Once the attack ceased, Fletcher’s top priority was to recycle fighters to reestablish a fresh combat air patrol before a second strike turned up. Radar detected unidentified planes to the northwest and even tracked a large southbound bogey passing fifty miles to westward. The Saratoga landed fighters from both carriers and hoped to recover her first strike before sunset, barely an hour off. The Enterprise, no longer billowing smoke, appeared to be doing fine. Asked if he could handle planes, Kinkaid replied: “Negative now,” but “possible later.” Knowing his strike faced a tough time getting back, Kinkaid detached the Grayson to stand forty miles north, “Pick up pilots who may be short on gas,” and rejoin TF-16 at noon 250 miles southward. Again Kinkaid’s judgment was questionable regarding fuel. The hard-luck Grayson, at 28 percent earlier that day, had the least oil of his destroyers. She was evidently picked because she wore the pennant of Commander Holcomb, Comdesdiv 22.1
At 1730 after the Saratoga recovered twenty-seven F4Fs, Fletcher reversed course northwest at twenty-five knots to close lagging TF-16. The Sara frantically respotted fighters for launch, even while Fletcher and Kinkaid warily tracked the suspected enemy group circling to the southwest. Kinkaid described the damage to the Enterprise: “Two bomb hits aft. Damage apparently light. Fires under control. Some underwater damage not yet located.” Oddly he did not mention the third bomb that exploded just abaft the island. At 1749 the Enterprise commenced landing fighters and patient SBDs and TBFs from the afternoon search and antisubmarine patrol. The Saratoga dispatched fifteen F4Fs to restore the combat air patrol. By 1805, when she started bringing in her first strike, the sun dipped to the horizon. The Sara began bringing in fuel-starved SBDs and TBFs “so quickly and skillfully,” the flight deck “quickly crowded with planes,” with many more “circling in the air.” Likewise by 1810, the Enterprise had with equal efficiency recovered thirty-three aircraft (twenty SBDs, four TBFs, and nine F4Fs). Davis called a brief halt and respotted the deck to enable five Wildcats to take off on combat air patrol.2
At 1813 the Enterprise’s steering motors shorted out. She lurched to starboard, nearly collided with the Balch, and circled at ten knots while Davis tried fruitlessly to steer with engines. The Big E was in a terrible fix. Fortunately at 1827, just as twilight waned, the supposed enemy strike prowling seventy miles west abruptly turned away and disappeared northwest off radar. Pilots from the first Saratoga strike group informed Fletcher that at 1700 they saw at least eighteen dive bombers, nine torpedo planes, and three fighters 110 miles northwest of TF-61. Radar had tracked this group that could have finished off the Big E. Fletcher radioed Ghormley to request a tug to take her under tow. The nearest, the Navajo, was at Nouméa. Davis fortuitously regained control and gingerly pointed his hurt carrier southeast. Two SBDs and one TBF ditched from lack of fuel.3
Capt. Arthur C. Davis, USS Enterprise (CV-6), 22 July 1942.
Courtesy of National Archives (80-G-13038)
By 1915 as Kinkaid gratefully regained twenty-five knots, he saw TF-11 seven miles east. The Saratoga had recovered all planes overhead, twenty-eight SBDs, eight TBFs, and thirty-one F4Fs, that jammed her flight deck. During the landings, Ramsey, to make more room, even risked striking aircraft below via the one slowly cycling amidships elevator. Inevitably one F4F hurtled the barrier, flipped over, and pitched into the open elevator well. Only the snagging of its wings on the edge of the flight deck prevented the Wildcat from plunging into the hangar and catastrophically disabling the elevator in the down position. Despite gridlock, the Sara opened enough space to bring in three TBFs and two SBDs from her second strike. At 1930 the Sara closed up shop. Her air losses were surprisingly light: three F4Fs, one SBD, and two TBFs. With the Enterprise strike expected to stage to Guadalcanal, Fletcher did not think anyone else would come back that night.4
Following the attack, Fletcher received new information on the Japanese forces, but it did not clarify much. At 1815 a B-17 strike leader (Maj. Ernest R. Manierre) reported bombing a “large” carrier near the target assigned to the Saratoga’s first strike but gave no damage estimate. Felt described to Fletcher how he plastered a carrier at 1600 within twenty miles of the one Manierre’s B-17s later attacked. He never copied Fletcher’s 1435 message redirecting him north toward what was in fact a false contact, but flew the mission as briefed. At 1536 he had sighted one small carrier, one heavy cruiser, and two destroyers and deployed twenty-two SBDs and five VT-8 TBFs against the flattop, six SBDs and two TBFs after the cruiser. Appalled no bombs hit the swiftly turning flattop, Felt dived in, followed by the six SBDs that fortuitously shifted over from the cruiser. Several 1,000-pound bombs and at least one torpedo were seen to strike the carrier, and one torpedo supposedly hit the cruiser. No Saratoga planes fell. Felt stayed until 1620, watching in fascination as the flattop ran “in circles to the right pouring forth black smoke which would die down and belch forth in great volume again.” The attackers vowed they had sunk the distinctive Ryūjō. Slonim recalled “quite an argument” whether the carrier succumbed, but no one in the Saratoga could say for certain one way or the other.5
In turn Larsen’s tiny second Saratoga strike brought news of ships discovered north, rather than northwest, of TF-61. The five TBFs and Elder’s two SBDs flew northwest after the false 3V37 contact at 1405 of a single carrier and escorts, actually the Ryūjō force ninety miles to the west. At 1755 after searching through heavy clouds, Larsen and Elder attacked a formidable force of one Mutsu-class battleship, four heavy cruisers, and a dozen or more destroyers steaming southeast at fifteen to twenty knots. Elder claimed setting the battleship on fire, and Larsen one torpedo hit on a cruiser. Two TBFs failed to rejoin. Larsen and Elder never saw a carrier, but Fletcher soon learned that at 1815 a different B-17 strike group (Maj. J. Allan Sewart) bombed a small carrier, three heavy cruisers, and two destroyers about thirty miles east of where Larsen and Elder attacked.6
