Fletcher’s TF-17 hustled through the New Hebrides to meet Fitch’s TF-11 on the afternoon of 1 May in the Coral Sea at Point Butternut, three hundred miles northwest of New Caledonia and four hundred miles southeast of Guadalcanal.1 On the evening of 30 April Fletcher read a Cincpac message that confirmed the principal objective was Port Moresby. It identified Deboyne, an atoll strategically placed in the central Louisiades, as site for a rendezvous between the “Saipan Base Force” and two merchantmen (“Marus”) about to sail from Rabaul. Japan also showed interest in Samarai at the east tip of New Guinea and Cape Rodney on the south Papuan coast.2
Before Fletcher could fight, however, he must assemble his scattered forces and refuel, a process he anticipated might take several days. Cincpac ordered him to unite the task forces at Butternut, but circumstances militated against it. At dawn on 1 May, the Lexington’s unmistakable profile materialized a dozen miles west of TF-17. Fletcher took advantage of Fitch’s early arrival to send TF-11 alone to Butternut and gather the Chicago, Perkins, and oiler Tippecanoe coming out from Nouméa. Fletcher would join them there at dawn on 2 May. He changed the plan because he hoped to fuel TF-17 from the Neosho, not easy given the bad weather that barred fueling on 30 April. May Day likewise shaped up raw and rainy. Rough seas and gusty southeast winds forced TF-17 to steer away from Butternut to ease conditions. Heavy seas smashing over the Neosho’s main deck injured several crewmen. It took until almost dark just to fuel the Yorktown and Astoria, while the Portland and Chester provided oil to the destroyers. In the meantime, Fitch marked time near Butternut. At noon unescorted Tippecanoe hove into view, followed by the Chicago and Perkins. Fitch did not start immediately fueling TF-11, but had a destroyer shepherd the slow oiler around the area and rejoin the next dawn. According to Poco Smith, Fletcher refused to permit both carriers to refuel simultaneously at low speed in possible sub-infested waters. Fletcher, Fitch, and the Tippecanoe group all joined at sunrise on 2 May, where the oilers fueled their respective forces. Fletcher hoped by the evening of the third to get as much fuel as possible out of the Tippecanoe before releasing her to Efate and to hold the balance of Neosho’s oil in immediate reserve. He constantly worried about uncertain logistics and preferred to keep a fast oiler with him until just before going into action.3
Fletcher received no fresh information on 1 May either from intelligence or aircraft sightings, indicating the enemy offensive must still be in an early stage. The next afternoon while fueling he learned of increasing pressure against Tulagi. Leary’s summary remarked of “ships sighted afternoon 1 May,” as well as raids by shipborne aircraft on Tulagi that had to come from “a possible cruiser or seaplane carrier.” The reference to “ships sighted afternoon 1 May” caught Fletcher’s attention, for Leary hitherto had said nothing about them. Two ships, types not given, were spotted on the first about thirty-five miles southwest of Gizo in New Georgia, two hundred miles northwest of Tulagi. An RAAF Catalina was to have shadowed them the night of 1–2 May, but Leary offered no updates. He also told Fletcher that Area C (from New Guinea east to Bougainville) and Area E (the east flank of Solomons) were not searched on 1 May. Presumably they would be on the second.4
Carrier operations, Coral Sea, 1–4 May 1942
Enemy activity in the Solomons did not as yet appear strong or focused. Unfortunately on the second Fitch’s fueling did not go well. He started with his cruisers, and the Tippecanoe’s outmoded fueling rig carried away once that afternoon. The Chicago did not finish until near to evening, leaving the destroyers and the Lex’s vast maw to fill. Fitch ruefully advised Fletcher that he did not expect to wrap up until noon on 4 May. Although “disappointed,” Fletcher believed that “reports of enemy forces precluded remaining so far to the southward.” He once again separated the carriers, this time for thirty-six hours. TF-17 would head west along with the Chicago and Perkins. Fitch would continue fueling to southward, release the Tippecanoe and a destroyer to Efate, then rejoin TF-17 on the morning of 4 May at the same rendezvous set for Crace’s TF-44, which was coming up from Sydney. That was three hundred miles northwest of their present position.5
Captain Bates criticized Fletcher for not immediately uniting TF-17 and TF-11, or at least keeping in visual contact. He recognized Fletcher’s “desire to be ready for immediate service—his desire for freedom of action should an emergency arise—his anxiety over the developing situation—his desire to be more to the westward and northward.” Nevertheless, he should have combined the two task forces. Bates understood it was unwise to loiter in a small fueling area because of subs and speculated that Fletcher could have expedited fueling by swapping oilers with Fitch. Actually the ancient Tippecanoe gave Fletcher few options. To stay together greatly restricted mobility, especially considering Fitch’s lengthy estimate for fueling. Fletcher could not simply give Fitch the Neosho, because he counted on her oil to tide TF-17 over for the next week. Instead, he planned to patrol west of his present position, watching the situation develop to the northwest in the Louisiades. Intelligence pointed to the Rabaul-Louisiade-Moresby axis where the enemy carriers would turn up to support the Invasion Force. Fletcher obviously expected nothing major to break loose until after Fitch rejoined on 4 May.6
The Sowespac air search network poorly served Fletcher, but not through the fault of the aviators. The problem lay with higher headquarters not correctly interpreting search results or disseminating them in a timely manner. On 1 May RAAF flying boats shuttling in and out of Tulagi discovered not only “two ships” (which they properly identified as merchant types), but also a convoy of five “merchant vessels” west of New Georgia. One Catalina shadowed the little convoy for five hours and later amplified the contact as one ship of twenty-five hundred tons and numerous armed trawlers. That activity occurred in Area C, which Leary mistakenly said was not even searched. In fact by 2 May Japanese ships dotted the northern and central Solomons. The “two ships” were small converted gunboats from Marumo’s Support Force involved in building seaplane bases. He left New Ireland on 29 April with two light cruisers and the seaplane tender Kamikawa Maru, stopped at Queen Carola Harbor on Buka, and ventured south on 1 May through Area C. By the next dawn he was in Blanche Channel off southern New Georgia. Shima’s Tulagi Invasion Force departed Rabaul on the thirtieth and passed west of the Solomons, again through Area C. The little convoy belatedly reported late on 2 May to Fletcher was Shima’s Patrol Force of small minesweepers. The whole force was to be off Tulagi by midnight on 2–3 May. On the evening of 2 May Gotō’s MO Main Force (four heavy cruisers, light carrier Shōhō, and a destroyer) cut through Bougainville Strait between Bougainville and Choiseul, then turned southeast toward Tulagi. These extensive enemy movements within the area supposedly being searched by Brett’s flyers far exceeded the “two ships” relayed to Fletcher.7
