of rebroadcast messages to grasp the appalling defeat of the Screening Group. At 1150 he radioed Ghormley: “Am receiving dispatches from Turner describing heavy losses in cruisers in engagement which continues to westward. Are you receiving them?” Once the air threat ended at the eight-hundred-mile mark from Rabaul, the entire staff, less Fletcher and the duty officer, assembled in the flag quarters for a somber lunch. That afternoon Maas recorded in his diary, “Things going quite badly at Tulagi. Japs attacked & sank or damaged four cruisers etc.” No one could fathom how a few cruisers and destroyers wrought such havoc against a superior number of radar-equipped warships. At 1345 Comsopac warned of a “considerable force sighted east of Bougainville heading southeast.” Maas deplored that “latest report [of ] Jap troops assaulting Guadalcanal.” The marines were “now without air support.” By that time the carriers were 250 miles southeast of Lunga. The best way they could help Vandegrift was to expedite refueling.36
That afternoon Ghormley complained to Fletcher, McCain, and Nimitz that he had received no situation summary from Turner, except the one for 7 August, and asked them to forward any pertinent information. At 1415 Fletcher relayed everything he had. It did not make pleasant reading:
Following summary of messages delivered from Turner, quote: At 081645 [0345 on the ninth] surface attack on screen coordinated with use aircraft flares Chicago hit torpedo Canberra on fire. At 2100Z [0800] heavy actions continue to westward. More of our ships in trouble. Submarines in area. At 2152Z [0852] Quincy sunk by torpedoes and gunfire. Air attack enroute. At 2325Z [1025] Vincennes sunk by gunfire and torpedoes 0245 [1325] casualties heavy. At 2350 [1050] Astoria has fire in wardroom destroyer ordered to pump and [cargo ship] Alchiba to tow through Lunga channel to Roses [Efate] as chance to save her. Movements require protection which I am unable to provide. Unquote.
Fletcher concluded by asking Ghormley to “direct Turner to make reports direct to you info to me.” This message is perhaps the most telling contemporaneous evidence that Fletcher had no inkling of the existence, much less the scope, of the Savo disaster until long after sunrise on 9 August.37
Bates and Innis deplored that Fletcher showed “little concern” for Turner’s plight and so ruthlessly abdicated his command responsibility for the Expeditionary Force. Their accusation is valid to the extent that Fletcher weighed priorities and determined his best course of action was to swiftly restore full mobility to the carrier task force by refueling. However, they also found it “surprising” neither Ghormley nor Fletcher advised Turner that Fletcher departed with the carriers, as if Noyes might be expected to drop him off somewhere. They also charged Turner was “not advised of his succession to command or of his superior’s plans.” Again it must be noted the confusion in the minds of the critics about the nature of Fletcher’s command. They failed to grasp that he was tied to the carriers despite having the whole Expeditionary Force. That was an error Nimitz did not repeat. In July 1943 after reading the Hepburn report, Capt. George L. Russell, King’s flag secretary, commented, “When Rear Admiral Noyes [sic] decided to retire, it may or may not have been with the concurrence of Vice Admiral Fletcher, but whether it was or not, he took Vice Admiral Fletcher with him.” Furthermore, “I don’t know what Vice Admiral Fletcher contributed to the operation.” It is unfortunate—or incredible—that neither Fletcher’s superiors nor Hepburn ever officially asked him to describe his command relationships and decisions during the Guadalcanal landings. Suffice it here to note that by the morning of 9 August Fletcher believed he had no choice but to move the carriers south for fuel to prepare for the big carrier action that must be in the offing. As noted, the noon fuel state of the destroyers averaged 35 percent. The heavy ships were also getting low. It was high time to begin the lengthy refueling process.38
Fletcher’s message to Ghormley crossed one from Nimitz to Ghormley again offering well meant but unwittingly ill-timed praise. “Your shipmates in Pacific Ocean Areas are watching your progress with great admiration.” Soon afterward a shocked Nimitz passed Fletcher’s dire news to Cominch. The dispatch reached Washington after midnight on 9 August. Russell violated King’s injunction against interrupting his sleep by bringing the word that burst the bubble of excessive optimism regarding the prospects for Watchtower’s easy success. King recalled it as the “blackest day of the whole war.”39
At 1415 MacArthur placed four enemy cruisers northeast of Bougainville and headed northwest. They looked like the culprits that struck Savo the previous night. However, Ghormley cautioned, “Indications enemy landing forces [are] proceeding against Cactus.” Radio intelligence hinted of Mikawa’s plans to send a small convoy with naval infantry to the west tip of Guadalcanal. Once Mikawa realized the strength of naval forces in the area, he had recalled the convoy, but the S-38 picked off one of the transports in St. George’s Channel. Another venerable, valiant southwest Pacific S-boat won the only immediate revenge for Savo. On the morning of 10 August, in an action reminiscent of the I-168 at Midway, the S-44 sank the heavy cruiser Kako as she was about to enter Kavieng harbor on New Ireland. No one on the Allied side knew that good news for two weeks.