Fletcher deduced the carrier and land-based strikes must have hit the two light carriers (the Ryūjō to the west, another in the east) that evidently flanked the enemy main force. Nothing else was forthcoming from the two large carriers that Enterprise searcher 9V395 (Lieutenant Davis) spotted nearly five hours before. Kinkaid’s search summary at 2013 finally did identify two big carriers, four heavy cruisers, six light cruisers, and eight destroyers that, as of 1500, were only two hundred miles northwest of TF-61 (and surprisingly close to the erroneous Kellam contact). Fletcher figured Larsen had encountered part of its screen, reckoned at 1740 about 185 miles north of TF-61. Had the main force continued southeast at twenty knots, it could, by 1930, be within 160 miles of TF-61. If Fletcher then turned northwest at twenty to twenty-five knots, he might bring about a night surface battle in four or five hours. The staff debated that audacious course of action, but Fletcher soon resolved to withdraw south at fifteen knots in the wake of wounded TF-16. Slonim described him “collapsing into a chair,” joking: “Boys, I’m going to get two dispatches tonight, one from Admiral Nimitz telling me what a wonderful job we did, and one from King saying ‘Why in hell didn’t you use your destroyers and make torpedo attacks?’ and by God, they’ll both be right.”7
Retiring TF-11 on the evening of 24 August became Fletcher’s most controversial decision of the battle. He never explained why he rejected night battle, but one can note some factors that very likely influenced him at that time. Comparing his available surface strength (three heavy cruisers, a light cruiser, and seven destroyers) with the ships reported to the north demonstrated the enemy’s distinct advantage in numbers, not to mention superior night fighting tactics. Extracting the North Carolina from TF-16, even if Fletcher wanted to risk the inexperienced battleship in such uncertain circumstances, would cost valuable time. Rapidly decreasing destroyer fuel also limited his options for high-speed steaming. The Saratoga’s overcrowded flight deck perhaps also figured in his decision not to seek a night battle. He also had no real idea of the losses the big strike sustained in the attack on TF-16, but good reason to believe only one enemy carrier suffered any significant damage in return.
At 2214 Fletcher advised Ghormley that TF-11 and TF-16 headed south “toward the fueling rendezvous.” The Enterprise sustained “some underwater damage,” but with steering restored, she no longer needed the tug. Fletcher regretted not learning of the two large carriers “in time” to attack, but reported Saratoga planes smashed the Ryūjō, “Left burning badly,” and roughed up cruisers and a battleship in a separate force. He could not say what damage the Enterprise attack group might have caused but stated those planes were to remain at Cactus “until further orders.” TF-11 and 16 “must fuel tomorrow,” but Noyes’s “Task Force 18 is fueled today [and] will stand toward Cactus to support that place in accordance operation order.” At least Fletcher assumed that was where Noyes would go, but Noyes would do much more.8
Kinkaid provided Fletcher more sobering news. The Enterprise suffered seventy-four killed and one hundred wounded, the center elevator was inoperable, and an unspecified number of “after compartments” had flooded. Twenty-six aircraft were “missing,” but presumably included planes on the Saratoga and at Cactus. At 2040 the Sara’s radar showed approaching “friendly planes,” Enterprise aircraft that did not go to Guadalcanal. In the light of the nearly full moon, six TBFs appeared above TF-16. Kinkaid courageously illuminated his ships for landing. The second TBF plowed spectacularly into the crane abaft the Enterprise’s island and fouled the deck, forcing Davis to ask the Saratoga at 2134 whether she could take four more planes. Ramsey replied, “Affirmative, in about 15 minutes.” Fletcher also authorized lights to be shown. Following another furious respot, the Saratoga recovered the four lucky TBFs by 2205. Thus Fletcher learned the Enterprise strike never rendezvoused. In rapidly worsening weather the TBFs sighted ship wakes in the dwindling twilight about 275 miles out, only to recognize Roncador Reef north of Santa Isabel. Unaware of the option to fly to Henderson Field, the Avengers dumped their fish and fortuitously found home plate. Late that evening Vandegrift radioed that the eleven Enterprise SBDs set down safely without contacting the enemy. A half hour after the last TBF landed on the Saratoga, Davis entreated Ramsey to take another orphan on board. “Affirmative,” Ramsey gamely replied, “but it will be a tight fit.” At 2303 Leslie’s TBF thumped down on the Sara after a 6.5 hour flight. He explained to Fletcher he, too, never sighted Japanese ships. It was fortunate, given the Sara’s crammed flight deck, the eleven SBDs never showed. She had ninety-three aircraft (forty-three F4F-4s, one F4F-7, thirty-two SBDs, and seventeen TBFs) from both groups. Things got so bad that plane handlers even tipped some F4Fs onto their noses to stack them close together and open more room forward. The air department performed magnificently in compensating for the Sara’s real drawbacks in battle.9
Following TF-16 by an hour, Fletcher monitored some, but not all, the contacts transmitted by two radar-equipped PBY night searchers from Ndeni. At 2133 Lt. (jg) Norman K. Brady encountered “unidentified vessels in rain” about two hundred miles north of TF-61, roughly at the location of the light carrier group that flanked the main body to eastward. A half hour later Ens. William C. Corbett warned of five ships only 145 miles northwest of TF-11 and closing at twenty knots. The Japanese were pushing hard in pursuit. McCain directed the Mackinac to get her PBYs up an hour before first light on 25 August and clear out. Ghormley agreed with Fletcher that as many as four carriers chased TF-61 from “northward Malaita,” and only one had sustained damage. He pointedly reminded Fletcher of Cincpac’s 20 August message warning of increased numbers of enemy destroyers and the “Jap boast of [night] torpedo technique,” hardly a ringing endorsement for a night surface fight. Ghormley also imparted to Vandegrift cheery news of “enemy carriers heading toward northern Malaita,” that “may hit your positions at day light.” He told the Hornet task force, staging south from Pearl, to refuel “as soon as possible” and be ready for “offensive operations” to westward of the rendezvous with TF-61. TF-17 still only neared Fiji and was one thousand miles from the battle area. Much could happen in the three days it required to get there.10