On the afternoon of 2 May, before TF-17 and TF-11 separated, an SBD dropped a message informing Fletcher of a sub running on the surface twenty-three miles north. Three Yorktown TBD torpedo bombers plastered the I-boat with depth charges as it crash-dived. The pilots were so positive about the kill that Fletcher thought it “difficult to understand how the submarine could have escaped.” To make sure, he dispatched eleven dive bombers and sent for good measure the Anderson and Sims with orders to rejoin TF-17 to westward the next dawn. Biard’s radio intelligence team heard no contact report from a sub, which again led credence to its destruction. In fact the I-21 was crossing the Coral Sea to scout Nouméa when two planes forced her to dive. Her lookouts failed to recognize the assailants as carrier-based, nor did her skipper attach importance to the attack, for he was within range of Allied air bases.8
The evening of 2 May Fletcher shoved off to the west. No intelligence he received that evening offered any reason to change his plans. Tulagi was again bombed, which indicated “either enemy intention destroy base or preparation for occupation.” The Japanese seemed interested also in South Cape on New Guinea and Australian air bases at Cooktown and Townsville. The 6th Destroyer Squadron was believed still en route from Truk to Rabaul, another hint that the offensive had not yet begun. The sighting reports, all from early morning flights on the second, again placed a few merchant ships, mostly small, off New Georgia. As usual those reports only appeared after long delays, especially inexplicable considering one of those Catalinas reached Port Moresby at noon. Early on 3 May Leary advised that the RAAF was evacuating the Tulagi base. By dawn TF-17 was one hundred miles west of TF-11. The Sims and Anderson reappeared on schedule after finding no trace of the sub bombed the previous afternoon. During the day Fletcher worked his way northwest toward next morning’s meeting with Fitch and Crace. By 1530 he had efficiently topped off all seven destroyers from the Neosho, in line with his wise policy of keeping TF-17 “in readiness for action on short notice.” He endeavored to fuel his destroyers from heavy ships or oilers, “whenever they could receive as much as five hundred barrels of fuel.” Mortified by the delay in fueling, Fitch meanwhile ordered the Tippecanoe and four destroyers to fuel through the night. Favorable wind and sea conditions facilitated the process. Escorted by the Worden, the Tippecanoe left for Efate to fuel the convoy after providing TF-11 more than half her oil. At the same time Fitch, having completed fueling twenty-two hours ahead of his revised schedule, proceeded northwest to join TF-17 the next morning.9
Once again the intelligence provided to Fletcher depicted limited enemy activity. The daily Cincpac bulletin opined, but could not affirm, that Fourth Fleet, with the expected cast of Cardiv 5, Crudiv 5, Crudiv 18, and the 5th Air Attack Corps, had started its offensive against southeast New Guinea and “outlying islands.” The Kaga was believed to have left Japan, whereas the Sōryū was ready to sail in a few days. Unknown to Fletcher, Shima seized undefended Tulagi before dawn on 3 May. The Shōhō from Gotō’s MO Main Force launched an unneeded air strike from just south of New Georgia and 120 miles west of Tulagi and by noon started back to Bougainville to fuel. The MO Operation appeared to be going very well, but its most important component was out of position. The MO Striking Force (Shōkaku and Zuikaku) had departed Truk on 1 May. The next day Takagi sent off the ferry flight of nine Zeros and seven carrier attack planes from northeast of Rabaul. Storms compelled their return to the carriers. Because Inoue considered those nine fighters vital to gain air superiority over Moresby, Takagi postponed his advance toward the waters east of Tulagi and maneuvered northward in a wide box to return to the same launching point on the morning of 3 May. Things went no better the second time. Disgusted, Takagi deferred flying off the Zeros to Rabaul until he could do so from the Coral Sea flank of the Solomons. He belatedly resumed his southward advance to the east of the Solomons. This setback would prove extremely detrimental to Japanese fortunes.10
At dusk on 3 May Fletcher anticipated an easy overnight run to next morning’s rendezvous with Fitch and Crace just 130 miles northwest. At 1830, however, the radio resounded with “the kind of report we had been waiting two months to receive.” MacArthur tardily placed five or six ships (the biggest five to eight thousand tons) off southern Santa Isabel Island as of 1630, 2 May—twenty-six hours before—and suggested they might be bound for Tulagi. The source for this intelligence must have been an Australian coast watcher. None of the available aircraft search reports mention this sighting. Fletcher’s report noted another message received about that time of two transports unloading troops into barges at Tulagi. Again coast watchers must have provided it, but the details are not clear. With Tulagi less than four hundred miles north, Fletcher considered what, if anything, to do about it. The absence of TF-11 was “regrettable,” but he was determined if at all possible to hit such “juicy” targets at dawn. TF-17 would need to work eastward into the prevailing wind to conduct air operations, so Arnold advised approaching Tulagi from the southwest. Fletcher did not believe that the Moresby offensive had yet begun in earnest or that hostile carriers prowled nearby and remained convinced the Japanese were still unaware of his presence. Surprising Tulagi seemed a real possibility. The strength at hand (one carrier, four heavy cruisers, and six destroyers) was adequate for that task. Due to careful husbanding of fuel, TF-17 had oil to spare for high-speed steaming. Fletcher detached the Neosho and Russell to keep the appointment with Fitch and Crace and redirect them to a new meeting 180 miles east at dawn on 5 May at familiar Point Corn, 325 miles south of Guadalcanal.11
Bates again faulted Fletcher for not keeping TF-17 and TF-11 together. Unknown to Fletcher, Fitch had finished refueling well ahead of schedule and at 2000 was only sixty miles east of TF-17. Strict radio silence prevented either admiral from advising the other. Bates likewise questioned Fletcher’s decision to set the 5 May rendezvous at Corn, rather than closer to Tulagi, where he could have united the task forces more quickly after the strike. Hindsight gave Bates the advantage of knowing Japanese dispositions and movements, although neither he nor Morison ever actually understood Inoue’s plan. The MO Striking Force was to sweep around the Solomons and westward into the Coral Sea right enough, but to attack Townsville, not lurking U.S. carriers. Bates also surmised that the Shōhō force was to come down from the north and help trap the U.S. carrier force west of the Solomons. He chided Fletcher for “fail[ing] to discern the Japanese plan of encirclement from the east.” Not only did Bates unjustly focus the lens of hindsight, but he also falsely accused Fletcher of being duped by a crafty plan that never existed. Inoue, in fact, had no clue an enemy carrier was even in the area, and certainly he entertained no plan to “encircle” it. And if Fletcher was lax in not anticipating danger from the east, he was not alone. No one at Pearl or Washington thought the Japanese carriers would use the back door either. Fletcher set the 5 May meeting at Point Corn because he did not want to end up too far to the northeast. He saw sighting reports of seventeen transports waiting in Rabaul and assumed, correctly, the Moresby invasion convoy was about to sail. That evening Leary (echoing a Cincpac message Fletcher could not read) advised that two “marus” with escort were expected to reach Deboyne about noon on 5 May, probably to set up a seaplane base. Indeed Fletcher believed, and with good reason, that Tulagi was peripheral to the main event. If he wanted to attack it, he must do so quickly before defending Port Moresby. The most logical employment of the enemy carriers, which Allied air had yet to spot, was to stick close to the invasion convoy once it sailed from Rabaul and guard against the threat from southward in the Coral Sea.12
