The CXAM Goes to War - Chapter 6 of Radar and the Fighter Directors
By David L. Boslaugh, Capt USN, Retired
On the Day of Infamy, three of the eight battleships at Pearl Harbor had a CXAM radar installed. Of the twenty CXAMs ordered, nineteen had been installed in heavy combatants throughout the U.S. Navy, and twenty-seven out of the forty-eight SA and SC-1 radars had been installed in smaller combatants. Although the battleship California was damaged so badly that she would not be ready again for active duty until early 1944, her CXAM suffered no damage and was removed from the ship on 8 December. It was set up on a hill overlooking Honolulu and served to extend the Army’s radar coverage. The other two ships at Pearl Harbor having the CXAM were USS West Virginia (badly damaged), and USS Pennsylvania (moderate damage; sailed on 20 December). [1, p.181] [36, II: pp.14-15, V: pp.251-252, VIII: pp.223-224] [61, p.240)
Three days after the Pearl Harbor attack, Japanese torpedo planes attacked the British battleship HMS Prince of Wales and the battlecruiser HMS Repulse while steaming in the Gulf of Siam within fighter range of Singapore. They had no air cover of their own. Both ships were equipped with search radars, and Repulse’s search radar operators detected a Japanese reconnaissance airplane at about 1040 on the morning of 10 December. The reconnaissance plane was in the lead of a flight of ninety-four twin-engine Mitsubishi G3M and G4M bombers. They had risen from airfields around Saigon, and were equipped with bombs and torpedoes. The bombers were vectored in to the task force, and first concentrated on the larger Prince of Wales. In spite of putting up heavy AA fire, Prince of Wales succumbed to six torpedoes. Repulse came under attack next, and at that time, the battlecruisr’s commanding officer finally broke radio silence to request fighter support from Singapore. Repulse then absorbed five torpedoes, and early in the afternoon both ships went under, Prince of Wales going last. A few minutes before the battleship went under, Brewster Buffalo fighters arrived from Singapore but there was nothing they could do. Repulse’s radar officer survived the encounter, and later wrote in his report that if the officer in tactical command had requested fighter protection at the time they first detected the reconnaissance airplane, the two ships would have survived. It was a telling example of the vulnerability of even heavily armored and armed capital ships to air attack. [29, pp.122-26] [15, pp.35-38]
On 20 February 1942 the carrier Lexington was operating 400 miles east of New Britain Island preparing to attack the Japanese base at Rabaul the next day. Lexington’s CXAM radar’s flying bedspring antenna was mounted high up on the leading edge of the smokestack, and the radar set itself was in a small shack below the antenna. The radar set was six feet high, five feet wide, and two feet deep, and the radar compartment was just big enough to hold the set and an operator’s chair. The only communication from the radar shack to the outside world was a phone line leading down to Air Plot on the aft side of the bridge. Only eleven people in ship’s company knew what the flying bedspring was for, and that number included the skipper, CAPT Frederick C. Sherman and his executive officer. The remainder were fighter director officer Lieutenant Frank F. “Red” Gill, his two ensign assistant FDOs, two chief petty officer operators, and four maintenance technicians. Before taking on fighter direction duties LT Gill had been a naval aviator for ten years. The two ensigns and the two chiefs took turns either manning the set in the radar shack or plotting the reported ranges and bearings on the dead reckoning tracer (DRT) in Air Plot. This narrow compartment, holding the desk-sized DRT, radio equipment, and a small plotting team, had been a tight squeeze even before the advent of radar. The plotters task, before radar, was to keep track of all of the carrier’s airborne airplanes. This they did from pilot’s reports and dead reckoning the tracks of the aircraft. If the planes were over land, the pilots would tell Air Plot where they were with respect to landmarks. After the CXAM was installed, the plotters added radar ranges and bearings to the plot. Those few in the know about radar were not allowed to tell any one else what the rotating bedspring was for, or what their job was. It is no wonder that bridge officers found it hard to believe their reports of aircraft detections at more than 100 miles. [33, p.89][6, p.37-38]
It was about 1045 on the morning of 20 February 1942. One of the assistant fighter directors was in Lexington’s radar shack slowly turning the CXAM’s antenna train crank and gazing at the A-scope. A blip shot up out of the grass and he stopped the antenna. Then he cranked the antenna back to where the blip was at its highest. He reached up and punched the ‘trip’ button on the console to verify whether the blip was in the fifty mile range scale, or was a second-time-around range that should have fifty miles added to displayed range. He read off range and then read antenna bearing from the train indicator. Then pressing the button on his sound powered phone headset, he said “plot—radar”, and came the response, “plot aye”. “Air target bearing two five one degrees, range fifty-two miles.” The plotters noted there were no friendly aircraft at that location, classified the target “probable bandit,” and called the sighting to the bridge who told the fighter director in air plot to send some of the airborne CAP to investigate. [6, p.38]