If the situation on the evening of 24 August confused the U.S. commanders, their Japanese counterparts smugly contemplated their glorious victory. Kondō’s own Advance Force (five heavy cruisers, one light cruiser, five destroyers, and seaplane carrier Chitose) and Nagumo’s Kidō Butai (Shōkaku, Zuikaku, two battleships, three heavy cruisers, one light cruiser, and eight destroyers) weathered disconcerting attacks by a few dive bombers, but took their revenge by hurling seventy-three aircraft against the U.S. carriers. Soon joined by Abe’s vanguard, Kondō raced south to destroy the cripples. The top commanders were unaware of the sinking of the Ryūjō in Hara’s Support Force. Cornered that afternoon by the Saratoga’s first strike, the Ryūjō absorbed one torpedo, which knocked out steering and propulsion, and severe blast damage from near misses by 1,000-pound bombs. Hara detailed the destroyers to sink the burning carrier while he left in the Tone in accordance with Nagumo’s orders. Following an unsuccessful attack by Manierre’s three B-17s, the Ryūjō went down after dark.11
Nagumo knew his first wave found the U.S. carriers. At 1650 a fresh shadower, the Chikuma Number Five seaplane, corrected the erroneous position report of the Hiei searcher. Yet Takahashi, leading the second wave, copied neither that vital message nor Nagumo’s rebroadcast telling him the U.S. carriers were east of where he was headed. Some second wave aircrews heard both messages but wrongly assumed Takahashi did, too. Nor did the second wave notice the returning Saratoga first strike cut southeast across its flight path. Consequently Takahashi fetched up too far west and south to sight Fletcher’s carriers. At 1730 Takahashi again missed another golden opportunity from Seki’s post strike report, also rebroadcast by the Shōkaku, that others in his group did copy. At 1743 he reached where he expected to find enemy carriers, but saw only open water. The second strike was eighty-three miles southwest of TF-16. Takahashi searched westward in vain until sunset and reluctantly turned for home. His failure to find TF-61 and finish off the stricken Enterprise became the crucial Japanese mistake of the battle.12
About the same time Takahashi finally gave up, Sewart’s four high-flying B-17s encountered Abe’s vanguard thirty miles ahead of Nagumo’s main body. Sewart claimed multiple bomb hits on a nonexistent carrier but actually missed the destroyer Maikaze. Later his Fortresses skirmished with Nagumo’s combat air patrol but failed in the setting sun to discern nearby Shōkaku and Zuikaku. Neither Army Air Force navigation nor its ship identification was on the mark. Sewart’s 1810 report of another small carrier, three heavy cruisers and two destroyers was what led Fletcher to believe there was a carrier thirty miles east of the force Larsen struck.13
For his own part Kondō kept east of Nagumo’s carriers as both task forces aimed for the gap between Malaita and the Santa Cruz Islands. The Chitose (Elder’s “battleship”), while stopped to recover aircraft, noticed Larsen’s tiny strike at 1740. Near misses from the two SBDs inflicted severe damage. The five VT-8 TBFs scored no hits against Kondō’s four swift heavy cruisers. Two Avengers ditched (their crews were rescued) after encountering far-ranging combat air patrol Zeros. That evening Kondō also sought the U.S. carriers to the south and absorbed Abe’s vanguard. Special night search float planes fanned out looking for targets. From Truk, Vice Adm. Komatsu Teruhisa, leading the submarine fleet, roused his ten I-boats in pursuit.14
Only at 1900 did just twelve of Seki’s original thirty-seven planes reach home. Nagumo announced at 1928 after debriefing Seki the first strike scored more than three bomb hits on an Enterprise-class carrier. A Zuikaku report claimed two bomb hits on a different carrier. The enemy remnant was fleeing at twenty-five knots. Nagumo decided to head north to refuel after recovering his second wave. He got an argument from his flag Capt. Arima Masafumi, who advised vigorous pursuit, but Nagumo declared the battle was already won. His final assessment, transmitted at 2040, chalked up six bomb hits on a “new-type” carrier (the Essex) and another carrier and battleship set afire. Kondō independently reached the same optimistic conclusion, radioing at 2005 that he would pursue only until midnight to the last known enemy position to finish off disabled carriers.15
Retrieving the second wave proved exceptionally difficult. Starting at 2015 the carriers recovered twenty-eight of thirty-six planes. Takahashi learned to his great regret of the bungled opportunities to finish off the U.S. carriers. Seven carrier bombers went missing, and another ditched. Nagumo turned on searchlights and broadcast radio homing messages until all hope was lost. Incredibly two hours later, three errant carrier bombers turned up flying on fumes. Slonim’s radio intelligence team in the Saratoga eavesdropped the homing efforts Fletcher correctly attributed to the strike group that never found TF-61. “These planes were later heard in the evening trying to home on their carrier but received no answer,” Pederson recalled. As will be shown, Hypo analysts listening in at Pearl wrongly decided many enemy aircraft splashed that night vainly seeking their own carriers. The Shōkaku and Zuikaku in fact lost nine fighters and twenty-four dive bombers, leaving one hundred aircraft (forty-one fighters, twenty-five dive bombers, and thirty-four torpedo planes) available the next day.16
Kondō’s lookouts made no contacts before the stipulated time. His night air search overtook only a lone “light cruiser” (the Grayson) running away. The bloodied U.S. carriers obviously fled beyond reach. Kondō called off the pursuit at 2330 and led the surface force north at twenty-four knots to refuel. At the same time Yamamoto’s Combined Fleet Main Force (one battleship, one converted carrier, and three destroyers) approached the battle area from the northwest. He exulted the “Second Solomons Sea Battle” (Savo was the first) left two U.S. carriers in flames and sent the whole enemy force packing. It is fascinating to consider, especially in contrast to his vehemence after the Coral Sea battle, that Yamamoto endorsed the decision of his top commanders not to harass the beaten enemy. Kondō broke off the battle when it was “appropriate under the circumstances,” but Nagumo was to be ready to attack any “damaged ships” that turned up the next morning. Learning of the loss of the Ryūjō, Yamamoto theorized that carriers skulking southeast of Guadalcanal had hoped to ambush the Second Echelon convoy, only to encounter the unlucky Hara and suffer a devastating surprise assault. Ugaki called the Ryūjō a splendid, if unwitting, “decoy” (as did historians) and strongly regretted not actually finishing off the hurt U.S. carriers. Their temporary escape made no difference in the short run, because the Allied defense of Guadalcanal was doomed. To the top brass only severe communication failures and poorly framed contact reports prevented an even more decisive Japanese victory.17