Late on the afternoon of 9 August Ghormley queried Fletcher about the radio frequency of Turner’s messages. Fletcher replied 2122 kilocycles, but “reception [is] very poor; am missing most of his transmissions.” At 1850 Ghormley ordered Turner to withdraw all naval forces to Nouméa. Unsure whether the Second Marines ever landed at Ndeni, he gave Turner three options. If the regiment was still at Guadalcanal it was to remain there. If it was already at Ndeni, it was to stay there. Otherwise he should drop off the Second Marines at Espíritu Santo. McCain was to withdraw the Mackinac and her PBYs from Malaita, although she had been supposed to stay until relieved by the Ballard on 13 August. Likewise Ghormley directed the McFarland and her PBYs pull out of Ndeni if the Second Marines were not there. Thus McCain’s long-range search suffered crippling blows. Ghormley told Fletcher, “Cover movement as practicable without interference with fueling.”40
Indeed Turner was already in the midst of pulling out his forces after a rough day cleaning up after the loss of the cruisers. The appearance of only a few SBDs demonstrated that Fletcher’s carriers had gone. After the first coast-watcher warning Turner suspended unloading at 0840 and herded the transports and cargo ships into defensive formation. Air attack seemed imminent, and he was painfully aware he lacked fighter support. By 1100 when no strike turned up, he recalled his ships. Vandegrift told him that vital equipment and supplies for Tulagi remained in three Yoke transports, and one in the X-Ray Group held similar necessary cargo for Lunga. Turner directed those ships to resume unloading, but the others were to retrieve their landing boats and prepare to depart. He detailed the barely unloaded cargo ship Alchiba to take the Astoria in tow, but at 1215, before that could occur, the cruiser sank. Turner ordered all survivors placed in transports. Early that afternoon he notified the X-Ray Group (six transports and five cargo ships, the damaged Chicago, three destroyers, and the five high-speed minesweepers) to get under way at 1500 and withdraw. In the meantime he took the McCawley over to Tulagi and personally led the Yoke Group (six transports, one cargo ship, and four APDs, the Australia, San Juan, Hobart, and nine destroyers) eastward at 1700. That afternoon only a “small quantity” of additional matériel was landed at Lunga, although Tulagi did get ten days’ worth of supplies, that only just, according to Ashe, averted “catastrophe.” Sunset on 9 August found Vandegrift’s marines totally on their own, minus much of their equipment and supplies (cargo ships Betelgeuse and Alchiba left with 50 and 75 percent respectively of their cargoes), not to mention fourteen hundred men (mostly Second Marines) still in transports. Vandegrift certainly did “stand alone” at Guadalcanal.41
On 12 August as TF-62 neared Nouméa, Turner transmitted the first detailed summary of Savo. He outlined the split deployment of the Screening Group that intercepted “enemy CAs DDs or torpedo boats probably subs,” with “heavy running actions” continuing about forty minutes. He knew of “no knowledge of damage to enemy except one sub probably sunk.” Late on the night of 11–12 August copies went to King and the White House. On the thirteenth King explained to Roosevelt that battleships South Dakota and Washington and light cruiser Juneau would more than compensate for the sunken heavy cruisers. In 1949 Morison opined the loss of the four cruisers at Savo “delayed the completion of Operation ‘Watchtower’ for several months,” a judgment Turner called an “exaggeration.” Savo inflicted its greatest blow to the navy’s pride. Initially it was thought only the misreported handful of cruisers and destroyers first sighted on 8 August off Bougainville sprang the colossal surprise. Not until October did prisoners of war reveal there were five heavy cruisers, two light cruisers, and one destroyer.42
The main causes of the Savo disaster were inadequate night battle training, inexperience, insufficient number of destroyers, over-reliance on radar, and a lack of readiness by very tired captains and crews, greatly abetted by the belief of the top commanders that no surface threat existed that night. Other factors certainly applied, but in truth the Screening Group failed. Crutchley honestly told Turner: “The fact must be faced that we had adequate force placed with the very purpose of repelling surface attack and when that surface attack was made, it destroyed our force.” On 8 September the third Cominch-Cincpac conference discussed Savo at length and singled out Crutchley’s night disposition for criticism. The long delay by