LCDR John S. “Jimmy” Thatch was on combat air patrol, in charge of six Wildcat fighters orbiting Lexington’s task group. They were in strict radio silence and the air-waves had been totally quiet. Thatch later wrote,” I almost jumped out of my seat when the loud voice of the Lexington’s fighter director [LT Red Gill] came on giving me a vector to course 240 degrees and saying there was apparently a snooper about thirty-five miles away.” Thatch and his wingman, Ensign Edward Sellstrom, started out on the vector, advancing their throttles from maximum endurance power to about two-thirds power. Thatch radioed back to LT Gill, at 1056 that they would soon be in heavy clouds, and a few minutes later radioed that they were on station and about to enter a rain cloud. He asked Gill if the target was ahead of him in the cloud. Given an affirmative reply, he and Sellstrom switched to instrument flight and bored into the fog. Gill had put him right on top of his quarry. Below him was a very large four-engine Kawanishi H8K flying boat patrol bomber. Thatch sent his tallyho report at 1100, and both fighters charged their guns, but lost sight of their target as their momentum carried them out of the other side of the cloud. At the same time Gill advised them that the big boat had disappeared from the CXAM’s scope, meaning it had dived. Minutes later they saw the craft break out of the cloud at about 1,000-foot altitude, at high power and heading for home. Thatch intended to approach from the port quarter to stay out of the cone of fire of the boat’s suspected twenty mm cannon tail gun, but Sellstrom misunderstood his instructions and crossed to the other side exposing himself to the explosive projectiles, but by violent maneuvering took no hits. They closed from both sides firing at the engines, with the final result that streaming gasoline set the whole wing afire. They saw eight bombs fall free and soon after, the big boat hit the water in a thunderclap. Thatch noted that it made a cloud of smoke that could be seen all the way back to the Lexington. They were also pretty sure that the snooper’s crew had been able to radio a sighting report back to Rabaul before impact. [44, p.52] [33, pp.91-92]
About the time Thatch and Sellstrom were landing on Lexington, Gill’s radar operators reported another suspected snooper to the north. This time LT Gill selected the two-plane “Orange Section” of LTJG Onia B. Stanley and ENS Leon W. Haynes to check out the intruder. To illustrate the terse and compact fighter direction vocabulary, the entire vector order was as follows, “Orange Section from Romeo—Vector three four three—Buster—Angels six.” Orange Section has already been explained, and “Romeo” was Gill’s call sign. The rest meant take a heading of magnetic 343 degrees at maximum sustained power at an altitude of 6,000 feet. The only higher power setting would be “Gate” meaning full military power, which could only be used for a short time without engine damage. They proceeded out about twenty miles when ENS Haynes spotted another “BIG” silver four-engine flying boat, which apparently also spotted the two fighters because it immediately jettisoned its bombs to gain flying speed. LTJG Stanley tallyhoed back to Gill, and the two fighters set up their attack while sparkling flashes and tracers coming from their intended target showed that gunners with twenty mm cannons on the big boat were seeking their range. On their first pass Stanley’s four fifty caliber machine guns failed to fire, but Haynes had the satisfaction of putting the tail cannon out of action. In the meantime Stanley realized he had not flipped on the gun master arm switch because it was located in a different position than in the Wildcat he usually flew. In the next pass his guns worked and he could see his tracers penetrating through the wing and in to the cockpit. The engine area of the wing was soon a sheet of fire and gray smoke, and the huge craft went into a steep dive that ended up as a circle of flame on the ocean’s surface. The task force could also see the smoke of the second burning Kawanishi. [33, pp.93-94]
In confirmation that at least one of the snoopers had managed to send a sighting report, the assistant fighter director at the CXAM console, at 1611 that afternoon picked up a large formation at seventy-six miles headed toward Lexington. Gill vectored six Wildcats of the airborne CAP under LT Noel A. M. Gaylor out to intercept these land based craft of the Japanese 25th Air Flotilla, and by 1630 the bogey was in to twenty-four miles. A few minutes later, Gaylor radioed back his tallyho report: nine twin-engine bombers, and no fighter escort because the distance of Task Force 11 from Rabaul was beyond the combat radius of Japanese fighters. By 1643, Gaylor’s section of six fighters had splashed five of the nine bombers, and both the bombers and the darting fighters could be seen by the men of the task force. As the four remaining bombers released their bombs amid bursting AA projectiles, CAPT Sherman was busy twisting and turning Lexington at thirty knots to avoid the falling missiles. None struck the ship and the bombers turned back toward their base with the fighters after them. One of the Wildcat pilots, ENS John W. Wilson was having engine troubles that required him to attack his target from directly astern, which brought him under the fatal fire of a tail gunner that managed to put an explosive twenty mm round into the fighter’s cockpit, bringing the Wildcat down. A rescue destroyer found neither airplane nor survivor. During the melee, Gill launched four more fighters under the charge of LCDR Jimmy Thatch, and sent Thatch and his wingman after the retreating bombers. They soon reduced the force to one surviving bomber. That bomber, however, fell prey to the machine guns of a patrolling Douglas Dauntless dive bomber. All nine of the attackers had been downed. At 1649, while the CAP pilots were still chasing the retreating raiders, the CXAM operator detected another large raid approaching from zero one five degrees, the opposite direction of the first raid, and at thirty miles. As he had been trained, Red Gill had left two CAP fighters, led by Lieutenant Edward H. “Butch” O’Hare, stationed in reserve over task force center. Gill vectored the two Wildcats toward the nine oncoming bandits, which by this time could be plainly seen by men on Lexington’s deck. [6, p.39] [33, pp.97-101]