Now that his forces swept away all significant naval and air opposition, Yamamoto directed the KI Operation to proceed and deposit the Ichiki Second Echelon at Guadalcanal on the evening of the twenty-fifth. Near to midnight on 24 August, when 190 miles north of Guadalcanal, Tanaka received orders once more to go in. Having seen smoke from the burning Ryūjō, he understandably entertained “grave doubts about this slow convoy’s chances of reaching its goal.” Yamamoto rejected the vehement recommendation of Tanaka’s superior Tsukahara for Nagumo’s carriers to bomb the Guadalcanal airfield. Evidently Yamamoto thought the Ryūjō, before expiring, destroyed Lunga’s planes on the ground. Three destroyers rushing down from Shortland would finish the job that night. Yamamoto and his two fleet commanders exhibited an astonishing level of wishful thinking and were in for sore disappointment.18
By following Kinkaid’s TF-16 south, Fletcher ensured TF-11 would not fight on 25 August unless the Japanese advanced within range of the Saratoga’s planes. Thus he passed the torch to Noyes’s TF-18 with the Wasp (sixty-two planes). Noyes started north late on the afternoon of 24 August after rapidly refueling. TF-18 worked up to twenty knots, as fast as he dared push the Wasp’s tender propulsion plant. He soon intercepted radio messages that portrayed a full-blown carrier battle in which the Enterprise was hit and Fletcher also withdrawing. Rather than simply close Cactus 350 miles northwest, he resolutely reinforced Fletcher and Kinkaid east of Malaita. In the early hours of 25 August TF-18 swept past TF-16, TF-11, and finally the Grayson moving hell-bent in the opposite direction. Now only the Japanese carrier fleet loomed ahead of Noyes, with every expectation of renewing battle at dawn.19
Committed to refueling, Fletcher followed the action by radio, but it is uncertain what messages the Saratoga succeeded in copying. Thus it is not known whether he (or Noyes, for that matter) ever received Vandegrift’s warning of warships bombarding Lunga after midnight. The second message stated that seven attacking destroyers now drew off to the northeast. In fact three destroyers breezed into the anchorage off Lunga Point looking in vain for U.S. ships and loosed a few rounds toward the airfield for ten minutes before turning north to swell Tanaka’s antisubmarine escort. The shelling inflicted little damage (the leader of the destroyers thought as much), and only served to stir up Vandegrift and his flyers. Six marine and navy SBDs harassed the retreating destroyers. Vandegrift certainly counted on Fletcher’s close support, but Ghormley knew better. He urged MacArthur to commit his subs against the ships assailing Guadalcanal, because there were “no friendly vessels [in] that area.”20
That night the two Airsopac PBYs probed the dark waters north of Malaita and Guadalcanal, trusting their radar to locate targets and the bright moonlight to identify what turned up. At 0149 McCain relayed an “unauthenticated aircraft contact report,” first transmitted at 2105, of an “enemy carrier group” discovered only 120 miles north of Lunga and speeding south at twenty-five knots. Fortunately at 0225 Ensign Corbett, who made that contact, corrected its position to 180 miles north of Guadalcanal and reduced its speed to fifteen knots. Around 0340 five float planes from Mikawa’s cruisers bombed Lunga, again without noticeable result. Vandegrift went all out against the carriers north of Malaita. Well before sunrise he sent a strike of eight marine and navy SBDs and ten F4Fs. Shortly thereafter a wary Noyes launched the Wasp dawn search from east of Malaita and 250 miles from Lunga. Likewise Fletcher, then 120 miles south of Noyes, wrung out ten Saratoga SBDs to sweep two hundred miles north.21
Lt. (jg) Charles P. Muckenthaler, the first PBY from the dawn search to make contact, squawked of attacks by aircraft. His reported position was 170 miles north of TF-18 and three hundred miles north of TF-11. Those Japanese planes, if carrier-based, had to originate from a different group than the one bearing down on Cactus from the north. Nimitz read an ominous estimate that “from all appearances an aerial attack will be carried out against our holdings in Tulagi-Guadalcanal at daybreak followed by an attempted landing.” To Ghormley, Fletcher, and McCain he stressed, “Situation [is] still critical,” but the “exchange of damage to date seems to be in our favor. Let’s finish off those carriers.” In the next several hours carriers from both sides failed to materialize, while Lunga-based planes blasted the invasion convoy. Beginning at 0930 messages from Cactus and various Airsopac searchers mentioned at least one transport seen afire. Finally at 1055 Lt. James Murphy’s PBY confirmed, after carefully checking the area, that the convoy of seven destroyers and two cruisers (but no carriers) had left a burning transport behind and slunk away. Later, PBYs encountered scattered groups of battleships, cruisers, and destroyers likewise retiring northward, the nearest more than three hundred miles northeast of Lunga and three hundred miles north of TF-18.22
The immediate threat to Guadalcanal had passed. The ships Corbett first sighted that night (and mistakenly identified as a carrier force) were Tanaka’s Second Echelon convoy, the only force actually steaming toward Guadalcanal. Two of the three destroyers that shelled Lunga joined after dawn, giving Tanaka one light cruiser, seven destroyers, and four patrol boats escorting three slow transports. By 0800 he was within 150 miles of Guadalcanal, as close as he would get. After seeking the nonexistent carrier group, the Cactus SBDs settled for the convoy so conveniently served up without air cover. They bombed the big transport Kinryū Maru (which burned), hit another transport, and roughed up Tanaka’s flagship Jintsu. At 1015 three B-17s sank the destroyer Mutsuki, which had stopped to rescue survivors. Totally isolated, a dazed Tanaka ordered a retreat. By preventing Japanese reinforcements from reaching Guadalcanal, the valiant Cactus flyers transformed an indecisive clash of carriers into an Allied strategic success.23