Fletcher increased speed to twenty-seven knots, hastening toward a dawn launch point southwest of Guadalcanal. Relaxing in the Astoria’s flag quarters, Smith happened to be reading a novel when the sudden vibration of the ship startled him. His spirits lifted at the prospective attack. Likewise cheers resounded on the Yorktown when Buckmaster announced that the air group would strike Tulagi at dawn. TF-17 soon ran into bad weather caused by a moderate cold front moving slowly northward toward the Solomons—a characteristic Coral Sea weather pattern that time of year. Aerologist Strange forecast generally bad weather on 4 May, with an overcast, squalls, and gusty southeast winds from the cold front stalled on the southern fringe of the Solomons. Fletcher expected no fighter opposition at Tulagi and retained the eighteen F4Fs to defend the task force. That evening Arnold devised the air attack plan for all forty SBDs and TBDs. The Yorktown’s first deck load comprised six F4Fs for combat air patrol, thirteen VS-5 SBDs under Commander Burch, and Commander Taylor’s twelve TBDs. Lieutenant Short’s fifteen VB-5 SBDs were poised on the hangar deck. With group leader Pederson stuck on board as FDO, Arnold directed Burch, the senior squadron commander, to coordinate his dive bombers with Taylor’s torpedo strike. Working in darkness, the air department earned praise for efficiently spotting the flight deck and arming each SBD with a 1,000-pound bomb and each TBD with a torpedo. Burch would lead the strike, but not as ranking officer. That night Cdr. Butch Schindler, the Crupac staff gunnery officer, asked Fletcher for permission to ride in the rear seat of an SBD. Qualified in aerial observation, he was eager to study the attacks. Reluctant to risk him, Fletcher finally gave in to his earnest pleas. Fletcher felt dubious about the ability of the aviators to recognize ship types and depended on Schindler’s judgment in that regard. Schindler in fact would fly several combat missions with VS-5.13
Prior to dawn on 4 May, TF-17 closed within one hundred miles of Guadalcanal and 150 miles southwest of Tulagi. The stormy skies matched Strange’s prediction, but the Yorktown aviators were eager to go. By 0702 the last attacker was airborne. Continuing northeast to shorten the range to Tulagi, Fletcher prepared for a possible air counterattack. By 0945 the first wave had returned. Burch, the first to land, raced to the bridge to tell Buckmaster they had “hit the Japs but didn’t do any good.” Like all the SBD pilots, he was extremely frustrated that his windshield and telescopic sight fogged while diving from high altitude into much warmer humid air near the sea. “It’s like putting a white sheet in front of you and you have to bomb from memory.” Burch wanted to go back and finish the job. “All right,” Buckmaster replied, “get in your plane.” Fletcher had decided on a second strike well before Burch suggested it. Radio chatter revealed Tulagi crawled with targets. Biard could not decipher the frantic messages, but at least he recognized no obvious sighting reports of TF-17. Nor did the Yorktown’s radar detect bogeys that might be enemy searchers. The weather proved an effective shield. The Yorktown recovered VS-5 and VT-5 first, struck them below for fuel and ordnance, then landed Short’s VB-5 SBDs (the last at 0954) and quickly serviced them on the flight deck. All planes returned safely, and the pilots hastily filled out Burch’s account. The skies providentially cleared near Guadalcanal, leaving only patchy clouds over Tulagi harbor. Three cruisers were moored together, and nearby were three transports or cargo ships, a large seaplane tender, four gunboats, and numerous small craft. The nested “cruisers” drew the most attention. Burch reported four “sure” bomb hits and one probable. Taylor, splitting his attack between cruisers and transports, claimed hits on all three merchantmen. Two of the “cruisers” then left port, but the third beached itself. Ten minutes later VB-5 bombed a cargo ship and the large seaplane tender but could not confirm any hits.14
The Yorktown Air Group spoiled a beautiful morning for the Tulagi Invasion Force. Shima’s plans proceeded like clockwork as he regrouped his forces. Three small minesweepers had just left for the Port Moresby Invasion Force, the old destroyers Kikuzuki and Yūzuki had moored alongside the fast minelayer Okinoshima (Shima’s flagship) to fuel, and two large transports (Azumasan Maru and Kōei Maru) were unloading base equipment. Two small patrol craft were under way in the harbor. Only the Kikuzuki suffered damage in the first attack. A torpedo in her engine room forced her to go aground before she sank. As flagship light cruiser Kashima arrived at Rabaul, Inoue heard Shima’s alarm. Shocked a U.S. carrier was already in the area, he ordered Takagi’s MO Striking Force and the land-based 5th Air Attack Force to destroy the assailants within two hundred miles of Tulagi. Fearing additional strikes in the meantime, he told Shima to retire north. Fortunately for Fletcher, Takagi was far behind schedule due to the two fruitless attempts to deliver fighters to Rabaul. Instead of 120 miles north of Tulagi, from where he could have provided fighter support and possibly attacked TF-17 that afternoon (if he found it), Takagi languished 350 miles northward and was refueling to boot. Unlike the U.S. Navy, Japanese battleships, carriers, and cruisers did not as yet regularly fuel destroyers, a considerable drawback. Leaving most of his destroyers to finish with the Tōhō Maru, Takagi hastened southeast with the rest of his ships. He searched 250 miles ahead in the waters east of the Solomons. Fletcher, of course, was south not east of Tulagi, so Takagi’s flyers sighted nothing. The three Shortland flying boats got the word too late to search. That was bad for Inoue, because only they might have actually located TF-17.15
Scouting west and north of Guadalcanal for ships fleeing Tulagi, Short’s fourteen VB-5 SBDs, leading the second wave, sank three “gunboats,” actually the No. 1 and No. 2 special duty minesweepers and the Tama Maru, which departed Tulagi prior to the first attack. A heavy cruiser or a large seaplane tender proved much more elusive. She was actually the minelayer Okinoshima, accompanied by the Yūzuki. Her twenty-knot speed and adroit shiphandling snookered not only VB-5 but also Burch’s thirteen VS-5 SBDs (which claimed two nonexistent hits) and Taylor’s eleven TBDs. The SBDs ran afoul of pugnacious float planes over Tulagi harbor. Buckmaster got wind of trouble and asked permission for a fighter sweep to Tulagi. Nothing hostile appeared in the immediate vicinity, so Fletcher released four F4Fs. In a sharp engagement over Tulagi, they claimed three seaplanes and later strafed the Yūzuki.