O’Hare and his wingman, LTJG Marion W. Dufilho, tallyhoed the bandits, again with no fighter escort, only a few miles astern of the task force. The two Wildcats had a 3,000 foot altitude advantage which allowed them to make a high side run, and O’Hare was able to mortally wound two of the bombers by hitting one in an engine nacelle, destroying the engine. He then slightly shifted his aim to the second bomber, giving its right engine the same treatment. O’Hare would remark later that his main worry at that time was staying clear of the first falling Mitsubishi G4M1 bomber. By this time he was aware that all four of his wingman’s guns had jammed, and he was on his own. Next, O’Hare shifted his attention to three bombers in a vee formation on the port side of the group, and again in another high deflection run concentrated on an engine nacelle. By this time there were five G4M1s left, they were directly over the task force, and every gun that could bear on the five bombers was concentrating on the group, including the lone Wildcat. It was probably his skill in high deflection shooting that enabled him to attack from abeam of the bombers rather than having to traverse the very dangerous heavily defended area at the rear of the formation, and that probably kept him alive. With five-inch projectiles bursting around both the bombers and his fighter, O’Hare next selected the flight leader’s aircraft to disable the master bombardier and hopefully destroy the aim of the remaining bombardiers. Again the port engine nacelle in the lead bomber received the fifty caliber slugs of four guns, and the hit was so concentrated that the engine seized and tore itself off its mounts. The lead bomber started down, out of control. Simultaneously the remaining four bombers released their payloads which missed the wildly maneuvering Lexington’s fantail by 100 feet. Crewmen on the decks of task force ships cheered O’Hare on, and even the ensign operating the CXAM momentarily stepped out of the radar shack to add to the cheering. O’Hare had clearly destroyed three G4M1s and mortally wounded two others that would have to ditch; giving him a score of five, at a price of one enemy bullet hole and two “friendly” flack holes in his Wildcat. Thanks to CAPT Sherman’s radical maneuvering, Lexington remained undamaged, and O‘Hare became the Navy’s first ace of WW II. Later, after reading the battle report, Chief of Naval Operations, Admiral Ernest J. King expressed his displeasure that the first CAP had spent time and ammunition chasing a retreating formation, stating, “..fighter patrol primary function is protection of surface units and not pursuing planes retiring.” [11, pp.18-19] [44, p.52] [6, p39] [15, p.55]
On 27 February 1942, the seaplane tender Langley, formerly the U.S. Navy’s first aircraft carrier, was ferrying thirty-two P-40 fighters to Java when she was attacked without air escort in the Java Sea by Japanese dive bombers. She was so badly damaged that her surface escorts had to sink her. Then on 5 April, LCDR Takshiye Egusa, commander of the Japanese carrier Soryu’s Air Group, found the British cruisers HMS Cornwall and HMS Dorsetshire without air cover in the Laccadive Sea off Columbo, Ceylon. Egusa attacked with eighty dive bombers and sank both ships in nineteen minutes. Then on 9 April, the British carrier HMS Hermes, having no serviceable aircraft on board, was sunk by Japanese dive bombers in the Bay of Bengal. These incidents again showing the vulnerability of capital ships to air attack. [15, pp.48-50]
Engineer Robert M. Page at the Naval Research Lab realized early in his radar research that the capability to identify friendly aircraft should be incorporated with radar, and he began work on radar recognition equipment in 1937. By early 1940 he had working prototypes of both aircraft-to-ship and ship-to-ship identification devices. At this time the General Electric Company was brought in to participate in production design as the prospective large scale producer. In mid 1940 navy and army representatives agreed that the new IFF sets would be used jointly by both services, and the designs were adapted to accommodate army requirements. By summer 1941 GE was producing engineering development models, with the the air/ship version being designated “IFF ABA” and the ship-ship version “IFF BI.”
The Anglo-American cooperation in radar that started with the Tizard Mission soon cast a cloud over ABA and BI production. It had been agreed that both nations would use the same recognition devices to prevent misidentification of the other’s ships and aircraft, and Page had argued strongly for the NRL system. The British were concerned, however, that the ABA/BI, which operated on a fixed frequency independent of radar frequency, had its frequency set too close to the operating frequency of the German Wurzburg radar. This could result in the Germans soon discovering the existence of Allied IFF and devising countermeasures. It was thus agreed that the British system, with modifications, would be used.
The British Mark II IFF did not use a single frequency, but rather scanned through the frequencies of all radars and responded on the frequency of the radar that had interrogated it. The Americans were concerned that as more radars with diverse frequencies were added to inventory, the Mark II could become overloaded, especially in the heat of battle. All agreed that the Mark II system would be modified to work at a fixed frequency, and would be designated IFF Mark III. It was also agreed that the American ABA/BI would be produced in limited quantities for testing so that it could be ready as a backup system if the Mark III were compromised. On 1 December 1941, the Chief of Naval Operations authorized American production of both the ABA/BI sets and the Mark II and Mark III with a priority second only to radar itself. More than 3000 of the ABA/BI sets were delivered for testing by the Army and Navy starting in early 1942. For both U.S. and U.K. use, the U.S. agreed to start large scale production of Mark II sets as a stopgap, pending development and test of the Mark III; then Mark III sets would be manufactured in the U.S. for both sides. Mark III sets started trickling into the fleet by May 1942. [1, pp.179-180] [23, p.255] [27, p.22]
Lexington Goes Down