Comsowespac in forwarding the cruiser sighting report on 8 August “made no difference,” because it was still “received in time to be acted upon.” Ghormley bore “overall responsibility,” although the “immediate responsibility” was Turner’s. “Final judgment” must await “receipt of completed reports.” On 8 October Nimitz queried Ghormley for his “opinion as to the responsibility for the dispositions and actions” at Savo. “Such a blow cannot be lightly passed over, and we owe it to the country to do our best to fix the responsibility for that disaster, and to take the action necessary to prevent a recurrence.” The “final reports” proved unavailing. In December King assigned Hepburn the task of investigating Savo and assessing fault.43
Turner was adamant he would accept no blame for the disaster. Nothing he did—nor any action of Crutchley’s that he previously approved—contributed in any way to the debacle. Marine Gen. Bill Twining put it well. “In [Dyer’s biography of Turner] there is a particularly illuminating passage of intemperate denials wherein [Turner] places blame for the disaster on just about everyone in the world except himself and Mother Teresa.” Turner deflected complaints concerning the inadequate radar picket and the divided deployment of the Screening Group (important factors in the defeat) onto the faulty search that misidentified the Japanese ships. He also castigated the precipitate withdrawal of the carriers that forced him to convene the conference, pull Crutchley and the Australia off line, and fatally weaken the western cruiser group. Ultimately, though, it was all Fletcher’s fault. “It was expected that the covering force would prevent the arrival of superior enemy forces in the vicinity of Savo Island, but that expectation was not well founded.” Moreover, “When seven enemy cruisers and seven [sic] destroyers arrived to combat the five cruisers and six destroyers on outpost duty, no help was given by other naval forces either to support our attack or to destroy the enemy naval forces during their retirement.” Thus Turner cast aspersions not only on Fletcher, but also Rear Adm. Norman Scott, who led the nearby group of two light cruisers and two destroyers. Scott, who died a hero’s death at Guadalcanal in November 1942, could not answer such a challenge. Turner’s shifting of the blame worked brilliantly. Morison’s Two Ocean War condescendingly decreed the “worst of all blunders that night,” above any failure by the Screening Group, was “‘Frank Jack’” pulling out the carriers. This interpretation is an astonishing tribute to Turner’s ability to mask his own culpability.44
Critics roundly condemned the search effort by carriers and shore-based air alike.45 Divided between Sowespac, Aircraft, South Pacific Force (Airsopac), and TG-61.1, the search was badly coordinated, with poor ship identification and tardy reporting of results. Yet the area was huge and remote from the few friendly air bases with only a handful of search planes. At the same time the stronger than expected response of enemy air power constrained Fletcher to shift the carriers south of Guadalcanal and unfortunately reduced that part of the central Solomons his planes could examine. None of this should have surprised anyone. Nevertheless, no enemy force reached Guadalcanal totally undetected. The search network, feeble as it was, did sight Mikawa on 8 August. Regretfully the word got to Fletcher too late to strike Mikawa that day, while Turner perceived no immediate surface threat. Turner recalled in his 1945 history that Fletcher promised if land-based air could not cover any sectors, the carriers would “fill in for short-range scouting, both morning and late afternoon, to protect against the approach of surface forces.” Yet “so far as is known,” Fletcher “never sent his scouting planes in any direction, but this fact was learned only considerably later.” Where Turner “learned” that “fact”is not known, because it is demonstrably false. Obviously, though, it was just one instance of how he blackened Fletcher’s reputation. The following year he told Baldwin the carriers never scouted the Slot on 8 August. In fact Fletcher’s comprehensive afternoon search only missed Mikawa by thirty miles.46