Fletcher’s morning search was negative, as was Noyes’s, but Wasp SBDs destroyed three ship-based float planes. With no other targets reported within reach, Noyes angled northwest toward the invasion convoy being assailed north of Lunga. At 1326, when about 275 miles distant, he loosed twenty-four SBDs and ten TBFs—every available attack plane—at extreme range and continued northwest to shorten their return flight. The Wasp strike carefully searched the designated area and shot down a flying boat that trailed them from the vicinity of TF-18 but sighted no ships. On the way back they checked out Rekata Bay, the suspected float plane lair on the north coast of Santa Isabel, but found nothing. The disappointed strike crews turned southeast for home. Just before sundown the Wasp landed them and shaped course southeast toward TF-11. Noyes did everything he could under the circumstances to harm the enemy and support Fletcher, but Nimitz and King never accorded him any credit.24
Fletcher’s TF-11 followed Kinkaid’s TF-16 with the hurt Enterprise toward the oiler rendezvous between San Cristóbal and Espíritu Santo. When air searches showed no sign of pursuit, the chief concern became subs. At 0910 Ens. George G. Estes, flying intermediate air patrol from the Enterprise, caught two big subs on the surface only fifteen miles ahead of TF-11. Certain he destroyed one, he dropped a message onto the Saratoga’s deck describing the encounter. On his way back south to TF-16, he discovered a surfaced sub in the same location as the previous attack, strafed, then returned to the Sara to land. Lookouts sighted yet another I-boat to the southwest. Fletcher detached three destroyers to attack and keep it down. At 1330 a cluster of ships emerged ahead. TF-16 guzzled oil from the three oilers. Kinkaid revised his damage estimate of the Enterprise to three “medium” bombs and three damaging near misses. Sprung hull plating flooded one large aft compartment, which “prevents maneuvering at high speed.” Fortunately her propulsion system was untouched. The Enterprise had thirty-one flyable planes (six F4Fs, twenty SBDs, and five TBFs) and could operate fifty if necessary. Kinkaid recommended proceeding to Nouméa “when practicable” to “effect temporary repairs,” but obviously the Enterprise required extended time in a proper ship repair facility.25
With oilers in abundance, Fletcher filled TF-11 to near capacity. From TF-16 he drew the North Carolina, light cruiser Atlanta, and destroyers Grayson and Monssen, giving the Saratoga a screen of one battleship, three heavy cruisers, two light cruisers, and nine destroyers. Kinkaid retained heavy cruiser Portland and three destroyers. Another destroyer would join him after transferring personnel and matériel. Victory claims by U.S. aircraft totaled fifty-seven planes, and antiaircraft chalked up twenty-four more victims. The two carriers sorted their air groups. Plane losses on 24 August came to only sixteen aircraft: five F4Fs and four TBFs shot down or missing, one F4F, two SBDs and one TBF ditched, and two F4Fs and one TBF jettisoned. The Saratoga absorbed seventeen Enterprise F4Fs and six TBFs, raising her air complement to eighty-nine planes (forty fighters, thirty-one dive bombers, seventeen torpedo planes, and one reconnaissance aircraft). The Enterprise retained eight fighters, twenty-two dive bombers, five torpedo planes, and one reconnaissance plane but was to leave most of them behind prior to departing Sopac. The Saratoga took her turn with the Sabine that evening. Only numerous sub contacts interrupted a quiet day.26
Contact reports that Fletcher copied revealed all enemy ships, including the battered invasion convoy, had retired north beyond air search range. The daily Cincpac intelligence bulletin belatedly acknowledged the Zuikaku, Shōkaku, and Ryūjō in the Solomons area, while the Zuihō, Junyō, and Hiyō “have all earmarks of departing Japan,” perhaps “within last 48 hours.” If so, they could reach the Solomons in the next week. The converted carrier Kasuga Maru was thought to be near Truk but bound for Rabaul. Among other fleet units in the New Britain area or headed there were four cruiser divisions and two destroyers squadrons. “Cinc Combined Fleet” flagship had left home waters for Truk. The proximity of such powerful enemy forces to Sopac hardly came as news to Ghormley and Fletcher, but at least there was welcome respite from battle.27
Fletcher intended to send Kinkaid south to between Efate and New Caledonia to await further orders that would likely mean a return to Pearl. He also provided a message for Ghormley that the Enterprise was to fly to Efate on 26 August as TF-16 passed by. It outlined the current situation and explained that TF-11 expected to complete fueling the night of 25 August and rejoin Noyes’s TF-18 east of Malaita. Fletcher recommended the fragile Wasp remain “until situation clears,” and desired TF-17 (Hornet) to join him “as soon as practicable.” In any event TF-11 and TF-18 must provision in ten days. That entailed visiting an island base, because big ships could not easily draw large amounts of food stores at sea. Well aware Ghormley might require a separate surface force, Fletcher offered Crutchley’s TF-44 (Australia, Hobart, and three destroyers), plus the Salt Lake City. TF-44 should come under direct Comsopac control, “As radio silence precludes proper command by me.” That night Fletcher released Kinkaid’s TF-16 to southward. After the fast, highly efficient refueling session, augmented TF-11 turned north toward San Cristóbal to find TF-18. The empty Platte and Cimarron left for Nouméa, while the Sabine, which retained some oil, started for Espíritu Santo.28