Mulling the reports, Fletcher considered sending in two cruisers and two destroyers to “mop up damaged ships.” Asked by Fletcher to nominate two cruisers for Tulagi, Smith swiftly offered his flagship Astoria along with the Chester. Although Tulagi was only seventy-five air miles distant, the big island of Guadalcanal blocked the way, meaning no chance to reach there before darkness. He planned to approach from the west, shell Tulagi harbor at sunrise on 5 May, then retire eastward at high speed. Fletcher was not yet ready to unleash Smith but waited to see how the situation developed. At 1330 the Yorktown started landing SBDs from the second wave, which claimed sinking two destroyers or small gunboats and possibly a cargo ship. Seeking firsthand information, Fletcher spoke with one of the pilots, Lt. (jg) Floyd E. Moan of VB-5, who described bombing the gunboats and the big seaplane tender. Moan suggested a follow-up attack. Fletcher appeared somewhat skeptical, but by the time Moan got down to the VB-5 ready room, he learned there was to be a third wave. Burch also desired another go. Twenty-one SBDs left at 1400 to deliver the coup de grâce. Likely it was then Fletcher gave up the idea of sending Smith to bombard Tulagi. That was a good thing. Every ship less the stricken Kikuzuki and a small auxiliary soon decamped, and Takagi’s carriers would have made short work of Smith’s cruisers on 5 May. On their third foray Burch’s dozen SBDs incorrectly claimed one hit on a cargo ship, although just the blast of near misses harmed the crew of the Azumasan Maru and sailors of the Tulagi garrison. Short’s nine SBDs took another unsuccessful shot at the Okinoshima and Yūzuki. With their departure, the Tulagi air battle ended.16
The Yorktown recovered all but one Devastator missing from the second wave. Two fighters set down on the beach at Cape Henslow on Guadalcanal less than fifty miles north of TF-17. Fletcher acted swiftly to retrieve the downed aviators. He detached the Perkins to seek the TBD’s crew and the Hammann to Cape Henslow. The two destroyers were to rejoin TF-17 at sunrise at Corn 325 miles south of Guadalcanal. The Perkins did not find the TBD crew, who luckily fell in with loyal Solomon islanders, but the Hammann’s motor whaleboat effected a courageous rescue of the two VF-42 pilots. Morison chose this occasion to give Fletcher “great credit for initiating efforts to rescue aviators downed in combat,” but he had already done so on the 1 February raid.17
By sunset with all aircraft retrieved, it was high time to leave. Fletcher turned southeast at twenty-three knots toward Point Corn. Morale was high. One SBD pilot wrote in his diary that “the men were proud of the way Admiral Fletcher and Captain Buckmaster ran the affair.” After reviewing reports and aerial photographs, Yorktown assessed two destroyers, a cargo ship, and four small gunboats (one thousand to fifteen hundred tons) sunk and a light cruiser (possibly Jintsu-class) forced aground. One nine-thousand-ton seaplane tender (or perhaps heavy cruiser) sustained heavy damage; a light cruiser and a large cargo ship (eight to ten thousand tons) were also damaged to a lesser degree. Yorktown aircraft shot down five small seaplanes and strafed numerous small boats. Australian coast watchers later stated that nine ships went down off Tulagi, including three cruisers (one beached) and three transports. Actual Japanese losses were much less. Only the destroyer Kikuzuki (beached at Tulagi) and three small minesweepers (No. 1, No. 2, and Tama Maru) sank, whereas the minelayer Okinoshima and the destroyer Yūzuki sustained minor damage. Four observation seaplanes were lost.18
Even according to the Yorktown’s exaggerated claims, the lack of air opposition and heavy antiaircraft certainly rendered the results, to use Nimitz’s own word, “disappointing.” Only five of twenty-two torpedoes and eleven of seventy-six 1,000-pound bombs were adjudged hits. Fogging of telescopic sights hindered bombing, but Schindler, who flew three missions with VS-5, candidly described overall accuracy as “poor.” Peacetime standards of marksmanship, he thought, would have tripled enemy losses. However, Tulagi served as a valuable warm-up for aviators gone stale during a long and quiet, but nonetheless nerve-wracking, cruise. It marked the first time in a year that VT-5 had actually dropped torpedoes. Nimitz praised the “very creditable willingness” of the Yorktown aviators to “to keep after their enemy objective until it was destroyed,” but diagnosed their obvious need for regular target practice. He enjoined the task force commanders to increase training at sea. Yet the risky strategic situation and the Yorktown’s lengthy cruise offered few chances for the kind of elaborate air group practices held off Pearl.19
The Cincpac daily intelligence that Fletcher received on the afternoon of 4 May added the Kaga, some fast battlewagons, and two more heavy cruisers to the ominous list of reinforcements possibly bound for Truk. The 2nd Carrier Division (Sōryū and Hiryū) might also come south in mid May. All this conformed to Nimitz’s expectation of a widespread offensive in the South Pacific. During 4 May Comsowespac air sighting reports were remarkably sparse. Several search areas were not even covered. That evening Fletcher belatedly learned that at 1135 a U.S. B-25 bomber in Area C spotted a carrier (“probably Kaga class”) and two battleships or heavy cruisers off the west coast of Bougainville, approximately two hundred miles southeast of Rabaul and more than five hundred miles northwest of TF-17. That sighting did not affect Fletcher’s immediate plans. Sowespac still placed twenty to twenty-five transports at Rabaul, obviously the Moresby invasion convoy. Fletcher held off until that convoy and its covering force had moved south. In fact Sowespac finally discovered the light carrier Shōhō that for two days crisscrossed the northern Solomons with Gotō’s MO Main Force. While returning to Bougainville to fuel, Gotō learned that Shima was in trouble and turned back.20