Hepburn indicted Fletcher for contributing to the “disproportionate damage” sustained by the ships at Savo. “The withdrawal of the carrier groups on the evening before the battle” was “responsible for Admiral Turner’s conference, which in turn was responsible for the absence of the Australia from the action,” and that Turner’s “need to confer with his senior commanders cannot be questioned.” However, it is strongly debatable whether that particular conference was really justified. It is vital to stress once again if Turner truly believed that Fletcher said all along the carriers would only remain two days, he should have planned accordingly and not been surprised. Instead, he insisted on an immediate face-to-face meeting with his two top subordinates. Vandegrift’s presence might have been necessary for Turner to gain a better idea of conditions ashore, but Crutchley personally led one of two groups of cruisers and destroyers that guarded Savo. Although Crutchley had queried Turner that morning for a summary of the plan—logical in that he was second in command of the Amphibious Force and standing orders existed for a major withdrawal that night—he did not necessarily require or desire to meet directly with his boss. Turner claimed to Hepburn he never intended to pull the Australia off her patrol station but did not explain how Crutchley would otherwise reach the conference. Then he amended his statement to the effect that he thought things would be quiet at least until midnight, because no surface force should arrive before Crutchley and the Australia could resume guard duty. Of course, neither Turner nor Crutchley thought anything was going to happen that night, so much that Crutchley felt justified not returning on station.47
Bates and Innis concurred with Hepburn. Turner’s decision to confer with his subordinates was “sound” and of “extreme importance.” They based their conclusion on the false premise that Turner had intended all along to keep the transports and the cargo ships off Guadalcanal for a full five days. Hence the prospect of being forced to withdraw “two days ahead of time,” with “serious implications for the marines ashore,” mandated that Turner summon Crutchley and Vandegrift to coordinate the revised plan. Although in general agreement concerning Vandegrift’s attendance, Morison still wondered why Crutchley, too, had to be present. In contrast, Twining called the conference a “tragic folly” and decried efforts to whitewash Turner’s actions, particularly with regard to Crutchley. Turner could have just sent a message or even detailed staff officers to brief his two subordinates. “Nothing could have been accomplished at such a meeting, and nothing was accomplished,” although Twining conceded, “Turner was left uncovered by Fletcher’s flight and had to pull out through no fault of his own.”48
The more realistic explanation, however, is that by 2000 on 8 August, Turner already decided to abandon incremental withdrawals and instead pull out the entire force by the next dawn, despite not having landed all the men and supplies even from the transports. He desired his two top commanders ratify what was certain to be viewed as a controversial move. In keeping with Turner’s bullheaded reputation as someone not disposed to “consult” anyone, the meeting by all indications constituted more a lecture than any actual “consultation.” Vandegrift hated to see all the ships depart but acquiesced mainly because he thought the supply situation at Guadalcanal was “all right,” although he could not say the same for Tulagi. Because Turner steadfastly supported Crutchley’s night combat dispositions, it is understandable Crutchley never complained about the odd circumstances that kept him out of the most important battle of his career. In turn his staff officer Gatacre did not “deny” that Turner was “right” in summoning his boss, but strongly deplored the resulting absence of flagship Australia, considered the most battle-ready ship in the Screening Group. He believed her presence might have made all the difference.49
The leaders of the Amphibious Force and the First Marine Division all closed ranks to condemn Fletcher. It was in their interest to do so. However, there remained the question of whom to blame for the unloading woes and the consequent failure to leave adequate supplies, despite having kept the transports and cargo ships off Guadalcanal through most of 9 August. All that could not be laid solely at Fletcher’s door. Maas and Pye independently questioned why Turner withdrew without completing the unloading. Bates and Innis admitted the marines had only half their sixty-day allotment of supplies and four instead of ten units of fire. Yet they excused the poor performance of the Amphibious Force on 9 August by stating that Turner, upon “realizing the difficulty of getting supplies ashore” and that “practically no supplies had been landed all day,” decided “nothing was to be gained by remaining.” That begs the question why Turner did not do more to organize and expedite the unloading, although in truth much of the chaos arose from everyone’s gross inexperience. It is interesting that Turner told Hepburn in 1943 that he felt “certain the transports could not be unloaded in two days,” but on 20 August 1942 he directed the Seventh Marines and Fifth Defense Battalion at Samoa to embark “in such manner as will ensure AP’s unloading in 48 hours AK’s [cargo ships] in 72 hours unloading day and night.” That schedule accorded with his pre-7 August prognostications that the feat could be done, but once confronted with the responsibility for the breakdown at Tulagi and Guadalcanal, Turner averred such speed in unloading the transports was never possible.50