The morning of 25 August brought bitter disillusionment for a hitherto ebullient Japanese high command. By 1100 when Yamamoto’s Main Force met the Kidō Butai 350 miles northwest of Guadalcanal, he knew the mauling of the Second Echelon convoy prevented any troops from landing. He called off the KI Operation and directed the Zuikaku to cover Tanaka’s retreat. Even worse, he soon understood the U.S. carriers survived their ordeal surprisingly well. Around midnight while redeploying southward, the I-17 and I-15, operating independently 125 miles east of San Cristóbal, had simultaneously sighted carrier forces retiring at twenty knots. Neither sub could maneuver into attack position. Komatsu ordered all boats in pursuit. After sunrise several subs piped up with fresh contacts of task forces 130–150 miles from the sightings reported the previous night. At 0800 the I-9 reported a battleship 330 miles southeast of Lunga, actually Fletcher’s TF-11. After running afoul of an enemy plane and some destroyers, the I-9 limped to Truk. The I-15 encountered a southbound force at 0915 that comprised a battleship, an Enterprise-class carrier, two cruisers, five destroyers, and two oilers supposedly forty miles northeast of the I-9 contact. That obviously was TF-16, but the I-15 placed it too far north. The rest of daylight brought no new contacts by subs, although they eagerly sought carriers. That evening Komatsu ordered all of them back to their previous picket lines. During the day Yamamoto took his Main Body to 330 miles of Guadalcanal, before relenting and rejoining Kondō and Nagumo to the north. “We must be on alert,” Ugaki decreed. Apparently one U.S. carrier and a battleship had refueled southeast of San Cristóbal and kept watch by “sending out search planes, some of which our search planes met” to their detriment between Malaita and Ndeni. The report by the missing flying boat that afternoon of a carrier five hundred miles southeast of Shortland led Ugaki to believe it located “yesterday’s damaged carrier.” Actually the flying boat had, as noted above, sighted TF-18 with the fully battle-ready Wasp.29
With U.S. air power in the region, carrier and land-based, manifestly active and quite dangerous, Yamamoto revised his plans. The battle for the Solomons would be “a prolonged one.” The recapture of Guadalcanal must now precede the destruction of the U.S. fleet, but Lunga’s air force had to disappear before the transports could be risked again. Destroyers would resume infiltrating troops into Guadalcanal. In Rabaul Tsukahara reiterated his call for Nagumo’s carriers to strike Lunga, but Ugaki urged careful deliberation before undertaking that extreme measure. Combined Fleet needed to regroup and bring in more land-based aircraft. Beginning 26 August, Kondō and Nagumo were to refuel at sea. Lacking oil, two fast battleships had to retire to Truk to await the arrival of another tanker from Japan. Nagumo did not consider the slower but more powerful Mutsu an adequate stand-in. The remainder of Kondō’s Support Force should be ready to fight in a few days. In early September light carrier Zuihō and two more fast battleships were to come south from Japan, followed by the Hiyō and Junyō in early October as soon as their air groups were ready.30
The Eastern Solomons became the most intensively studied carrier action yet. The single overriding complaint concerned communications: poorly functioning radios, delayed contact reports, and other shortcomings in passing vital information. Captain Davis described communications as “weak to the point of danger,” and Kinkaid stated, “Communication failures are the primary cause of many tactical errors.” Fletcher submitted an entire report on the failure of communications. The problem did not admit to a quick fix and dogged the U.S. Navy for the rest of the year.31
Several commentators described a visible deterioration of enemy pilot skills, a premature assessment that the next carrier battle would show. Fletcher called fighter direction “not entirely satisfactory although much better than previously experienced.” The intercept of lone shadowers continued to excel, but the carriers still seemed at a loss concentrating combat air patrol fighters to defeat strike groups. Yet ship antiaircraft improved considerably, and the fast battleship had proven invaluable. At the same time fighter escorts remained weak, particularly for the vulnerable torpedo planes. Some leaders, notably Davis, recommended limiting torpedo planes to moonlight strikes and finishing off cripples, or using them as horizontal bombers. Fletcher strongly disagreed. The aerial torpedo was the most powerful antiship weapon in the air arsenal. His fears for the misuse of torpedo planes soon came true.32
The carriers appeared more satisfied with the Grumman F4F-4 Wildcat, although acknowledging its inferiority to the Zero. As noted previously, the only immediate action in the absence of better fighters was to have more Wildcats. Fletcher even desired a dedicated “fighter carrier,” if enough other carriers were available to accomplish the offensive mission. Kinkaid concurred. Up to this point the carriers depended chiefly on surprise to compensate for numerical or matériel inferiority, but now, in the face of extensive shore-based air, that no longer pertained. Kinkaid rated fighter strength according to the anticipated opposition. For “normal operations” the carriers should each have forty F4Fs instead of the current thirty-six, but against tougher defenses there should be sixty. To conduct amphibious offensives, such as Watchtower, where the carriers themselves must support ground operations, one carrier should have one hundred or more fighters. Increasing fighter strength, however, correspondingly reduced offensive power in the form of dive bombers and torpedo planes. The whole issue needed careful consideration.33
Fletcher continued to oppose the single carrier task forces he was ordered to use on 24 August. His 24 September letter on carrier tactics called for carrier task forces assigned to the same mission to remain together for “mutual support and protection” and not separate “unless there is some strong tactical reason.” In this he followed the iconoclastic Rear Adm. Ted Sherman. Fletcher elaborated these views in his 25 September endorsement of the Enterprise Eastern Solomons report. Davis urged the individual carrier task forces draw at least fifteen miles apart in the event of air attack. Fletcher strongly disagreed. “To an attacking air group, it makes little difference whether the carriers are separated by 5 or 20 miles but to the defenders it makes a great deal. By keeping the carriers separated 15–20 miles there is always the danger that the full fighter force may not be brought to bear decisively against the enemy as happened at Midway.” Davis opined, “The joint operation of more than two carrier forces is too unwieldy.” Fletcher countered that according to “our recent experiences” a task force commander could handle three carrier task forces “almost as easily as two.” Even four could operate together “without too much difficulty.” The “advantages to be gained from such a concentration of air power would more than offset any disadvantages.” He prophesied the “tendency will be to operate more and more carriers together as our offensive gains momentum.”34