In early 1942, The Japanese high command decided to strengthen their South Pacific defensive perimeter by invading and establishing bases at Port Moresby in New Guinea and the island of Tulagi, just north of Guadalcanal Island. U.S. Navy codebreakers, having broken the Japanese Navy JN-25 code, learned of the invasion plans, and notified the staff of Fleet Admiral Nimitz, Commander in Chief, Pacific (CINCPAC). In the Pacific Fleet, Task Force 17, commanded by Rear Admiral Frank J. Fletcher, was already enroute to the Coral Sea. This force consisted of the carrier Yorktown, three cruisers, and six destroyers supported by two replenishment oil tankers. Also in the South Pacific, Nimitz had the carrier Lexington, two cruisers and five destroyers comprising Task Force 11 under the command of Rear Admiral Aubrey Fitch. On 29 April, CINCPAC ordered the two task forces and a squadron of three Australian and American cruisers and one destroyer to proceed toward Tulagi. A Japanese fleet of troop transports, cruisers, and destroyers with air support by the light carrier Shoho had sent troops ashore on Tulagi without opposition on 3 May. By the next day, most of the invasion force had departed, but Yorktown was close enough to stage a raid on Tulagi, sinking a destroyer, destroying five seaplanes and shooting up Japanese forces ashore. In the process the Japanese shot down two Wildcat fighters providing air cover. [15, p.58] [33, p.73, p.190]
The total complement of aircraft assigned to the two U.S. carriers numbered 130, and of that number, only six so far had been equipped with IFF sets. Unfortunately, the two downed Wildcats carried two of the sets. The lack of IFF sets caused the FDOs considerable difficulties because every new radar contact had to be assumed hostile until it was identified. Many times it could only be positively identified by sending a CAP fighter out to check it over. The only other means of identifying returning strike aircraft required the flight leader to break radio silence using a prearranged code word and to give his direction of approach. [9, p.241]
While the Tulagi campaign was under way, Admiral Takagi had assembled an invasion force at Rabaul, to be shepherded by the heavy aircraft carriers, Shokaku and Zuikaku. This force was steaming south from Rabaul, and would soon join up with Shoho and the invasion force retiring from Tulagi. Together they would make the attack on Port Moresby, planned for 10 May. Their plan was interrupted on the morning of seven May when both sides began suspecting that heavy enemy forces were in the area. At 0833 LT Red Gill’s radar operators picked up a bogey at 270 degrees and only twenty-five miles away. Further plotting of the bogey showed it was circling over a fixed position. CAPT Sherman denied Gill’s request to send a fighter to check it out because the skipper thought the circling contact was probably friendly. Bear in mind that at this time only a small percentage of American planes were IFF equipped. When the contact persisted in staying put rather than approaching for landing, Gill convinced his captain that it should be checked out, and he was allowed to send out a two-plane section that never found the bogey because of heavy concealing clouds. Gill was not only sure the bogey was a hostile snooper, but also that it had reported Task Force 17’s position, and he so advised his seniors. In actuality, it and a few other Japanese cruiser-launched floatplanes had been shadowing and reporting on the task force. In the mean time, increasing cloud cover was getting so bad that Gill had to caution his CAP not to get beyond visual range of the task force. [15, pp.58-59] [33, pp.192-193]
Meanwhile at 0815, a patrolling SBD-5 (Scout Bomber Douglas-5) “Dauntless” dive bomber reported sighting two Japanese carriers with four cruiser escorts located about 200 miles northwest of Task Force 17. Admiral Fletcher was sure that the two Japanese fleet carriers Shokaku and Zuikaku had been located, and issued orders for an immediate air strike with all available attack aircraft accompanied by covering fighters. A fleet carrier normally carried four aircraft squadrons: one of fighters, one of torpedo bombers, one of dive bombers called “Scouting”, and one of dive bombers called “Bombing”. Both scouting and bombing were equipped with Dauntlesses, and they would usually divide the work of scouting and bombing. The senior squadron commander of the four squadrons was called Commander Air Group (CAG) and he fit in the chain of command between the carrier’s Air Officer, who reported to the carrier’s skipper as a department head, and the squadron commanders. The Air Officer, also called the ‘Air Boss’ had other division heads reporting to him including the: flight deck officer, ordnance officer, landing signals officer, supply officer, gasoline officer, hangar deck officer, engineering officer, catapult officer, photographic officer, and arresting gear officer. One of the principal jobs of Commander Air Group was flying on torpedo, bombing, and fighter strikes as the strike coordinator, usually riding in the back seat of a “command” dive bomber. Aboard Yorktown, Lieutenant Commander Oscar Pederson, Commanding Officer of Fighter Squadron 42 was CAG, but the carrier’s commanding officer, CAPT Elliot C. Buckmaster, had other work in mind for Pederson than flying with strike groups, or even flying at all. Buckmaster reasoned that Pederson knew more about flight operations than any other officer aboard, and that made him most qualified to be fighter director. To him, adroit fighter direction was far more important than leading a strike group. Pederson’s unhappy protests were to no avail. [52, p.239] [34, p.7] [33, p.54, p.56, p.169]
RADM Fitch, on board Lexington, was senior naval aviator present, and RADM Fletcher assigned him as “Commander, Air” meaning he would be the officer in tactical command (OTC) until the air action was over; also meaning that LT Red Gill, Lexington’s senior fighter director, would be running the combined task forces’ air defense. This meant LCDR Pederson would be taking direction in deploying Yorktown’s fighters from LT Gill as representative of the OTC. That morning (7 May) the aircraft order of battle on the two U.S. carriers was: Yorktown —thirty-five SBD dive bombers, ten TBD torpedo bombers and seventeen F4F fighters; Lexington—thirty-five SBD, twelve TBD, and nineteen F4F. Admiral Fitch launched ninety-three of these from the two carriers, including eighteen escorting Wildcats. In the mean time, the Dauntless pilot who had sent in the “two carriers” sighting report returned to Yorktown, and when asked for further details about the sighting, stated he had reported only two cruisers and two destroyers. A quick check of his cipher pad showed he had made a coding error in ship type, but within minutes another sighting report was received of a carrier, sixteen other combatants, and ten transports about thirty miles from the four ships originally sighted. At 1050, the American strike force found the Japanese light carrier Shoho, its escorts, and transports; which turned out to be the part of the Port Moresby invasion force. In the ensuing attack, fifty-three dive bombers and Douglas TBD torpedo bombers went after the light carrier which absorbed thirteen bombs and seven torpedoes and was soon on the bottom. [33, pp.193-205]