Anticipating criticism for the bungled Watchtower supply situation, Turner attempted to shift the responsibility to Vandegrift. He blamed the marine pioneers at Red Beach for not unloading the boats and shifting supplies inland and explained to the commander of the Seventh Marines, “There were two primary reasons for failure to completely unload: first the vast amount of unnecessary impediments taken, and second a failure on the part of the 1st Division to provide adequate and well organized unloading details at the beach.” The marines did not quietly abide Turner’s self-exoneration. Citing the division supply officer, Twining charged, “Very little unloading was accomplished after the enemy torpedo plane attack on the afternoon of 8 August.” Moreover, on 9 August the “disorder” in the Amphibious Force again became manifest when some transports never even retrieved all their landing boats before hauling out. Vandegrift deplored the intransigence of Captain Ashe, the Yoke Group transport commander, as a major cause for the unloading fiasco at Tulagi, while Twining laid the responsibility for the mess at Guadalcanal on Capt. Lawrence F. Reifsnider, Ashe’s X-Ray Group counterpart. Twining also thought Turner should have landed all the marines still on board the transports. In what stands as a fair assessment given all the circumstances, Dyer stated the Amphibious Force did a “commendable job” unloading despite inexperience and highly adverse circumstances.51
Any risk that Turner would suffer relief or even rebuke over Savo swiftly faded. He was too valuable to the war effort and enjoyed a strong ally in Spruance, who knew him well from the Naval War College. On 8 September Nimitz even told King that should Ghormley’s health fail, he preferred Turner as Comsopac. By early 1943 when Hepburn actually began his inquiry, the Guadalcanal campaign was won, and Turner played a pivotal role. Hepburn granted him and Crutchley full opportunity to describe the situation at Savo and explain their actions. As usual, Turner proved a persuasive advocate on his own behalf and received the benefit of every doubt. Indeed Russell explained to King in July 1943 that Turner emerged from Hepburn’s Savo investigation with “pretty much a clean bill of health.” That verdict was not merely expedient but just. As one of Nimitz’s inner circle, Turner led the amphibious forces during some of the toughest invasions in the Pacific and contributed significantly to the defeat of Japan. But that does not erase the fact that Turner, to salvage his reputation and career, relentlessly misrepresented events to malign Fletcher and others.52
“It does not necessarily follow that because we took a beating,” Russell sagely suggested, “somebody must be the goat.” Turner and Crutchley were exonerated; Bode and Riefkohl were censured; and Ghormley, Fletcher, and Noyes were found wanting. However, Hepburn never accorded Fletcher or Noyes the same thorough hearing as Turner and Crutchley before condemning their actions, indeed never even extended them the courtesy of consulting them. In March 1943 as his investigation neared its end, Hepburn passed through San Francisco. Fletcher was nearby in Seattle, and Noyes in San Diego. Nor did Turner or the marines ever acknowledge their obligation to Fletcher to seize their objectives swiftly and expedite unloading to clear out as quickly as possible and reduce the risk to the carriers of being caught in a static covering position. Fletcher explained to Dyer, “A defensive decision was in order on 8 August, although perhaps not exactly the one I made at the time.” He, unlike Turner, was prepared to consider that his decisions may have erred—even when those actions, at least until after he actually learned of the defeat at Savo, were not only fully defendable but, as Maas stressed on 8 August, also sound.53
On 13 August Crutchley’s action report offered perhaps the best perspective of events at that moment. “The success or failure of this operation cannot yet be judged as it depends on our ability to hold what we have taken and to make use of it for further offensive operations.”54