Despite intensive analysis, the battle as a whole remained a mystery. The exceptionally poor communications and numerous failures by the search to report correctly the enemy’s strength and position rendered Fletcher, as Slonim explained to Morison, “Thoroughly confused throughout the action.” Fletcher thought there were at least two and very possibly three separate enemy carrier task forces with four flattops. Due though to the “incompleteness of last contact reports,” he could not “fix with any degree of accuracy” their positions. Fletcher’s 6 September 1942 preliminary report designated “Task Force A” the group of ships he thought included the light carrier Ryūjō, which the Saratoga’s aviators claimed to have sunk. He speculated the Ryūjō launched the evening strike that fortuitously missed TF-61. East of Task Force A was “Task Force B” with the big carriers Shōkaku and Zuikaku, which likely dispatched the huge strike of eighty aircraft that hit TF-16. Larsen’s mini-strike very likely targeted B’s screen, possibly damaging a battleship and two cruisers. Fletcher supposed another light carrier force, “Task Force C,” operated southeast of Task Force B and was the source of the Cactus-bound strike detected early that afternoon on radar. The confusion persisted long after the battle. The Ryūjō remained in the Cincpac intelligence bulletins for a week. Ghormley told Nimitz on 3 September it was likely Fletcher’s and McCain’s bombers hit different carriers and that the Ryūjō was just badly damaged. Pederson remarked in late 1943, “[The] consensus of those of us who were there is that there were 4 Jap carriers.” Of course, only three Japanese carriers fought in the Eastern Solomons. The phantom Task Force C resulted from misidentification and incorrect navigation by search planes and Sewart’s B-17s. Even nailing down the loss of the Ryūjō proved difficult. The cryptanalysts were not certain until January 1943, when they deciphered a naval message that deleted her from the navy list.35
The other fundamental question raised by the battle was how the Japanese carriers could reach the Solomons without radio intelligence predicting and confirming their movements. Fletcher discussed the intelligence failure with Nimitz when they next met. He brought up the Cincpac daily intelligence bulletin, received on 23 August, placing all Japanese carriers north of Truk, where instead as many as four flattops had secretly closed within a few hundred miles of TF-61. According to Layton, Fletcher “cussed so loudly about the intelligence being bad,” but he was not the only one deeply concerned. Even while the battle raged, Captain Steele’s daily summary of radio intelligence had noted: “It now seems probable that Jap CVs are about to enter Truk. If this is so they can hardly arrive in the Solomons before August 27th, local date.” The day after the battle Rochefort’s Hypo analysts admitted their bafflement. “Success of a large task force including carriers in reaching the Solomons without detection by R.I. indicates that radio security practices of the Japs are effective insofar as concealing actual movements is concerned.” The experts finally deduced the initial inclusion of the carrier forces in the local communication net really meant they were already present instead of just en route. Layton later decided the reason the expected carrier arrival report at Truk had not been intercepted was that, unknown to the Allies, the striking force bypassed Truk and kept going south. In the end, “You can’t always be right,” and “that’s what RI is all about.”36
Despite an at least temporary strategic victory in the Battle of the Eastern Solomons, many thought Fletcher should have done much better. He should have taken advantage of what they perceived was a highly advantageous tactical position on the evening of 24 August. The extremely exaggerated claims of aircraft destroyed in the attack on TF-16 (at least eighty-four planes shot down by combat air patrol and antiaircraft), as well as tantalizing hints from radio intelligence of additional heavy plane losses, had mesmerized Nimitz. His preliminary report postulated that 24 August cost the enemy “the best part of three or four carrier groups destroyed in action,” or “most of the Japanese aerial striking force available that day.” The eighty planes that were annihilated assailing TF-16 represented “at least one large and one small carrier group.” Moreover, “The second attack group that missed locating Task Force 61 late in the afternoon could have been the other large carrier air group,” whose “planes were later in the evening heard trying to home their carrier.” Thus, Nimitz speculated, “Some or all of the group were lost.” Given confusing reports of twin-engine bombers attacking Cactus on 24 August, he was uncertain whether carrier planes also struck Henderson Field, whose defenders claimed eleven single and twin-engine bombers and seven fighters. “If so, the Japanese lost the best part of three carrier groups in the action of 24 August, and perhaps much of a fourth if a carrier group joined in the attack on Guadalcanal.” Consequently, “Not only was the enemy landing expedition turned back, but we had definitely won control of the air.” At the same time Fletcher’s own air losses (sixteen planes) were extraordinarily low. Despite damage to the Enterprise, “We still had intact practically two full carrier air groups.” At the same time, the “Wasp Task Force was fueled and proceeding North to join TF-61 as it retired southward. The Hornet Task Force was nearby enroute.” Therefore Nimitz strongly regretted that Fletcher withdrew the Saratoga group and wasted all 25 August refueling. At the same time Noyes’s TF-18 merely “took position in the general area southeast of Guadalcanal prepared to repel further Japanese attack.”37