Admiral Takagi was now sure that American aircraft carriers were in the area, and turned his troop and cargo ships back toward the safety of Rabaul, canceling the Port Moresby invasion. It was far more important for his battle force to find and destroy the U.S. carriers. He immediately launched searchers and an attack force that by noon on the seventh thought they had found the American main body. What they found, however, was the supporting oiler USS Neosho and escorting destroyer Sims. The destroyer, taking three bomb hits, broke in half and sank, Neosho was badly damaged, left dead in the water, and unable to get under way again. Four days later the U.S. destroyer Henley found the oiler, and after taking survivors off both Sims and Neosho, sank the oiler to prevent capture. On the evening of the seventh, as the sun was setting, ADM Takagi launched another air strike of twenty-six planes that did not find the American carriers, most likely because of the very bad weather, low overcast, and growing darkness. But it did make for an interesting evening for the fighter directors. At 1648 Yorktown’s CXAM operator detected a bogey at eighteen miles range on bearing 250 degrees. On Lexington, LT Gill’s operators could not find the contact, so Yorktown’s LCDR Pederson took over fighter direction to check it out. He vectored out a division of four fighters, but clouds and heavy weather prevented them from finding it. Postwar analysis of Japanese records does not reveal whether the snooper sent in a sighting report. [33, p.210] [15, pp.58-59] [5, pp.176-177]
Next, at 1747, Lexington’s radar operator detected what appeared to be a fairly large enemy formation approaching the task force on a bearing of 144 degrees and range forty-eight miles. LT Gill had a CAP of eight Wildcats overhead, but a quick check on their fuel status told him they each had only about sixty gallons remaining which would require him to put them exactly on target with no fuel allowance for searching. He sent all eight out, but soon realized that only four of them had IFF, and he needed to know exactly where each of his friendlies was. He recalled the four without IFF to stay over TF center, and he also got permission to scramble more Wildcats from both carriers because the admirals had a feeling this was going to be the expected major Japanese attack. Soon there were thirty fighters aloft. Gill had to make two minor course changes that brought the first fighters right on top of nine torpedo bombers arrayed in a scouting line. In a few minutes the Wildcat pilots shot down three of the bombers and badly damaged one more. A few minutes later a torpedo bomber and a Wildcat were both consumed in a huge explosion. It was thought that the pursuing pilot, LTJG Paul G. Baker, had either made a direct hit on the plane’s torpedo warhead or had collided with his intended victim. The rest of the shattered torpedo formation turned back to their carrier. [33, pp.210-212]
While the first four fighters, now very low on fuel, returned to the task force, Gill vectored out seven of the recently airborne Wildcats on a heading of 240 degrees to investigate a new bogey. Meanwhile, as a reserve, he stationed eight CAP over Yorktown and ten over Lexington. Gill figured the new contact was another snooper, but suddenly six torpedo planes momentarily exited a cloud bank and flew directly below the seven fighters. Gill had put them right on the target again and they were headed directly for the task force. The Wildcat pilots attacked and downed one of the six on their first pass, causing the Japanese formation to break up, with the seven fighters also separating in the growing darkness and heavy weather to pursue the departing torpedo planes. The final score was two enemy torpedo bombers shot down and one ditched on return to his carrier; at the cost one Wildcat pilot, Ensign Leslie L. B. Knox, who never returned. It was conjectured that he had either been hit by a rear seat gunner, or had gotten lost in the bad weather with malfunctioning radio and homing receiver. Next, four of the remaining six fighters came across a group of enemy dive bombers groping to find the task force in the almost dark sky and low clouds. In the resulting encounter, where the targets were nothing more than dark shadows, one more Japanese dive bomber went into the water with its tail blown off. In all, nine Japanese attackers had been shot down at the cost of two Wildcats and their pilots.
The Japanese attack force had been dispersed into small groups that were now more interested in finding their own carriers than finding Task Force 17. To get better mileage from their remaining fuel, they jettisoned bombs and torpedoes while searching for their roosts. In the gathering darkness some Japanese pilots did find the U.S. carriers, but thought they were their own carriers. As Yorktown and Lexington were retrieving fighters, Gill’s CXAM operators kept finding unidentified bogies milling around the edges of the task force, but CAP planes sent out to check them out could not find them, thanks to clouds and growing dark. One of the Japanese pilots actually joined a landing circle and flashed a request to land. Another lined up to land aboard Yorktown, but gunners on screening ships finally realized there were Japanese planes mixed in with friendly planes, and repulsed it with AA fire. Their bursts also came uncomfortably close to some of the Grummans, causing one pilot to radio “What are you shooting at me for? What have I done now?” as LT Gill recorded in his FDO log. Now aware that the carriers below were not theirs, but having rid themselves of their weapons, the enemy pilots decided to search elsewhere for their floating homes. Most finally did, and eleven crash landed on their own carriers. In all, twenty-one of the original twenty-six Japanese attackers were lost or disabled. Both sides planned to launch air attacks early the next morning: 8 May. [33, p.190, pp.211-214]