On 4 May Fitch was surprised to find only an oiler and a destroyer waiting at the rendezvous. The Neosho’s Capt. John S. Phillips duly advised that Fletcher had gone north to Tulagi and would be at Point Corn, 180 miles east, at dawn on the fifth. Crace arrived on schedule with the Australia and Hobart. Fitch led the combined force southeast. Crace wondered why that direction instead of east or even northeast toward Tulagi. Later so did Bates, who again scolded both U.S. commanders for keeping TF-17 and TF-11 separated. When Crace joined, TF-11 was nearly three hundred miles southwest of Fletcher, “unable,” according to Bates, “to support TF 17 should a sudden need arise.” Bates blamed Fletcher. Yet Fitch steered southeast, “Even though such a course would probably place him in a poor supporting position.” Bates thought if Fitch had only turned northeast, he could have been 250 miles nearer to TF-17 by sundown, but still out of enemy search range. Instead he maintained his southeasterly course until sundown, then turned eastward toward dawn rendezvous. Fitch later told Bates that he preferred the southeast because he did not know how Fletcher fared at Tulagi and judged it best to keep TF-11 well south of the rendezvous. Six Lexington SBDs searched two hundred miles northwest to ensure TF-11 itself was not being stalked. Fitch had access to the same intelligence sources as Fletcher. Had it seemed Fletcher was in peril, he would not have hesitated to rush to Tulagi. Instead like Fletcher, he anticipated little real danger to the north or northeast. The main threat to Port Moresby must come via the Louisiades.21
At dawn on 5 May Fitch’s TF-11 and Crace’s TF-44 made the scheduled meeting at Point Corn. The Perkins showed up alone after her fruitless search off Guadalcanal. Crace, for one, thought Fletcher was off chasing the enemy carrier sighted near Bougainville the previous day, but actually TF-17 loomed over the eastern horizon. On 3 May he received another special Cincpac message transmitted in a cipher he did not hold. After dawn on the fifth he broke radio silence—believing he could finally do so safely, for he had certainly revealed his presence if not his exact position—to inform Pearl of his inability to decipher certain systems, a good thing because another unreadable message came through in the interval.22
TF-17 emerged from thick overcast only to detect a bogey thirty miles west. Four Wildcats ambushed a Type 97 flying boat low over the water. Both TF-17 and TF-11 noticed the smoke that announced its fiery end, and forty minutes later they sighted each other as well. The Kawanishi went down fifteen miles from TF-11 and twenty-seven miles from TF-17. The victim was one of three flying boats searching out of Shortland; six others flew from Tulagi. It never established radio contact with base, and only later did its absence lead to the conclusion a U.S. carrier prowled south of the Solomons. Soon afterward a Yorktown search SBD discovered a sub 150 miles northwest and moving toward TF-17. Fletcher wondered whether it responded to a contact report from the flying boat. Three TBDs hunted the sub without success. Had Fletcher known it was only one of four I-boats of the Eastern Detachment, he would have been more perturbed. They cut through the Solomons bound for a deployment line farther south in the Coral Sea to catch Allied fleet units expected to rush north from Australia. The center of that line was about two hundred miles southwest of Fletcher’s current position. On the fifth Inoue instructed the Eastern Detachment to shift 150 miles northeast, but even that fell short of TF-17.23
Waiting for the enemy offensive to develop, Fletcher radioed Cincpac a summary of the Tulagi strike and noted the three task forces had now joined. He refueled TF-17 to make up for the past day’s high-speed run. Fueling required steaming slowly southeast into the wind during the day, but Fletcher planned to run northwest overnight to regain position. The Neosho fueled the Astoria and Yorktown, while six destroyers drew much-needed oil from the Chester and Portland. Cincpac promptly responded to Fletcher’s report. “Congratulations and well done to you and your force. Hope you can exploit your success with your augmented force.” The afternoon was quieter than the morning. At 1341 Leary provided a rare timely sighting report of a seaplane tender discovered at 1313 southwest of New Georgia, though the Sowespac morning search missed the massive Port Moresby invasion convoy that finally departed Rabaul the previous evening. MacArthur did advise Fletcher that he had ordered his aircraft to relay contact reports in plain language on certain frequencies, but that did not solve the problem of passing on information expeditiously. The daily Cincpac bulletin warned that the Hiryū might depart Japan soon for Truk and the Sōryū might already be on her way there, according to “unconfirmed information.”24
Carrier operations, Coral Sea, 5–6 May 1942
Fitch flew over to the Yorktown that afternoon for a brief reunion with Fletcher. Naturally mistaking the stocky man in the back seat of the Lex SBD for a radioman, one of the deck crew teased: “Well, chief, you guys kinda missed out on some fun yesterday.” Fitch laughed, revealed his two stars, and replied: “Yes, son, I guess we did.” Frustrated with shore duty, he was more than ready to fight. Fitch’s own brand of dynamic leadership complemented the Lex’s fiery Capt. Ted Sherman, himself no shrinking violet. He named Sherman ex officio TF-11 chief of staff for the present cruise and requested the assignment be permanent. Bunav, however, offered its regrets. Sherman was due for promotion and another billet. Like their boss nearly all of Fitch’s staff were naval aviators. Confident of their expertise, aviation and otherwise, they were eager to run a carrier battle if Fletcher let them.25
Fitch received his copy of TF-17 Op-Ord 2–42 and the job of air task group commander. He also informed Fletcher that on 2 May his own radio intelligence team, led by Lt. Cdr. Ranson Fullinwider, indeed detected a contact report from the bombed sub (I-21), along with appropriate reverberations within the enemy theater command. Fletcher relayed these findings to an astonished Biard, who heard nothing from that sub and had advised Fletcher to that effect. Fullinwider was a Japanese language specialist whom Fletcher knew and respected. Of the 2 May incident, Fletcher declared in his action report that the “proximity of the submarine to our surface forces and radio interceptions pointed to the probability of our position having been reported to the enemy,” and that “in over two months of operating in the Coral Sea, this was the first definite indication of our presence having become known to the enemy.” Biard, though, was completely correct. There was no report of the incident, although it is possible the I-21 sent some other routine message. The relationship between Fletcher and Biard was already troubled, and according to Biard this only magnified Fletcher’s growing distrust.26