In his ignorance of actual events, Nimitz was extremely unfair. The final Cincpac battle report, dated 24 October 1942, not only repeated the earlier comments regarding heavy enemy air losses, but also strongly questioned Fletcher’s actions: “The withdrawal of the Saratoga Task Force on the night of 24 August broke off the engagement. At this time the enemy carrier air forces were depleted. Based on hindsight it appears that had the Saratoga group remained in the vicinity occupied on the 24th, they might have been able to strike the enemy with air and probably surface forces, since the Japanese heavy units continued to close during the night and did not reverse course until the forenoon (afternoon for some groups) of 25 August.” Cincpac described Japanese vulnerability on 25 August not only to carrier air attacks, but also surface combat from Fletcher’s task groups. “We must use our surface ships more boldly as opportunity warrants.” However, “The distances involved, the problem of coordinating widely scattered forces, along with inadequate communication facilities and training, together prevented the victory from being as decisive as it should have been had planes and surface forces come to grips with the enemy after his air forces were largely destroyed on the 24th of August.” The Cominch 1943 Battle Experience Bulletin echoed Nimitz’s opinion of poor leadership at Eastern Solomons. “Task forces must anticipate repeated attacks and be prepared to repel them and not be in an unprepared and disorganized condition,” referring to TF-11 on the evening of 24 August. “Our task force took the defensive more than the offensive,” whereas “effort should be made to follow up initial successes in order to completely annihilate the enemy whether it is daylight or dark.” On 25 August, however, the “attitude was too defensive,” especially because “an offensive by Task Force 18 might have resulted in complete destruction of the enemy.”38
Of course, the tactical situation Fletcher actually faced on the evening of 24 August was far less sanguine than Nimitz and King ever appreciated.39 They considered the supposedly decimated Japanese easy pickings for the Saratoga and Wasp flyers had Fletcher only mustered the courage to stay and fight. In fact, Cincpac vastly overestimated Japanese carrier plane losses. The actions on 24 August cost the Shōkaku and Zuikaku just thirty-three aircraft. Far from having his air groups “largely destroyed,” Nagumo enjoyed his full aerial torpedo capability of thirty-four carrier attack planes and retained half his dive bombers, as well as forty Zero fighters. The battleships, cruisers, and destroyers led by Kondō and Abe were too formidable for Fletcher to take on alone in surface combat. As so often, the operative word is “hindsight.” Just a spectator to the attack on TF-16, all Fletcher had to go on that evening, while he deliberated whether or not to withdraw TF-11, were the initial impressions of tired fighter pilots who landed on the Saratoga. That night it looked as if three of four enemy carriers remained untouched, and that powerful surface forces closed fast. He did not gain a fuller picture of the defense of TF-16 until after contacting Kinkaid the next afternoon. The compilation of victory claims only came after collating the action reports. Criticism of Noyes as merely standing toward Guadalcanal instead of seeking to engage carriers to the north was extremely unjustified. Noyes went where he thought the enemy’s main body would be. He did not find it, because the Japanese themselves pulled out. Nimitz’s mention of the Hornet task force being nearby was especially misleading. During the battle TF-17 was fully one thousand miles away and could not possibly have intervened for several days.
The negative impression persisted of Fletcher’s failure to pursue.40 Morison partially heeded Slonim’s explication that the battle “exemplified the difficulties in the exercise of command at the point of contact.” Thus “by playing up these difficulties,” Slonim thought Morison could mitigate “some of the criticism of Admiral Fletcher which seems to be somewhat harsh.” Of his old commander, Slonim condescendingly stated: “We all know Mr. Fletcher was not the greatest of war leaders, but I think his problems were sufficiently great to justify some of the errors.” In truth Fletcher’s problems were far more profound than Slonim ever understood. Enjoying better, but by no means comprehensive, access to Japanese records than wartime analysts, Morison conceded when Fletcher “decided to call it quits for the day,” he was “amply justified” in escaping the charge of Kondō’s “sea-going cavalry force.” Morison favorably compared Fletcher’s decision to the wisdom Spruance demonstrated on the night of 4 June in backing out of harm’s way at Midway. “It was not known whether the Japanese had had enough, but it was clear that their available gunfire power was much greater than Fletcher’s, and night was no time to test the truth of this estimate.” Yet Morison condemned what he, too, perceived a lack of an aggressive spirit combined with Fletcher’s characteristic obsession with fuel. “How badly did his destroyers need that drink of fuel oil for which he took Saratoga completely out of the battle scene on the 25th?” Instead, “As long as the issue was in doubt, every available carrier aircraft should have been used to protect our tenuous lease on Guadalcanal. That is what they were there for. Fletcher won the battle to be sure; but only because the Japanese were more timid than he.”41
Fletcher’s decision to withdraw was tactically wise; he did need to fuel his destroyers and the rest of TF-11, for that matter, to permit the extended high-speed operations that renewed battle would demand. Vandegrift already had as many or more carrier planes as Cactus could handle. Fletcher always had to keep in mind that only the intervention of the Japanese heavy fleet forces would ensure in the end whether or not the marines held Cactus. Their carriers were his primary objective. Kondō and Nagumo in truth withdrew their fleets not from timidity, but overconfidence. They would be back, and Sopac ready to meet them, in part due to Fletcher’s well-measured responses during the Battle of the Eastern Solomons. Morison’s final verdict, “American movements too were unaggressive, largely from want of intelligence about the enemy,” is a fair assessment. So, however, is Nimitz’s considered judgment that despite all Fletcher’s perceived failures to take advantage of the situation, the Battle of the Eastern Solomons was “a victory that turned back the first major assault of the Japanese to regain Guadalcanal-Tulagi and gave us several weeks of valuable time to consolidate positions there.”42