The American flyers most likely would never had made such a mistake as trying to land on an enemy carrier because their planes were fitted with a special radio receiver, designated “ZB,” that worked in conjunction with a “YE” homing transmitter aboard the aircraft carriers. The rotating YE antenna was synchronized with the ships gyrocompass so that it sent out in geographically fixed thirty-degree sectors a series of letters in Morse code. The pilots had a chart on their navigation board that told them which of the twelve letters they would hear in each sector. Thus by hearing a particular letter in Morse code in heir headset they knew the approximate direction back to their carrier. Most Japanese dive and torpedo bombers were equipped with a similar homing receiver, however, fighters were not, most likely to reduce weight. It is surmised that their homing system did not work as well as the American system. Also, Japanese pilots later noted that American radio communications seemed to jam their homers. [25, p.15] [23, pp.273-274] [52, pp.181-182]
On the morning of 8 May RADM Fletcher estimated that the Japanese main force was about 170 miles somewhere to the north of his task force. On Lexington at 0807, LT Red Gill’s radar operators found a contact at range twenty-two miles, bearing 330 degrees, and they estimated it was flying quite low. It was proceeding at high speed on a heading of 240 degrees, and Gill vectored fighters after it, but with no sighting. By 0816 it was no longer on the CXAM, but Gill advised his seniors he was sure it was a snooper, and that the Japanese had found TF 17. Both sides had launched scouting planes at dawn, and at 0820 Lexington received a messagefrom one of the patrolling SBDs that the Japanese main body was almost due north of Task Force 17 at 175 miles and steaming south at fifteen knots. A few minutes later Lexington’s CO, CAPT Sherman, was handed an interpreted voice radio transmission from one of the Japanese scouts. It gave TF 17s location, course, and speed; LT Gill was right, Admiral Takagi knew where they were! CAPT Sherman estimated the Japanese attack would arrive about 1100. It would turn out that the two forces were almost evenly matched in air power, the Americans having 122 airplanes, and the Japanese 121. [5, p.177] [28, pp.52-53] [34, p.222]
At 0915, Yorktown competed launching her attacking group, and Lexington had all her planes in the air by 0925. Total USN aircraft in the two groups were forty-six dive bombers, fifteen fighters, and twenty one torpedo planes. In the meantime, Tagaki’s carriers, by 0915, had launched thirty-three dive bombers, eighteen fighters, and and eighteen torpedo planes. Total attacking aircraft: eighty-two American, sixty-nine Japanese. Yorktown’s group found the enemy formation first and concentrated on the carrier Shokaku, as Zuikaku was hidden under a rain squall. The nine TBD Devastators launched their torpedoes, but made no hits. They apparently launched from too great a distance, and the torpedoes were so slow that the carrier could turn and outrun the torpedoes. The fifteen dive bombers did better. Making two bomb hits on Shokaku, they ignited fires, but it could still recover aircraft. Lexington’s dive bombers soon fixed that by scoring another hit that knocked out ability to handle aircraft. Badly wounded, the Japanese carrier had no choice but to begin the return trip to Japan where she would be out of action for a critical month. [5, pp.177-178] [28, pp.53-56] [15, pp.59-60]
Admiral Fletcher was expecting the arrival of the Japanese at any time, and at 1008 lookouts aboard spotted a large four-engine flying boat at very low altitude, low enough to apparently have avoided the task force’s CXAMs. CAP Wildcats immediately spotted the intruder, and as had been done to other Kawanishi flying boats, they riddled the engine nacelles and fuel tanks with tracers until the huge craft exploded in mid air. Again, strong evidence that the Japanese knew where they were and would soon be on the scene. At 1012, CAPT Sherman, skipper of Lexington, had LT Gill launch ten Dauntless dive bombers to circle the task force as a torpedo plane screen. With their two forward firing fifty caliber machine guns plus two flexible thirty calibers in the rear cockpit, Sherman thought they made pretty good fighters in a pinch, although the dive bomber pilots did not much like the duty. At 1015 LT Gill advised the air bosses on both carriers to bring all deck-bound fighters up to standby. [33, p.244]
As Captain Sherman had predicted, the blip of the main attacking group appeared on the CXAM A-scope at 1048. They were to the northwest at 68 miles. LT Gill noted in his log that, once detected, they never again disappeared from the screen, indicating they had for some time been obscured by a big fade zone. Commander Jim Dudley was Lexington’s navigation officer and served as officer of the deck during the battle. He later noted that the CXAM that had done so many great things for them in the past, really let them down when needed most because of the blind zone that had allowed the enemy to get so close without detection. At the time of detection, Gill had eight Wildcats on various patrol stations aloft. There was also the lower patrol of Dauntlesses circling the task force watching for torpedo planes, but they were not under Gill’s control. One of the four-plane divisions had been up for seventy minutes and the other for ninety, and he did not think either had enough fuel to make a long-range intercept. Instead, Gill got permission to launch nine of the fighters he had on standby: five from Lexington and four from Yorktown. Then he sent out the “Hey Rube” call to the eight airborne fighters, meaning they should return to their carriers and orbit overhead, waiting instructions. At 1101 he vectored the five Lexington fighters out on a heading of 020 degrees and told them to climb to 10,000 feet. After a few minutes, Gill told the five to go to “Buster”, meaning maximum sustainable power. There was a low cloud layer that would prevent fighters at 10,000 feet from spotting low flying torpedo planes, so at 1107 he ordered two of the five to a lower altitude to look for torpedo planes. [28, pp.65-68, p.70, p.77, pp.80-81]