By the afternoon of 5 May as Fletcher fueled TF-17, it seemed he eluded any immediate retribution from the Tulagi raid. His dominating concern remained the Japanese carriers. More information from the Sowespac air search reached him that afternoon. For the second time in two days MacArthur’s flyers located the Kaga or a “Kaga” class carrier near Bougainville. The search summary on 5 May placed a Kaga-class carrier and a battleship (“Yamashiro or Haruna type”) 125 miles southwest of Bougainville. War Plans at Pearl assumed the presence of warships of that type in the Solomons, and that was now confirmed. Layton, the Cincpac intelligence officer, postulated the carrier sighted on 4 May off the west coast of Bougainville could be one of two Cardiv 5 flattops in the Moresby Striking Force. Comsowespac only reinforced that impression by amending the original report, calling the carrier sighted the previous day “Ihokauu” class, a garble for Shōkaku. In truth on both days aircraft sighted only the Shōhō. On the morning of 5 May the Shōhō and destroyer Sazanami continued north to cover the invasion convoy, while the 6th Cruiser Division ducked into Shortland to refuel. The 5th Carrier Division was nowhere nearby. On 4 May, as noted above, Shima’s cry for help surprised Takagi’s MO Striking Force busy fueling well north of the central Solomons. Takagi raced southeast with most of his ships expecting to find the enemy carrier east of Tulagi but never sighted it. Directed by Inoue to resume the plan of entering the Coral Sea from south of the Solomons, Takagi rescheduled the rendezvous with his oiler to dawn on 6 May in the Coral Sea 180 miles west of Tulagi. On the afternoon of the fifth he circled west around San Cristóbal, too late in the day for the Tangier’s flying boats to sight him. Nor did his own search planes peering three hundred miles ahead spot any Allied ships. That night he passed west of Guadalcanal headed for his oiler.27
Thus on the afternoon of 5 May TF-17 and MO Striking Force unwittingly drew much closer to each other than either realized. Fletcher based his perception of the situation largely on a series of Cincpac special intelligence messages. The first, sent on 3 May in the wrong cipher and rebroadcast the afternoon of 5 May, noted that “good information indicates following Orange Operation Order for Southwest Pacific offensive now starting.” The Japanese would operate “to utmost to restrict Blue fleet movements by attacks on outlying units and various areas along north coast of Australia.” They might even occupy Allied air bases “to wipe out shore based aircraft.” What Fletcher made of that particular message is not precisely known, except that he could be expected to believe that carrier strikes against air bases was a high Japanese priority. The Cincpac analysts judged its meaning: “To keep our forces from interfering, the Japanese plan to raid such places as Cooktown, Townsville, and Horn Island, and may raid as far east as [Efate], Nouméa, Fiji, and Samoa.” That, of course, reflected Nimitz’s prediction of a wider Japanese South Pacific offensive and was the main reason he committed Halsey’s two carriers along with Fletcher and Fitch. The seizure of Tulagi only heightened that impression. Hypo had extracted this decrypt from a Japanese message that proved “difficult to read due to lack of code groups.” In truth Rochefort’s assessment of the meaning of that message was exactly the opposite of its true intent. It was actually the 29 April signal in which Yamamoto called off carrier raids on Australia and directed Takagi to be prepared to hunt down U.S. carriers.28
The code breakers provided further erroneous interpretations of partially deciphered Japanese messages. Two additional Cincpac special messages that Fletcher received on the afternoon of 5 May dealt specifically with the “Moresby Striking Force” of the 5th Cruiser and 5th Carrier divisions. According to the first message, “reliable indications on 3 May” showed the Moresby Striking Force had orders to strike Port Moresby on X-3 or X-2 Days apparently from the southeast. Uncertain of X-Day, Pearl correctly guessed 10 May. That meant carrier raids against Moresby could occur on the seventh or eighth. Cincpac asserted to Fletcher, “above attacks [are] to be carried out until successful completion by Orange.” For the most part the U.S. cryptanalysts at Pearl and Melbourne recovered the sense of that particular Japanese message but never realized that it was not even relevant. On 3 May Inoue attempted to revive the controversial MO Striking Force raids on troublesome Allied air bases, in this case Port Moresby. He radioed Takagi it had been “agreed informally” that, depending on the enemy air strength, he was to attack Port Moresby on 7 or 8 May from the southeast. Takagi, though, chose to ignore the “informal agreement,” as was his right. Unlike the Allied code breakers, he well understood the terms of Yamamoto’s 29 April order that canceled carrier raids on shore bases until the U.S. carriers were destroyed. Fletcher’s Tulagi strike ended any doubt on that score. Consequently Takagi had no intention of attacking a shore target. However, Cincpac’s categorical rendering of the “order” to attack Port Moresby appeared to offer the Japanese carrier commander little discretion. Thus Fletcher learned that the Moresby Striking Force was to bomb Port Moresby either on 7 or 8 May. Geography made the southeast a virtual certainty due to the remote reef-strewn waters northeast of Papua. For the Moresby Striking Force, the voyage westward from the Bougainville Strait via the Louisiades to within strike range (two hundred miles) southeast of Port Moresby would total six hundred miles—thirty hours at a steady twenty knots. To get there on time the Japanese carriers must get cracking through the Louisiades.29
The last Cincpac special message vouchsafed Fletcher on 5 May appeared to describe how his Tulagi strike might affect future movements of the enemy carriers. According to Cincpac, the commander of the Moresby Striking Force “indicates 4 May” that if he determines the U.S. carrier force is in the Coral Sea (“questionable location”), he would “proceed north northeast of Bougainville, thence southward.” Then at 0630 local time on 5 May, after reaching someplace not known to Cincpac, he would continue in accordance with “further orders.” If no further orders were received, the Moresby Striking Force was to make for Tulagi. If it proved necessary to search to the south and some other sector, the 5th Carrier Division was to send its bombers to Tulagi at daybreak. What did Fletcher and the others deduce from this message? At this long remove without specific documentation one can only speculate. One possible rendering is that if the Japanese discovered the U.S. carrier force south of Tulagi, as opposed to