At 1108 Gill contacted the four Wildcats that had just launched from Yorktown and sent them out at 1,000 feet on heading 020 degrees to also look for low flying torpedo planes. At 1109 the three fighters at 10,000 feet sighted the raid that they estimated to number about sixty aircraft and stacked in altitude from 10,000 to 13,000 feet. Covering fighters were on top, next below were dive bombers, then another layer of fighters, and torpedo bombers at the bottom. At intercept, they were only twenty miles from formation center, and the Wildcat flight leader elected to climb to intercept the dive bombers as the torpedo bombers began their dive on the task force. In the excitement, there was apparently no “tallyho” report so Gill did not get the altitude information he badly needed, but he could tell from radio calls that the section was attacking. Gill notified the CAP stationed over the carriers at 1111 that the three extended fighters were attacking and that CAP needed to keep a sharp lookout. Meanwhile, the six fighters at the lower altitude saw nothing but some of their own Dauntlesses on anti-torpedo plane patrol. [33, pp.247-248]
The Japanese torpedo planes were under the command of Lieutenant Commander Shigekazu Shimazaki who deployed fourteen to attack Lexington and four to Yorktown. By this time the nine Wildcats that Gill had vectored out on 020 degrees were coming back to fleet center, and one Wildcat did away with one of the torpedo bombers; and the eight CAP over the carriers got three more. Next, the Zeros escorting the torpedo bombers shot down four of the patrolling SBDs. CAPT Sherman observed that the torpedo attack was very well coordinated. The attackers dispersed themselves all around Lexington and came in from all points of the compass so that if he turned the ship to avoid one torpedo he would have to cross the track of another. He also noted that the two planes that came in from ahead placed themselves on either side of the bow so that avoiding one torpedo would bring him into the path of the other torpedo. By this time the defending Wildcats were mixing it up with the Japanese dive bombers directly over the task force with AA projectiles bursting all around them. At 1118 Lexington took a torpedo hit on the port side, and two minutes later another on the same side. At 1122 a third torpedo hit, also on the port side. The damage control officer called the bridge and told them that if they took any more torpedoes they should take them on the starboard side. [28, pp.66-81] [11, p.19] [5, p. 178]
Meanwhile, four torpedo planes had launched torpedoes at Yorktown, but thanks to CAPT Buckmaster’s adept maneuvering, all missed. A few minutes later, however, an Aichi D3A dive bomber managed to score with a 500 pound bomb that penetrated the flight deck and exploded beneath it, killing 66 men. About this time LT Gill was having troubles with his fighter net radio and was about to recommend switching fighter direction to Yorktown. However, when he heard of the bomb hit on the companion carrier, he decided to wait for radio repair. During the battle the two carriers maneuvered separately and separated to the extent that LCDR Pederson on Yorktown had to take over his own CAP. However, at 1131 the CXAM on Yorktown went down. Because LT Gill’s fighter direction radio was not working, Pederson could not ask Gill to take over his CAP; all he could do was tell his CAP to “Protect the force!” Then he instructed the radar officer aboard the cruiser Chester to take over task force fighter direction. By 1141 his CXAM was back up and Pederson resumed TF fighter direction for a few minutes until LT Gill returned to the air. Then they busied themselves with reorganizing the CAPs and getting ready to recover strike aircraft. [15, p.60] [28, pp.84-85] [33, p.269]
By 1132 the attack was over and Lexington was still afloat. In total, Lexington had taken at least three torpedoes and three bomb hits. There were some localized fires, but they seemed to be under control. In spite of the damaged flight deck the ship was able to recover returning aircraft, many of which were severely damaged. Around noon, LCDR Bill Burch was leading a group of twelve Dauntless dive bombers returning from the air strike, when he spied six Zero fighters returning from the strike on the American carriers, and heading right toward him. He radioed LT Gill at 1203 asking for some fighter cover. Gill’s CXAM operators found the dive bomber group out at thirty miles north of the task force, but he had only eight cap protecting the two carriers. He could not afford to send fighters that far away from the carriers, but he did move some of the CAP ten miles further north so they could meet the bombers and cover their return. Fortunately, it seems the Zeros never saw the bomber formation, or possibly they were out of ammunition.
At 1223 LT Gill entered in his log that the ship was making twenty-five knots without trouble. However, at 1241 the bridge was notified that there was a very strong smell of aviation gasoline in the chief petty officer messing spaces and the general workshop. CAPT Sherman later wrote that at 1247, by his watch, the ship was violently shaken by a “terrific internal explosion.” It was gasoline fumes exploding, which then ignited gasoline fires from ruptured tanks and piping throughout the ship. By 1406 Lexington CXAM operators could no longer train the flying bedspring antenna, and Gill transferred fighter direction duties to LCDR Pederson on Yorktown. At 1452 hours, LT Gill logged that the fires were not under control, and around 1500 Yorktown's CXAM again began acting up, causing Pederson to direct the cruiser Chester to assume the radar watch. Around 1700, Pederson’s technicians had the radar working again to the extent that operators were sure they had a large unidentified bogey. Pederson had seven Wildcats launched from Yorktown, and warned all task force ships to be ready to repel another air attack. Chester offered to direct the fighters, but Pederson said his CXAM seemed to be working reasonably well and he would take control. Fortunately, the sighting was a false alarm. At 1710 CAPT Sherman ordered the Lexington crew to abandon ship.
When he heard the abandon ship order over the speaker in Air Plot, assistant fighter director Ensign Stan Foote removed the dead reckoning tracer plotting sheet on which Lexington’s track had been recorded before and during the battle. He rolled it up and slid it into a waterproof cardboard tube. Then he reported to his abandon ship station and was soon over the side. The tube helped buoy him up until a destroyer crew hauled him out of the water. The destroyer transferred Foote to the cruiser Astoria, and from 8 to 12 May he picked position data from Astoria's DRT to add to his plot. He then drew up a duplicate copy of the sheet that he provided to CAPT Sherman who incorporated the plotting sheet into his Coral Sea battle report. Many years later Captain Foote showed this writer the original of Lexington’s water stained DRT plotting sheet. To prevent Lexington from falling into enemy hands, the destroyer Phelps had to hit her with five torpedoes before she would go under. [36, IV:p.105] [28, pp100-169] [6, p40] [33, pp.271-272, pp.279-280]
RADM Fletcher and CAPTs Sherman and Buckmaster all indicated in their battle reports their dissatisfaction with fighter direction during the battle. They generally concluded that Lexington’s loss was primarily due to inadequate fighter direction, however their criticism fell mainly on plainly deficient fighter direction facilities and equipment. They did note that the fighter directors sent out too few fighters to intercept such a large raid, and that the fighters were positioned too low. Their reports also noted that the first fighters to encounter the raid did not send a complete ‘tallyho’ report giving raid size, composition, and altitude. Further, the fighter radio net was so crowded with fighter-to-fighter talk that the FDOs could barely get in their directions. We should also remember that Lexington’s fighter net radio was experiencing difficulties during part of the battle. CAPT Sherman recommended that when an enemy dive bomber attack is expected, the CAP should be directed to 20,000 feet, and CAPT Buckmaster recommended the same altitude when the raid altitude is unknown. The problem, however, with following these recommendations is a CAP at that altitude would probably not see aircraft at a much lower altitude. LT Gill later stated that this is why he had sent his initial fighter thrust out at 10,000 feet so that they could spot low fliers and high fliers equally well. The need for height finding radar was made abundantly clear. They also recommended that FDO should attempt to make interceptions at least thirty miles from task force center. In defense of the fighter directors, it was realized that this was the most complex fighter direction operation yet experienced, and that conditions of continuous radio silence had never allowed FDO/pilot practice at any where near this scale of complexity. [33, pp.303-304] [11, pp.20-21]
It was also noted that, in general, when there were only a few raids and a few groups of CAP, fighter direction was usually very good, but as the number of aircraft and their dispersion increased it became very difficult to maintain an accurate air plot of all the friendlies and bogeys. A lot of the difficulty was in the processing of CXAM radar data. It must be remembered that the antenna train control was by hand, and that the antenna had to be stopped each time a blip showed on the A-scope, and then the antenna had to be slewed slightly back and forth to get a good bearing on the maximum return. The operator then had to read target range and target bearing from two different indicators before he phoned, or radioed, the data to Air Plot. Furthermore, the operator had to make an estimate of raid size from the configuration of the pip, or multiple pips close together. Even with multiple radars throughout the task force providing data, obtaining and plotting track information was a laborious, time consuming job, and it can be seen why radar plots could fall critically behind time in an intense air battle. Finally, once the tracks were plotted, there was still the need to estimate target altitude based on range gaps where the tracks faded out. Lack of IFF equipment in most of the airplanes made the FDO’s work even more challenging, often necessitating sending precious CAP out to check on bogies that more often than not turned out to be friendlies.
CAPT Buckmaster concluded his report with the following:
The makeshift Radar Plot in this vessel, wherein all functions of Radar Plot are attempted to be accomplished in a corner of Air Plot, again showed itself, during the air attack on May 8, to be woefully inadequate to enable complete use to be made of all the information which the combined radars of own and other ships are capable of furnishing, or even to use with full effectiveness information which can be furnished by this one vessel’s radar. In order that Radar Plot may properly perform its function it must:
a. Be, in itself, a complete unit
b. Have sufficient room to allow the Fighter Director and his plotting and communication assistants to perform their functions without mutual interference.
c. Be so isolated as to be relatively free from spectator interference and from noise interference from other activities. Both spectator and noise interference are unacceptably great with radar installed in a corner of Air Plot.
d. Have its own radio communications, capable of transmitting or receiving on any aircraft circuit, and on a super-frequency circuit with other search-radar equipped vessels and with other Fighter Directors.
e. Be provided with interior communication channels connecting to signal bridge, lookouts, flag, and important ship and fire control stations.
f. Be contiguous to Air Plot and should have means for actual physical conversational communications with Air Plot.
g. Be provided with plotting facilities sufficient to allow two simultaneous radar plots to be run: one search plot, and a Fighter Director (tracking) plot.
h. Have sufficient blackboard and extra chart board space to allow a complete picture to be maintained of the situation of our own aircraft and of the general and immediate tactical situation.
Lastly, Capt Buckmaster noted, “Distribution of aircraft radar recognition equipment [IFF] should be continued with highest priority.” In item d. above when the captain refers to a “super-frequency circuit” he is writing of what today is called ‘very high frequency (VHF)’ radio. Most navy voice communication radios of the early 40s used ‘medium frequency’ waves that had the property of long ranges caused by their following the curvature of the earth by repeated reflections off of the ionosphere. This made them vulnerable to enemy radio direction finding and to locating the transmitting ship by triangulation. On the other hand, VHF waves do not follow earth’s curvature but travel out into space following optical line of sight. This makes them much less liable to radio direction finding. It can be noted that concern about position disclosure by enemy radio direction finding had a severe dampening effect on live fighter direction practice which would have benefited both the FDOs and the pilots, and might have made a difference in the outcome of the battle. Yorktown’s CXAM had been inoperable for a period during the attack and the skipper proposed that all carriers should have two radars equivalent to the CXAM. Admiral Nimitz seconded the recommendation, and he also recommended that a carrier's fighter allowance be increased to thirty-six. [33, p.53] [11, pp.20-21]
On reading Buckmaster’s report, the Chief of Naval Operations, Admiral Ernest J. King, wrote, “This project is well in hand for new construction.” He also expressed, “...the urgent need for fighter director doctrine and experienced directors.” Reflecting on this comment, former FDO CAPT “Nick” Hammond wrote in 1988, “Looking back on those days, one wonders where those ‘experienced directors’ were supposed to come from. The most experienced people were those that just went through the Battle of the Coral Sea.” [11, pp.20-21] [25, pp.35-36]
At Coral Sea, the Japanese felt they had won a great tactical victory by sinking Lexington, the oiler Neosho, and the destroyer Sims, experiencing only the loss of the light carrier Shoho and one destroyer. But in many respects, the Americans had won a strategic victory. They had not only thwarted the invasion of Port Moresby, but had also prevented two Japanese fleet carriers and one light carrier from being available for the next critical sea battle at Midway Island. Shoho was gone forever. Shokaku was so badly damaged it would not make the battle, and Zuikaku had lost so many aircraft and aircrew that she could not participate.