east of the Solomons, the Moresby Striking Force was to withdraw northeast of Bougainville, then turn back south. That could keep its carriers clear of Allied land-based air search until the proper moment to resume the southward advance. As for the Moresby Striking Force proceeding to Tulagi if it received no “further orders,” one possible explanation is that should the U.S. carrier be found east of the Solomons, the Japanese carrier force was to deploy toward Tulagi to guard the flank. As mentioned before, Fletcher knew of sightings of Japanese carriers in the Bougainville area. He had every reason to think from the supposed sighting of TF-17 and TF-11 on the morning of 5 May that the Japanese well knew he was on the Coral Sea side of Tulagi. That could induce the Moresby Striking Force to mark time a short while northeast beyond Bougainville before advancing southwest toward the Louisiades. Unfortunately U.S. naval intelligence (and thus Nimitz and Fletcher) did not understand this last order never even applied to Takagi’s MO Striking Force but instead concerned Gotō’s MO Main Force (6th Cruiser Division, Shōhō, and destroyer Sazanami), the separate existence of which Hypo and Belconnen were totally unaware.30
Despite the seemingly comprehensive intelligence, the Allies actually had no clue where the 5th Carrier Division was or where it had been. Fletcher did receive a tangible indication that evening of the apparently reliability of this intelligence. For several days Nimitz and Leary had forecast the arrival at noon on 5 May of two merchant ships (“Marus”) at Deboyne near the northern approach to Jomard Passage. That evening Fletcher learned Allied aircraft indeed located “two merchant vessels about 2,000 tons with other units” about one hundred miles northeast of Deboyne and obviously bound there. Although the Japanese ran a bit late, the contact report appeared to give dramatic proof of the astonishing accuracy of Cincpac’s prognostications.31
On the evening of 5 May Fletcher instructed Fitch and Crace to keep within ten miles of TF-17 while steaming overnight to the northwest. By dawn on 6 May they were about 150 miles west of Point Corn. Fletcher swung back southeast. The Lexington loosed a dozen SBDs for a routine search northward to 275 miles, nearly to the southern Solomons. Fletcher executed Op-Ord 2–42 that merged the three task forces into TF-17, with Kinkaid’s Attack Group (TG-17.2), Crace’s Support Group (TG-17.3), Fitch’s Air Task Group (TG-17.5), Phillips’s Fueling Group (TG-17.6) and Cdr. George H. DeBaun’s Search Group (TG-17.9) with the Tangier. Fletcher predicated his plans for the sixth on his belief, based on the intelligence provided him, that the battle would not break out until at least the next day. He welcomed the lull to replace the oil his two cruisers furnished the TF-17 destroyers and, if time and the enemy permitted, to fuel Crace’s cruisers as well. That again meant heading slowly southeast into the prevailing wind. The weather, fine on 5 May, decayed as clouds closed in and the wind picked up. Rough seas hindered fueling all day. The SBDs that searched in the clear weather up near the Solomons filed no contact reports. It was Fletcher’s ill fortune they did not locate the MO Striking Force around 0815 as it fueled south of New Georgia. By good luck Takagi loitered just beyond the limit of visibility of the nearest Lexington SBD. A sighting would have precipitated a carrier battle, with Fletcher hoping to surprise the enemy as he had at Tulagi. Unfortunately the Sowespac air search concentrated on the Bougainville-Louisiades area, so Fletcher gained no help from that quarter.32
The Japanese shore-based search had better luck. About 1015 TF-17 radars discovered an intruder lurking in nearby clouds. Yorktown F4Fs failed to trap their quarry, although at least one ship glimpsed it. As a precaution should an enemy strike follow on its heels, Fletcher drew all of TF-17, except the Fueling Group, into the tight “Victor” defensive formation. Nothing showed up, but the bogey did not go away either. It was a near certainty that a snooper was shadowing TF-17, although neither Fullinwider nor Biard intercepted any sighting reports. By noon flying conditions became so bad that the SOC float planes flying antisubmarine patrol over the Fueling Group were soon recalled. The Chester finally finished fueling about then, and the Portland took her place alongside the oiler. At 1252 radar warned of another contact twenty-eight miles south.33
That afternoon as Fletcher warily fueled, he learned from Melbourne and Pearl that the Japanese were indeed tracking something. Cincpac’s message quoted a deciphered message warning of “enemy one carrier one BB two cruisers five destroyers and four unidentified vessels course 190 speed 20 at 0830 Item (minus 9) [1030, Z-11].” Unfortunately neither Belconnen nor Hypo could read the position, but Fletcher correctly understood that with Tulagi only four hundred miles north the message must refer to TF-17. At 1030 a Japanese flying boat had revealed a U.S. carrier force 420 miles southwest of Tulagi, but erred in placing it fifty miles south of Fletcher’s actual position. It also wrongly reported course and speed as south (190 degrees) at twenty knots, whereas TF-17 slowly fueled on a southeasterly course. Fletcher soon received more evidence of carriers prowling off Bougainville. At 1207 Comsowespac placed a carrier, two cruisers, and two destroyers only thirty miles off the southwest coast of Bougainville and headed south. Forty minutes later MacArthur advised that B-17s sought a carrier in the “Misima” (eastern Louisiades) and “Bougainville areas.” Fletcher took the precaution that afternoon of having Lexington SBDs again search 275 miles north and northwest.34
The TF-17 afternoon search found no enemy ships, but that did not rule out any hiding under the clouds. At 1050 Takagi was again embarrassed to be caught fueling when he received word of the U.S. carrier force discovered 350 miles south and headed almost directly away. At noon he released Hara’s 5th Carrier Division and two fueled destroyers to chase the carrier. He soon followed with his two cruisers, leaving the rest of the destroyers behind with the Tōhō Maru to continue fueling. Racing southward into increasingly hostile weather, Hara perceived dim prospects for a long-range air strike that afternoon and elected not to launch his own air search to pinpoint the enemy. He was unsure given the poor visibility whether it would be worthwhile, and he most especially did not want to alert the Americans to his presence.35
Actually TF-17 was less than three hundred miles away. Because the Lexington SBDs missed both the 5th Carrier Division and 5th Cruiser Division in the murk, Fletcher had no idea the scattered elements of MO Striking Force bore down on him from the north. By 1500 the Neosho finished fueling the Portland. To give Crace’s cruisers oil would require continuing southeast during the evening. All day, though, Fletcher copied Sowespac sighting reports of activity north of the Louisiades. Just to draw within striking distance (170 miles) of Deboyne by the next dawn already demanded an overnight run of nearly three hundred miles at twenty-plus knots. Therefore Fletcher brought TF-17 around to the northwest. To Crace, disappointed not to fuel, it looked “like business.” Fletcher intended by dawn on 7 May to be 170 miles southeast of Deboyne to strike forces reportedly concentrating off Misima just east of Deboyne in the Louisiade Archipelago. At 1730 he stepped up to twenty-one knots. The Neosho and Sims peeled off to the south in order to shuttle every other day between Points Corn and Rye. Fletcher committed a major error by failing to modify his operations order after the fall of Tulagi. Corn was only 360 miles south and Rye 440 miles southwest of Tulagi. Both were easily within range of flying boats he knew were now based there. It would have been much wiser to park the Fueling Group beyond the range of any Japanese shore-based aircraft.36
Thus by the evening of 6 May Fletcher rated TF-17 ready for battle. His attention to fueling had and would continue to pay huge dividends.37 The tactical situation he faced was not for the faint of heart. He had no recourse to subterfuge or the luxury of the indirect approach. The enemy air searches blanketing the critical area pinpointed TF-17 on the fifth and the sixth. Once the Invasion Force showed its hand, he might have to bull his way through three carriers, perhaps more, not to mention land-based air. However, he reckoned on one special advantage. Air search reports had located numerous Japanese ships, “practically every type,” within the triangle formed by Papua, New Britain, and the Solomon Islands. They appeared “scattered” and to lack a “common direction of movement.” Yet the radio intelligence furnished by Nimitz and Leary seemed to reveal where they were going and when they might get there. As early as 7 May the invasion convoy should traverse the center of the Louisiades at Jomard Passage. Aircraft sightings on 6 May of several groups of transports and auxiliaries northeast of Deboyne and farther north off Woodlark further confirmed the Jomard shortcut. That fitted Fletcher’s estimate that the final advance on Moresby would commence on the seventh or the eighth. The numerous Sowespac sightings in the Louisiades on 5–6 May included the 18th Cruiser Division (old light cruisers Tenryū and Tatsuta), the seaplane tender Kamikawa Maru, and several auxiliaries from Marumo’s Support Force converging on Deboyne to build the seaplane base. As yet the Allies had not seen Port Moresby Invasion Force, except for a few ships coming up from Tulagi.38
The convoy was one thing, the carriers another. Fletcher never stated exactly what he made of the enemy carrier deployment other than he anticipated action off the Louisiades on 7 or 8 May. Fitch offered the clearest estimate: “Naval units, including one carrier and the invasion fleet were reported converging on Deboyne Island. It was expected that the enemy would use Jomard Entrance to enter the Coral Sea. Two additional CV’s [aircraft carriers], probably Carrier Division Five, were reported in the vicinity of Bougainville Island, on May 6 . . . Carrier Division Five [was expected to] run southward from the vicinity of Bougainville and might be within striking distance on the morning of May 7.” Fitch had access to the same special Cincpac intelligence as Fletcher, but Crace only received Comsowespacfor messages. Crace’s diary on 6 May placed “about 10 transports off Deboyne Islands & one or possibly two carriers with cruiser or battleship escort to the west of Bougainville Island steering south & presumably acting in support.” Thus two senior admirals in TF-17 located Japanese forces, including their carriers, far to the northwest, perhaps four to five hundred miles away.39
That assessment was deeply flawed. Of the three carriers, only the Shōhō (Ryūkaku to the Allies) was or had ever been anywhere near Bougainville. After his cruisers fueled at Shortland, Gotō reassembled MO Main Force on the morning of 6 May west of Bougainville, then steamed southwest to meet the Port Moresby Invasion Force approaching the northern Louisiades. That morning three B-17s bombed a cruiser and then spotted a carrier nearby. At 1800 on 6 May, close to sundown, the Shōhō was about 425 miles northwest of TF-17, but Hara’s 5th Carrier Division was already only 150 miles north. By turning northwest Fletcher unwittingly increased the rate of closure between himself and Hara. Fortunately for him, Hara still thought his opponent was much farther south. Even so Hara persisted in his pursuit until 1930, some seventy-five minutes after sunset, when he was only sixty or seventy miles north of TF-17, but never knew it. Takagi had long given up and reversed course to reassemble his force. They both decided it would be wiser to abandon their futile pursuit. The U.S. carrier force must be gunning for the invasion convoy and MO Main Force converging on the Louisiades. That seemed to present an excellent opportunity for MO Striking Force to surprise the American carrier from the rear.40
Had the air searches of either side been a little sharper, the first carrier clash in the Coral Sea should have occurred on 6 May rather than the seventh. Even so, the results of the day had armed Takagi and Hara with a general notion of where their opponent was and what he might do the next day. That augured well for a devastating ambush. Fletcher enjoyed no similar benefit. He wrongly perceived that his most dangerous adversaries mustered far to the northwest. In the broadest sense the comint furnished to Fletcher was reliable and potentially decisive at the strategic level. Yet for immediate tactical purposes, it created an incomplete and dangerously false portrait. Allied radio intelligence depended on partially decrypted, therefore incomplete, Japanese naval messages. There was the great danger that code breakers and intelligence officers would misinterpret their contents or fail to understand the context in which they were sent. Their errors unwittingly placed TF-17 in grave danger on 6 May and set up the strong possibility of Fletcher being surprised, perhaps fatally, on the morning of 7 May.41